Re: [Kicad-developers] make install problems..
Thanks! I will try this some day, but will use nightlies for now. PS. Would it be possible to have a barebones kicad nightly package that only includes the app(s) and not the footprints, symbols and 3D models? I already have those cloned separately. 1.3GB is quite a lot to download each time one want to try the latest build. On Mon, Jun 14, 2021 at 8:35 PM Jon Evans wrote: > > Hi Jonatan, > > Using your system Python is no longer supported after the recent Python > changes. > > Please see the instructions here: https://dev-docs.kicad.org/en/build/macos/ > and obtain kicad-mac-builder for the dependencies here: > https://gitlab.com/kicad/packaging/kicad-mac-builder > > Adam has indicated that he thinks it may be possible to support system > Python in the future, but there are other higher priorities right now > to get 6.0 ready for release. > > Best, > Jon > > On Mon, Jun 14, 2021 at 2:32 PM Jonatan Liljedahl wrote: > > > > So, it's been a while since I rebuilt so I updated my branch to > > current master and tried. Build went without errors, but make install > > fails: > > > > ... > > -- Analyzing > > app='/Users/lijon/Coding/kicad/build/install_py3/gerbview.app/Contents/MacOS/gerbview' > > -- bundle='/Users/lijon/Coding/kicad/build/install_py3/gerbview.app' > > -- > > executable='/Users/lijon/Coding/kicad/build/install_py3/gerbview.app/Contents/MacOS/gerbview' > > -- valid='1' > > -- executable file 1: > > /Users/lijon/Coding/kicad/build/install_py3/gerbview.app/Contents/MacOS/gerbview > > -- executable file 2: > > /Users/lijon/Coding/kicad/build/install_py3/gerbview.app/Contents/MacOS/Python.framework/Versions/3.9/Resources/Python.app/Contents/MacOS/Python > > -- verified='0' > > -- info='external prerequisites found: > > f='/Users/lijon/Coding/kicad/build/install_py3/gerbview.app/Contents/MacOS/Python.framework/Versions/3.9/Resources/Python.app/Contents/MacOS/Python' > > external_prereqs='/usr/local/Cellar/python@3.9/3.9.4/Frameworks/Python.framework/Versions/3.9/Python' > > ' > > -- > > CMake Error at > > /Applications/CMake.app/Contents/share/cmake-3.20/Modules/BundleUtilities.cmake:1128 > > (message): > > error: verify_app failed > > Call Stack (most recent call first): > > > > /Applications/CMake.app/Contents/share/cmake-3.20/Modules/BundleUtilities.cmake:995 > > (verify_app) > > gerbview/cmake_install.cmake:84 (fixup_bundle) > > cmake_install.cmake:67 (include) > > > > > > make: *** [install] Error 1 > > > > And trying to launch the app fails: > > > > lijon@lijon-mbp master_py3 % open ../install_py3/KiCad.app > > LSOpenURLsWithRole() failed with error -10810 for the file > > /Users/lijon/Coding/kicad/build/install_py3/KiCad.app. > > > > Any ideas? > > -- > > /Jonatan > > http://kymatica.com > > > > ___ > > Mailing list: https://launchpad.net/~kicad-developers > > Post to : kicad-developers@lists.launchpad.net > > Unsubscribe : https://launchpad.net/~kicad-developers > > More help : https://help.launchpad.net/ListHelp -- /Jonatan http://kymatica.com ___ Mailing list: https://launchpad.net/~kicad-developers Post to : kicad-developers@lists.launchpad.net Unsubscribe : https://launchpad.net/~kicad-developers More help : https://help.launchpad.net/ListHelp
[Kicad-developers] make install problems..
So, it's been a while since I rebuilt so I updated my branch to current master and tried. Build went without errors, but make install fails: ... -- Analyzing app='/Users/lijon/Coding/kicad/build/install_py3/gerbview.app/Contents/MacOS/gerbview' -- bundle='/Users/lijon/Coding/kicad/build/install_py3/gerbview.app' -- executable='/Users/lijon/Coding/kicad/build/install_py3/gerbview.app/Contents/MacOS/gerbview' -- valid='1' -- executable file 1: /Users/lijon/Coding/kicad/build/install_py3/gerbview.app/Contents/MacOS/gerbview -- executable file 2: /Users/lijon/Coding/kicad/build/install_py3/gerbview.app/Contents/MacOS/Python.framework/Versions/3.9/Resources/Python.app/Contents/MacOS/Python -- verified='0' -- info='external prerequisites found: f='/Users/lijon/Coding/kicad/build/install_py3/gerbview.app/Contents/MacOS/Python.framework/Versions/3.9/Resources/Python.app/Contents/MacOS/Python' external_prereqs='/usr/local/Cellar/python@3.9/3.9.4/Frameworks/Python.framework/Versions/3.9/Python' ' -- CMake Error at /Applications/CMake.app/Contents/share/cmake-3.20/Modules/BundleUtilities.cmake:1128 (message): error: verify_app failed Call Stack (most recent call first): /Applications/CMake.app/Contents/share/cmake-3.20/Modules/BundleUtilities.cmake:995 (verify_app) gerbview/cmake_install.cmake:84 (fixup_bundle) cmake_install.cmake:67 (include) make: *** [install] Error 1 And trying to launch the app fails: lijon@lijon-mbp master_py3 % open ../install_py3/KiCad.app LSOpenURLsWithRole() failed with error -10810 for the file /Users/lijon/Coding/kicad/build/install_py3/KiCad.app. Any ideas? -- /Jonatan http://kymatica.com ___ Mailing list: https://launchpad.net/~kicad-developers Post to : kicad-developers@lists.launchpad.net Unsubscribe : https://launchpad.net/~kicad-developers More help : https://help.launchpad.net/ListHelp
Re: [Kicad-developers] Odp: Re: SPICE_VALUE behavior
I'm not sure this applies also for symbol Spice_Model fields (or Value field of Spice_Model is empty), but if it does then please make sure that one can still pass extra parameters there, such as "pot k=0.5 r=100k". On Wed, May 19, 2021 at 6:45 PM Sylwester Kocjan wrote: > Hi Wayne, > > Dnia 19 maja 2021 00:32 Wayne Stambaugh napisał(a): > > The test cases look correct so I don't see any issues here. > > > Could you please check again with better formatting below? Hertz is a > valid SI unit, so strings: "10m" and "10mHz" are valid, but > values ending with "Hzz" are invalid: > > > > > > *Scenario:* > *Input:* > *Actual behavior:* > *Expected behavior:* > *Result:* > Positive > 10m > Value parsed correctly > Value parsed correctly > PASS > Positive > 10mHz > Value parsed correctly > Value parsed correctly > PASS > Negative > 10mHzz > Value parsed correctly > Exception thrown > FAIL > Positive > 10Hz > Exception thrown > Value parsed correctly > FAIL > Negative > 10Hzz > Exception thrown > Exception thrown > PASS > > Parsing strings in V7 wouldn't hurt, that's no problem, but the easiest > solution is to reject all extra characters: > > > > *Input:* > *Actual behavior:* > 10m > Value parsed correctly > 10mHz > Exception thrown > 10mHzz > Exception thrown > 10Hz > Exception thrown > 10Hzz > Exception thrown > > Best regards, > Sylwester > > Dnia 19 maja 2021 00:32 Wayne Stambaugh napisał(a): > > Hi Sylwester, > > I don't know that we need to support SI units in spice strings but I > suppose it wouldn't hurt. That would have to wait until V7 as it is a > new feature. The test cases look correct so I don't see any issues here. > > - Wayne > > On 5/18/21 2:06 PM, Sylwester Kocjan wrote: > > Hello! > > I've noticed a minor bug in SPICE_VALUE class. Could you please let me > know if we should parse SI units in SPICE strings or ignore them, so I > can fix it? I think ngspice accepts units like, V, I, Ohm, but this may > be an overkill. The easiest solution is to ignore everything after > number and prefix. > > See below the testcases I made on AC tab of Simulation Settings dialog: > > Input: Actual behavior:Expected behavior: > 10m Value parsed correctly Value parsed correctly > 10mHz Value parsed correctly Value parsed correctly > 10mHzz Value parsed correctly Exception thrown > 10HzException thrownValue parsed correctly > 10Hzz Exception thrownException thrown > > Best regards, > Sylwester > > ___ > Mailing list: https://launchpad.net/~kicad-developers > Post to : kicad-developers@lists.launchpad.net > Unsubscribe : https://launchpad.net/~kicad-developers > More help : https://help.launchpad.net/ListHelp > > > ___ > Mailing list: https://launchpad.net/~kicad-developers > Post to : kicad-developers@lists.launchpad.net > Unsubscribe : https://launchpad.net/~kicad-developers > More help : https://help.launchpad.net/ListHelp > > > ___ > Mailing list: https://launchpad.net/~kicad-developers > Post to : kicad-developers@lists.launchpad.net > Unsubscribe : https://launchpad.net/~kicad-developers > More help : https://help.launchpad.net/ListHelp > -- /Jonatan http://kymatica.com ___ Mailing list: https://launchpad.net/~kicad-developers Post to : kicad-developers@lists.launchpad.net Unsubscribe : https://launchpad.net/~kicad-developers More help : https://help.launchpad.net/ListHelp
Re: [Kicad-developers] Building with local Python3 etc
That's good news Adam! I'll stop trying here then and just wait for the next step. Let me know when there's anything to test. FYI, I installed wxpython via pip since the homebrew'd one didn't install any 'wx' module in site-packages. Cheers On Thu, Apr 29, 2021 at 3:34 PM Adam Wolf wrote: > > Thanks Jon and Jonatan! > > I think we are only a little bit away from having this work well for > developers who want to have fast local incremental builds, packagers who need > robust from-scratch relocatable redistributable builds, and users who want > signed and notarized builds with Python 3. > > The aim is that developers will be able to use a system Python and not need > to make install or run wrangle-bundle, and that the CI builder will produce > signed, notarized builds and upload them to the CERN download servers. > > The CI builder part is already producing signed builds, and I've uploaded > notarized and signed ones to osx/testing. > > The next steps on improving the developer experience is to make sure folks > can use a system Python, and then to make sure that devs don't need to make > install. > > Adam > > On Thu, Apr 29, 2021, 7:30 AM Jon Evans wrote: >> >> Hi Jonatan, >> >> Adam and I have been bouncing ideas back and forth and he helped me get to a >> working setup, most recently with Seth's new_python branch but I think this >> should work with current master too. >> >> The key point is that we don't yet have a working way to use your system >> Python (but Adam has ideas on that and is working on it) >> >> So, the steps that worked for me are: >> >> 1) Use kicad-mac-builder to set up the dependencies. >> As described in the readme: >> https://gitlab.com/kicad/packaging/kicad-mac-builder >> run ./build.py --target setup-kicad-dependencies >> >> Note: this will install everything, including wxwidgets, wxpython, etc. >> There isn't currently a way to skip some if you already have them built >> outside kicad-mac-builder. >> Really you and I mostly need Python out of this, as it's the only dependency >> that is hard to manage right now. >> >> 2) At the end of its run, build.py will spit out a lot of suggested CMake >> flags. Save these, and then do a merge with your current CMake flags that >> you are building with. >> In particular, you'll want to carry over the flags that point at the Python >> framework created by build.py. >> >> 3) Use Adam's wrangle-bundle script to fixup the installed bundle. >> Obtain that from https://gitlab.com/adamwwolf/dyldstyle and install it (I >> used the venv method suggested by Adam in the readme) >> Run something like wrangle-bundle --python-version 3.8 --fix >> path/to/KiCad.app >> >> 4) At this point, your build should work and make install should also work >> (at least it did for me). Please let us know if you don't have success >> >> 5) There are additional steps and complications if you want to run and debug >> targets without first doing make install, as is possible on other platforms. >> I have figured out a workaround that makes this work, but it is not >> awesome, so Adam and I are discussing an easier route. >> >> We will update the documentation accordingly, things are just moving quickly >> at the moment and we're trying to have them settle down in a way that is >> easy to follow for developers. >> >> Best, >> Jon >> >> On Thu, Apr 29, 2021 at 7:16 AM Jonatan Liljedahl wrote: >>> >>> I now tried manually copying my system site-packages into the bundle: >>> >>> cp -a /usr/local/lib/python3.9/site-packages/* >>> /Users/lijon/Coding/kicad/build/install_py3/KiCad.app/Contents/Frameworks/Python.framework/Versions/3.9/lib/python3.9/site-packages/ >>> >>> Now when starting pcbnew, console gives a lot of: >>> >>> Class wxSTCPopupBaseView is implemented in both >>> /Users/lijon/Coding/kicad/build/install_py3/KiCad.app/Contents/Frameworks/libwx_osx_cocoau-3.1.5.0.0.dylib >>> (0x10f13c8d8) and >>> /Users/lijon/Coding/kicad/build/install_py3/KiCad.app/Contents/Frameworks/Python.framework/Versions/3.9/lib/python3.9/site-packages/wx/libwx_osx_cocoau_stc-3.1.5.0.0.dylib >>> (0x1198ddbf8). One of the two will be used. Which one is undefined. >>> >>> And trying to open scripting console says: >>> >>> Traceback (most recent call last): >>> File "", line 5, in >>> File >>> "/Users/lijon/Coding/kicad/build/install_py3/KiCad.app/Contents/SharedSupport/scripting/kic
Re: [Kicad-developers] Building with local Python3 etc
Hi, So I continued to try building with my local python3 (installed via homebrew), and wxpython installed via pip. Running wrangle-bundle first gave this output: lijon@lijon-mbp master_py3 % wrangle-bundle --python-version 3.9 --fix ../install_py3/KiCad.app Looking for issues in ../install_py3/KiCad.app, and fixing what we can... ERROR: Dependency /usr/local/Cellar/python@3.9/3.9.4/Frameworks/Python.framework/Versions/3.9/Python of /Users/lijon/Coding/kicad/build/install_py3/KiCad.app/Contents/MacOS/Python.framework/Versions/3.9/Resources/Python.app/Contents/MacOS/Python isn't inside the app directory. ERROR: Dependency /usr/local/Cellar/python@3.9/3.9.4/Frameworks/Python.framework/Versions/3.9/Python of /Users/lijon/Coding/kicad/build/install_py3/KiCad.app/Contents/Frameworks/Python.framework/Versions/3.9/Resources/Python.app/Contents/MacOS/Python isn't inside the app directory. ERROR: Dependency /usr/local/opt/opencascade/lib/libTKVCAF.7.dylib of /Users/lijon/Coding/kicad/build/install_py3/KiCad.app/Contents/PlugIns/3d/libTKVCAF.7.dylib isn't inside the app directory. ERROR: Dependency /usr/local/opt/freeimage/lib/libfreeimage.dylib of /Users/lijon/Coding/kicad/build/install_py3/KiCad.app/Contents/PlugIns/3d/libTKVCAF.7.dylib isn't inside the app directory. ...snip... Issues found: Found dependency issues. Still no 'wx' inside the bundle, so pcbnew scripting console can't import wx. Tried this: cp -a /usr/local/lib/python3.9/site-packages/* /Users/lijon/Coding/kicad/build/install_py3/KiCad.app/Contents/Frameworks/Python.framework/Versions/3.9/lib/python3.9/site-packages/ And then re-ran wrangle-bundle, which now crashes: lijon@lijon-mbp master_py3 % wrangle-bundle --python-version 3.9 --fix ../install_py3/KiCad.app Looking for issues in ../install_py3/KiCad.app, and fixing what we can... Traceback (most recent call last): File "/usr/local/bin/wrangle-bundle", line 8, in sys.exit(main()) File "/usr/local/lib/python3.9/site-packages/dyldstyle/bundle_wrangler.py", line 505, in main handle_bundle(app_path, args.python_version, fix=True) File "/usr/local/lib/python3.9/site-packages/dyldstyle/bundle_wrangler.py", line 166, in handle_bundle macho_types = utils.get_macho_types(app_path) File "/usr/local/lib/python3.9/site-packages/dyldstyle/utils.py", line 231, in get_macho_types macho_type = get_macho_type(p) File "/usr/local/lib/python3.9/site-packages/dyldstyle/utils.py", line 202, in get_macho_type magic = get_file_magic(path) File "/usr/local/lib/python3.9/site-packages/dyldstyle/utils.py", line 151, in get_file_magic subprocess.check_output(cmd, stderr=subprocess.STDOUT) UnicodeDecodeError: 'utf-8' codec can't decode byte 0xd8 in position 194: invalid continuation byte Any ideas? On Thu, Apr 29, 2021 at 2:30 PM Jon Evans wrote: > > Hi Jonatan, > > Adam and I have been bouncing ideas back and forth and he helped me get to a > working setup, most recently with Seth's new_python branch but I think this > should work with current master too. > > The key point is that we don't yet have a working way to use your system > Python (but Adam has ideas on that and is working on it) > > So, the steps that worked for me are: > > 1) Use kicad-mac-builder to set up the dependencies. > As described in the readme: > https://gitlab.com/kicad/packaging/kicad-mac-builder > run ./build.py --target setup-kicad-dependencies > > Note: this will install everything, including wxwidgets, wxpython, etc. > There isn't currently a way to skip some if you already have them built > outside kicad-mac-builder. > Really you and I mostly need Python out of this, as it's the only dependency > that is hard to manage right now. > > 2) At the end of its run, build.py will spit out a lot of suggested CMake > flags. Save these, and then do a merge with your current CMake flags that > you are building with. > In particular, you'll want to carry over the flags that point at the Python > framework created by build.py. > > 3) Use Adam's wrangle-bundle script to fixup the installed bundle. > Obtain that from https://gitlab.com/adamwwolf/dyldstyle and install it (I > used the venv method suggested by Adam in the readme) > Run something like wrangle-bundle --python-version 3.8 --fix path/to/KiCad.app > > 4) At this point, your build should work and make install should also work > (at least it did for me). Please let us know if you don't have success > > 5) There are additional steps and complications if you want to run and debug > targets without first doing make install, as is possible on other platforms. > I have figured out a workaround that makes this work, but it is not awesome, > so Adam and I are discussing an easier route. > > We will update the documentation accordingly, things are just movin
Re: [Kicad-developers] Building with local Python3 etc
I now tried manually copying my system site-packages into the bundle: cp -a /usr/local/lib/python3.9/site-packages/* /Users/lijon/Coding/kicad/build/install_py3/KiCad.app/Contents/Frameworks/Python.framework/Versions/3.9/lib/python3.9/site-packages/ Now when starting pcbnew, console gives a lot of: Class wxSTCPopupBaseView is implemented in both /Users/lijon/Coding/kicad/build/install_py3/KiCad.app/Contents/Frameworks/libwx_osx_cocoau-3.1.5.0.0.dylib (0x10f13c8d8) and /Users/lijon/Coding/kicad/build/install_py3/KiCad.app/Contents/Frameworks/Python.framework/Versions/3.9/lib/python3.9/site-packages/wx/libwx_osx_cocoau_stc-3.1.5.0.0.dylib (0x1198ddbf8). One of the two will be used. Which one is undefined. And trying to open scripting console says: Traceback (most recent call last): File "", line 5, in File "/Users/lijon/Coding/kicad/build/install_py3/KiCad.app/Contents/SharedSupport/scripting/kicad_pyshell/__init__.py", line 230, in makePcbnewShellWindow pyshell = PcbnewPyShell(parent, id=-1, title=INTRO) File "/Users/lijon/Coding/kicad/build/install_py3/KiCad.app/Contents/Frameworks/Python.framework/Versions/3.9/lib/python3.9/site-packages/wx/py/editor.py", line 265, in __init__ EditorFrame.__init__(self, parent, id, title, pos, File "/Users/lijon/Coding/kicad/build/install_py3/KiCad.app/Contents/Frameworks/Python.framework/Versions/3.9/lib/python3.9/site-packages/wx/py/editor.py", line 29, in __init__ frame.Frame.__init__(self, parent, id, title, pos, size, style) File "/Users/lijon/Coding/kicad/build/install_py3/KiCad.app/Contents/Frameworks/Python.framework/Versions/3.9/lib/python3.9/site-packages/wx/py/frame.py", line 69, in __init__ wx.Frame.__init__(self, parent, id, title, pos, size, style) wx._core.PyNoAppError: The wx.App object must be created first! Hopefully this helps! On Thu, Apr 29, 2021 at 1:09 PM Jonatan Liljedahl wrote: > > So, I updated my CMake and tweaked some variables, and now 'make > install' succeeds without "app verification failed". Also I can start > pcbnew without crashing. However, trying to open the scripting console > in pcbnew gives an error popup. The console reveals that it couldn't > import the 'wx' module. > > During make install, I saw this: > > cp: cannot overwrite directory > /Users/lijon/Coding/kicad/build/install_py3/KiCad.app/Contents/Frameworks/Python.framework/Versions/3.9/lib/python3.9/site-packages > with non-directory > /usr/local/Cellar/python@3.9/3.9.2_1/Frameworks/Python.framework/Versions/3.9/lib/python3.9/site-packages > > It's non-directory because it's a symlink. It points to > /usr/local/lib/python3.9/site-packages, which does contain the 'wx' > module. > > I've tried rerunning cmake with > -DPYTHON_SITE_PACKAGE_PATH=/usr/local/lib/python3.9/site-packages but > I get the same error at make install. Perhaps it's cached somewhere, > how can I clear that without doing the whole build again from scratch? > (which takes a couple of hours on my slow machine...) > > -- > /Jonatan > http://kymatica.com -- /Jonatan http://kymatica.com ___ Mailing list: https://launchpad.net/~kicad-developers Post to : kicad-developers@lists.launchpad.net Unsubscribe : https://launchpad.net/~kicad-developers More help : https://help.launchpad.net/ListHelp
[Kicad-developers] Building with local Python3 etc
So, I updated my CMake and tweaked some variables, and now 'make install' succeeds without "app verification failed". Also I can start pcbnew without crashing. However, trying to open the scripting console in pcbnew gives an error popup. The console reveals that it couldn't import the 'wx' module. During make install, I saw this: cp: cannot overwrite directory /Users/lijon/Coding/kicad/build/install_py3/KiCad.app/Contents/Frameworks/Python.framework/Versions/3.9/lib/python3.9/site-packages with non-directory /usr/local/Cellar/python@3.9/3.9.2_1/Frameworks/Python.framework/Versions/3.9/lib/python3.9/site-packages It's non-directory because it's a symlink. It points to /usr/local/lib/python3.9/site-packages, which does contain the 'wx' module. I've tried rerunning cmake with -DPYTHON_SITE_PACKAGE_PATH=/usr/local/lib/python3.9/site-packages but I get the same error at make install. Perhaps it's cached somewhere, how can I clear that without doing the whole build again from scratch? (which takes a couple of hours on my slow machine...) -- /Jonatan http://kymatica.com ___ Mailing list: https://launchpad.net/~kicad-developers Post to : kicad-developers@lists.launchpad.net Unsubscribe : https://launchpad.net/~kicad-developers More help : https://help.launchpad.net/ListHelp
Re: [Kicad-developers] ngspice-34
Ah, yes that sounds reasonable. I guess /CMakeModules/Findngspice.cmake could be modified to use pkg-config to get the version and paths needed? On Tue, Mar 23, 2021 at 4:20 PM Jon Evans wrote: > > It sounds like the preference is to move away from using a config.h at all, > which means all the KiCad packaging systems would need to update to use > pkg-config to generate a header which can then be included in kicad's build > process. Is that correct? > > On Tue, Mar 23, 2021 at 11:15 AM Jonatan Liljedahl wrote: >> >> If ngspice-35 is not ready to be tagged soon, perhaps Holger could >> make a branch called "ngspice-34-config-fix" or something, based on >> the ngspice-34 tag but only adding the simple fix of installing >> config.h into /include/ngspice/ instead? Then KiCad packagers could >> build from this branch? Unless some packages are not using locally >> built ngspice? Just some ideas of possible solutions... >> >> Cheers >> >> On Tue, Mar 23, 2021 at 3:22 PM Holger Vogt wrote: >> > >> > The intention is to not at all install config.h. >> > >> > Installing it into ./include was a bug in ngspice-34 which has already >> > been removed in the current ngspice master branch. >> > >> > Perhaps you may make use of this bug by automatically moving config.h >> > from <...>/include/ to <...>/include/ngspice/ when building KiCad, >> > before we will have a solution without config.h in ngspice-35? >> > >> > >> > ___ >> > Mailing list: https://launchpad.net/~kicad-developers >> > Post to : kicad-developers@lists.launchpad.net >> > Unsubscribe : https://launchpad.net/~kicad-developers >> > More help : https://help.launchpad.net/ListHelp >> >> >> >> -- >> /Jonatan >> http://kymatica.com >> >> ___ >> Mailing list: https://launchpad.net/~kicad-developers >> Post to : kicad-developers@lists.launchpad.net >> Unsubscribe : https://launchpad.net/~kicad-developers >> More help : https://help.launchpad.net/ListHelp -- /Jonatan http://kymatica.com ___ Mailing list: https://launchpad.net/~kicad-developers Post to : kicad-developers@lists.launchpad.net Unsubscribe : https://launchpad.net/~kicad-developers More help : https://help.launchpad.net/ListHelp
Re: [Kicad-developers] ngspice-34
If ngspice-35 is not ready to be tagged soon, perhaps Holger could make a branch called "ngspice-34-config-fix" or something, based on the ngspice-34 tag but only adding the simple fix of installing config.h into /include/ngspice/ instead? Then KiCad packagers could build from this branch? Unless some packages are not using locally built ngspice? Just some ideas of possible solutions... Cheers On Tue, Mar 23, 2021 at 3:22 PM Holger Vogt wrote: > > The intention is to not at all install config.h. > > Installing it into ./include was a bug in ngspice-34 which has already > been removed in the current ngspice master branch. > > Perhaps you may make use of this bug by automatically moving config.h > from <...>/include/ to <...>/include/ngspice/ when building KiCad, > before we will have a solution without config.h in ngspice-35? > > > ___ > Mailing list: https://launchpad.net/~kicad-developers > Post to : kicad-developers@lists.launchpad.net > Unsubscribe : https://launchpad.net/~kicad-developers > More help : https://help.launchpad.net/ListHelp -- /Jonatan http://kymatica.com ___ Mailing list: https://launchpad.net/~kicad-developers Post to : kicad-developers@lists.launchpad.net Unsubscribe : https://launchpad.net/~kicad-developers More help : https://help.launchpad.net/ListHelp
Re: [Kicad-developers] ngspice-34
I've built KiCad on macOS with ngspice built from the "ngspice-34" git tag, but there is one problem: KiCad is looking for /include/ngspice/config.h, but ngspice installs it in /include/config.h: ./include ./include/config.h ./include/ngspice ./include/ngspice/sharedspice.h ... So after copying/linking config.h into include/ngspice/ KiCad finds it and builds as expected. I can only assume that it's an error in ngspice build system to put config.h directly in the include directory! Quite high chance that it would collide with some other package. Perhaps easier to just wait for ngspice-35 instead of working around it here? Cheers On Tue, Mar 23, 2021 at 9:51 AM Holger Vogt wrote: > > sharedspice.h (the only header to be distributed with ngspice) does > contain something like > > #ifndef NGSPICE_PACKAGE_VERSION > #define NGSPICE_PACKAGE_VERSION "34+" > #endif > > in the ngspice git master branch, and available later in ngspice-35. > > ___ > Mailing list: https://launchpad.net/~kicad-developers > Post to : kicad-developers@lists.launchpad.net > Unsubscribe : https://launchpad.net/~kicad-developers > More help : https://help.launchpad.net/ListHelp -- /Jonatan http://kymatica.com ___ Mailing list: https://launchpad.net/~kicad-developers Post to : kicad-developers@lists.launchpad.net Unsubscribe : https://launchpad.net/~kicad-developers More help : https://help.launchpad.net/ListHelp
Re: [Kicad-developers] compile warnings about copying loop variable
Got it. I'll just leave it as it is then and try to remember to set -Wno-range-loop-analysis the next time I build. On Wed, Mar 10, 2021 at 3:20 PM Seth Hillbrand wrote: > Almost certainly not. These warnings are that the string is going to be > copied one way or another, so we shouldn't be marking them for reference > because it is misleading. There may be a way for us to improve the code in > places, knowing that the values cannot be referenced but just removing the > references will make the code do the same thing, just without the warning. > > -Seth > > On Wed, Mar 10, 2021 at 6:13 AM Jonatan Liljedahl > wrote: > >> Would it make any noticeable performance improvements, avoiding all those >> string copying? >> >> On Wed, Mar 10, 2021 at 2:58 PM Seth Hillbrand >> wrote: >> >>> That's fun! New warning flags. Those are from -Wrange-loop-analysis >>> that is specific to clang. Those of us using gcc won't see them. >>> >>> They are suggesting minor improvements. You should feel free to submit >>> a MR to correct these cases or you can set your flags to >>> -Wno-range-loop-analysis to disable them. >>> >>> Best- >>> Seth >>> >>> On Wed, Mar 10, 2021 at 5:26 AM Jonatan Liljedahl >>> wrote: >>> >>>> I get a lot of compile warnings like this: >>>> >>>> In file included from >>>> /Users/lijon/Coding/kicad/include/dialog_shim.h:29: >>>> /Users/lijon/Coding/kicad/include/hashtables.h:87:26: warning: loop >>>> variable 'c' is always a copy because the range of type 'const >>>> wxString' does not return a reference >>>> [-Wrange-loop-analysis] >>>> for( const auto& c : aString ) >>>> ^ >>>> /Users/lijon/Coding/kicad/include/hashtables.h:87:14: note: use >>>> non-reference type 'wxUniChar' >>>> for( const auto& c : aString ) >>>> ^~~ >>>> >>>> >>>> -- >>>> /Jonatan >>>> http://kymatica.com >>>> >>>> ___ >>>> Mailing list: https://launchpad.net/~kicad-developers >>>> Post to : kicad-developers@lists.launchpad.net >>>> Unsubscribe : https://launchpad.net/~kicad-developers >>>> More help : https://help.launchpad.net/ListHelp >>>> >>> >>> >>> -- >>> [image: KiCad Services Corporation Logo] >>> Seth Hillbrand >>> *Lead Developer* >>> +1-530-302-5483 >>> Long Beach, CA >>> www.kipro-pcb.comi...@kipro-pcb.com >>> >> >> >> -- >> /Jonatan >> http://kymatica.com >> > > > -- > [image: KiCad Services Corporation Logo] > Seth Hillbrand > *Lead Developer* > +1-530-302-5483 > Long Beach, CA > www.kipro-pcb.comi...@kipro-pcb.com > -- /Jonatan http://kymatica.com ___ Mailing list: https://launchpad.net/~kicad-developers Post to : kicad-developers@lists.launchpad.net Unsubscribe : https://launchpad.net/~kicad-developers More help : https://help.launchpad.net/ListHelp
Re: [Kicad-developers] compile warnings about copying loop variable
Would it make any noticeable performance improvements, avoiding all those string copying? On Wed, Mar 10, 2021 at 2:58 PM Seth Hillbrand wrote: > That's fun! New warning flags. Those are from -Wrange-loop-analysis that > is specific to clang. Those of us using gcc won't see them. > > They are suggesting minor improvements. You should feel free to submit a > MR to correct these cases or you can set your flags to > -Wno-range-loop-analysis to disable them. > > Best- > Seth > > On Wed, Mar 10, 2021 at 5:26 AM Jonatan Liljedahl > wrote: > >> I get a lot of compile warnings like this: >> >> In file included from /Users/lijon/Coding/kicad/include/dialog_shim.h:29: >> /Users/lijon/Coding/kicad/include/hashtables.h:87:26: warning: loop >> variable 'c' is always a copy because the range of type 'const >> wxString' does not return a reference >> [-Wrange-loop-analysis] >> for( const auto& c : aString ) >> ^ >> /Users/lijon/Coding/kicad/include/hashtables.h:87:14: note: use >> non-reference type 'wxUniChar' >> for( const auto& c : aString ) >> ^~~ >> >> >> -- >> /Jonatan >> http://kymatica.com >> >> ___ >> Mailing list: https://launchpad.net/~kicad-developers >> Post to : kicad-developers@lists.launchpad.net >> Unsubscribe : https://launchpad.net/~kicad-developers >> More help : https://help.launchpad.net/ListHelp >> > > > -- > [image: KiCad Services Corporation Logo] > Seth Hillbrand > *Lead Developer* > +1-530-302-5483 > Long Beach, CA > www.kipro-pcb.comi...@kipro-pcb.com > -- /Jonatan http://kymatica.com ___ Mailing list: https://launchpad.net/~kicad-developers Post to : kicad-developers@lists.launchpad.net Unsubscribe : https://launchpad.net/~kicad-developers More help : https://help.launchpad.net/ListHelp
Re: [Kicad-developers] build failure
On Wed, Mar 10, 2021 at 2:34 PM Jon Evans wrote: > > > As I said earlier, it links fine for me now. It might have been some > > combination of installing CommandLineTools via xcode-select and > > uninstalling and reinstalling OCC. > > Well, I did those things too and was still having the same problem :( > > I guess I can try blowing up homebrew and the tools/sdk and starting from > scratch. I have OSX_SYSROOT as follows: CMAKE_OSX_SYSROOT:PATH=/Applications/Xcode.app/Contents/Developer/Platforms/MacOSX.platform/Developer/SDKs/MacOSX11.0.sdk I attach my full CMakeCache.txt in case it helps. Before, it was set to MacOSX10.15.sdk but recent Xcode does not ship with such an SDK, and I tried working around it by symlinking MacOSX10.15.sdk->MacOSX11.0.sdk. I think it all just made a mess, so I removed the symlink, installed CommandLineTools via xcode-select, reinstalled opencascade, and started with a clean build directory, and then it worked. -- /Jonatan http://kymatica.com # This is the CMakeCache file. # For build in directory: /Users/lijon/Coding/kicad/build/master # It was generated by CMake: /Applications/CMake.app/Contents/bin/cmake # You can edit this file to change values found and used by cmake. # If you do not want to change any of the values, simply exit the editor. # If you do want to change a value, simply edit, save, and exit the editor. # The syntax for the file is as follows: # KEY:TYPE=VALUE # KEY is the name of a variable in the cache. # TYPE is a hint to GUIs for the type of VALUE, DO NOT EDIT TYPE!. # VALUE is the current value for the KEY. # EXTERNAL cache entries //Dependencies for the target 3d-viewer_LIB_DEPENDS:STATIC=general;gal;general;kimath;general;-L/Users/lijon/Coding/wxWidgets/wxwidgets-dest/lib;general;-framework IOKit;general;-framework Carbon;general;-framework Cocoa;general;-framework AudioToolbox;general;-framework System;general;-framework OpenGL;general;-lwx_osx_cocoau_gl-3.1;general;-lwx_osx_cocoau-3.1;general;/Applications/Xcode.app/Contents/Developer/Platforms/MacOSX.platform/Developer/SDKs/MacOSX11.0.sdk/System/Library/Frameworks/OpenGL.framework;general;/Applications/Xcode.app/Contents/Developer/Platforms/MacOSX.platform/Developer/SDKs/MacOSX11.0.sdk/System/Library/Frameworks/OpenGL.framework;general;kicad_3dsg; //In debug build: create smaller binaries. BUILD_SMALL_DEBUG_FILES:BOOL=OFF //Build the testing tree. BUILD_TESTING:BOOL=OFF //Path to a program. BZRCOMMAND:FILEPATH=BZRCOMMAND-NOTFOUND //The directory containing a CMake configuration file for Boost. Boost_DIR:PATH=Boost_DIR-NOTFOUND //Boost filesystem library (debug) Boost_FILESYSTEM_LIBRARY_DEBUG:FILEPATH=/usr/local/lib/libboost_filesystem-mt.dylib //Boost filesystem library (release) Boost_FILESYSTEM_LIBRARY_RELEASE:FILEPATH=/usr/local/lib/libboost_filesystem-mt.dylib //Path to a file. Boost_INCLUDE_DIR:PATH=/usr/local/include //Boost library directory DEBUG Boost_LIBRARY_DIR_DEBUG:PATH=/usr/local/lib //Boost library directory RELEASE Boost_LIBRARY_DIR_RELEASE:PATH=/usr/local/lib //Boost system library (debug) Boost_SYSTEM_LIBRARY_DEBUG:FILEPATH=/usr/local/lib/libboost_system-mt.dylib //Boost system library (release) Boost_SYSTEM_LIBRARY_RELEASE:FILEPATH=/usr/local/lib/libboost_system-mt.dylib //Boost unit_test_framework library (debug) Boost_UNIT_TEST_FRAMEWORK_LIBRARY_DEBUG:FILEPATH=/usr/local/lib/libboost_unit_test_framework-mt.dylib //Boost unit_test_framework library (release) Boost_UNIT_TEST_FRAMEWORK_LIBRARY_RELEASE:FILEPATH=/usr/local/lib/libboost_unit_test_framework-mt.dylib //Path to a file. CAIRO_INCLUDE_DIR:PATH=/usr/local/include/cairo //Path to a library. CAIRO_LIBRARY:FILEPATH=/usr/local/Cellar/cairo/1.16.0_5/lib/libcairo.dylib //Path to a program. CMAKE_AR:FILEPATH=/Applications/Xcode.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/bin/ar //Set to either "Release" or "Debug" CMAKE_BUILD_TYPE:STRING=Debug //Enable/Disable color output during build. CMAKE_COLOR_MAKEFILE:BOOL=ON //CXX compiler CMAKE_CXX_COMPILER:FILEPATH=/Applications/Xcode.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/bin/c++ //Flags used by the CXX compiler during all build types. CMAKE_CXX_FLAGS:STRING= //Flags used by the CXX compiler during DEBUG builds. CMAKE_CXX_FLAGS_DEBUG:STRING=-g //Flags used by the CXX compiler during MINSIZEREL builds. CMAKE_CXX_FLAGS_MINSIZEREL:STRING=-Os -DNDEBUG //Flags used by the CXX compiler during RELEASE builds. CMAKE_CXX_FLAGS_RELEASE:STRING=-O3 -DNDEBUG //Flags used by the CXX compiler during RELWITHDEBINFO builds. CMAKE_CXX_FLAGS_RELWITHDEBINFO:STRING=-O2 -g -DNDEBUG //C compiler CMAKE_C_COMPILER:FILEPATH=/Applications/Xcode.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/bin/cc //Flags used by the C compiler during all build types. CMAKE_C_FLAGS:STRING= //Flags used by the C compiler during DEBUG builds. CMAKE_C_FLAGS_DEBUG:STRING=-g //Flags used by
[Kicad-developers] compile warnings about copying loop variable
I get a lot of compile warnings like this: In file included from /Users/lijon/Coding/kicad/include/dialog_shim.h:29: /Users/lijon/Coding/kicad/include/hashtables.h:87:26: warning: loop variable 'c' is always a copy because the range of type 'const wxString' does not return a reference [-Wrange-loop-analysis] for( const auto& c : aString ) ^ /Users/lijon/Coding/kicad/include/hashtables.h:87:14: note: use non-reference type 'wxUniChar' for( const auto& c : aString ) ^~~ -- /Jonatan http://kymatica.com ___ Mailing list: https://launchpad.net/~kicad-developers Post to : kicad-developers@lists.launchpad.net Unsubscribe : https://launchpad.net/~kicad-developers More help : https://help.launchpad.net/ListHelp
Re: [Kicad-developers] build failure
Hi, As I said earlier, it links fine for me now. It might have been some combination of installing CommandLineTools via xcode-select and uninstalling and reinstalling OCC. I did not force a 10.14 bottle. However, 'make install' still fails for me so I'm running kicad from the build dir (which works mostly). I think it might be the same issue as https://gitlab.com/kicad/code/kicad/-/issues/3718 Would be great if we could find a fix for this.. -- fixup_bundle -- app='/Users/lijon/Coding/kicad/build/install/KiCad.app/Contents/MacOS/kicad' -- libs='/Users/lijon/Coding/kicad/build/install/KiCad.app/Contents/PlugIns/_cvpcb.kiface;/Users/lijon/Coding/kicad/build/install/KiCad.app/Contents/PlugIns/_eeschema.kiface;/Users/lijon/Coding/kicad/build/install/KiCad.app/Contents/PlugIns/_gerbview.kiface;/Users/lijon/Coding/kicad/build/install/KiCad.app/Contents/PlugIns/_pcb_calculator.kiface;/Users/lijon/Coding/kicad/build/install/KiCad.app/Contents/PlugIns/_pcbnew.kiface;/Users/lijon/Coding/kicad/build/install/KiCad.app/Contents/PlugIns/_pl_editor.kiface;/Users/lijon/Coding/kicad/build/install/KiCad.app/Contents/PlugIns/3d/libs3d_plugin_idf.so;/Users/lijon/Coding/kicad/build/install/KiCad.app/Contents/PlugIns/3d/libs3d_plugin_oce.so;/Users/lijon/Coding/kicad/build/install/KiCad.app/Contents/PlugIns/3d/libs3d_plugin_vrml.so;/Users/lijon/Coding/kicad/build/install/KiCad.app/Contents/PlugIns/sim/libngspice.0.dylib;/Users/lijon/Coding/kicad/build/install/KiCad.app/Contents/Frameworks/python/site-packages/_pcbnew.so' -- dirs=' /usr/local/opt/opencascade/lib' -- ignoreItems='' -- fixup_bundle: preparing... -- warning: embedded item does not exist '/Users/lijon/Coding/kicad/build/install/KiCad.app/Contents/PlugIns/3d/libTKVCAF.7.dylib' -- warning: cannot resolve item '@loader_path/libTKVCAF.7.dylib' possible problems: need more directories? need to use InstallRequiredSystemLibraries? run in install tree instead of build tree? CMake Error at /Applications/CMake.app/Contents/share/cmake-3.15/Modules/BundleUtilities.cmake:452 (message): otool -l failed: 1 /Applications/Xcode.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/bin/objdump: error: '@loader_path/libTKVCAF.7.dylib': No such file or directory Call Stack (most recent call first): /Applications/CMake.app/Contents/share/cmake-3.15/Modules/BundleUtilities.cmake:521 (get_item_rpaths) /Applications/CMake.app/Contents/share/cmake-3.15/Modules/BundleUtilities.cmake:616 (set_bundle_key_values) /Applications/CMake.app/Contents/share/cmake-3.15/Modules/BundleUtilities.cmake:939 (get_bundle_keys) kicad/cmake_install.cmake:101 (fixup_bundle) cmake_install.cmake:67 (include) On Wed, Mar 10, 2021 at 3:48 AM Adam Wolf wrote: > > I haven't seen that. I can take a look at a bigger build log if you want. > > To be clear for others, I'm not advocating carrying around a 10.14 > bottle of opencascade; this is troubleshooting in progress :) > > Adam > > On Tue, Mar 9, 2021 at 7:02 PM Jon Evans wrote: > > > > Hi Jonatan, > > > > I hit the exact same issue (I'm also on 10.15), and after chatting with > > Adam about it, decided to try using the 10.14 bottle manually: > > > > Download > > https://bintray.com/homebrew/bottles/download_file?file_path=opencascade-7.5.0.mojave.bottle.tar.gz > > > > Install with: > > > > brew install -f --force-bottle opencascade-7.5.0.mojave.bottle.tar.gz > > brew link --overwrite opencascade > > > > After this, I am able to get past the opencascade link errors. > > > > I do have another error after this, which I haven't figured out yet: > > > > make[6]: *** No rule to make target > > `/usr/local/Cellar/cairo/1.16.0_3/lib/libcairo.dylib', needed by > > `kicad/KiCad.app/Contents/PlugIns/_eeschema.kiface'. Stop. > > > > @Adam or anyone else, have you seen this before? > > > > -Jon > > > > On Sat, Mar 6, 2021 at 3:47 AM Jonatan Liljedahl wrote: > >> > >> On Fri, Mar 5, 2021 at 8:45 PM Adam Wolf > >> wrote: > >> > > >> > It is certainly possible that Homebrew is distributing bottles that > >> > are linked a little weird, and you'd be getting the MacOS 10.14 > >> > reference from that. We've had this happen before. > >> > >> Yes, I think this was the case with my OCE install, OCEConfig.cmake > >> referenced 10.14. After unbrewing OCE, I tried to brew install it > >> again but only got a 404. However, after that I reinstalled OCC and it > >> built fine so maybe OCE and OCC was in conflict or something. > >> > >> > Regarding the libTKVCAF error, it looks like something's not quite > >> > right between the library and the fixup_bundle call. > >> >
Re: [Kicad-developers] clang format failure
Ok, never mind. A "brew upgrade clang-format" solved it. On Mon, Mar 8, 2021 at 10:26 AM Jonatan Liljedahl wrote: > > Hi, > > Trying to commit a fix for > https://gitlab.com/kicad/code/kicad/-/issues/7823 so I can make an MR, > clang format fails to run: > > ``` > lijon@lijon-mbp kicad % git commit > YAML:11:32: error: invalid boolean > AllowShortBlocksOnASingleLine: Never >^ > Error reading /Users/lijon/Coding/kicad/_clang-format: Invalid argument > error: `clang-format -style=file -lines=302:303 > eeschema/dialogs/dialog_sim_settings.cpp` failed > ERROR: you need to run clang-format (e.g. using tools/check_coding.sh) > on your commit > Running clang-format on the following files: > eeschema/dialogs/dialog_sim_settings.cpp > YAML:11:32: error: invalid boolean > AllowShortBlocksOnASingleLine: Never >^ > Error reading /Users/lijon/Coding/kicad/_clang-format: Invalid argument > error: `clang-format -style=file -lines=302:303 > eeschema/dialogs/dialog_sim_settings.cpp` failed > Hook /Users/lijon/Coding/kicad/.githooks/pre-commit.d/50-check-format > failed with error code 1 > To commit anyway, use --no-verify > ``` > > If I change that setting to "false" instead of "Never", it fails here instead: > > ``` > YAML:55:11: error: unknown enumerated scalar > Standard: c++11 > ^ > Error reading /Users/lijon/Coding/kicad/_clang-format: Invalid argument > error: `clang-format -style=file -lines=302:303 > eeschema/dialogs/dialog_sim_settings.cpp` failed > ``` > > -- > /Jonatan > http://kymatica.com -- /Jonatan http://kymatica.com ___ Mailing list: https://launchpad.net/~kicad-developers Post to : kicad-developers@lists.launchpad.net Unsubscribe : https://launchpad.net/~kicad-developers More help : https://help.launchpad.net/ListHelp
[Kicad-developers] clang format failure
Hi, Trying to commit a fix for https://gitlab.com/kicad/code/kicad/-/issues/7823 so I can make an MR, clang format fails to run: ``` lijon@lijon-mbp kicad % git commit YAML:11:32: error: invalid boolean AllowShortBlocksOnASingleLine: Never ^ Error reading /Users/lijon/Coding/kicad/_clang-format: Invalid argument error: `clang-format -style=file -lines=302:303 eeschema/dialogs/dialog_sim_settings.cpp` failed ERROR: you need to run clang-format (e.g. using tools/check_coding.sh) on your commit Running clang-format on the following files: eeschema/dialogs/dialog_sim_settings.cpp YAML:11:32: error: invalid boolean AllowShortBlocksOnASingleLine: Never ^ Error reading /Users/lijon/Coding/kicad/_clang-format: Invalid argument error: `clang-format -style=file -lines=302:303 eeschema/dialogs/dialog_sim_settings.cpp` failed Hook /Users/lijon/Coding/kicad/.githooks/pre-commit.d/50-check-format failed with error code 1 To commit anyway, use --no-verify ``` If I change that setting to "false" instead of "Never", it fails here instead: ``` YAML:55:11: error: unknown enumerated scalar Standard: c++11 ^ Error reading /Users/lijon/Coding/kicad/_clang-format: Invalid argument error: `clang-format -style=file -lines=302:303 eeschema/dialogs/dialog_sim_settings.cpp` failed ``` -- /Jonatan http://kymatica.com ___ Mailing list: https://launchpad.net/~kicad-developers Post to : kicad-developers@lists.launchpad.net Unsubscribe : https://launchpad.net/~kicad-developers More help : https://help.launchpad.net/ListHelp
Re: [Kicad-developers] Can't show 3D models (was: build failure)
Yes. This seems to be the case. I've commented on the issue. I've also commented on https://gitlab.com/kicad/code/kicad/-/issues/3718 regarding "make install" failing with OCC from homebrew. Cheers On Sat, Mar 6, 2021 at 11:47 PM Ian McInerney wrote: > > Hmm, I wonder if this is another instance of a pathing issue. We had a > similar bug on Linux that was caused by the 3d viewer looking for the plugin > to load OCE files in the wrong location, so it just wouldn't load any files. > That was https://gitlab.com/kicad/code/kicad/-/issues/7750, which I have > reopened so we can take a look at it again on macOS. > > -Ian > > On Sat, Mar 6, 2021 at 10:31 PM Jonatan Liljedahl wrote: >> >> Hi, >> >> I tried the nightly 20210306, and actually no, it shows no models >> either! See attached screenshots. >> As you see, I have KICAD6_3DMODEL_DIR = >> /Users/lijon/Coding/kicad-packages3D, which is where I've cloned the >> kicad-packages3D gitlab repo. The model is there: >> >> % ls -l /Users/lijon/Coding/kicad-packages3D/LED_THT.3dshapes/LED_D3.0mm.wrl >> -rw-r--r-- 1 lijon staff 36129 Aug 16 2020 >> /Users/lijon/Coding/kicad-packages3D/LED_THT.3dshapes/LED_D3.0mm.wrl >> >> On Sat, Mar 6, 2021 at 6:56 PM Adam Wolf >> wrote: >> > >> > Do the release versions work for you? >> > >> > Assuming it does, this is almost certainly an issue with dyld and >> > fixup_bundle. >> > >> > Let me know if the release/nighties work on your system, and I can walk >> > folks through how to solve it. >> > >> > >> > Adam >> > >> > On Sat, Mar 6, 2021, 11:09 AM Jonatan Liljedahl wrote: >> >> >> >> If I manually copy libTKVCAF.7.dylib into >> >> /kicad/KiCad.app/Contents/PlugIns/3d/ then "make install" >> >> succeeds. >> >> >> >> Also, libTKVCAF and all its friends was copied into >> >> $(CMAKE_INSTALL_PREFIX)/KiCad.app/Contents/Frameworks/ >> >> so they all exist there. >> >> >> >> But still, no 3D models at all are shown. >> >> >> >> I tried moving them to KiCad.app/Contents/PlugIns/3d/ but that didn't >> >> help either. >> >> >> >> Ideas? >> >> >> >> On Sat, Mar 6, 2021 at 9:47 AM Jonatan Liljedahl >> >> wrote: >> >> > >> >> > On Fri, Mar 5, 2021 at 8:45 PM Adam Wolf >> >> > wrote: >> >> > > >> >> > > It is certainly possible that Homebrew is distributing bottles that >> >> > > are linked a little weird, and you'd be getting the MacOS 10.14 >> >> > > reference from that. We've had this happen before. >> >> > >> >> > Yes, I think this was the case with my OCE install, OCEConfig.cmake >> >> > referenced 10.14. After unbrewing OCE, I tried to brew install it >> >> > again but only got a 404. However, after that I reinstalled OCC and it >> >> > built fine so maybe OCE and OCC was in conflict or something. >> >> > >> >> > > Regarding the libTKVCAF error, it looks like something's not quite >> >> > > right between the library and the fixup_bundle call. >> >> > > >> >> > > Does libTKVCAF.7.dylib exist on your system? >> >> > >> >> > Yes, that and all other OCC libs exist in >> >> > /usr/local/Cellar/opencascade/7.5.0_1/lib as well as symlinked into >> >> > /usr/local/lib (by homebrew). >> >> > >> >> > So I assume the problem here is that it's not finding all these libs >> >> > at runtime? How can I check if this is actually the issue here? >> >> > >> >> > During "make install" I get all these warnings: >> >> > >> >> > -- warning: embedded item does not exist >> >> > '/Users/lijon/Coding/kicad/build/install/KiCad.app/Contents/PlugIns/3d/libTKCAF.7.dylib' >> >> > -- >> >> > warning: cannot resolve item '@loader_path/libTKCAF.7.dylib' >> >> > >> >> > possible problems: >> >> > need more directories? >> >> > need to use InstallRequiredSystemLibraries? >> >> > run in install tree instead of build tree? >> >> > >> >> > -- warning: embedded item does not exist >> >> > '/Users/lijon/Coding/kicad/build/install/KiCad.app/Contents/PlugIns/3d/libTKV3d.
[Kicad-developers] Can't show 3D models (was: build failure)
If I manually copy libTKVCAF.7.dylib into /kicad/KiCad.app/Contents/PlugIns/3d/ then "make install" succeeds. Also, libTKVCAF and all its friends was copied into $(CMAKE_INSTALL_PREFIX)/KiCad.app/Contents/Frameworks/ so they all exist there. But still, no 3D models at all are shown. I tried moving them to KiCad.app/Contents/PlugIns/3d/ but that didn't help either. Ideas? On Sat, Mar 6, 2021 at 9:47 AM Jonatan Liljedahl wrote: > > On Fri, Mar 5, 2021 at 8:45 PM Adam Wolf > wrote: > > > > It is certainly possible that Homebrew is distributing bottles that > > are linked a little weird, and you'd be getting the MacOS 10.14 > > reference from that. We've had this happen before. > > Yes, I think this was the case with my OCE install, OCEConfig.cmake > referenced 10.14. After unbrewing OCE, I tried to brew install it > again but only got a 404. However, after that I reinstalled OCC and it > built fine so maybe OCE and OCC was in conflict or something. > > > Regarding the libTKVCAF error, it looks like something's not quite > > right between the library and the fixup_bundle call. > > > > Does libTKVCAF.7.dylib exist on your system? > > Yes, that and all other OCC libs exist in > /usr/local/Cellar/opencascade/7.5.0_1/lib as well as symlinked into > /usr/local/lib (by homebrew). > > So I assume the problem here is that it's not finding all these libs > at runtime? How can I check if this is actually the issue here? > > During "make install" I get all these warnings: > > -- warning: embedded item does not exist > '/Users/lijon/Coding/kicad/build/install/KiCad.app/Contents/PlugIns/3d/libTKCAF.7.dylib' > -- > warning: cannot resolve item '@loader_path/libTKCAF.7.dylib' > > possible problems: > need more directories? > need to use InstallRequiredSystemLibraries? > run in install tree instead of build tree? > > -- warning: embedded item does not exist > '/Users/lijon/Coding/kicad/build/install/KiCad.app/Contents/PlugIns/3d/libTKV3d.7.dylib' > -- > warning: cannot resolve item '@loader_path/libTKV3d.7.dylib' > > possible problems: > need more directories? > need to use InstallRequiredSystemLibraries? > run in install tree instead of build tree? > > -- warning: embedded item does not exist > '/Users/lijon/Coding/kicad/build/install/KiCad.app/Contents/PlugIns/3d/libTKLCAF.7.dylib' > -- > warning: cannot resolve item '@loader_path/libTKLCAF.7.dylib' > > possible problems: > need more directories? > need to use InstallRequiredSystemLibraries? > run in install tree instead of build tree? > > -- warning: embedded item does not exist > '/Users/lijon/Coding/kicad/build/install/KiCad.app/Contents/PlugIns/3d/libTKCDF.7.dylib' > -- > warning: cannot resolve item '@loader_path/libTKCDF.7.dylib' > > possible problems: > need more directories? > need to use InstallRequiredSystemLibraries? > run in install tree instead of build tree? > > -- warning: embedded item does not exist > '/Users/lijon/Coding/kicad/build/install/KiCad.app/Contents/PlugIns/3d/libTKBO.7.dylib' > -- > warning: cannot resolve item '@loader_path/libTKBO.7.dylib' > > possible problems: > need more directories? > need to use InstallRequiredSystemLibraries? > run in install tree instead of build tree? > > -- warning: embedded item does not exist > '/Users/lijon/Coding/kicad/build/install/KiCad.app/Contents/PlugIns/3d/libTKPrim.7.dylib' > -- > warning: cannot resolve item '@loader_path/libTKPrim.7.dylib' > > possible problems: > need more directories? > need to use InstallRequiredSystemLibraries? > run in install tree instead of build tree? > > -- warning: embedded item does not exist > '/Users/lijon/Coding/kicad/build/install/KiCad.app/Contents/PlugIns/3d/libTKService.7.dylib' > -- > warning: cannot resolve item '@loader_path/libTKService.7.dylib' > > possible problems: > need more directories? > need to use InstallRequiredSystemLibraries? > run in install tree instead of build tree? > > -- warning: embedded item does not exist > '/Users/lijon/Coding/kicad/build/install/KiCad.app/Contents/PlugIns/3d/libTKMesh.7.dylib' > -- > warning: cannot resolve item '@loader_path/libTKMesh.7.dylib' > > possible problems: > need more directories? > need to use InstallRequiredSystemLibraries? > run in install tree instead of build tree? > > -- warning: embedded item does not exist > '/Users/lijon/Coding/kicad/build/install/KiCad.app/Contents/PlugIns/3d/libTKShHealing.7.dylib' > -- > warning: cannot resolve item '@loader_path/libTKShHealing.7.dylib' > > possible problems: > need
Re: [Kicad-developers] build failure
On Fri, Mar 5, 2021 at 8:45 PM Adam Wolf wrote: > > It is certainly possible that Homebrew is distributing bottles that > are linked a little weird, and you'd be getting the MacOS 10.14 > reference from that. We've had this happen before. Yes, I think this was the case with my OCE install, OCEConfig.cmake referenced 10.14. After unbrewing OCE, I tried to brew install it again but only got a 404. However, after that I reinstalled OCC and it built fine so maybe OCE and OCC was in conflict or something. > Regarding the libTKVCAF error, it looks like something's not quite > right between the library and the fixup_bundle call. > > Does libTKVCAF.7.dylib exist on your system? Yes, that and all other OCC libs exist in /usr/local/Cellar/opencascade/7.5.0_1/lib as well as symlinked into /usr/local/lib (by homebrew). So I assume the problem here is that it's not finding all these libs at runtime? How can I check if this is actually the issue here? During "make install" I get all these warnings: -- warning: embedded item does not exist '/Users/lijon/Coding/kicad/build/install/KiCad.app/Contents/PlugIns/3d/libTKCAF.7.dylib' -- warning: cannot resolve item '@loader_path/libTKCAF.7.dylib' possible problems: need more directories? need to use InstallRequiredSystemLibraries? run in install tree instead of build tree? -- warning: embedded item does not exist '/Users/lijon/Coding/kicad/build/install/KiCad.app/Contents/PlugIns/3d/libTKV3d.7.dylib' -- warning: cannot resolve item '@loader_path/libTKV3d.7.dylib' possible problems: need more directories? need to use InstallRequiredSystemLibraries? run in install tree instead of build tree? -- warning: embedded item does not exist '/Users/lijon/Coding/kicad/build/install/KiCad.app/Contents/PlugIns/3d/libTKLCAF.7.dylib' -- warning: cannot resolve item '@loader_path/libTKLCAF.7.dylib' possible problems: need more directories? need to use InstallRequiredSystemLibraries? run in install tree instead of build tree? -- warning: embedded item does not exist '/Users/lijon/Coding/kicad/build/install/KiCad.app/Contents/PlugIns/3d/libTKCDF.7.dylib' -- warning: cannot resolve item '@loader_path/libTKCDF.7.dylib' possible problems: need more directories? need to use InstallRequiredSystemLibraries? run in install tree instead of build tree? -- warning: embedded item does not exist '/Users/lijon/Coding/kicad/build/install/KiCad.app/Contents/PlugIns/3d/libTKBO.7.dylib' -- warning: cannot resolve item '@loader_path/libTKBO.7.dylib' possible problems: need more directories? need to use InstallRequiredSystemLibraries? run in install tree instead of build tree? -- warning: embedded item does not exist '/Users/lijon/Coding/kicad/build/install/KiCad.app/Contents/PlugIns/3d/libTKPrim.7.dylib' -- warning: cannot resolve item '@loader_path/libTKPrim.7.dylib' possible problems: need more directories? need to use InstallRequiredSystemLibraries? run in install tree instead of build tree? -- warning: embedded item does not exist '/Users/lijon/Coding/kicad/build/install/KiCad.app/Contents/PlugIns/3d/libTKService.7.dylib' -- warning: cannot resolve item '@loader_path/libTKService.7.dylib' possible problems: need more directories? need to use InstallRequiredSystemLibraries? run in install tree instead of build tree? -- warning: embedded item does not exist '/Users/lijon/Coding/kicad/build/install/KiCad.app/Contents/PlugIns/3d/libTKMesh.7.dylib' -- warning: cannot resolve item '@loader_path/libTKMesh.7.dylib' possible problems: need more directories? need to use InstallRequiredSystemLibraries? run in install tree instead of build tree? -- warning: embedded item does not exist '/Users/lijon/Coding/kicad/build/install/KiCad.app/Contents/PlugIns/3d/libTKShHealing.7.dylib' -- warning: cannot resolve item '@loader_path/libTKShHealing.7.dylib' possible problems: need more directories? need to use InstallRequiredSystemLibraries? run in install tree instead of build tree? -- warning: embedded item does not exist '/Users/lijon/Coding/kicad/build/install/KiCad.app/Contents/PlugIns/3d/libTKHLR.7.dylib' -- warning: cannot resolve item '@loader_path/libTKHLR.7.dylib' possible problems: need more directories? need to use InstallRequiredSystemLibraries? run in install tree instead of build tree? -- warning: embedded item does not exist '/Users/lijon/Coding/kicad/build/install/KiCad.app/Contents/PlugIns/3d/libTKTopAlgo.7.dylib' -- warning: cannot resolve item '@loader_path/libTKTopAlgo.7.dylib' possible problems: need more directories? need to use InstallRequiredSystemLibraries? run in install tree instead of build tree? -- warning: embedded item does not exist '/Users/lijon/Coding/kicad/build/install/KiCad.app/Contents/PlugIns/3d/libTKGeomAlgo.7.dylib' -- warning: cannot resolve item '@loader_path/libTKGeomAlgo.7.dylib' possible
Re: [Kicad-developers] build failure
I tried "make install" in case something wasn't in the right place, but now that fails (which used to work fine): -- fixup_bundle -- app='/Users/lijon/Coding/kicad/build/install/KiCad.app/Contents/MacOS/kicad' -- libs='/Users/lijon/Coding/kicad/build/install/KiCad.app/Contents/PlugIns/_cvpcb.kiface;/Users/lijon/Coding/kicad/build/install/KiCad.app/Contents/PlugIns/_eeschema.kiface;/Users/lijon/Coding/kicad/build/install/KiCad.app/Contents/PlugIns/_gerbview.kiface;/Users/lijon/Coding/kicad/build/install/KiCad.app/Contents/PlugIns/_pcb_calculator.kiface;/Users/lijon/Coding/kicad/build/install/KiCad.app/Contents/PlugIns/_pcbnew.kiface;/Users/lijon/Coding/kicad/build/install/KiCad.app/Contents/PlugIns/_pl_editor.kiface;/Users/lijon/Coding/kicad/build/install/KiCad.app/Contents/PlugIns/3d/libs3d_plugin_idf.so;/Users/lijon/Coding/kicad/build/install/KiCad.app/Contents/PlugIns/3d/libs3d_plugin_oce.so;/Users/lijon/Coding/kicad/build/install/KiCad.app/Contents/PlugIns/3d/libs3d_plugin_vrml.so;/Users/lijon/Coding/kicad/build/install/KiCad.app/Contents/PlugIns/sim/libngspice.0.dylib;/Users/lijon/Coding/kicad/build/install/KiCad.app/Contents/Frameworks/python/site-packages/_pcbnew.so' -- dirs=' /usr/local/lib' -- ignoreItems='' -- fixup_bundle: preparing... -- warning: embedded item does not exist '/Users/lijon/Coding/kicad/build/install/KiCad.app/Contents/PlugIns/3d/libTKVCAF.7.dylib' -- warning: cannot resolve item '@loader_path/libTKVCAF.7.dylib' possible problems: need more directories? need to use InstallRequiredSystemLibraries? run in install tree instead of build tree? CMake Error at /Applications/CMake.app/Contents/share/cmake-3.15/Modules/BundleUtilities.cmake:452 (message): otool -l failed: 1 /Applications/Xcode.app/Contents/Developer/Toolchains/XcodeDefault.xctoolchain/usr/bin/objdump: error: '@loader_path/libTKVCAF.7.dylib': No such file or directory Call Stack (most recent call first): /Applications/CMake.app/Contents/share/cmake-3.15/Modules/BundleUtilities.cmake:521 (get_item_rpaths) /Applications/CMake.app/Contents/share/cmake-3.15/Modules/BundleUtilities.cmake:616 (set_bundle_key_values) /Applications/CMake.app/Contents/share/cmake-3.15/Modules/BundleUtilities.cmake:939 (get_bundle_keys) kicad/cmake_install.cmake:101 (fixup_bundle) cmake_install.cmake:67 (include) make: *** [install] Error 1 ___ Mailing list: https://launchpad.net/~kicad-developers Post to : kicad-developers@lists.launchpad.net Unsubscribe : https://launchpad.net/~kicad-developers More help : https://help.launchpad.net/ListHelp
Re: [Kicad-developers] build failure
Ok, I finally made it build with OCC, after uninstalling and reinstalling OCC with homebrew. Not sure what happened.. However, now there are *no* 3D models showing up. I noticed that there's a new KICAD6_3DMODEL_DIR path, which defaults to "/usr/local/3dmodels/" (where I have no 3d models). Is this new variable used instead of KISYS3DMOD? On Fri, Mar 5, 2021 at 1:08 PM Nick Østergaard wrote: > > @Jonatan Liljedahl Please share your cmake commandss > > On Fri, 5 Mar 2021 at 12:42, Jeff Young wrote: > > > > I never managed to get this to work (but my kung fu with build systems is > > notoriously weak). > > > > Anyway, my current build flags are: > > > > -DCMAKE_C_COMPILER=clang > > -DCMAKE_CXX_COMPILER=clang++ > > -DCMAKE_OSX_DEPLOYMENT_TARGET=10.14 > > -DwxWidgets_CONFIG_EXECUTABLE=/Users/jeff/kicad_dev/wxWidgets/wx-bin/bin/wx-config > > -DKICAD_STDLIB_LIGHT_DEBUG=OFF > > -DKICAD_SANITIZE=0 > > -DKICAD_SCRIPTING=OFF > > -DKICAD_SCRIPTING_MODULES=OFF > > -DKICAD_SCRIPTING_WXPYTHON=OFF > > -DKICAD_USE_OCE=OFF > > -DMAINTAIN_PNGS=OFF > > -DCMAKE_INSTALL_PREFIX=./bin > > -DCMAKE_BUILD_TYPE=Debug > > -DPYTHON_SITE_PACKAGE_PATH=/Users/jeff/kicad_dev/wxWidgets/wx-bin/lib/python2.7/site-packages > > > > I think the breakage in 3D model rendering is elsewhere. > > > > Cheers, > > Jeff. > > > > > > On 5 Mar 2021, at 11:28, Jonatan Liljedahl wrote: > > > > I've tried with a fresh build dir, still getting this: > > Undefined symbols for architecture x86_64: > > "Standard_Type::Register(char const*, char const*, unsigned long, > > opencascade::handle const&)", referenced from: > > opencascade::type_instance::get() in > > libkicad2step_lib.a(oce_utils.cpp.o) > > opencascade::type_instance::get() in > > libkicad2step_lib.a(oce_utils.cpp.o) > > opencascade::type_instance::get() in > > libkicad2step_lib.a(oce_utils.cpp.o) > > opencascade::type_instance::get() in > > libkicad2step_lib.a(oce_utils.cpp.o) > > opencascade::type_instance::get() in > > libkicad2step_lib.a(oce_utils.cpp.o) > > opencascade::type_instance::get() in > > libkicad2step_lib.a(oce_utils.cpp.o) > > "Quantity_Color::valuesOf(Quantity_NameOfColor, > > Quantity_TypeOfColor)", referenced from: > > PCBMODEL::transferModel(opencascade::handle&, > > opencascade::handle&, TRIPLET) in > > libkicad2step_lib.a(oce_utils.cpp.o) > > "BRepLib_Command::~BRepLib_Command()", referenced from: > > BRepLib_MakeShape::~BRepLib_MakeShape() in > > libkicad2step_lib.a(oce_utils.cpp.o) > > "BRepAlgoAPI_Algo::Shape()", referenced from: > > PCBMODEL::CreatePCB() in libkicad2step_lib.a(oce_utils.cpp.o) > > "Geom_BezierCurve::Geom_BezierCurve(NCollection_Array1 > > const&)", referenced from: > > OUTLINE::addEdge(BRepBuilderAPI_MakeWire*, KICADCURVE&, > > DOUBLET&) in libkicad2step_lib.a(oce_utils.cpp.o) > > "Standard_Failure::~Standard_Failure()", referenced from: > > Standard_ConstructionError::~Standard_ConstructionError() in > > libkicad2step_lib.a(oce_utils.cpp.o) > > Standard_ConstructionError::~Standard_ConstructionError() in > > libkicad2step_lib.a(oce_utils.cpp.o) > > Standard_OutOfMemory::~Standard_OutOfMemory() in > > libkicad2step_lib.a(oce_utils.cpp.o) > > Standard_OutOfRange::~Standard_OutOfRange() in > > libkicad2step_lib.a(oce_utils.cpp.o) > > Standard_OutOfRange::~Standard_OutOfRange() in > > libkicad2step_lib.a(oce_utils.cpp.o) > > "XCAFDoc_ShapeTool::AddComponent(TDF_Label const&, TDF_Label const&, > > TopLoc_Location const&)", referenced from: > > PCBMODEL::AddComponent(std::__1::basic_string > std::__1::char_traits, std::__1::allocator > const&, > > std::__1::basic_string, > > std::__1::allocator > const&, bool, DOUBLET, double, TRIPLET, > > TRIPLET, TRIPLET) in libkicad2step_lib.a(oce_utils.cpp.o) > > "XCAFDoc_ShapeTool::UpdateAssemblies()", referenced from: > > PCBMODEL::CreatePCB() in libkicad2step_lib.a(oce_utils.cpp.o) > > "NCollection_BaseMap::Destroy(void (*)(NCollection_ListNode*, > > opencascade::handle&), bool)", referenced > > from: > > BRepTools_Modifier::~BRepTools_Modifier() in > > libkicad2step_lib.a(oce_utils.cpp.o) > > NCollection_Map > TopTools_ShapeMapHasher>::~NCollection_Map() in > > libkicad2step_lib.a(oce_utils.cpp.o) > > NCollection_
Re: [Kicad-developers] build failure
scade::handle::~handle() in libkicad2step_lib.a(oce_utils.cpp.o) PCBMODEL::~PCBMODEL() in libkicad2step_lib.a(oce_utils.cpp.o) PCBMODEL::AddPadHole(KICADPAD const*) in libkicad2step_lib.a(oce_utils.cpp.o) std::__1::vector >::push_back(TopoDS_Shape const&) in libkicad2step_lib.a(oce_utils.cpp.o) ... "Standard_Transient::IncrementRefCounter() const", referenced from: PCBMODEL::AddPadHole(KICADPAD const*) in libkicad2step_lib.a(oce_utils.cpp.o) std::__1::vector >::push_back(TopoDS_Shape const&) in libkicad2step_lib.a(oce_utils.cpp.o) PCBMODEL::CreatePCB() in libkicad2step_lib.a(oce_utils.cpp.o) NCollection_List::Append(TopoDS_Shape const&) in libkicad2step_lib.a(oce_utils.cpp.o) PCBMODEL::WriteSTEP(wxString const&) in libkicad2step_lib.a(oce_utils.cpp.o) PCBMODEL::transferModel(opencascade::handle&, opencascade::handle&, TRIPLET) in libkicad2step_lib.a(oce_utils.cpp.o) OUTLINE::addEdge(BRepBuilderAPI_MakeWire*, KICADCURVE&, DOUBLET&) in libkicad2step_lib.a(oce_utils.cpp.o) ... ld: symbol(s) not found for architecture x86_64 clang: error: linker command failed with exit code 1 (use -v to see invocation) make[2]: *** [kicad/KiCad.app/Contents/MacOS/kicad2step] Error 1 make[1]: *** [utils/kicad2step/CMakeFiles/kicad2step.dir/all] Error 2 On Fri, Mar 5, 2021 at 11:57 AM Nick Østergaard wrote: > > You need to make sure you have a clean buid dir and try yo explicitly disable > oce and enable occt on your cmake configure line. > > fre. 5. mar. 2021 11.48 skrev Jonatan Liljedahl : >> >> Ok, I'm now trying to build against OCE instead, as I'm sure that used >> to work before. >> I managed to have CMake find my homebrew installed OCE by setting >> OCE_DIR, however it fails here: >> >> make[2]: *** No rule to make target >> `/Applications/Xcode.app/Contents/Developer/Platforms/MacOSX.platform/Developer/SDKs/MacOSX10.14.sdk/System/Library/Frameworks/OpenGL.framework', >> needed by `kicad/KiCad.app/Contents/MacOS/kicad2step'. Stop. >> >> Because I don't have MacOSX10.14.sdk, but 10.15. The weird thing is >> that I have set CMake build variables to the correct path: >> >> CMAKE_OSX_DEPLOYMENT_TARGET:STRING=10.15 >> CMAKE_OSX_SYSROOT:PATH=/Library/Developer/CommandLineTools/SDKs/MacOSX10.15.sdk >> >> But even then, kicad2step still has 10.14: >> >> utils/kicad2step/CMakeFiles/kicad2step.dir/build.make:kicad/KiCad.app/Contents/MacOS/kicad2step: >> /Applications/Xcode.app/Contents/Developer/Platforms/MacOSX.platform/Developer/SDKs/MacOSX10.14.sdk/System/Library/Frameworks/OpenGL.framework >> >> Removing utils/kicad2step/CMakeFiles (and plugins/3d/oce/CMakeFiles), >> a recursive grep in my build directory tells me that there's NO >> mention of "MacOSX10.14" anywhere. But after I've run cmake, it shows >> up again in the above mentioned places! >> >> So where is this "MacOSX10.14" reference coming from? >> >> lijon@lijon-mbp kicad % grep -R --include CMakeLists.txt 10.14 . >> >> ...show nothing, so it must come outside the kicad source tree. >> Any ideas? >> >> Cheers >> /Jonatan >> >> ___ >> Mailing list: https://launchpad.net/~kicad-developers >> Post to : kicad-developers@lists.launchpad.net >> Unsubscribe : https://launchpad.net/~kicad-developers >> More help : https://help.launchpad.net/ListHelp -- /Jonatan http://kymatica.com ___ Mailing list: https://launchpad.net/~kicad-developers Post to : kicad-developers@lists.launchpad.net Unsubscribe : https://launchpad.net/~kicad-developers More help : https://help.launchpad.net/ListHelp
Re: [Kicad-developers] build failure
Ok, I'm now trying to build against OCE instead, as I'm sure that used to work before. I managed to have CMake find my homebrew installed OCE by setting OCE_DIR, however it fails here: make[2]: *** No rule to make target `/Applications/Xcode.app/Contents/Developer/Platforms/MacOSX.platform/Developer/SDKs/MacOSX10.14.sdk/System/Library/Frameworks/OpenGL.framework', needed by `kicad/KiCad.app/Contents/MacOS/kicad2step'. Stop. Because I don't have MacOSX10.14.sdk, but 10.15. The weird thing is that I have set CMake build variables to the correct path: CMAKE_OSX_DEPLOYMENT_TARGET:STRING=10.15 CMAKE_OSX_SYSROOT:PATH=/Library/Developer/CommandLineTools/SDKs/MacOSX10.15.sdk But even then, kicad2step still has 10.14: utils/kicad2step/CMakeFiles/kicad2step.dir/build.make:kicad/KiCad.app/Contents/MacOS/kicad2step: /Applications/Xcode.app/Contents/Developer/Platforms/MacOSX.platform/Developer/SDKs/MacOSX10.14.sdk/System/Library/Frameworks/OpenGL.framework Removing utils/kicad2step/CMakeFiles (and plugins/3d/oce/CMakeFiles), a recursive grep in my build directory tells me that there's NO mention of "MacOSX10.14" anywhere. But after I've run cmake, it shows up again in the above mentioned places! So where is this "MacOSX10.14" reference coming from? lijon@lijon-mbp kicad % grep -R --include CMakeLists.txt 10.14 . ...show nothing, so it must come outside the kicad source tree. Any ideas? Cheers /Jonatan ___ Mailing list: https://launchpad.net/~kicad-developers Post to : kicad-developers@lists.launchpad.net Unsubscribe : https://launchpad.net/~kicad-developers More help : https://help.launchpad.net/ListHelp
[Kicad-developers] build failure
I'm having trouble building on mac after enabling OCC. I had both OCC and OCE disabled and I assume that's why 3D viewer didn't show STEP models any more. I installed OCC 7.5.0 using homebrew 'opencascade' formula. Is it possible to disable just kicad2step? [ 30%] Linking CXX static library libkicad2step_lib.a [ 30%] Built target kicad2step_lib [ 30%] Building CXX object 3d-viewer/3d_cache/sg/CMakeFiles/kicad_3dsg.dir/ifsg_faceset.cpp.o [ 30%] Building CXX object libs/kimath/CMakeFiles/kimath.dir/src/geometry/shape_poly_set.cpp.o [ 30%] Building CXX object utils/kicad2step/CMakeFiles/kicad2step.dir/kicad2step.cpp.o [ 30%] Building CXX object 3d-viewer/3d_cache/sg/CMakeFiles/kicad_3dsg.dir/ifsg_normals.cpp.o [ 30%] Building CXX object 3d-viewer/3d_cache/sg/CMakeFiles/kicad_3dsg.dir/ifsg_shape.cpp.o [ 30%] Building CXX object libs/kimath/CMakeFiles/kimath.dir/src/geometry/shape_rect.cpp.o [ 30%] Linking CXX executable ../../kicad/KiCad.app/Contents/MacOS/kicad2step Undefined symbols for architecture x86_64: "Standard_Type::Register(char const*, char const*, unsigned long, opencascade::handle const&)", referenced from: opencascade::type_instance::get() in libkicad2step_lib.a(oce_utils.cpp.o) opencascade::type_instance::get() in libkicad2step_lib.a(oce_utils.cpp.o) opencascade::type_instance::get() in libkicad2step_lib.a(oce_utils.cpp.o) opencascade::type_instance::get() in libkicad2step_lib.a(oce_utils.cpp.o) opencascade::type_instance::get() in libkicad2step_lib.a(oce_utils.cpp.o) opencascade::type_instance::get() in libkicad2step_lib.a(oce_utils.cpp.o) "Quantity_Color::valuesOf(Quantity_NameOfColor, Quantity_TypeOfColor)", referenced from: PCBMODEL::transferModel(opencascade::handle&, opencascade::handle&, TRIPLET) in libkicad2step_lib.a(oce_utils.cpp.o) "BRepLib_Command::~BRepLib_Command()", referenced from: BRepLib_MakeShape::~BRepLib_MakeShape() in libkicad2step_lib.a(oce_utils.cpp.o) "BRepAlgoAPI_Algo::Shape()", referenced from: PCBMODEL::CreatePCB() in libkicad2step_lib.a(oce_utils.cpp.o) "Geom_BezierCurve::Geom_BezierCurve(NCollection_Array1 const&)", referenced from: OUTLINE::addEdge(BRepBuilderAPI_MakeWire*, KICADCURVE&, DOUBLET&) in libkicad2step_lib.a(oce_utils.cpp.o) "Standard_Failure::~Standard_Failure()", referenced from: Standard_ConstructionError::~Standard_ConstructionError() in libkicad2step_lib.a(oce_utils.cpp.o) Standard_ConstructionError::~Standard_ConstructionError() in libkicad2step_lib.a(oce_utils.cpp.o) Standard_OutOfMemory::~Standard_OutOfMemory() in libkicad2step_lib.a(oce_utils.cpp.o) Standard_OutOfRange::~Standard_OutOfRange() in libkicad2step_lib.a(oce_utils.cpp.o) Standard_OutOfRange::~Standard_OutOfRange() in libkicad2step_lib.a(oce_utils.cpp.o) "XCAFDoc_ShapeTool::AddComponent(TDF_Label const&, TDF_Label const&, TopLoc_Location const&)", referenced from: PCBMODEL::AddComponent(std::__1::basic_string, std::__1::allocator > const&, std::__1::basic_string, std::__1::allocator > const&, bool, DOUBLET, double, TRIPLET, TRIPLET, TRIPLET) in libkicad2step_lib.a(oce_utils.cpp.o) "XCAFDoc_ShapeTool::UpdateAssemblies()", referenced from: PCBMODEL::CreatePCB() in libkicad2step_lib.a(oce_utils.cpp.o) "NCollection_BaseMap::Destroy(void (*)(NCollection_ListNode*, opencascade::handle&), bool)", referenced from: BRepTools_Modifier::~BRepTools_Modifier() in libkicad2step_lib.a(oce_utils.cpp.o) NCollection_Map::~NCollection_Map() in libkicad2step_lib.a(oce_utils.cpp.o) NCollection_DataMap::~NCollection_DataMap() in libkicad2step_lib.a(oce_utils.cpp.o) NCollection_DataMap::~NCollection_DataMap() in libkicad2step_lib.a(oce_utils.cpp.o) NCollection_DataMap::~NCollection_DataMap() in libkicad2step_lib.a(oce_utils.cpp.o) NCollection_Map::~NCollection_Map() in libkicad2step_lib.a(oce_utils.cpp.o) NCollection_DataMap::~NCollection_DataMap() in libkicad2step_lib.a(oce_utils.cpp.o) ... "NCollection_BaseList::PClear(void (*)(NCollection_ListNode*, opencascade::handle&))", referenced from: PCBMODEL::CreatePCB() in libkicad2step_lib.a(oce_utils.cpp.o) NCollection_List::~NCollection_List() in libkicad2step_lib.a(oce_utils.cpp.o) BRepBuilderAPI_MakeShape::~BRepBuilderAPI_MakeShape() in libkicad2step_lib.a(oce_utils.cpp.o) NCollection_List::Assign(NCollection_List const&) in libkicad2step_lib.a(oce_utils.cpp.o) NCollection_DataMap, TopTools_ShapeMapHasher>::DataMapNode::~DataMapNode() in libkicad2step_lib.a(oce_utils.cpp.o) BRepLib_MakeShape::~BRepLib_MakeShape() in libkicad2step_lib.a(oce_utils.cpp.o) NCollection_List::~NCollection_List() in libkicad2step_lib.a(oce_utils.cpp.o) ... "Standard_OutOfMemory::Standard_OutOfMemory(char const*)", referenced from: OUTLINE::addEdge(BRepBuilderAPI_MakeWire*, KICADCURVE&, DOUBLET&) in libkicad2step_lib.a(oce_utils.cpp.o)
[Kicad-developers] question about XML BOM output
Hi, When using jlcpcb-kicad-tools, it's tedious and error prone to fill in the LCSC_Part for all components, so I'm thinking how to automate this from a database/CSV file of preferred parts. I'm thinking one could make a BOM plugin that processes the XML and matches value and footprint (package) to insert the correct LCSC_Part. Looking at the XML, it looks for example like this: 5819 kymatica:D_SOD-123 ~ C8598 /Users/lijon/Dropbox/Electronics/Spice-Models/easyeda.lib 0 N 2 1 R ----5d77ea7f How come each custom field is repeated both in as well as having a ? Which one is used? Is there any way to read such an XML file back after processing it? Or is there a plugin API to process the schematic in place to fill in those LCSC_Part fields? -- /Jonatan http://kymatica.com ___ Mailing list: https://launchpad.net/~kicad-developers Post to : kicad-developers@lists.launchpad.net Unsubscribe : https://launchpad.net/~kicad-developers More help : https://help.launchpad.net/ListHelp
Re: [Kicad-developers] eeschema for IC design
On Wed, Oct 14, 2020 at 8:00 PM pepijn de vos wrote:
...
> For example, two other things that would be great to add is a way to plot
> algebraic expressions of different signals, to plot things such as impedance
> and transconductance, and a way to annotate operating points in the
> schematic. But these are probably a lot more work, and also require some more
> in-depth design discussion.
Actually you can plot algebraic expressions. Here's an example from my
experiments with checking phase margin of op-amp circuits etc:
.ac dec 20 10 10Meg
.control
set controlswait
* calculate loop gain, noise gain, open loop gain
let AB=V("/m2")/V("/m1")
let B1=V(out)/V("/m2")
let Aol=V(out)/V("/m1")
* print out cross-over frequency and phase margin
let dbAB=db(AB)
let phAB=180/pi*ph(AB)
display
meas ac unitf when dbAB=0
meas ac phm find phAB at=unitf
.endc
Some things to note here are that local labels in eeschema will have
"/" in their names and thus must be quoted, also the new signals AB,
B1 and Aol won't show up in the signal list, but you can still add
them manually by entering their name in the textbox below the list!
(Actually I added that textbox specifically for this purpose, but it
also works for quickly adding a signal with known name instead of
having to scroll around in the list)
Cheers
--
/Jonatan
http://kymatica.com
___
Mailing list: https://launchpad.net/~kicad-developers
Post to : kicad-developers@lists.launchpad.net
Unsubscribe : https://launchpad.net/~kicad-developers
More help : https://help.launchpad.net/ListHelp
Re: [Kicad-developers] running gerbview from local build on macOS
Yes, exactly this. On Wed, Aug 26, 2020 at 11:36 PM Jon Evans wrote: > I won't speak for Jonatan, but installing after a full build on MacOS > takes significant time. In fact, if I'm changing a single file and > recompiling, doing the install probably takes ~10x longer than the > compile. So, for iterating quickly, it's not great. > > On Wed, Aug 26, 2020 at 5:22 PM Nick Østergaard wrote: > >> Why don't you just install it? >> >> On Wed, 19 Aug 2020 at 13:04, Jonatan Liljedahl >> wrote: >> > >> > I see! But isn't gerbview always run standalone? Running pcbnew and >> > eeschema from the main kicad project manager works fine from the build >> > directory, as long as you don't open a pcb or schematic which are not >> > named the same as the project - then it will launch a standalone >> > instance. >> > >> > On Wed, Aug 19, 2020 at 12:14 PM Ian McInerney < >> ian.s.mciner...@ieee.org> wrote: >> > > >> > > My solution was not to use anything other than the main kicad >> launcher ;). So I never used standalone pcbnew/eeschema/gerbview/... on >> OSX. I did figure out what it would take to fix the kiface detection issue, >> and started https://gitlab.com/kicad/code/kicad/-/merge_requests/82 to >> try to make it so everything will work from the build directory, but I >> haven't gotten around to figuring out the Python pathing yet (that is the >> one that is most troublesome for me now that I switched back to a Linux >> daily driver machine - the kifaces just work for me now :) ). >> > > >> > > -Ian >> > > >> > > On Wed, Aug 19, 2020 at 10:39 AM Jeff Young wrote: >> > >> >> > >> It’s a “thing” on OSX. I run the attached script which fixes things >> up. Sadly it has to be run after each build. >> > >> >> > >> I know Jon has the same issue, but I haven’t heard about it from >> Ian. Perhaps he has a better solution…. >> > >> >> > >> Cheers, >> > >> Jeff. >> > >> >> > >> >> > >> >> > >> On 19 Aug 2020, at 10:32, Jonatan Liljedahl >> wrote: >> > >> >> > >> I now tried deleting the old system-wide /Applications/KiCad, and now >> > >> when clicking the gerbview button in project manager it gives the >> same >> > >> error about not finding _gerbview.kiface >> > >> >> > >> On Wed, Aug 19, 2020 at 11:25 AM Jonatan Liljedahl < >> li...@kymatica.com> wrote: >> > >> >> > >> >> > >> Hi! I'm having difficulties launching gerbview from my local build. >> > >> >> > >> I'm opening kicad.app like this, while standing in the build >> directory >> > >> (/Users/lijon/Coding/kicad/build/master) >> > >> >> > >>open kicad/kicad.app >> > >> >> > >> ...which works fine. However, clicking the gerbview button in the >> > >> project manager launches an old gerbview installed in systemwide >> > >> /Applications/KiCad/ >> > >> >> > >> Trying to open it manually by >> > >> >> > >>open gerbview/gerbview.app >> > >> >> > >> fails with "Failed to load kiface library >> > >> “/Users/lijon/Coding/kicad/build/Contents/PlugIns/_gerbview.kiface” >> > >> and then crashes. >> > >> >> > >> Same with: >> > >> >> > >>./gerbview/gerbview.app/Contents/MacOS/gerbview >> > >> >> > >> except now it tries to look in >> > >> /Users/lijon/Coding/kicad/build/master/Contents/PlugIns/ >> > >> >> > >> I've also tried to stand in the kicad.app directory and open it from >> > >> there, but with similar results. >> > >> >> > >> Any ideas how to run a locally built gerbview on macOS? >> > >> >> > >> Cheers >> > >> -- >> > >> /Jonatan >> > >> http://kymatica.com >> > >> >> > >> >> > >> >> > >> >> > >> -- >> > >> /Jonatan >> > >> http://kymatica.com >> > >> >> > >> ___ >> > >> Mailing list: https://launchpad.net/~kicad-developers >> > >> Post to :
Re: [Kicad-developers] running gerbview from local build on macOS
I see! But isn't gerbview always run standalone? Running pcbnew and eeschema from the main kicad project manager works fine from the build directory, as long as you don't open a pcb or schematic which are not named the same as the project - then it will launch a standalone instance. On Wed, Aug 19, 2020 at 12:14 PM Ian McInerney wrote: > > My solution was not to use anything other than the main kicad launcher ;). So > I never used standalone pcbnew/eeschema/gerbview/... on OSX. I did figure out > what it would take to fix the kiface detection issue, and started > https://gitlab.com/kicad/code/kicad/-/merge_requests/82 to try to make it so > everything will work from the build directory, but I haven't gotten around to > figuring out the Python pathing yet (that is the one that is most troublesome > for me now that I switched back to a Linux daily driver machine - the kifaces > just work for me now :) ). > > -Ian > > On Wed, Aug 19, 2020 at 10:39 AM Jeff Young wrote: >> >> It’s a “thing” on OSX. I run the attached script which fixes things up. >> Sadly it has to be run after each build. >> >> I know Jon has the same issue, but I haven’t heard about it from Ian. >> Perhaps he has a better solution…. >> >> Cheers, >> Jeff. >> >> >> >> On 19 Aug 2020, at 10:32, Jonatan Liljedahl wrote: >> >> I now tried deleting the old system-wide /Applications/KiCad, and now >> when clicking the gerbview button in project manager it gives the same >> error about not finding _gerbview.kiface >> >> On Wed, Aug 19, 2020 at 11:25 AM Jonatan Liljedahl >> wrote: >> >> >> Hi! I'm having difficulties launching gerbview from my local build. >> >> I'm opening kicad.app like this, while standing in the build directory >> (/Users/lijon/Coding/kicad/build/master) >> >>open kicad/kicad.app >> >> ...which works fine. However, clicking the gerbview button in the >> project manager launches an old gerbview installed in systemwide >> /Applications/KiCad/ >> >> Trying to open it manually by >> >>open gerbview/gerbview.app >> >> fails with "Failed to load kiface library >> “/Users/lijon/Coding/kicad/build/Contents/PlugIns/_gerbview.kiface” >> and then crashes. >> >> Same with: >> >>./gerbview/gerbview.app/Contents/MacOS/gerbview >> >> except now it tries to look in >> /Users/lijon/Coding/kicad/build/master/Contents/PlugIns/ >> >> I've also tried to stand in the kicad.app directory and open it from >> there, but with similar results. >> >> Any ideas how to run a locally built gerbview on macOS? >> >> Cheers >> -- >> /Jonatan >> http://kymatica.com >> >> >> >> >> -- >> /Jonatan >> http://kymatica.com >> >> ___ >> Mailing list: https://launchpad.net/~kicad-developers >> Post to : kicad-developers@lists.launchpad.net >> Unsubscribe : https://launchpad.net/~kicad-developers >> More help : https://help.launchpad.net/ListHelp >> >> >> ___ >> Mailing list: https://launchpad.net/~kicad-developers >> Post to : kicad-developers@lists.launchpad.net >> Unsubscribe : https://launchpad.net/~kicad-developers >> More help : https://help.launchpad.net/ListHelp -- /Jonatan http://kymatica.com ___ Mailing list: https://launchpad.net/~kicad-developers Post to : kicad-developers@lists.launchpad.net Unsubscribe : https://launchpad.net/~kicad-developers More help : https://help.launchpad.net/ListHelp
Re: [Kicad-developers] running gerbview from local build on macOS
I now tried deleting the old system-wide /Applications/KiCad, and now when clicking the gerbview button in project manager it gives the same error about not finding _gerbview.kiface On Wed, Aug 19, 2020 at 11:25 AM Jonatan Liljedahl wrote: > > Hi! I'm having difficulties launching gerbview from my local build. > > I'm opening kicad.app like this, while standing in the build directory > (/Users/lijon/Coding/kicad/build/master) > > open kicad/kicad.app > > ...which works fine. However, clicking the gerbview button in the > project manager launches an old gerbview installed in systemwide > /Applications/KiCad/ > > Trying to open it manually by > > open gerbview/gerbview.app > > fails with "Failed to load kiface library > “/Users/lijon/Coding/kicad/build/Contents/PlugIns/_gerbview.kiface” > and then crashes. > > Same with: > > ./gerbview/gerbview.app/Contents/MacOS/gerbview > > except now it tries to look in > /Users/lijon/Coding/kicad/build/master/Contents/PlugIns/ > > I've also tried to stand in the kicad.app directory and open it from > there, but with similar results. > > Any ideas how to run a locally built gerbview on macOS? > > Cheers > -- > /Jonatan > http://kymatica.com -- /Jonatan http://kymatica.com ___ Mailing list: https://launchpad.net/~kicad-developers Post to : kicad-developers@lists.launchpad.net Unsubscribe : https://launchpad.net/~kicad-developers More help : https://help.launchpad.net/ListHelp
[Kicad-developers] running gerbview from local build on macOS
Hi! I'm having difficulties launching gerbview from my local build. I'm opening kicad.app like this, while standing in the build directory (/Users/lijon/Coding/kicad/build/master) open kicad/kicad.app ...which works fine. However, clicking the gerbview button in the project manager launches an old gerbview installed in systemwide /Applications/KiCad/ Trying to open it manually by open gerbview/gerbview.app fails with "Failed to load kiface library “/Users/lijon/Coding/kicad/build/Contents/PlugIns/_gerbview.kiface” and then crashes. Same with: ./gerbview/gerbview.app/Contents/MacOS/gerbview except now it tries to look in /Users/lijon/Coding/kicad/build/master/Contents/PlugIns/ I've also tried to stand in the kicad.app directory and open it from there, but with similar results. Any ideas how to run a locally built gerbview on macOS? Cheers -- /Jonatan http://kymatica.com ___ Mailing list: https://launchpad.net/~kicad-developers Post to : kicad-developers@lists.launchpad.net Unsubscribe : https://launchpad.net/~kicad-developers More help : https://help.launchpad.net/ListHelp
[Kicad-developers] migrating to new symbol library format
Hi, It's been a while since I synced my git clone, and now I noticed that my symbol library is read-only because it's in the old file format. When trying to save it, it asks me to save it as a new file using the new format, which is great. However, it's not very clear that I then also need to go into "Manage symbol libraries" in eeschema and remove my old lib and add my new one manually. Perhaps it could detect that the lib is in my table and ask to replace it with the new one? Cheers -- /Jonatan http://kymatica.com ___ Mailing list: https://launchpad.net/~kicad-developers Post to : kicad-developers@lists.launchpad.net Unsubscribe : https://launchpad.net/~kicad-developers More help : https://help.launchpad.net/ListHelp
Re: [Kicad-developers] Rebase the tree?
Perhaps the git commit hook can include a check for file size and complain if larger than some-arbitrary-threshold? On Tue, Jan 21, 2020 at 12:43 AM Wayne Stambaugh wrote: > > No problem. Hopefully it wont cause too many issues for other devs who > have to rebase any local changes. > > Cheers, > > Wayne > > On 1/20/20 6:15 PM, Jeff Young wrote: > > Thanks, Wayne! > > > > Sorry for causing such a mess. > > > > Cheers, > > Jeff. > > > > > >> On 20 Jan 2020, at 22:32, Wayne Stambaugh wrote: > >> > >> I found it. I had to create a new protection for the master branch. I > >> pushed the changes and enabled the GitLab protection for the master > >> branch so we should be good to resume normal development. Thank you for > >> the help and patience to work through this. > >> > >> Cheers, > >> > >> Wayne > >> > >> On 1/20/20 5:27 PM, Nick Østergaard wrote: > >>> It should be there, I am sure you are just confused by the gitlab webui. > >>> > >>> It should be something along the lines of: > >>> Settings -> Repository -> Protected branches, click expand. Set the > >>> proper settings in the "Protect a branch" section. > >>> > >>> On Mon, 20 Jan 2020 at 23:26, Wayne Stambaugh > >>> wrote: > > Well this is a kick in the teeth. I just unprotected it and I don't see > an option to re-enable the protection after I force the changes. > > On 1/20/20 5:18 PM, Nick Østergaard wrote: > > There is a "protected branches" section in the settings of the repo. > > > > On Mon, 20 Jan 2020 at 23:18, Nick Østergaard wrote: > >> > >> You probably need to disable the option to disable force pushing on > >> the repo. > >> > >> On Mon, 20 Jan 2020 at 23:20, Wayne Stambaugh > >> wrote: > >>> > >>> GitLab rejected the forced push using this method. Anyone else have > >>> any > >>> ideas. Until we get this resolved, please do not push any commits to > >>> the master branch. > >>> > >>> Thanks, > >>> > >>> Wayne > >>> > >>> On 1/20/20 5:13 PM, Wayne Stambaugh wrote: > Please do not push anything to the master branch or perform any merge > requests until I push the rebase the master branch to prevent any > commit > losses because forcing a push will wipe out any changes. I ran the > git > command suggested by Simon and it seems to have the desired results > but > I have no idea how this is going to play out so I'm making a backup > clone of master just in case things go sideways. I'll ping everyone > once I have pushed the rebase. Thank you for your cooperation and I > apologize for whatever pain and agony this causes. Hopefully we will > never make this mistake again. > > Cheers, > > Wayne > > On 1/18/20 7:29 AM, Simon Richter wrote: > > Hi, > > > > On 17.01.20 19:14, Simon Richter wrote: > > > >> 1. "git rebase -i origin/master" > >> 2. in the editor, if they are present, remove the lines > >> > >>pick ea31730b4 Handle error returns from lstat. > >>pick e83420f19 Remove file accidentally commited in ea31730b4 > >>pick e27e6ee16 Also catch null dereference in case wxASSERT was > >> skipped. > >>pick e1925b89c Remove file accidentally added in e27e6ee1 > >> > >> 3. save and exit > > > > Even less interactive: > > > >git filter-branch \ > >--prune-empty \ > >--index-filter \ > >'git rm --cached --ignore-unmatch common/libcommon.a.*' \ > >9df2cfb32..HEAD > > > > This rewrites the current branch to a state where the files were > > never > > added, and removes the now-empty correction commits. The SHA1 sums > > in > > the new branch are different, but as the commit contents are > > identical, > > rebasing feature branches then goes smoothly even from the gitlab > > GUI. > > > > If you rebased a branch containing the offending commits on top of a > > cleaned one, this generates four commits adding and removing the > > files > > with no further changes, and the filter-branch commit above then > > reduces > > these to no-ops and removes the commits. > > > > A simple test in gitlab *merge request is descended from e1925b89c" > > could identify merge requests that would need to be rewritten. > > > > Simon > > > > > > ___ > > Mailing list: https://launchpad.net/~kicad-developers > > Post to : kicad-developers@lists.launchpad.net > > Unsubscribe :
Re: [Kicad-developers] Illegal instruction 4 when running eeschema after executing kicad-mac-builder
I'm also on mac. I also started with the kicad-mac-builder, but now I simply build kicad directly since it's much faster. I still use the built libs etc from kicad-mac-builder, like so: mkdir -p build/master cd build/master cmake -DNGSPICE_INCLUDE_DIR=/Users/lijon/Coding/kicad-mac-builder/build/ngspice-dest/include/ -DNGSPICE_LIBRARY=/Users/lijon/Coding/kicad-mac-builder/build/ngspice-dest/lib/libngspice.dylib -DKICAD_USE_OCE=OFF -DKICAD_SCRIPTING=OFF -DwxWidgets_CONFIG_EXECUTABLE=/Users/lijon/Coding/kicad-mac-builder/build/wxwidgets-dest/bin/wx-config -DKICAD_BUILD_QA_TESTS=OFF -DCMAKE_OSX_DEPLOYMENT_TARGET=10.14 ../.. make -j4 mkdir kicad/kicad.app/Contents/PlugIns/sim ln -s /Users/lijon/Coding/kicad-mac-builder/build/ngspice-dest/lib/libngspice.0.dylib kicad/kicad.app/Contents/PlugIns/sim/ open kicad/kicad.app On Tue, Jan 14, 2020 at 9:35 AM Heath Raftery wrote: > > No issues with downloaded Kicad package. Version info: > > Application: Eeschema > Version: (5.1.5-0-10_14), release build > Libraries: > wxWidgets 3.0.4 > libcurl/7.54.0 LibreSSL/2.6.5 zlib/1.2.11 nghttp2/1.24.1 > Platform: Mac OS X (Darwin 18.7.0 x86_64), 64 bit, Little endian, wxMac > Build Info: > wxWidgets: 3.0.4 (wchar_t,STL containers,compatible with 2.8) > Boost: 1.69.0 > OpenCASCADE Community Edition: 6.9.1 > Curl: 7.54.0 > Compiler: Clang 9.0.0 with C++ ABI 1002 > > Build settings: > USE_WX_GRAPHICS_CONTEXT=ON > USE_WX_OVERLAY=ON > KICAD_SCRIPTING=ON > KICAD_SCRIPTING_MODULES=ON > KICAD_SCRIPTING_PYTHON3=OFF > KICAD_SCRIPTING_WXPYTHON=ON > KICAD_SCRIPTING_WXPYTHON_PHOENIX=OFF > KICAD_SCRIPTING_ACTION_MENU=ON > BUILD_GITHUB_PLUGIN=ON > KICAD_USE_OCE=ON > KICAD_USE_OCC=OFF > KICAD_SPICE=ON > > > > On 14 Jan 2020, at 3:48 pm, Adam Wolf wrote: > > > > Before I dig too deep into that, do you have any issues running a > > downloaded Eeschema? > > > ___ > Mailing list: https://launchpad.net/~kicad-developers > Post to : kicad-developers@lists.launchpad.net > Unsubscribe : https://launchpad.net/~kicad-developers > More help : https://help.launchpad.net/ListHelp -- /Jonatan http://kymatica.com ___ Mailing list: https://launchpad.net/~kicad-developers Post to : kicad-developers@lists.launchpad.net Unsubscribe : https://launchpad.net/~kicad-developers More help : https://help.launchpad.net/ListHelp
Re: [Kicad-developers] Connection graph question
On Thu, Dec 12, 2019 at 12:01 AM Wayne Stambaugh wrote: > Now that the LIB_PART of the SCH_COMPONENT is no longer a pointer link > but an actual copy of the library symbol, any change to the underlying > library symbol will result in stale pin map pointers. I'm just curious to know, since I'm not at all familiar with this part of the code, why does it need to be a copy instead of a pointer? And is this per symbol instance in the schematic? -- /Jonatan http://kymatica.com ___ Mailing list: https://launchpad.net/~kicad-developers Post to : kicad-developers@lists.launchpad.net Unsubscribe : https://launchpad.net/~kicad-developers More help : https://help.launchpad.net/ListHelp
Re: [Kicad-developers] suggestion: git branching model
On Wed, Dec 11, 2019 at 5:35 PM Seth Hillbrand wrote: > > On 2019-12-11 08:17, Jonatan Liljedahl wrote: > > Ok, good to hear! > > > > How about moving the new symbol inheritance stuff into such a feature > > branch until it's working? (See > > https://gitlab.com/kicad/code/kicad/issues/3658) I understand it would > > mean less testing, but in its current state it's impossible to test or > > work on anything else since kicad keeps crashing. > > > > This will happen sometimes. We try to avoid it but we can't always > foresee the crashes, especially with large features. Frequently the > crashes are actually due to the difficulty of rebasing and merging the > branch. > ... > > Asking the development team to keep the development branch bug-free is > not realistic. This issue will get resolved. Until then, please use > the older revision. Of course, I'm not expecting the dev branch to be bug-free. I was just thinking about git branching models and wondering if things could be improved, specifically about larger feature branches. I didn't know that the symbol inheritance branch was available for testing from Waynes personal repo before merge, possibly because I'm new here and just recently joined the dev-list. I'm sorry if I stepped on any toes! :) Hopefully all this will be much easier now with GitLab, where everyone can see and access available merge requests in a simple way. I think the trick with feature branches is to rebase against master very frequently, to avoid a pile of merge conflicts. > If you are interested in helping test feature branches to minimize the > crashes in the future, you might test and review the feature branch at > https://gitlab.com/kicad/code/kicad/merge_requests/15 . It will be > merged soon but has not had any testing outside of my own at the moment. > This is the downside of feature branches in a volunteer-based > development model. I will take a look! However, since it's rebased on current master, which is crashing a lot, it's hard to test and know what broke because of which change. When the current issues on master are sorted out and !15 rebased against it, it will be easier. Cheers -- /Jonatan http://kymatica.com ___ Mailing list: https://launchpad.net/~kicad-developers Post to : kicad-developers@lists.launchpad.net Unsubscribe : https://launchpad.net/~kicad-developers More help : https://help.launchpad.net/ListHelp
Re: [Kicad-developers] suggestion: git branching model
Ok, good to hear! How about moving the new symbol inheritance stuff into such a feature branch until it's working? (See https://gitlab.com/kicad/code/kicad/issues/3658) I understand it would mean less testing, but in its current state it's impossible to test or work on anything else since kicad keeps crashing. On Wed, Dec 11, 2019 at 4:26 PM Jon Evans wrote: > > We already have used temporary branches for major features in the past. I'm > sure we'll continue to do this (through gitlab merge requests) where it makes > sense - it's a lot of work to keep a feature branch in mergeable state. > > -Jon > > On Wed, Dec 11, 2019, 10:16 Jonatan Liljedahl wrote: >> >> Good points. However, I still think it might make sense to have at >> least one more level of granularity, and use temporary branches for >> major changes etc. so that they can be tested before being merged into >> master. Simply, to make a merge request for big changes even if done >> by one of the core developers. Going back to the 5.1 branch just to >> have a non-crashing version of kicad is quite a leap. >> >> On Wed, Dec 11, 2019 at 3:41 PM Wayne Stambaugh wrote: >> > >> > We have discussed this before and given that very few users would ever >> > test the development branch(es), I'm not going to change our branching >> > policy. I don't think it's unfair to ask users who are aware that the >> > master branch (which is the KiCad development branch) is always in a >> > state of flux to deal with a bit of temporary instability in exchange >> > for some comprehensive testing of new features. Most users seem willing >> > to help with the testing in spite of some minor and sometimes some not >> > so minor inconveniences. I think have development branches would just >> > slow down how quickly new feature bugs would get fixed. >> > >> > Cheers, >> > >> > Wayne >> > >> > On 12/11/19 9:21 AM, Jonatan Liljedahl wrote: >> > > Hi, >> > > >> > > Perhaps it would make sense to adopt something like this? >> > > https://nvie.com/posts/a-successful-git-branching-model/#the-main-branches >> > > >> > > In short, all development happens on 'develop' branch and only when >> > > this is stable it's merged back to 'master'. One doesn't have to >> > > follow the above model strictly, for example a merge into master >> > > doesn't need to mean "new version to be released". >> > > >> > > Another nice thing is that stuff that are work in progress and not yet >> > > stable can live in a feature branch until it's stable enough to merge >> > > into 'develop'. (For example the new symbol inheritance stuff, which >> > > currently makes the master branch a bit unusable) >> > > >> > > Maybe some of this makes sense, and some not? Just some thoughts while >> > > trying to find a point in the master branch history that doesn't crash >> > > all the time :) >> > > >> > > Cheers >> > > >> > >> > ___ >> > Mailing list: https://launchpad.net/~kicad-developers >> > Post to : kicad-developers@lists.launchpad.net >> > Unsubscribe : https://launchpad.net/~kicad-developers >> > More help : https://help.launchpad.net/ListHelp >> >> >> >> -- >> /Jonatan >> http://kymatica.com >> >> ___ >> Mailing list: https://launchpad.net/~kicad-developers >> Post to : kicad-developers@lists.launchpad.net >> Unsubscribe : https://launchpad.net/~kicad-developers >> More help : https://help.launchpad.net/ListHelp -- /Jonatan http://kymatica.com ___ Mailing list: https://launchpad.net/~kicad-developers Post to : kicad-developers@lists.launchpad.net Unsubscribe : https://launchpad.net/~kicad-developers More help : https://help.launchpad.net/ListHelp
Re: [Kicad-developers] suggestion: git branching model
Good points. However, I still think it might make sense to have at least one more level of granularity, and use temporary branches for major changes etc. so that they can be tested before being merged into master. Simply, to make a merge request for big changes even if done by one of the core developers. Going back to the 5.1 branch just to have a non-crashing version of kicad is quite a leap. On Wed, Dec 11, 2019 at 3:41 PM Wayne Stambaugh wrote: > > We have discussed this before and given that very few users would ever > test the development branch(es), I'm not going to change our branching > policy. I don't think it's unfair to ask users who are aware that the > master branch (which is the KiCad development branch) is always in a > state of flux to deal with a bit of temporary instability in exchange > for some comprehensive testing of new features. Most users seem willing > to help with the testing in spite of some minor and sometimes some not > so minor inconveniences. I think have development branches would just > slow down how quickly new feature bugs would get fixed. > > Cheers, > > Wayne > > On 12/11/19 9:21 AM, Jonatan Liljedahl wrote: > > Hi, > > > > Perhaps it would make sense to adopt something like this? > > https://nvie.com/posts/a-successful-git-branching-model/#the-main-branches > > > > In short, all development happens on 'develop' branch and only when > > this is stable it's merged back to 'master'. One doesn't have to > > follow the above model strictly, for example a merge into master > > doesn't need to mean "new version to be released". > > > > Another nice thing is that stuff that are work in progress and not yet > > stable can live in a feature branch until it's stable enough to merge > > into 'develop'. (For example the new symbol inheritance stuff, which > > currently makes the master branch a bit unusable) > > > > Maybe some of this makes sense, and some not? Just some thoughts while > > trying to find a point in the master branch history that doesn't crash > > all the time :) > > > > Cheers > > > > ___ > Mailing list: https://launchpad.net/~kicad-developers > Post to : kicad-developers@lists.launchpad.net > Unsubscribe : https://launchpad.net/~kicad-developers > More help : https://help.launchpad.net/ListHelp -- /Jonatan http://kymatica.com ___ Mailing list: https://launchpad.net/~kicad-developers Post to : kicad-developers@lists.launchpad.net Unsubscribe : https://launchpad.net/~kicad-developers More help : https://help.launchpad.net/ListHelp
[Kicad-developers] suggestion: git branching model
Hi, Perhaps it would make sense to adopt something like this? https://nvie.com/posts/a-successful-git-branching-model/#the-main-branches In short, all development happens on 'develop' branch and only when this is stable it's merged back to 'master'. One doesn't have to follow the above model strictly, for example a merge into master doesn't need to mean "new version to be released". Another nice thing is that stuff that are work in progress and not yet stable can live in a feature branch until it's stable enough to merge into 'develop'. (For example the new symbol inheritance stuff, which currently makes the master branch a bit unusable) Maybe some of this makes sense, and some not? Just some thoughts while trying to find a point in the master branch history that doesn't crash all the time :) Cheers -- /Jonatan http://kymatica.com ___ Mailing list: https://launchpad.net/~kicad-developers Post to : kicad-developers@lists.launchpad.net Unsubscribe : https://launchpad.net/~kicad-developers More help : https://help.launchpad.net/ListHelp
Re: [Kicad-developers] Development branch symbol behavior change
Hi Wayne, I noticed one can no longer edit some fields when pressing E on a symbol in the schematic editor. For example reference and footprint are greyed out. /Jonatan On Fri, Dec 6, 2019 at 5:41 PM Wayne Stambaugh wrote: > > For those of you who have not been following recent development, I have > been working on the underlying schematic and symbol editor code in > preparation for the new file formats. I just pushed a major change to > how symbols are handled. I posted an announcement[1] on the KiCad forum > with all of the details should you be interested. > > Cheers, > > Wayne > > [1]: https://forum.kicad.info/t/symbol-library-editor-changes/20183 > > > ___ > Mailing list: https://launchpad.net/~kicad-developers > Post to : kicad-developers@lists.launchpad.net > Unsubscribe : https://launchpad.net/~kicad-developers > More help : https://help.launchpad.net/ListHelp -- /Jonatan http://kymatica.com ___ Mailing list: https://launchpad.net/~kicad-developers Post to : kicad-developers@lists.launchpad.net Unsubscribe : https://launchpad.net/~kicad-developers More help : https://help.launchpad.net/ListHelp
Re: [Kicad-developers] [PATCH] Eeschema: Add option for selection highlight thickness
Hi, Here's the merge request: https://gitlab.com/kicad/code/kicad/merge_requests/5 Cheers On Mon, Dec 2, 2019 at 4:03 PM Seth Hillbrand wrote: > > On 2019-12-02 04:30, Jonatan Liljedahl wrote: > > ADDED new option to set selection highlight thickness. > > Also change selection shadow width constants to > > make the selection thickness change less drastically > > with the zoom level. > > > > Hi Jonatan- > > I need to test out the new GitLab Merge Request for people who are not > yet members of the lead development team. Would you mind submitting > this as a merge request on GitLab and let me know how the process goes > for you? > > Thanks- > Seth > > Seth Hillbrand > KiCad Services Corporation > https://www.kipro-pcb.com > +1 530 302 5483 | +1 212 603 9372 > > ___ > Mailing list: https://launchpad.net/~kicad-developers > Post to : kicad-developers@lists.launchpad.net > Unsubscribe : https://launchpad.net/~kicad-developers > More help : https://help.launchpad.net/ListHelp -- /Jonatan http://kymatica.com ___ Mailing list: https://launchpad.net/~kicad-developers Post to : kicad-developers@lists.launchpad.net Unsubscribe : https://launchpad.net/~kicad-developers More help : https://help.launchpad.net/ListHelp
[Kicad-developers] [PATCH] Eeschema: Add option for selection highlight thickness
ADDED new option to set selection highlight thickness. Also change selection shadow width constants to make the selection thickness change less drastically with the zoom level. -- /Jonatan http://kymatica.com 0001-Eeschema-Add-option-for-selection-highlight-thicknes.patch Description: Binary data ___ Mailing list: https://launchpad.net/~kicad-developers Post to : kicad-developers@lists.launchpad.net Unsubscribe : https://launchpad.net/~kicad-developers More help : https://help.launchpad.net/ListHelp
Re: [Kicad-developers] [PATCH] Selection appearance options
Thanks! I will revisit patch #2 and make the selection thickness configurable as well, with a default value that keeps the current appearance. Cheers On Wed, Nov 27, 2019 at 5:06 AM Seth Hillbrand wrote: > > On 2019-11-23 05:46, Jonatan Liljedahl wrote: > > > Thanks Seth, > > > > I'm on Mac only (my linux box hasn't been powered on since 5 years or > > so). The code I touched should not look different depending on > > platform. I guess any difference would rather be in how colors are > > presented (considering the #ifdef MAC around default > > selection-highlight color), but even in that case I strongly suspect > > this is rather about the monitor rather than operating system! > > > > Cheers > > > > On Fri, Nov 22, 2019 at 6:25 PM Seth Hillbrand > > wrote: > > > > On 11/21/19 8:01 AM, Jonatan Liljedahl wrote: > > > > Hi, > > > > Here comes two patches regarding the selection appearance. > > > > Patch #1: Adds three new options in the settings dialog: > > > > - Draw selected text items as box > > Instead of drawing a stroked shadow behind the text, a rounded > > rectangle is drawn according to the text boundary box. > > > > - Draw selected child items > > When disabled, only the main selected item is drawn with the selection > > shadow, not the various fields and pin names etc. Default is enabled > > (current behaviour) > > > > - Fill selected shapes > > Any selected shapes has their selection shadow filled, instead of just > > drawing along the lines. > > > > See attached screenshots for a demonstration of some various > > combinations of these settings. > > > > Patch #2: Tweaks the amount of zoom-level impact on the selection > > shadow thickness, to get a more coherent look while zooming. You need > > to try this in action to know if it's good (I think it is). > > > > Cheers > > > > Hi Jonatan- > > I tested the functionality of these patches and they seem to work as > advertised. The dialog boxes look correct on my Linux box. Patch #2 > makes the selections very large to my eye, so I haven't pushed that one > but I did push the selection options patch. > > Thank you for your contribution to KiCad! > > -Seth > > Seth Hillbrand > KiCad Services Corporation > https://www.kipro-pcb.com > +1 530 302 5483 | +1 212 603 9372 -- /Jonatan http://kymatica.com ___ Mailing list: https://launchpad.net/~kicad-developers Post to : kicad-developers@lists.launchpad.net Unsubscribe : https://launchpad.net/~kicad-developers More help : https://help.launchpad.net/ListHelp
Re: [Kicad-developers] [PATCH] Eeschema: simulator: use dotted traces for current and phase.
Hi, How should I proceed with this patch? I think that regardless of future features, it would make sense to either just apply this patch so at least the traces for the second y-scale have a different visual style, or to add a section in the preferences for setting default solid/dashed and line width for left/right-scale traces. On Fri, Nov 22, 2019 at 2:35 PM Wayne Stambaugh wrote: > > This would be a great long term solution. You could also store it in > the project file since it is project specific if you don't want to go > through the effort of designing a new s-expression file format. > > On 11/21/19 6:45 PM, Ian McInerney wrote: > > I think it would be good to define a new S-expression config file that > > can be used to store the plot configurations of the simulator. This file > > should just contain the plot information, such as lines displayed, their > > style/color, axis configurations, colors, etc. (and maybe analysis > > parameters, I haven't thought about that part yet). This would allow us > > to add the functionality to open/save plot styles (similar to how a > > Modelsim .do file can be used to save/configure the waveform display, > > but ours won't be a script). > > > > -Ian > > > > On Thu, 21 Nov 2019, 22:11 Jon Evans, > <mailto:j...@craftyjon.com>> wrote: > > > > I can add the simulator to the scope of effort to do the whole > > settings files upgrade for 6.0 > > > > On Thu, Nov 21, 2019 at 5:06 PM Wayne Stambaugh > > mailto:stambau...@gmail.com>> wrote: > > > > > > On 11/21/19 1:15 PM, Kymatica wrote: > > > Sure, I can make it configurable. Where should this setting be > > available? In the main preferences? (Perhaps a new Simulator > > section there?) > > > > I would think the simulator will over time have enough user (not > > simulation control) configuration settings that it would have > > it's own > > preferences dialog and that the settings would be saved in a > > separate > > configuration file. I think at the moment the config settings > > are saved > > in the eeschema config file. > > > > > > > > /Jonatan > > > > > >> 21 nov. 2019 kl. 18:26 skrev Wayne Stambaugh > > mailto:stambau...@gmail.com>>: > > >> > > >> Hi Jonatan, > > >> > > >> I just tested this and it seems to work as expected although I'm > > >> thinking it might be a good idea to make this a configuration > > option > > >> rather than the only behavior. Users might not find this > > desirable. > > >> Anyone else have any thoughts on this? For future reference, > > when you > > >> are going to make behavioral changes to KiCad, please ping > > the dev > > >> mailing list first for feedback. I would have made the option > > >> suggestion before you began coding. > > >> > > >> Cheers, > > >> > > >> Wayne > > >> > > >>> On 11/19/19 11:30 AM, Jonatan Liljedahl wrote: > > >>> This patch makes it easier to visually keep apart traces for > > current > > >>> vs voltage and phase vs magnitude, by using dotted line > > style for > > >>> current and phase traces. > > >>> > > >>> To get the dotted style drawn correctly, I had to switch > > from drawing > > >>> each trace line segment individually to creating an array > > and then > > >>> using DrawLines() on it, so that the whole trace is one > > continuous > > >>> line. > > >>> > > >>> > > >>> ___ > > >>> Mailing list: https://launchpad.net/~kicad-developers > > >>> Post to : kicad-developers@lists.launchpad.net > > <mailto:kicad-developers@lists.launchpad.net> > > >>> Unsubscribe : https://launchpad.net/~kicad-developers > > >>> More help : https://help.launchpad.net/ListHelp > > >>> > > >> > > >> __
Re: [Kicad-developers] [PATCH] Selection appearance options
Thanks Seth, I'm on Mac only (my linux box hasn't been powered on since 5 years or so). The code I touched should not look different depending on platform. I guess any difference would rather be in how colors are presented (considering the #ifdef MAC around default selection-highlight color), but even in that case I strongly suspect this is rather about the monitor rather than operating system! Cheers On Fri, Nov 22, 2019 at 6:25 PM Seth Hillbrand wrote: > On 11/21/19 8:01 AM, Jonatan Liljedahl wrote: > > Hi, > > Here comes two patches regarding the selection appearance. > > Patch #1: Adds three new options in the settings dialog: > > - Draw selected text items as box > Instead of drawing a stroked shadow behind the text, a rounded > rectangle is drawn according to the text boundary box. > > - Draw selected child items > When disabled, only the main selected item is drawn with the selection > shadow, not the various fields and pin names etc. Default is enabled > (current behaviour) > > - Fill selected shapes > Any selected shapes has their selection shadow filled, instead of just > drawing along the lines. > > See attached screenshots for a demonstration of some various > combinations of these settings. > > Patch #2: Tweaks the amount of zoom-level impact on the selection > shadow thickness, to get a more coherent look while zooming. You need > to try this in action to know if it's good (I think it is). > > Cheers > > > ___ > Mailing list: https://launchpad.net/~kicad-developers > Post to : kicad-developers@lists.launchpad.net > Unsubscribe : https://launchpad.net/~kicad-developers > More help : https://help.launchpad.net/ListHelp > > Hi Jonatan- > > Thanks for this patch. I'll give this a test run this weekend. I know > that the Mac display of highlight was always a bit different than Linux, so > it would be good if one of our Mac devs also tested for issues. > > Best- > Seth > > -- > KiCad Services Corporation [image: KiCad Services Corporation Logo] > Seth Hillbrand > *Lead Developer* > +1-530-302-5483 <+12126039372> > Davis, CA > www.kipro-pcb.comi...@kipro-pcb.com > https://twitter.com/KiProEDA <https://twitter.com/KiProEDA> > https://www.linkedin.com/company/kicad > <https://www.linkedin.com/company/kicad> > ___ > Mailing list: https://launchpad.net/~kicad-developers > Post to : kicad-developers@lists.launchpad.net > Unsubscribe : https://launchpad.net/~kicad-developers > More help : https://help.launchpad.net/ListHelp > -- /Jonatan http://kymatica.com ___ Mailing list: https://launchpad.net/~kicad-developers Post to : kicad-developers@lists.launchpad.net Unsubscribe : https://launchpad.net/~kicad-developers More help : https://help.launchpad.net/ListHelp
Re: [Kicad-developers] [PATCH] Eeschema: simulator: use dotted traces for current and phase.
Also this message didn't reach the list: On Fri, Nov 22, 2019 at 12:14 AM Kymatica wrote: > > Hi, > With quick and easy I wasn’t actually referring to the coding involved, but > to the user experience: when I fire up the simulator I (as a user) just want > to see the results and look at the plots, with as little confusion as > possible regarding which trace is which, or which trace belongs to which > scale. I don’t want to configure anything. But perhaps I missunderstood, and > that the proposal wasn’t to have style config per trace/signal but per trace > type? (Volt, Current, Mag, Phase) Because traces are volatile, you add and > remove them many times depending on what you want to see. > > /Jonatan > > > 21 nov. 2019 kl. 21:55 skrev Wayne Stambaugh : > > > > On 11/21/19 2:36 PM, Kymatica wrote: > >> The idea here was to simply use a different visual style for the traces > >> that are plotted for the secondary y axis, to make it easy to see which > >> scale the trace belongs to without having to look it up in the signals > >> list. > >> > >> Per trace configuration could be nice, at least if preparing images for > >> presentation. But for everyday simulation work I’d just want it quick > >> and easy. > > > > I understand the temptation to do the quick and easy hack. We have all > > done it. From a project maintenance perspective, allowing developers to > > merge their personal quick hacks quickly makes a mess of the KiCad code > > base. This is why I would prefer that we carefully consider what trace > > drawing options we want to support so we can design a solution that > > works over the long term. > > > >> > >> /Jonatan > >> > >>>> 21 nov. 2019 kl. 18:34 skrev Ian McInerney : > >>> > >>> > >>> This is probably introducing major feature creep, but it would be nice > >>> to develop a dialog that allows setting the per-trace characteristics > >>> (such as color, line type, line width, etc) that this could go in. > >>> Where we put the accessors to it, I am not sure (it would be great if > >>> we could link it with the legend fields). Then we can add in more > >>> configuration options there as we implement them. > >>> > >>> -Ian > >>> > >>> On Thu, Nov 21, 2019 at 5:26 PM Wayne Stambaugh >>> <mailto:stambau...@gmail.com>> wrote: > >>> > >>>Hi Jonatan, > >>> > >>>I just tested this and it seems to work as expected although I'm > >>>thinking it might be a good idea to make this a configuration option > >>>rather than the only behavior. Users might not find this desirable. > >>>Anyone else have any thoughts on this? For future reference, when you > >>>are going to make behavioral changes to KiCad, please ping the dev > >>>mailing list first for feedback. I would have made the option > >>>suggestion before you began coding. > >>> > >>>Cheers, > >>> > >>>Wayne > >>> > >>>On 11/19/19 11:30 AM, Jonatan Liljedahl wrote: > >>>> This patch makes it easier to visually keep apart traces for current > >>>> vs voltage and phase vs magnitude, by using dotted line style for > >>>> current and phase traces. > >>>> > >>>> To get the dotted style drawn correctly, I had to switch from > >>>drawing > >>>> each trace line segment individually to creating an array and then > >>>> using DrawLines() on it, so that the whole trace is one continuous > >>>> line. > >>>> > >>>> > >>>> ___ > >>>> Mailing list: https://launchpad.net/~kicad-developers > >>>> Post to : kicad-developers@lists.launchpad.net > >>><mailto:kicad-developers@lists.launchpad.net> > >>>> Unsubscribe : https://launchpad.net/~kicad-developers > >>>> More help : https://help.launchpad.net/ListHelp > >>>> > >>> > >>>___ > >>>Mailing list: https://launchpad.net/~kicad-developers > >>>Post to : kicad-developers@lists.launchpad.net > >>><mailto:kicad-developers@lists.launchpad.net> > >>>Unsubscribe : https://launchpad.net/~kicad-developers > >>>More help : https://help.launchpad.net/ListHelp > >>> > >>> ___ > >>> Mailing list: https://launchpad.net/~kicad-developers > >>> Post to : kicad-developers@lists.launchpad.net > >>> Unsubscribe : https://launchpad.net/~kicad-developers > >>> More help : https://help.launchpad.net/ListHelp > > > > ___ > > Mailing list: https://launchpad.net/~kicad-developers > > Post to : kicad-developers@lists.launchpad.net > > Unsubscribe : https://launchpad.net/~kicad-developers > > More help : https://help.launchpad.net/ListHelp -- /Jonatan http://kymatica.com ___ Mailing list: https://launchpad.net/~kicad-developers Post to : kicad-developers@lists.launchpad.net Unsubscribe : https://launchpad.net/~kicad-developers More help : https://help.launchpad.net/ListHelp
[Kicad-developers] [PATCH] Selection appearance options
Oh, I just realized that when replying on my phone, it got sent from another email address and didn't reach the list! -- Forwarded message - From: Kymatica Date: Thu, Nov 21, 2019 at 7:23 PM Subject: Re: [Kicad-developers] [PATCH] Selection appearance options To: Kevin Cozens Cc: I think the screenshots show in what way this can improve visibility and reduce clutter. The first screenshot is the current appearance, which I find distracting. Not sure what you mean regarding reducing color saturation, I don’t see that happening here? Note this is eeschema, not pcbnew. Yes, it works fine also for symbols with yellow-body-fill. /Jonatan > 21 nov. 2019 kl. 17:45 skrev Kevin Cozens : > > On 2019-11-21 11:01 a.m., Jonatan Liljedahl wrote: >> Here comes two patches regarding the selection appearance. > > I'm just curious as to what is the problem that the proposed patches are > meant to address? (I may have missed the information in an earlier email). > > Is the idea to provide a much more visible indicator of what has been > selected than just reducing colour saturation, particularly when zoomed out? > > One question about the proposed patches. Have they been tested with symbols > containing the yellow infill? > > -- > Cheers! > > Kevin. > > http://www.ve3syb.ca/ | "Nerds make the shiny things that > https://www.patreon.com/KevinCozens | distract the mouth-breathers, and >| that's why we're powerful" > Owner of Elecraft K2 #2172 | > #include | --Chris Hardwick > > ___ > Mailing list: https://launchpad.net/~kicad-developers > Post to : kicad-developers@lists.launchpad.net > Unsubscribe : https://launchpad.net/~kicad-developers > More help : https://help.launchpad.net/ListHelp -- /Jonatan http://kymatica.com ___ Mailing list: https://launchpad.net/~kicad-developers Post to : kicad-developers@lists.launchpad.net Unsubscribe : https://launchpad.net/~kicad-developers More help : https://help.launchpad.net/ListHelp
Re: [Kicad-developers] [PATCH] Eeschema: simulator plot: allow standard mac pan and pinch to zoom
Hi, I found it. Here's an updated patch that makes it behave exactly as in the schematic editor, and uses the "trackpad pan" common config setting. This should work as expected on Linux. Note that in mathplot.h the coding style git hook complained, but I left it as is because it matches the current style in mathplot.h. Better to update the coding style for the whole file in a separate commit. Cheers On Thu, Nov 21, 2019 at 7:13 PM Kymatica wrote: > > It changes to ctrl-scrollwheel for zooming. Perhaps someone can point me in > the right direction for having the correct behavior on linux? How is it done > in the schematic editor? > > /Jonatan > > > 21 nov. 2019 kl. 18:02 skrev Wayne Stambaugh : > > > > Hi Jonatan, > > > > I tested this patch on my Linux laptop which has Ubuntu 19.04 with a > > stock gnome desktop and it breaks the two finger zoom touchpad behavior. > > With your patch zoom becomes a pan. I'm not sure if this is just a > > configuration issue or something else but I'm not seeing the same > > behavior in the board and schematic editors so I suspect something else > > is at play. Can someone else with a linux laptop test this? I don't > > want to merge this if it breaks linux systems. > > > > Cheers, > > > > Wayne > > > >> On 11/19/19 11:22 AM, Jonatan Liljedahl wrote: > >> This patch allows standard trackpad gestures for pan and zoom, just > >> like in eeschema and pcbnew etc. > >> > >> While working on this, I noticed mathplot.cpp is a bit messy, with a > >> lot of commented out or unused code, and some buggy behaviours. This > >> patch does not address any of that. > >> > >> > >> ___ > >> Mailing list: https://launchpad.net/~kicad-developers > >> Post to : kicad-developers@lists.launchpad.net > >> Unsubscribe : https://launchpad.net/~kicad-developers > >> More help : https://help.launchpad.net/ListHelp > >> > > > > ___ > > Mailing list: https://launchpad.net/~kicad-developers > > Post to : kicad-developers@lists.launchpad.net > > Unsubscribe : https://launchpad.net/~kicad-developers > > More help : https://help.launchpad.net/ListHelp -- /Jonatan http://kymatica.com 0001-Eeschema-simulator-plot-allow-standard-mac-pan-and-p.patch Description: Binary data ___ Mailing list: https://launchpad.net/~kicad-developers Post to : kicad-developers@lists.launchpad.net Unsubscribe : https://launchpad.net/~kicad-developers More help : https://help.launchpad.net/ListHelp
Re: [Kicad-developers] [PATCH] Eeschema: simulator: use dotted traces for current and phase.
Right, such configuration could be useful. However, note that a selection of which lines (signals?) to show would be tied to the actual current state of the schematic. Nets changes names, either because user changes the names of labels, or automatically when things get reconnected and re-annotated (Net-(R12-Pad2)) etc.. In any case, I think it makes sense to have a base setting of default trace drawing style, and I think it makes sense if this is separated into "left scale" (Voltage and Magnitude) and "right scale" (Current and Phase). You can only view Voltage and Current, or Magnitude and Phase, in the same plot, not any other combinations. On Fri, Nov 22, 2019 at 12:46 AM Ian McInerney wrote: > > I think it would be good to define a new S-expression config file that can be > used to store the plot configurations of the simulator. This file should just > contain the plot information, such as lines displayed, their style/color, > axis configurations, colors, etc. (and maybe analysis parameters, I haven't > thought about that part yet). This would allow us to add the functionality to > open/save plot styles (similar to how a Modelsim .do file can be used to > save/configure the waveform display, but ours won't be a script). > > -Ian > > On Thu, 21 Nov 2019, 22:11 Jon Evans, wrote: >> >> I can add the simulator to the scope of effort to do the whole settings >> files upgrade for 6.0 >> >> On Thu, Nov 21, 2019 at 5:06 PM Wayne Stambaugh wrote: >>> >>> >>> On 11/21/19 1:15 PM, Kymatica wrote: >>> > Sure, I can make it configurable. Where should this setting be available? >>> > In the main preferences? (Perhaps a new Simulator section there?) >>> >>> I would think the simulator will over time have enough user (not >>> simulation control) configuration settings that it would have it's own >>> preferences dialog and that the settings would be saved in a separate >>> configuration file. I think at the moment the config settings are saved >>> in the eeschema config file. >>> >>> > >>> > /Jonatan >>> > >>> >> 21 nov. 2019 kl. 18:26 skrev Wayne Stambaugh : >>> >> >>> >> Hi Jonatan, >>> >> >>> >> I just tested this and it seems to work as expected although I'm >>> >> thinking it might be a good idea to make this a configuration option >>> >> rather than the only behavior. Users might not find this desirable. >>> >> Anyone else have any thoughts on this? For future reference, when you >>> >> are going to make behavioral changes to KiCad, please ping the dev >>> >> mailing list first for feedback. I would have made the option >>> >> suggestion before you began coding. >>> >> >>> >> Cheers, >>> >> >>> >> Wayne >>> >> >>> >>> On 11/19/19 11:30 AM, Jonatan Liljedahl wrote: >>> >>> This patch makes it easier to visually keep apart traces for current >>> >>> vs voltage and phase vs magnitude, by using dotted line style for >>> >>> current and phase traces. >>> >>> >>> >>> To get the dotted style drawn correctly, I had to switch from drawing >>> >>> each trace line segment individually to creating an array and then >>> >>> using DrawLines() on it, so that the whole trace is one continuous >>> >>> line. >>> >>> >>> >>> >>> >>> ___ >>> >>> Mailing list: https://launchpad.net/~kicad-developers >>> >>> Post to : kicad-developers@lists.launchpad.net >>> >>> Unsubscribe : https://launchpad.net/~kicad-developers >>> >>> More help : https://help.launchpad.net/ListHelp >>> >>> >>> >> >>> >> ___ >>> >> Mailing list: https://launchpad.net/~kicad-developers >>> >> Post to : kicad-developers@lists.launchpad.net >>> >> Unsubscribe : https://launchpad.net/~kicad-developers >>> >> More help : https://help.launchpad.net/ListHelp >>> >>> ___ >>> Mailing list: https://launchpad.net/~kicad-developers >>> Post to : kicad-developers@lists.launchpad.net >>> Unsubscribe : https://launchpad.net/~kicad-developers >>> More help : https://help.launchpad.net/ListHelp >> >> ___ >> Mailing list: https://launchpad.net/~kicad-developers >> Post to : kicad-developers@lists.launchpad.net >> Unsubscribe : https://launchpad.net/~kicad-developers >> More help : https://help.launchpad.net/ListHelp > > ___ > Mailing list: https://launchpad.net/~kicad-developers > Post to : kicad-developers@lists.launchpad.net > Unsubscribe : https://launchpad.net/~kicad-developers > More help : https://help.launchpad.net/ListHelp -- /Jonatan http://kymatica.com ___ Mailing list: https://launchpad.net/~kicad-developers Post to : kicad-developers@lists.launchpad.net Unsubscribe : https://launchpad.net/~kicad-developers More help : https://help.launchpad.net/ListHelp
Re: [Kicad-developers] [PATCH] Eeschema: simulator: add keyboard shortcuts for add signal, etc
Oops, I forgot to add the updated wxFormBuilder file. Here's a fixed patch. On Thu, Nov 21, 2019 at 5:04 PM Jonatan Liljedahl wrote: > > This allows running the simulator, adding signals, etc via keyboard shortcuts. > > -- > /Jonatan > http://kymatica.com -- /Jonatan http://kymatica.com 0001-Eeschema-simulator-add-keyboard-shortcuts-for-add-si.patch Description: Binary data ___ Mailing list: https://launchpad.net/~kicad-developers Post to : kicad-developers@lists.launchpad.net Unsubscribe : https://launchpad.net/~kicad-developers More help : https://help.launchpad.net/ListHelp
[Kicad-developers] suggestion: draw text selection as simple rectangle
Hi list, Here's another suggestion, to draw the boundary box of selected text instead of drawing the text itself with the thick selection shadow. See attached screenshot. Is this a good idea? If yes, does it need to be configurable? Personally I find it a lot easier and visually less tiresome. -- /Jonatan http://kymatica.com ___ Mailing list: https://launchpad.net/~kicad-developers Post to : kicad-developers@lists.launchpad.net Unsubscribe : https://launchpad.net/~kicad-developers More help : https://help.launchpad.net/ListHelp
Re: [Kicad-developers] eeschema selection appearance
On Wed, Nov 20, 2019 at 5:55 PM Seth Hillbrand wrote: > On 11/20/19 7:11 AM, Jonatan Liljedahl wrote: > > Ok, I'll look into making it configurable. However, I find the current > default appearance hard to use, so maybe one could consider changing > the defaults in that case? > > Everyone likes to have their preferences as the default. This is not a > useful line of discussion for the project. Once the options are > configurable, it shouldn't be an issue. > My point was that maybe it's not only a matter of personal preference, but usability and good UX design. Options are good, but that's no reason not trying to find the best possible defaults for those options. Of course, if it's only me that thinks my proposal improves things in general, let me know :) > Additionally, I changed the shadow width algo constants as follows, to > avoid the feeling of the selection shadow changing size drastically as > you zoom: > > Please move this idea into a separate patch and propose it. The selection > highlight width works well at all zoom levels for me but if this idea helps > some people without negatively impacting the current situation, there > should be no issues. Please keep this distinct from the options patch, > however. > Ok, will do! BTW, this is an example where I don't think a user setting makes sense, which might be called something like "Amount of zoom-level impact on selection shadow thickness". Better to just find the values for those constants that looks good and coherent. -- /Jonatan http://kymatica.com ___ Mailing list: https://launchpad.net/~kicad-developers Post to : kicad-developers@lists.launchpad.net Unsubscribe : https://launchpad.net/~kicad-developers More help : https://help.launchpad.net/ListHelp
Re: [Kicad-developers] Building current master fails
Ah, yes I have switched between 5.1 and master so that would probably be it then! Thanks. On Wed, Nov 20, 2019 at 3:34 PM Ian McInerney wrote: > > Thats correct, all those lexer files should now be in the build folder. I > think that only the board file has had new keywords added, so it would be the > only one that is different. > > Ok, so there is probably another cause of the files appearing. Did you switch > branches between master and 5.1 and back? 5.1 looks like it uses the old > method of generating the lexer files, so it puts them in the source tree. > Master was switched over in July to use the new way that puts them in the > build folder. > > -Ian > > On Wed, Nov 20, 2019 at 11:51 AM Jonatan Liljedahl wrote: >> >> Hi, >> >> I see. So then I assume all these should also not exist? >> >> $ ls include/*_lexer.h >> include/lib_table_lexer.h include/netlist_lexer.h >> include/page_layout_reader_lexer.h include/pcb_lexer.h >> include/pcb_plot_params_lexer.h >> >> Of these files, only pcb_lexer.h differed between include/ and build/common/. >> >> This is a fresh clone, I'm not sure how these files ended up in >> include/. I see that the files are not in the git file tree of either >> 5.1 or master branch. Maybe it happened when I built using >> kicad-mac-builder pointing it to my kicad source tree? >> >> On Wed, Nov 20, 2019 at 12:09 PM Ian McInerney >> wrote: >> > >> > Jonatan, >> > >> > The pcb_lexer.h in include/ shouldn't exist anymore. There was a >> > switchover a few months ago in how that file was generated, and that >> > change moved it into the build directory. As a consequence of that >> > switchover, you need to clean out the stray files that existed from old >> > build. >> > >> > Try running this command in the repository (note, this command will remove >> > any files you are ignoring in the git, so make sure you have everything >> > you want to save copied before running): >> > 'git clean -fx' >> > >> > Out of curiosity, is this an older build system that you are just >> > updating, or is this a fresh clone since July? >> > >> > -Ian >> > >> > On Wed, Nov 20, 2019 at 10:46 AM Nick Østergaard wrote: >> >> >> >> It seems to build ok for linux on jenkins. >> >> >> >> Maybe double chevk you are not in a merge state or something. Also, >> >> please state the has of the version you are trying to build. >> >> >> >> ons. 20. nov. 2019 11.34 skrev Jonatan Liljedahl : >> >>> >> >>> Hi, I'm trying to build the current master, cmake configured like this: >> >>> >> >>> cmake >> >>> -DNGSPICE_INCLUDE_DIR=/Users/lijon/Coding/kicad-mac-builder/build/ngspice-dest/include/ >> >>> -DNGSPICE_LIBRARY=/Users/lijon/Coding/kicad-mac-builder/build/ngspice-dest/lib/libngspice.dylib >> >>> -DKICAD_USE_OCE=OFF -DKICAD_SCRIPTING=OFF >> >>> -DwxWidgets_CONFIG_EXECUTABLE=/Users/lijon/Coding/kicad-mac-builder/build/wxwidgets-dest/bin/wx-config >> >>> -DKICAD_BUILD_QA_TESTS=OFF -DCMAKE_OSX_DEPLOYMENT_TARGET=10.14 ../.. >> >>> >> >>> but it fails here: >> >>> >> >>> [ 42%] Building CXX object >> >>> common/CMakeFiles/pcbcommon.dir/__/pcbnew/pcb_parser.cpp.o >> >>> /Users/lijon/Coding/kicad/pcbnew/pcb_parser.cpp:996:18: error: use of >> >>> undeclared identifier 'T_copper_finish' >> >>> case T_copper_finish: >> >>> ^ >> >>> /Users/lijon/Coding/kicad/pcbnew/pcb_parser.cpp:1002:18: error: use of >> >>> undeclared identifier 'T_edge_plating' >> >>> case T_edge_plating: >> >>> ^ >> >>> /Users/lijon/Coding/kicad/pcbnew/pcb_parser.cpp:1008:18: error: use of >> >>> undeclared identifier 'T_dielectric_constraints' >> >>> case T_dielectric_constraints: >> >>> ^ >> >>> /Users/lijon/Coding/kicad/pcbnew/pcb_parser.cpp:1014:18: error: use of >> >>> undeclared identifier 'T_edge_connector'; did you mean >> >>> 'T_zone_connect'? >> >>> case T_edge_connector: >> >>> ^~~~ >> >>> T_zone_connect >> >>> /Users/lijon/Coding/kicad/include/p
Re: [Kicad-developers] eeschema selection appearance
Ok, I'll look into making it configurable. However, I find the current
default appearance hard to use, so maybe one could consider changing
the defaults in that case?
Not only does it make the text easily legible, but also the whole
symbol less cluttered when selected. You can still select individual
text fields as expected, but only the symbol and its pins are drawn as
selected when the whole component is selected.
These are the parts I've disabled for drawing of the selection shadow:
- pin labels/names
- component fields
- subsheet pins
- subsheet title/file
Additionally, I changed the shadow width algo constants as follows, to
avoid the feeling of the selection shadow changing size drastically as
you zoom:
float SCH_PAINTER::getShadowWidth()
{
const MATRIX3x3D& matrix = m_gal->GetScreenWorldMatrix();
// For best visuals the selection width must be a cross between
the zoom level and the
// default line width.
return (float) ( ( fabs( matrix.GetScale().x * 1.0 ) +
GetDefaultLineThickness() ) * 4.0 );
}
So now the question is at what granularity all this should be
configurable? Perhaps:
- selection draw child-items: bool
- selection thickness: float
- selection color (already there)
Cheers
On Wed, Nov 20, 2019 at 3:29 PM Ian McInerney wrote:
>
> I'm on the fence about the text highlighting, on the one hand not doing it
> does make it so the text is still easily legible when selected, but on the
> other it can be nice to show that it is part of the selected symbol. I think
> this would definitely be a case where making it a configurable option would
> allow people to experiment and see what they prefer.
>
> -Ian
>
> On Wed, Nov 20, 2019 at 2:18 PM Seth Hillbrand wrote:
>>
>> On 2019-11-20 05:48, Jonatan Liljedahl wrote:
>> > Hi,
>> >
>> > I'm tweaking the appearance of the new selection, what do you think?
>> > Except a change of color, transparency and width, it also skips
>> > drawing the fields and pin labels of components. I think it gives a
>> > much cleaner look with less clutter.
>>
>> This will always be a matter of opinion. Right now the color is
>> configurable. If you'd like to add additional configurable parameters
>> to the selection, you should place them in the Eeschema preferences and
>> allow the user to choose them. Then post the patch for review.
>>
>> Otherwise, we'll end up bike shedding on this which would be nice to
>> avoid.
>>
>> Best-
>> Seth
>>
>>
>> Seth Hillbrand
>> KiCad Services Corporation
>> https://www.kipro-pcb.com
>> +1 530 302 5483 | +1 212 603 9372
>>
>> ___
>> Mailing list: https://launchpad.net/~kicad-developers
>> Post to : kicad-developers@lists.launchpad.net
>> Unsubscribe : https://launchpad.net/~kicad-developers
>> More help : https://help.launchpad.net/ListHelp
>
> ___
> Mailing list: https://launchpad.net/~kicad-developers
> Post to : kicad-developers@lists.launchpad.net
> Unsubscribe : https://launchpad.net/~kicad-developers
> More help : https://help.launchpad.net/ListHelp
--
/Jonatan
http://kymatica.com
___
Mailing list: https://launchpad.net/~kicad-developers
Post to : kicad-developers@lists.launchpad.net
Unsubscribe : https://launchpad.net/~kicad-developers
More help : https://help.launchpad.net/ListHelp
[Kicad-developers] eeschema selection appearance
Hi, I'm tweaking the appearance of the new selection, what do you think? Except a change of color, transparency and width, it also skips drawing the fields and pin labels of components. I think it gives a much cleaner look with less clutter. -- /Jonatan http://kymatica.com ___ Mailing list: https://launchpad.net/~kicad-developers Post to : kicad-developers@lists.launchpad.net Unsubscribe : https://launchpad.net/~kicad-developers More help : https://help.launchpad.net/ListHelp
Re: [Kicad-developers] Building current master fails
Hi, I see. So then I assume all these should also not exist? $ ls include/*_lexer.h include/lib_table_lexer.h include/netlist_lexer.h include/page_layout_reader_lexer.h include/pcb_lexer.h include/pcb_plot_params_lexer.h Of these files, only pcb_lexer.h differed between include/ and build/common/. This is a fresh clone, I'm not sure how these files ended up in include/. I see that the files are not in the git file tree of either 5.1 or master branch. Maybe it happened when I built using kicad-mac-builder pointing it to my kicad source tree? On Wed, Nov 20, 2019 at 12:09 PM Ian McInerney wrote: > > Jonatan, > > The pcb_lexer.h in include/ shouldn't exist anymore. There was a switchover a > few months ago in how that file was generated, and that change moved it into > the build directory. As a consequence of that switchover, you need to clean > out the stray files that existed from old build. > > Try running this command in the repository (note, this command will remove > any files you are ignoring in the git, so make sure you have everything you > want to save copied before running): > 'git clean -fx' > > Out of curiosity, is this an older build system that you are just updating, > or is this a fresh clone since July? > > -Ian > > On Wed, Nov 20, 2019 at 10:46 AM Nick Østergaard wrote: >> >> It seems to build ok for linux on jenkins. >> >> Maybe double chevk you are not in a merge state or something. Also, please >> state the has of the version you are trying to build. >> >> ons. 20. nov. 2019 11.34 skrev Jonatan Liljedahl : >>> >>> Hi, I'm trying to build the current master, cmake configured like this: >>> >>> cmake >>> -DNGSPICE_INCLUDE_DIR=/Users/lijon/Coding/kicad-mac-builder/build/ngspice-dest/include/ >>> -DNGSPICE_LIBRARY=/Users/lijon/Coding/kicad-mac-builder/build/ngspice-dest/lib/libngspice.dylib >>> -DKICAD_USE_OCE=OFF -DKICAD_SCRIPTING=OFF >>> -DwxWidgets_CONFIG_EXECUTABLE=/Users/lijon/Coding/kicad-mac-builder/build/wxwidgets-dest/bin/wx-config >>> -DKICAD_BUILD_QA_TESTS=OFF -DCMAKE_OSX_DEPLOYMENT_TARGET=10.14 ../.. >>> >>> but it fails here: >>> >>> [ 42%] Building CXX object >>> common/CMakeFiles/pcbcommon.dir/__/pcbnew/pcb_parser.cpp.o >>> /Users/lijon/Coding/kicad/pcbnew/pcb_parser.cpp:996:18: error: use of >>> undeclared identifier 'T_copper_finish' >>> case T_copper_finish: >>> ^ >>> /Users/lijon/Coding/kicad/pcbnew/pcb_parser.cpp:1002:18: error: use of >>> undeclared identifier 'T_edge_plating' >>> case T_edge_plating: >>> ^ >>> /Users/lijon/Coding/kicad/pcbnew/pcb_parser.cpp:1008:18: error: use of >>> undeclared identifier 'T_dielectric_constraints' >>> case T_dielectric_constraints: >>> ^ >>> /Users/lijon/Coding/kicad/pcbnew/pcb_parser.cpp:1014:18: error: use of >>> undeclared identifier 'T_edge_connector'; did you mean >>> 'T_zone_connect'? >>> case T_edge_connector: >>> ^~~~ >>> T_zone_connect >>> /Users/lijon/Coding/kicad/include/pcb_lexer.h:232:9: note: >>> 'T_zone_connect' declared here >>> T_zone_connect, >>> ^ >>> /Users/lijon/Coding/kicad/pcbnew/pcb_parser.cpp:1020:35: error: use of >>> undeclared identifier 'T_bevelled' >>> else if( token == T_bevelled ) >>> ^ >>> /Users/lijon/Coding/kicad/pcbnew/pcb_parser.cpp:1026:18: error: use of >>> undeclared identifier 'T_castellated_pads' >>> case T_castellated_pads: >>> ^ >>> /Users/lijon/Coding/kicad/pcbnew/pcb_parser.cpp:1088:30: error: use of >>> undeclared identifier 'T_addsublayer' >>> if( token == T_addsublayer ) >>> ^ >>> /Users/lijon/Coding/kicad/pcbnew/pcb_parser.cpp:1100:26: error: use of >>> undeclared identifier 'T_type'; did you mean 'T_top'? >>> case T_type: >>> ^~ >>> T_top >>> /Users/lijon/Coding/kicad/include/pcb_lexer.h:196:9: note: 'T_top' declared >>> here >>> T_top, >>> ^ >>> /Users/lijon/Coding/kicad/pcbnew/pcb_parser.cpp:1123:26: error: use of >>> undeclared identifier 'T_material'; did you mean 'T_general'? >>> case T_material: >>> ^~~
Re: [Kicad-developers] Building current master fails
include/pcb_lexer.h does not include T_copper_finish etc. However, build/common/pcb_lexer.h does! How come? I'm guessing the one in common is generated during build, but for some reason it picks the other one when including the header. On Wed, Nov 20, 2019 at 11:34 AM Jonatan Liljedahl wrote: > > Hi, I'm trying to build the current master, cmake configured like this: > > cmake > -DNGSPICE_INCLUDE_DIR=/Users/lijon/Coding/kicad-mac-builder/build/ngspice-dest/include/ > -DNGSPICE_LIBRARY=/Users/lijon/Coding/kicad-mac-builder/build/ngspice-dest/lib/libngspice.dylib > -DKICAD_USE_OCE=OFF -DKICAD_SCRIPTING=OFF > -DwxWidgets_CONFIG_EXECUTABLE=/Users/lijon/Coding/kicad-mac-builder/build/wxwidgets-dest/bin/wx-config > -DKICAD_BUILD_QA_TESTS=OFF -DCMAKE_OSX_DEPLOYMENT_TARGET=10.14 ../.. > > but it fails here: > > [ 42%] Building CXX object > common/CMakeFiles/pcbcommon.dir/__/pcbnew/pcb_parser.cpp.o > /Users/lijon/Coding/kicad/pcbnew/pcb_parser.cpp:996:18: error: use of > undeclared identifier 'T_copper_finish' > case T_copper_finish: > ^ > /Users/lijon/Coding/kicad/pcbnew/pcb_parser.cpp:1002:18: error: use of > undeclared identifier 'T_edge_plating' > case T_edge_plating: > ^ > /Users/lijon/Coding/kicad/pcbnew/pcb_parser.cpp:1008:18: error: use of > undeclared identifier 'T_dielectric_constraints' > case T_dielectric_constraints: > ^ > /Users/lijon/Coding/kicad/pcbnew/pcb_parser.cpp:1014:18: error: use of > undeclared identifier 'T_edge_connector'; did you mean > 'T_zone_connect'? > case T_edge_connector: > ^~~~ > T_zone_connect > /Users/lijon/Coding/kicad/include/pcb_lexer.h:232:9: note: > 'T_zone_connect' declared here > T_zone_connect, > ^ > /Users/lijon/Coding/kicad/pcbnew/pcb_parser.cpp:1020:35: error: use of > undeclared identifier 'T_bevelled' > else if( token == T_bevelled ) > ^ > /Users/lijon/Coding/kicad/pcbnew/pcb_parser.cpp:1026:18: error: use of > undeclared identifier 'T_castellated_pads' > case T_castellated_pads: > ^ > /Users/lijon/Coding/kicad/pcbnew/pcb_parser.cpp:1088:30: error: use of > undeclared identifier 'T_addsublayer' > if( token == T_addsublayer ) > ^ > /Users/lijon/Coding/kicad/pcbnew/pcb_parser.cpp:1100:26: error: use of > undeclared identifier 'T_type'; did you mean 'T_top'? > case T_type: > ^~ > T_top > /Users/lijon/Coding/kicad/include/pcb_lexer.h:196:9: note: 'T_top' declared > here > T_top, > ^ > /Users/lijon/Coding/kicad/pcbnew/pcb_parser.cpp:1123:26: error: use of > undeclared identifier 'T_material'; did you mean 'T_general'? > case T_material: > ^~ > T_general > /Users/lijon/Coding/kicad/include/pcb_lexer.h:100:9: note: 'T_general' > declared here > T_general, > ^ > /Users/lijon/Coding/kicad/pcbnew/pcb_parser.cpp:1129:26: error: use of > undeclared identifier 'T_epsilon_r' > case T_epsilon_r: > ^ > /Users/lijon/Coding/kicad/pcbnew/pcb_parser.cpp:1135:26: error: use of > undeclared identifier 'T_loss_tangent' > case T_loss_tangent: > ^ > /Users/lijon/Coding/kicad/pcbnew/pcb_parser.cpp:1141:26: error: use of > undeclared identifier 'T_color'; did you mean 'T_bold'? > case T_color: > ^~~ > T_bold > /Users/lijon/Coding/kicad/include/pcb_lexer.h:60:9: note: 'T_bold' declared > here > T_bold, > ^ > /Users/lijon/Coding/kicad/pcbnew/pcb_parser.cpp:1431:14: error: use of > undeclared identifier 'T_stackup'; did you mean 'T_status'? > case T_stackup: > ^ > T_status > /Users/lijon/Coding/kicad/include/pcb_lexer.h:183:9: note: 'T_status' > declared here > T_status, > ^ > /Users/lijon/Coding/kicad/pcbnew/pcb_parser.cpp:1524:14: error: use of > undeclared identifier 'T_user_diff_pair' > case T_user_diff_pair: > ^ > /Users/lijon/Coding/kicad/pcbnew/pcb_parser.cpp:1571:14: error: use of > undeclared identifier 'T_defaults'; did you mean 'default'? > case T_defaults: > ^~ > default > /Users/lijon/Coding/kicad/pcbnew/pcb_parser.cpp:1571:14: error: > expected expression > /Users/lijon/Coding/kicad/pcbnew/pcb_parser.cpp:1638:14:
[Kicad-developers] Building current master fails
Hi, I'm trying to build the current master, cmake configured like this: cmake -DNGSPICE_INCLUDE_DIR=/Users/lijon/Coding/kicad-mac-builder/build/ngspice-dest/include/ -DNGSPICE_LIBRARY=/Users/lijon/Coding/kicad-mac-builder/build/ngspice-dest/lib/libngspice.dylib -DKICAD_USE_OCE=OFF -DKICAD_SCRIPTING=OFF -DwxWidgets_CONFIG_EXECUTABLE=/Users/lijon/Coding/kicad-mac-builder/build/wxwidgets-dest/bin/wx-config -DKICAD_BUILD_QA_TESTS=OFF -DCMAKE_OSX_DEPLOYMENT_TARGET=10.14 ../.. but it fails here: [ 42%] Building CXX object common/CMakeFiles/pcbcommon.dir/__/pcbnew/pcb_parser.cpp.o /Users/lijon/Coding/kicad/pcbnew/pcb_parser.cpp:996:18: error: use of undeclared identifier 'T_copper_finish' case T_copper_finish: ^ /Users/lijon/Coding/kicad/pcbnew/pcb_parser.cpp:1002:18: error: use of undeclared identifier 'T_edge_plating' case T_edge_plating: ^ /Users/lijon/Coding/kicad/pcbnew/pcb_parser.cpp:1008:18: error: use of undeclared identifier 'T_dielectric_constraints' case T_dielectric_constraints: ^ /Users/lijon/Coding/kicad/pcbnew/pcb_parser.cpp:1014:18: error: use of undeclared identifier 'T_edge_connector'; did you mean 'T_zone_connect'? case T_edge_connector: ^~~~ T_zone_connect /Users/lijon/Coding/kicad/include/pcb_lexer.h:232:9: note: 'T_zone_connect' declared here T_zone_connect, ^ /Users/lijon/Coding/kicad/pcbnew/pcb_parser.cpp:1020:35: error: use of undeclared identifier 'T_bevelled' else if( token == T_bevelled ) ^ /Users/lijon/Coding/kicad/pcbnew/pcb_parser.cpp:1026:18: error: use of undeclared identifier 'T_castellated_pads' case T_castellated_pads: ^ /Users/lijon/Coding/kicad/pcbnew/pcb_parser.cpp:1088:30: error: use of undeclared identifier 'T_addsublayer' if( token == T_addsublayer ) ^ /Users/lijon/Coding/kicad/pcbnew/pcb_parser.cpp:1100:26: error: use of undeclared identifier 'T_type'; did you mean 'T_top'? case T_type: ^~ T_top /Users/lijon/Coding/kicad/include/pcb_lexer.h:196:9: note: 'T_top' declared here T_top, ^ /Users/lijon/Coding/kicad/pcbnew/pcb_parser.cpp:1123:26: error: use of undeclared identifier 'T_material'; did you mean 'T_general'? case T_material: ^~ T_general /Users/lijon/Coding/kicad/include/pcb_lexer.h:100:9: note: 'T_general' declared here T_general, ^ /Users/lijon/Coding/kicad/pcbnew/pcb_parser.cpp:1129:26: error: use of undeclared identifier 'T_epsilon_r' case T_epsilon_r: ^ /Users/lijon/Coding/kicad/pcbnew/pcb_parser.cpp:1135:26: error: use of undeclared identifier 'T_loss_tangent' case T_loss_tangent: ^ /Users/lijon/Coding/kicad/pcbnew/pcb_parser.cpp:1141:26: error: use of undeclared identifier 'T_color'; did you mean 'T_bold'? case T_color: ^~~ T_bold /Users/lijon/Coding/kicad/include/pcb_lexer.h:60:9: note: 'T_bold' declared here T_bold, ^ /Users/lijon/Coding/kicad/pcbnew/pcb_parser.cpp:1431:14: error: use of undeclared identifier 'T_stackup'; did you mean 'T_status'? case T_stackup: ^ T_status /Users/lijon/Coding/kicad/include/pcb_lexer.h:183:9: note: 'T_status' declared here T_status, ^ /Users/lijon/Coding/kicad/pcbnew/pcb_parser.cpp:1524:14: error: use of undeclared identifier 'T_user_diff_pair' case T_user_diff_pair: ^ /Users/lijon/Coding/kicad/pcbnew/pcb_parser.cpp:1571:14: error: use of undeclared identifier 'T_defaults'; did you mean 'default'? case T_defaults: ^~ default /Users/lijon/Coding/kicad/pcbnew/pcb_parser.cpp:1571:14: error: expected expression /Users/lijon/Coding/kicad/pcbnew/pcb_parser.cpp:1638:14: error: use of undeclared identifier 'T_max_error'; did you mean 'T_mirror'? case T_max_error: ^~~ T_mirror /Users/lijon/Coding/kicad/include/pcb_lexer.h:122:9: note: 'T_mirror' declared here T_mirror, ^ /Users/lijon/Coding/kicad/pcbnew/pcb_parser.cpp:1639:58: error: use of undeclared identifier 'T_max_error'; did you mean 'T_mirror'? designSettings.m_MaxError = parseBoardUnits( T_max_error ); ^~~ T_mirror /Users/lijon/Coding/kicad/include/pcb_lexer.h:122:9: note: 'T_mirror' declared here T_mirror, ^ /Users/lijon/Coding/kicad/pcbnew/pcb_parser.cpp:1643:14: error: use of undeclared identifier
[Kicad-developers] [PATCH] Eeschema: simulator: use dotted traces for current and phase.
This patch makes it easier to visually keep apart traces for current vs voltage and phase vs magnitude, by using dotted line style for current and phase traces. To get the dotted style drawn correctly, I had to switch from drawing each trace line segment individually to creating an array and then using DrawLines() on it, so that the whole trace is one continuous line. -- /Jonatan http://kymatica.com 0002-Eeschema-simulator-use-dotted-traces-for-current-and.patch Description: Binary data ___ Mailing list: https://launchpad.net/~kicad-developers Post to : kicad-developers@lists.launchpad.net Unsubscribe : https://launchpad.net/~kicad-developers More help : https://help.launchpad.net/ListHelp
[Kicad-developers] [PATCH] Eeschema: simulator plot: allow standard mac pan and pinch to zoom
This patch allows standard trackpad gestures for pan and zoom, just like in eeschema and pcbnew etc. While working on this, I noticed mathplot.cpp is a bit messy, with a lot of commented out or unused code, and some buggy behaviours. This patch does not address any of that. -- /Jonatan http://kymatica.com 0001-Eeschema-simulator-plot-allow-standard-mac-pan-and-p.patch Description: Binary data ___ Mailing list: https://launchpad.net/~kicad-developers Post to : kicad-developers@lists.launchpad.net Unsubscribe : https://launchpad.net/~kicad-developers More help : https://help.launchpad.net/ListHelp
Re: [Kicad-developers] [PATCH] Eeschema: simulator: more trace colors, smarter color allocation
And here's a screenshot showing how (some of the) new colors looks like. On Fri, Nov 8, 2019 at 5:13 PM Jonatan Liljedahl wrote: > > NEW: The simulator plot now avoids colors already in use, if possible. > The palette is replaced with a larger one with less saturated colors. > > Fixes: lp:1851372 > * https://bugs.launchpad.net/kicad/+bug/1851372 > > > -- > /Jonatan > http://kymatica.com -- /Jonatan http://kymatica.com ___ Mailing list: https://launchpad.net/~kicad-developers Post to : kicad-developers@lists.launchpad.net Unsubscribe : https://launchpad.net/~kicad-developers More help : https://help.launchpad.net/ListHelp
Re: [Kicad-developers] [PATCH] simulation: allow entering signals (spice vectors) by name
Thank you Wayne! I would of course be very happy if you decide to merge it to 5.1 branch as well, since it's currently the only way to do math on simulation plots, and I understand 6.0 is quite far away. Regarding automatically populating the signal list with available spice vectors, I agree that this would be great and I might start working on that as well if no one else wants to go first. Maybe some internal spice vectors should be ignored? Specifically 'frequency' and 'time'. However, in my opinion I think it's a good idea to keep the "enter by name" even when automatically adding spice vectors to the signal list. Since I added the feature, I often use it *instead* of picking the signal in the list, because it's a lot faster to just write "/out" and hitting return instead of scrolling through this rather long list. And since I made it automatically select the text after adding the signal, one can simply just continue and enter a new signal and press return again, etc. I find it to be a very convenient way to add signals. Also, in the future I'd like to look into automatically generate "let" spice commands from expressions entered in this textbox, so one could write "=V(/in, /out)" to get the difference between in and out, etc. Cheers On Mon, Nov 11, 2019 at 3:47 PM Wayne Stambaugh wrote: > > Hi Jonaton, > > I pushed your patch into the master branch. I haven't made up my mind > on the 5.1 branch. Technically this is a new feature. We have an > unwritten policy to not backport new features from the master branch. > Given that this is a fairly trivial change, I may consider applying your > patch to the 5.1 branch after 5.1.5 is released. > > All that being said, I hope that this patch is a temporary fix and that > we use the information Holger provided on how to query ngspice to fetch > a list of vectors that can be plotted rather than manually entering > them. I'm sure someone is going to request this and even if they don't > I think this would provide a far better user experience. > > Thank you for your contribution to KiCad. > > Cheers, > > Wayne > > On 11/7/19 3:52 AM, Jonatan Liljedahl wrote: > > Hi, > > > > Here's an updated patch with fixed commit message. Also I removed some > > unrelated UI tweaks in sim_plot_frame that accidentally was part of > > the previous patch. This patch applies to both 5.1 and master. > > > > Cheers > > > > On Thu, Nov 7, 2019 at 12:04 AM Kymatica wrote: > >> > >> Oh, this was for 5.1. > >> I’ll look into rebasing on master tomorrow! Sleepy time here in Sweden... > >> > >> /Jonatan > >> > >>> 6 nov. 2019 kl. 23:57 skrev Wayne Stambaugh : > >>> > >>> This patch does not apply cleanly against master using `git am`. Did > >>> you forget to rebase against the master branch before creating your patch? > >>> > >>> Wayne > >>> > >>>> On 11/6/19 5:48 PM, Jonatan Liljedahl wrote: > >>>> Ok, now I think I've followed the policies correctly, here's the updated > >>>> patch. > >>>> > >>>> Cheers > >>>> > >>>>> On Wed, Nov 6, 2019 at 10:33 PM Wayne Stambaugh > >>>>> wrote: > >>>>> > >>>>> Hi Jonathon, > >>>>> > >>>>> I tested your original patch on the 5.1 branch and it seems to work as > >>>>> advertised. It would be nice if there was a way to query ngspice for a > >>>>> list of vectors rather than using just the net list to populate add > >>>>> signals dialog. Maybe Holger can weigh in on this. > >>>>> > >>>>> Before you submit your updated patch, please fix your code formatting to > >>>>> follow the KiCad Coding Policy[1] and your commit message to follow the > >>>>> KiCad Commit Message[2] policy. There is a git commit hook[3] to help > >>>>> you with coding policy issues. I saw some K curly brace placement in > >>>>> your patch. > >>>>> > >>>>> Thanks, > >>>>> > >>>>> Wayne > >>>>> > >>>>> [1]: > >>>>> http://docs.kicad-pcb.org/doxygen/md_Documentation_development_coding-style-policy.html > >>>>> [2]: http://docs.kicad-pcb.org/doxygen/commit_messages.html > >>>>> [3]: > >>>>> http://docs.kicad-pcb.org/doxygen/md_Documentation_development_coding-style-policy.html#tools > >>>>> > >>>>> On 11/6/19 4:0
[Kicad-developers] [PATCH] Eeschema: simulator: more trace colors, smarter color allocation
NEW: The simulator plot now avoids colors already in use, if possible. The palette is replaced with a larger one with less saturated colors. Fixes: lp:1851372 * https://bugs.launchpad.net/kicad/+bug/1851372 -- /Jonatan http://kymatica.com 0001-Eeschema-simulator-more-trace-colors-smarter-color-a.patch Description: Binary data ___ Mailing list: https://launchpad.net/~kicad-developers Post to : kicad-developers@lists.launchpad.net Unsubscribe : https://launchpad.net/~kicad-developers More help : https://help.launchpad.net/ListHelp
Re: [Kicad-developers] macos dev environment
On Fri, Nov 8, 2019 at 4:15 PM Bernhard Stegmaier wrote: > > I don’t know about ngspice or Python scripting, I don’t use them. > Maybe you just need to supply a suitable path to cmake when building? That ngspice path is correct, otherwise it wouldn't even build it with ngspice support as far as I know. However, I could get it to find ngspice by: $ export DYLD_FALLBACK_LIBRARY_PATH=/Users/lijon/Coding/kicad-mac-builder/build/ngspice-dest/lib $ kicad/kicad.app/Contents/MacOS/kicad Or, by placing a symlink to libngspice inside the kicad.app bundle: $ mkdir kicad/kicad.app/Contents/PlugIns/sim $ ln -s /Users/lijon/Coding/kicad-mac-builder/build/ngspice-dest/lib/libngspice.0.dylib kicad/kicad.app/Contents/PlugIns/sim/libngspice.0.dylib and then opening kicad.app as usual (in Finder or using the "open" shell command) > When you do a "make install" the binaries in the build folder are not usable > directly any more. > The bundling process of a “make install” rewrites the library paths in all > binaries/libs so that it only fits the layout of the bundle, but not the > layout of the build folders. > So, it is expected that it crashes or some other bad things happen, because > it won’t find libs any more. Yes, I see. -- /Jonatan http://kymatica.com ___ Mailing list: https://launchpad.net/~kicad-developers Post to : kicad-developers@lists.launchpad.net Unsubscribe : https://launchpad.net/~kicad-developers More help : https://help.launchpad.net/ListHelp
Re: [Kicad-developers] macos dev environment
On Thu, Nov 7, 2019 at 11:08 PM Adam Wolf wrote: > > Merging them in is blocked on the "eliminate symlinked bundles" task. Is this about the symlinked eeschema.app etc in the main kicad folder? Are the plans to copy instead of symlink? Or could we instead move them, and make kicad.app able to find those apps, so something like "APP_BUNDLE_DIR/../eeschema.app"? > I recently changed my employer and have less time for KiCad during the > transition but I hope to be back to normal soon. > > Bt, if you want to tackle any of the open tickets, it'd be a great helps. > Let me know if you are interested and I can write up some help if you want. I'm tempted, but currently I must limit my time spent on this because I have too much work to do in my own business (I make iOS music apps and a tiny bit of eurorack hardware modules). Cheers -- /Jonatan http://kymatica.com ___ Mailing list: https://launchpad.net/~kicad-developers Post to : kicad-developers@lists.launchpad.net Unsubscribe : https://launchpad.net/~kicad-developers More help : https://help.launchpad.net/ListHelp
Re: [Kicad-developers] macos dev environment
First time, open the built kicad.app directly did work, except that it couldn't find ngspice (which I need, because it's the simulator UI I'm working on...). After make install, opening the kicad.app directly (not the installed one), it crashes when I click the eeschema button in the project window. On Thu, Nov 7, 2019 at 11:05 PM Bernhard Stegmaier wrote: > > At least the last time I tried you don’t need to do a “make install” and > build the bundle if you want to test something during development. > Just run needed KiCad binary directly from the build folder… > > I don’t know about CLion, but that worked for me using Qt Creator without > problems. > > The only thing to remember is that if you do a “make install” once, it won’t > run from build folder any more. > > > Regards, > Bernhard > > On 7. Nov 2019, at 22:35, Jonatan Liljedahl wrote: > > Ok, I got it working by changing CMAKE_INSTALL_PREFIX to an absolute > path. So using "../bin" as it says in > http://docs.kicad-pcb.org/doxygen/md_Documentation_development_compiling.html#build_osx > does not actually work. > > Perhaps it would be a good idea to merge the kicad-mac-builder patches > instead of having to apply them each time? I know this was discussed > before, regarding the mismatch between KiCad version info string for > the macOS package, but I'm not sure what was decided. > > Cheers > > On Thu, Nov 7, 2019 at 10:07 PM Jonatan Liljedahl wrote: > > > Thanks. I managed to get it building directly in my kicad source dir > by pointing it to the custom wxWidgets etc that was built by > kicad-mac-builder. However, make install fails during fixup bundle > (see below). I have the patches from kicad-mac-builder applied. I've > attached my CMakeCache.txt. > > Install the project... > -- Install configuration: "Debug" > -- Removing existing application bundles... > -- Installing: > /Users/lijon/Coding/kicad/build/debug/../bin/../bin/kicad.app/Contents/Frameworks/libkicad_3dsg.2.0.0.dylib > -- Installing: > /Users/lijon/Coding/kicad/build/debug/../bin/../bin/kicad.app/Contents/Frameworks/libkicad_3dsg.dylib > -- Installing: > /Users/lijon/Coding/kicad/build/debug/../bin/../bin/eeschema.app > -- Installing: > /Users/lijon/Coding/kicad/build/debug/../bin/../bin/eeschema.app/Contents > -- Installing: > /Users/lijon/Coding/kicad/build/debug/../bin/../bin/eeschema.app/Contents/MacOS > -- Installing: > /Users/lijon/Coding/kicad/build/debug/../bin/../bin/eeschema.app/Contents/MacOS/eeschema > -- Installing: > /Users/lijon/Coding/kicad/build/debug/../bin/../bin/eeschema.app/Contents/Resources > -- Installing: > /Users/lijon/Coding/kicad/build/debug/../bin/../bin/eeschema.app/Contents/Resources/eeschema_doc.icns > -- Installing: > /Users/lijon/Coding/kicad/build/debug/../bin/../bin/eeschema.app/Contents/Resources/eeschema.icns > -- Installing: > /Users/lijon/Coding/kicad/build/debug/../bin/../bin/eeschema.app/Contents/Info.plist > -- fixup_bundle > -- app='../bin/eeschema.app/Contents/MacOS/eeschema' > -- libs='' > -- dirs='' > -- ignoreItems='' > -- fixup_bundle: preparing... > warning: target '../bin/eeschema.app/Contents/MacOS/eeschema' is not > absolute... > -- warning: gp_resolved_file_type expects absolute full path for first > arg original_file > -- warning: gp_resolved_file_type expects absolute full path for first > arg original_file > -- warning: gp_resolved_file_type expects absolute full path for first > arg original_file > -- warning: gp_resolved_file_type expects absolute full path for first > arg original_file > -- warning: gp_resolved_file_type expects absolute full path for first > arg original_file > -- warning: gp_resolved_file_type expects absolute full path for first > arg original_file > -- warning: gp_resolved_file_type expects absolute full path for first > arg original_file > -- warning: gp_resolved_file_type expects absolute full path for first > arg original_file > -- warning: gp_resolved_file_type expects absolute full path for first > arg original_file > -- warning: gp_resolved_file_type expects absolute full path for first > arg original_file > -- warning: gp_resolved_file_type expects absolute full path for first > arg original_file > -- warning: gp_resolved_file_type expects absolute full path for first > arg original_file > -- warning: gp_resolved_file_type expects absolute full path for first > arg original_file > -- fixup_bundle: copying... > -- 1/24: *NOT* copying > '/Users/lijon/Coding/kicad/build/bin/eeschema.app/Contents/MacOS/eeschema' > -- 2/24: linking > '/Users/lijon/Coding/kicad/build/bin/eeschema.app/Contents/Frameworks/libwx_osx_cocoau-3.0.0.4.0.dylib' > -> > '/Users/
Re: [Kicad-developers] macos dev environment
Ok, I got it working by changing CMAKE_INSTALL_PREFIX to an absolute path. So using "../bin" as it says in http://docs.kicad-pcb.org/doxygen/md_Documentation_development_compiling.html#build_osx does not actually work. Perhaps it would be a good idea to merge the kicad-mac-builder patches instead of having to apply them each time? I know this was discussed before, regarding the mismatch between KiCad version info string for the macOS package, but I'm not sure what was decided. Cheers On Thu, Nov 7, 2019 at 10:07 PM Jonatan Liljedahl wrote: > > Thanks. I managed to get it building directly in my kicad source dir > by pointing it to the custom wxWidgets etc that was built by > kicad-mac-builder. However, make install fails during fixup bundle > (see below). I have the patches from kicad-mac-builder applied. I've > attached my CMakeCache.txt. > > Install the project... > -- Install configuration: "Debug" > -- Removing existing application bundles... > -- Installing: > /Users/lijon/Coding/kicad/build/debug/../bin/../bin/kicad.app/Contents/Frameworks/libkicad_3dsg.2.0.0.dylib > -- Installing: > /Users/lijon/Coding/kicad/build/debug/../bin/../bin/kicad.app/Contents/Frameworks/libkicad_3dsg.dylib > -- Installing: > /Users/lijon/Coding/kicad/build/debug/../bin/../bin/eeschema.app > -- Installing: > /Users/lijon/Coding/kicad/build/debug/../bin/../bin/eeschema.app/Contents > -- Installing: > /Users/lijon/Coding/kicad/build/debug/../bin/../bin/eeschema.app/Contents/MacOS > -- Installing: > /Users/lijon/Coding/kicad/build/debug/../bin/../bin/eeschema.app/Contents/MacOS/eeschema > -- Installing: > /Users/lijon/Coding/kicad/build/debug/../bin/../bin/eeschema.app/Contents/Resources > -- Installing: > /Users/lijon/Coding/kicad/build/debug/../bin/../bin/eeschema.app/Contents/Resources/eeschema_doc.icns > -- Installing: > /Users/lijon/Coding/kicad/build/debug/../bin/../bin/eeschema.app/Contents/Resources/eeschema.icns > -- Installing: > /Users/lijon/Coding/kicad/build/debug/../bin/../bin/eeschema.app/Contents/Info.plist > -- fixup_bundle > -- app='../bin/eeschema.app/Contents/MacOS/eeschema' > -- libs='' > -- dirs='' > -- ignoreItems='' > -- fixup_bundle: preparing... > warning: target '../bin/eeschema.app/Contents/MacOS/eeschema' is not > absolute... > -- warning: gp_resolved_file_type expects absolute full path for first > arg original_file > -- warning: gp_resolved_file_type expects absolute full path for first > arg original_file > -- warning: gp_resolved_file_type expects absolute full path for first > arg original_file > -- warning: gp_resolved_file_type expects absolute full path for first > arg original_file > -- warning: gp_resolved_file_type expects absolute full path for first > arg original_file > -- warning: gp_resolved_file_type expects absolute full path for first > arg original_file > -- warning: gp_resolved_file_type expects absolute full path for first > arg original_file > -- warning: gp_resolved_file_type expects absolute full path for first > arg original_file > -- warning: gp_resolved_file_type expects absolute full path for first > arg original_file > -- warning: gp_resolved_file_type expects absolute full path for first > arg original_file > -- warning: gp_resolved_file_type expects absolute full path for first > arg original_file > -- warning: gp_resolved_file_type expects absolute full path for first > arg original_file > -- warning: gp_resolved_file_type expects absolute full path for first > arg original_file > -- fixup_bundle: copying... > -- 1/24: *NOT* copying > '/Users/lijon/Coding/kicad/build/bin/eeschema.app/Contents/MacOS/eeschema' > -- 2/24: linking > '/Users/lijon/Coding/kicad/build/bin/eeschema.app/Contents/Frameworks/libwx_osx_cocoau-3.0.0.4.0.dylib' > -> > '/Users/lijon/Coding/kicad/build/bin/eeschema.app/Contents/Frameworks/libwx_osx_cocoau-3.0.0.dylib' > -- 3/24: linking > '/Users/lijon/Coding/kicad/build/bin/eeschema.app/Contents/Frameworks/libwx_osx_cocoau_gl-3.0.0.4.0.dylib' > -> > '/Users/lijon/Coding/kicad/build/bin/eeschema.app/Contents/Frameworks/libwx_osx_cocoau_gl-3.0.0.dylib' > -- 4/24: copying '/usr/local/opt/cairo/lib/libcairo.2.dylib' > -- 5/24: copying '/usr/local/opt/fontconfig/lib/libfontconfig.1.dylib' > -- 6/24: copying '/usr/local/opt/freetype/lib/libfreetype.6.dylib' > -- 7/24: linking > '/Users/lijon/Coding/kicad/build/bin/eeschema.app/Contents/Frameworks/libGLEW.2.1.0.dylib' > -> > '/Users/lijon/Coding/kicad/build/bin/eeschema.app/Contents/Frameworks/libGLEW.2.1.dylib' > -- 8/24: copying '/usr/local/opt/libpng/lib/libpng16.16.dylib' > -- 9/24: copying '/usr/local/opt/pixman/lib/libpixman-1.0.dylib' > -- 10/24: copying > '/Users/lijon/Coding/kicad
Re: [Kicad-developers] macos dev environment
Utilities.cmake:989 (fixup_bundle_item) eeschema/cmake_install.cmake:71 (fixup_bundle) cmake_install.cmake:57 (include) make: *** [install] Error 1 On Thu, Nov 7, 2019 at 6:02 PM Adam Wolf wrote: > > We're going to do that in the simple way, which is to copy rather than > symlink. It costs a few megs, but that should be fine. > > On Thu, Nov 7, 2019 at 9:12 AM Jon Evans wrote: > > > > I also use CLion. It's okay but not perfect. It's been a while since I've > > been compiling on my Mac but from what I recall, the most frustrating part > > is the need to re-generate all of the application packages whenever > > anything is touched (which is a slow and non-parallel process). > > > > It's possible that implementing one of the possible fixes for > > https://bugs.launchpad.net/kicad/+bug/1826649 will also enable faster > > incremental builds on MacOS > > > > -Jon > > > > On Thu, Nov 7, 2019 at 10:07 AM Adam Wolf > > wrote: > >> > >> I use CLion, but I also would love to see kicad-mac-builder improved > >> to the point where it could do the right things for dev, as well. > >> > >> Adam > >> > >> On Thu, Nov 7, 2019 at 7:45 AM Jonatan Liljedahl > >> wrote: > >> > > >> > Are there any developers on macOS that would like to give me some > >> > hints on how to set up a good environment? Currently I'm using > >> > kicad-mac-builder, but it's not well suited for the kind of iterative > >> > rebuilding during development. > >> > > >> > 1. It applies the patches each time. > >> > 2. It resets the git repo each time. > >> > 3. It re-downloads all docs etc each time. > >> > 4. It does not rebuild the changed source files. > >> > > >> > So for now, I've just commented out the docs dependency and commands > >> > for applying the patches and resetting git, etc. > >> > > >> > But it still does not rebuild when I change some source file, so I need > >> > to do: > >> > > >> > make -f CMakeFiles/Makefile2 CMakeFiles/kicad.dir/clean > >> > > >> > and rebuild, which probably takes more time than needed. > >> > > >> > -- > >> > /Jonatan > >> > http://kymatica.com > >> > > >> > ___ > >> > Mailing list: https://launchpad.net/~kicad-developers > >> > Post to : kicad-developers@lists.launchpad.net > >> > Unsubscribe : https://launchpad.net/~kicad-developers > >> > More help : https://help.launchpad.net/ListHelp > >> > >> ___ > >> Mailing list: https://launchpad.net/~kicad-developers > >> Post to : kicad-developers@lists.launchpad.net > >> Unsubscribe : https://launchpad.net/~kicad-developers > >> More help : https://help.launchpad.net/ListHelp -- /Jonatan http://kymatica.com # This is the CMakeCache file. # For build in directory: /Users/lijon/Coding/kicad/build/debug # It was generated by CMake: /Applications/CMake.app/Contents/bin/cmake # You can edit this file to change values found and used by cmake. # If you do not want to change any of the values, simply exit the editor. # If you do want to change a value, simply edit, save, and exit the editor. # The syntax for the file is as follows: # KEY:TYPE=VALUE # KEY is the name of a variable in the cache. # TYPE is a hint to GUIs for the type of VALUE, DO NOT EDIT TYPE!. # VALUE is the current value for the KEY. # EXTERNAL cache entries //Dependencies for the target 3d-viewer_LIB_DEPENDS:STATIC=general;gal;general;polygon;general;-L/Users/lijon/Coding/kicad-mac-builder/build/wxwidgets-dest/lib;general;-framework IOKit;general;-framework Carbon;general;-framework Cocoa;general;-framework AudioToolbox;general;-framework System;general;-framework OpenGL;general;-lwx_osx_cocoau_gl-3.0;general;-lwx_osx_cocoau-3.0;general;/Applications/Xcode.app/Contents/Developer/Platforms/MacOSX.platform/Developer/SDKs/MacOSX10.14.sdk/System/Library/Frameworks/OpenGL.framework;general;/Applications/Xcode.app/Contents/Developer/Platforms/MacOSX.platform/Developer/SDKs/MacOSX10.14.sdk/System/Library/Frameworks/OpenGL.framework;general;kicad_3dsg; //Build the GITHUB_PLUGIN for pcbnew. BUILD_GITHUB_PLUGIN:BOOL=OFF //In debug build: create smaller binaries. BUILD_SMALL_DEBUG_FILES:BOOL=OFF //The directory containing a CMake configuration file for Boost. Boost_DIR:PATH=Boost_DIR-NOTFOUND //Boost filesystem library (debug) Boost_FILESYSTEM_LIBRARY_DEBUG:FILEPATH=/usr/local/
Re: [Kicad-developers] SIM_PLOT_TYPE
Indeed, it seems that in all cases the actual spice vector used for
the X axis will be "frequency", which as far as I can tell is always
log. Perhaps it would make sense to just replace SPT_LOG_FREQUENCY and
SPT_LIN_FREQUENCY with a single SPT_FREQUENCY enum?
I'm not sure how plotting linear frequency works in ngspice, is this
done with some command that alters the "frequency" vector (which would
then propagate to KiCad), or is it handled in the plot command?
Cheers
On Thu, Nov 7, 2019 at 2:13 AM Ian McInerney wrote:
>
> I would guess that is a mistake. It seems that the intention is to eventually
> have both linear and log scaling, so they would need to be different. Right
> now though it looks like log scaling isn't handled differently (a search for
> SPT_LOG_FREQUENCY returns a TODO comment and an if check that is also for the
> SPT_LIN_FREQUENCY).
>
> -Ian
>
> On Wed, Nov 6, 2019 at 9:24 PM Jonatan Liljedahl wrote:
>>
>> in sim_types.h:
>>
>> //> Possible plot types
>> enum SIM_PLOT_TYPE {
>> // Y axis
>> SPT_VOLTAGE = 0x01,
>> SPT_CURRENT = 0x02,
>> SPT_AC_PHASE= 0x04,
>> SPT_AC_MAG = 0x08,
>>
>> // X axis
>> SPT_TIME= 0x10,
>> SPT_LIN_FREQUENCY = 0x20,
>> SPT_LOG_FREQUENCY = 0x20,
>> SPT_SWEEP = 0x40
>> };
>>
>> Is it by purpose that the lin/log freq flags have the same value?
>>
>> --
>> /Jonatan
>> http://kymatica.com
>>
>> ___
>> Mailing list: https://launchpad.net/~kicad-developers
>> Post to : kicad-developers@lists.launchpad.net
>> Unsubscribe : https://launchpad.net/~kicad-developers
>> More help : https://help.launchpad.net/ListHelp
--
/Jonatan
http://kymatica.com
___
Mailing list: https://launchpad.net/~kicad-developers
Post to : kicad-developers@lists.launchpad.net
Unsubscribe : https://launchpad.net/~kicad-developers
More help : https://help.launchpad.net/ListHelp
[Kicad-developers] macos dev environment
Are there any developers on macOS that would like to give me some hints on how to set up a good environment? Currently I'm using kicad-mac-builder, but it's not well suited for the kind of iterative rebuilding during development. 1. It applies the patches each time. 2. It resets the git repo each time. 3. It re-downloads all docs etc each time. 4. It does not rebuild the changed source files. So for now, I've just commented out the docs dependency and commands for applying the patches and resetting git, etc. But it still does not rebuild when I change some source file, so I need to do: make -f CMakeFiles/Makefile2 CMakeFiles/kicad.dir/clean and rebuild, which probably takes more time than needed. -- /Jonatan http://kymatica.com ___ Mailing list: https://launchpad.net/~kicad-developers Post to : kicad-developers@lists.launchpad.net Unsubscribe : https://launchpad.net/~kicad-developers More help : https://help.launchpad.net/ListHelp
Re: [Kicad-developers] [PATCH] simulation: allow entering signals (spice vectors) by name
Hi,
Here's an updated patch with fixed commit message. Also I removed some
unrelated UI tweaks in sim_plot_frame that accidentally was part of
the previous patch. This patch applies to both 5.1 and master.
Cheers
On Thu, Nov 7, 2019 at 12:04 AM Kymatica wrote:
>
> Oh, this was for 5.1.
> I’ll look into rebasing on master tomorrow! Sleepy time here in Sweden...
>
> /Jonatan
>
> > 6 nov. 2019 kl. 23:57 skrev Wayne Stambaugh :
> >
> > This patch does not apply cleanly against master using `git am`. Did
> > you forget to rebase against the master branch before creating your patch?
> >
> > Wayne
> >
> >> On 11/6/19 5:48 PM, Jonatan Liljedahl wrote:
> >> Ok, now I think I've followed the policies correctly, here's the updated
> >> patch.
> >>
> >> Cheers
> >>
> >>> On Wed, Nov 6, 2019 at 10:33 PM Wayne Stambaugh
> >>> wrote:
> >>>
> >>> Hi Jonathon,
> >>>
> >>> I tested your original patch on the 5.1 branch and it seems to work as
> >>> advertised. It would be nice if there was a way to query ngspice for a
> >>> list of vectors rather than using just the net list to populate add
> >>> signals dialog. Maybe Holger can weigh in on this.
> >>>
> >>> Before you submit your updated patch, please fix your code formatting to
> >>> follow the KiCad Coding Policy[1] and your commit message to follow the
> >>> KiCad Commit Message[2] policy. There is a git commit hook[3] to help
> >>> you with coding policy issues. I saw some K curly brace placement in
> >>> your patch.
> >>>
> >>> Thanks,
> >>>
> >>> Wayne
> >>>
> >>> [1]:
> >>> http://docs.kicad-pcb.org/doxygen/md_Documentation_development_coding-style-policy.html
> >>> [2]: http://docs.kicad-pcb.org/doxygen/commit_messages.html
> >>> [3]:
> >>> http://docs.kicad-pcb.org/doxygen/md_Documentation_development_coding-style-policy.html#tools
> >>>
> >>> On 11/6/19 4:06 PM, Jonatan Liljedahl wrote:
> >>>> This patch does not implement all of pt 4 in that bug report, it only
> >>>> allows adding plots of vectors by name, including vectors created by
> >>>> LET statements in a spice text block in the schematic (which are not
> >>>> listed in the signals list, since that list is based on the net names,
> >>>> not the existing spice vectors).
> >>>>
> >>>> However, I just realized that there's a problem with my patch, it only
> >>>> works for V, not I. I'll take a look and update my patch.
> >>>>
> >>>> On Wed, Nov 6, 2019 at 9:08 PM Sylwester Kocjan wrote:
> >>>>>
> >>>>> Hi Jonatan,
> >>>>>
> >>>>> On 06/11/2019 13:29, Jonatan Liljedahl wrote:
> >>>>>> useful for plotting calculated vectors ("let AB=V(a)*V(b)"),
> >>>>>> see for example
> >>>>>> https://forum.kicad.info/t/ac-analysis-of-op-amp-loop-gain-and-phase/19661/6
> >>>>>>
> >>>>>
> >>>>> Just small remark to limit entropy in bug reports: it looks like you
> >>>>> have implemented pt. 4 from:
> >>>>>
> >>>>> https://bugs.launchpad.net/kicad/+bug/1814188
> >>>>>
> >>>>> Maybe it's worth to update that bug report if it will be merged.
> >>>>>
> >>>>> Best regards,
> >>>>> Sylwester
> >>>>>
> >>>>> ___
> >>>>> Mailing list: https://launchpad.net/~kicad-developers
> >>>>> Post to : kicad-developers@lists.launchpad.net
> >>>>> Unsubscribe : https://launchpad.net/~kicad-developers
> >>>>> More help : https://help.launchpad.net/ListHelp
> >>>>
> >>>>
> >>>>
> >>>
> >>> ___
> >>> Mailing list: https://launchpad.net/~kicad-developers
> >>> Post to : kicad-developers@lists.launchpad.net
> >>> Unsubscribe : https://launchpad.net/~kicad-developers
> >>> More help : https://help.launchpad.net/ListHelp
> >>
> >>
> >>
> >
> > ___
> > Mailing list: https://launchpad.net/~kicad-developers
> > Post to : kicad-developers@lists.launchpad.net
> > Unsubscribe : https://launchpad.net/~kicad-developers
> > More help : https://help.launchpad.net/ListHelp
--
/Jonatan
http://kymatica.com
0001-Eeschema-simulation-allow-adding-signals-by-name-lik.patch
Description: Binary data
___
Mailing list: https://launchpad.net/~kicad-developers
Post to : kicad-developers@lists.launchpad.net
Unsubscribe : https://launchpad.net/~kicad-developers
More help : https://help.launchpad.net/ListHelp
[Kicad-developers] SIM_PLOT_TYPE
in sim_types.h:
//> Possible plot types
enum SIM_PLOT_TYPE {
// Y axis
SPT_VOLTAGE = 0x01,
SPT_CURRENT = 0x02,
SPT_AC_PHASE= 0x04,
SPT_AC_MAG = 0x08,
// X axis
SPT_TIME= 0x10,
SPT_LIN_FREQUENCY = 0x20,
SPT_LOG_FREQUENCY = 0x20,
SPT_SWEEP = 0x40
};
Is it by purpose that the lin/log freq flags have the same value?
--
/Jonatan
http://kymatica.com
___
Mailing list: https://launchpad.net/~kicad-developers
Post to : kicad-developers@lists.launchpad.net
Unsubscribe : https://launchpad.net/~kicad-developers
More help : https://help.launchpad.net/ListHelp
Re: [Kicad-developers] [PATCH] simulation: allow entering signals (spice vectors) by name
This patch does not implement all of pt 4 in that bug report, it only
allows adding plots of vectors by name, including vectors created by
LET statements in a spice text block in the schematic (which are not
listed in the signals list, since that list is based on the net names,
not the existing spice vectors).
However, I just realized that there's a problem with my patch, it only
works for V, not I. I'll take a look and update my patch.
On Wed, Nov 6, 2019 at 9:08 PM Sylwester Kocjan wrote:
>
> Hi Jonatan,
>
> On 06/11/2019 13:29, Jonatan Liljedahl wrote:
> > useful for plotting calculated vectors ("let AB=V(a)*V(b)"),
> > see for example
> > https://forum.kicad.info/t/ac-analysis-of-op-amp-loop-gain-and-phase/19661/6
> >
>
> Just small remark to limit entropy in bug reports: it looks like you
> have implemented pt. 4 from:
>
> https://bugs.launchpad.net/kicad/+bug/1814188
>
> Maybe it's worth to update that bug report if it will be merged.
>
> Best regards,
> Sylwester
>
> ___
> Mailing list: https://launchpad.net/~kicad-developers
> Post to : kicad-developers@lists.launchpad.net
> Unsubscribe : https://launchpad.net/~kicad-developers
> More help : https://help.launchpad.net/ListHelp
--
/Jonatan
http://kymatica.com
___
Mailing list: https://launchpad.net/~kicad-developers
Post to : kicad-developers@lists.launchpad.net
Unsubscribe : https://launchpad.net/~kicad-developers
More help : https://help.launchpad.net/ListHelp
[Kicad-developers] [PATCH] simulation: allow entering signals (spice vectors) by name
useful for plotting calculated vectors ("let AB=V(a)*V(b)"),
see for example
https://forum.kicad.info/t/ac-analysis-of-op-amp-loop-gain-and-phase/19661/6
--
/Jonatan
http://kymatica.com
0001-simulation-allow-entering-signals-spice-vectors-by-n.patch
Description: Binary data
___
Mailing list: https://launchpad.net/~kicad-developers
Post to : kicad-developers@lists.launchpad.net
Unsubscribe : https://launchpad.net/~kicad-developers
More help : https://help.launchpad.net/ListHelp
Re: [Kicad-developers] doing math on plots
For convenience, I created the following schematic symbol, and this subcircuit:
.subckt DIFFMETER a b out1 out2
BV1 out1 GND V=V(a,b)
R1 out2 GND 1g
.ends
By putting the pins out1 and out2 in the same location, hiding out2, I
could trick KiCad to consider any wire/label there connected.
It would be nice if one could pass the actual expression (V value) as
a subcircuit parameter, but my attempts at that failed. Not sure if
it's possible, Holger?
Cheers
On Thu, Oct 31, 2019 at 11:14 AM Jonatan Liljedahl wrote:
>
> On Wed, Oct 30, 2019 at 6:42 PM Holger Vogt wrote:
> >
> > The current eeschema-ngspice interface is very limited.
>
> Are there any plans or roadmap for improving it?
>
> > > How would one plot, for example, the difference between two vectors?
> > > I tried this in a text box:
> > >
> > > .save foo=(‘v(/input)-v(/output2)’)
> > > .tran 10u 50m
> > >
> > > but "foo" does not show up in the list of vectors to display in the plot
> > > window.
> >
> > Here you might have a look at
> > https://forum.kicad.info/t/spice-plotting-difference-of-voltages/19545/2
>
> Thanks! Also I found this way: I added a symbol and a dummy resistor,
> setting the symbols Spice_Primitive and Value such that I get this in
> the netlist:
>
> BV1 /diff GND V=V(/input,/output)
> Rdummy1 NC_03 /diff 1g
>
> "V(/diff)" then shows up in the kicad plot menu. This also work for
> other operations than diffing, for example
>
> BV2 /mul GND V=V(/input)*V(/output)/100
>
> It would be nice if one could simply append stuff to the netlist in a
> textblock, is this possible?
>
> > > Another thing, I found that one can use parameters for values, for
> > > example {Rx} for a resistor value and then add a textbox with ".param
> > > Rx=100k". Would it be possible to simultaneous get plots for a set of
> > > different values of Rx?
> > >
> >
> > Here you might try external ngspice. KiCad 5.1.x has a direct
> > interface, where you generate a netlist from your circuit and then may
> > call ngspice. This will offer the full ngspice capabilities and plotting
> > via ngspice or gnuplot. I have described an example at
> > http://ngspice.sourceforge.net/ngspice-eeschema.html#external .
> >
> > Unfortunately this interface has disappeared in KiCad 5.9.9 . I still
> > will have to make a wish list bug report to get this back.
>
> Is this supposed to work on macOS as well? I downloaded the ngspice
> package but the binary fails to run:
>
> $ ./ngspice
> dyld: Library not loaded: /opt/X11/lib/libXaw.7.dylib
> Referenced from: /Applications/ngspice/bin/./ngspice
> Reason: image not found
> Abort trap: 6
>
>
> --
> /Jonatan
> http://kymatica.com
--
/Jonatan
http://kymatica.com
___
Mailing list: https://launchpad.net/~kicad-developers
Post to : kicad-developers@lists.launchpad.net
Unsubscribe : https://launchpad.net/~kicad-developers
More help : https://help.launchpad.net/ListHelp
Re: [Kicad-developers] doing math on plots
On Wed, Oct 30, 2019 at 6:42 PM Holger Vogt wrote:
>
> The current eeschema-ngspice interface is very limited.
Are there any plans or roadmap for improving it?
> > How would one plot, for example, the difference between two vectors?
> > I tried this in a text box:
> >
> > .save foo=(‘v(/input)-v(/output2)’)
> > .tran 10u 50m
> >
> > but "foo" does not show up in the list of vectors to display in the plot
> > window.
>
> Here you might have a look at
> https://forum.kicad.info/t/spice-plotting-difference-of-voltages/19545/2
Thanks! Also I found this way: I added a symbol and a dummy resistor,
setting the symbols Spice_Primitive and Value such that I get this in
the netlist:
BV1 /diff GND V=V(/input,/output)
Rdummy1 NC_03 /diff 1g
"V(/diff)" then shows up in the kicad plot menu. This also work for
other operations than diffing, for example
BV2 /mul GND V=V(/input)*V(/output)/100
It would be nice if one could simply append stuff to the netlist in a
textblock, is this possible?
> > Another thing, I found that one can use parameters for values, for
> > example {Rx} for a resistor value and then add a textbox with ".param
> > Rx=100k". Would it be possible to simultaneous get plots for a set of
> > different values of Rx?
> >
>
> Here you might try external ngspice. KiCad 5.1.x has a direct
> interface, where you generate a netlist from your circuit and then may
> call ngspice. This will offer the full ngspice capabilities and plotting
> via ngspice or gnuplot. I have described an example at
> http://ngspice.sourceforge.net/ngspice-eeschema.html#external .
>
> Unfortunately this interface has disappeared in KiCad 5.9.9 . I still
> will have to make a wish list bug report to get this back.
Is this supposed to work on macOS as well? I downloaded the ngspice
package but the binary fails to run:
$ ./ngspice
dyld: Library not loaded: /opt/X11/lib/libXaw.7.dylib
Referenced from: /Applications/ngspice/bin/./ngspice
Reason: image not found
Abort trap: 6
--
/Jonatan
http://kymatica.com
___
Mailing list: https://launchpad.net/~kicad-developers
Post to : kicad-developers@lists.launchpad.net
Unsubscribe : https://launchpad.net/~kicad-developers
More help : https://help.launchpad.net/ListHelp
[Kicad-developers] doing math on plots
How would one plot, for example, the difference between two vectors?
I tried this in a text box:
.save foo=(‘v(/input)-v(/output2)’)
.tran 10u 50m
but "foo" does not show up in the list of vectors to display in the plot window.
Another thing, I found that one can use parameters for values, for
example {Rx} for a resistor value and then add a textbox with ".param
Rx=100k". Would it be possible to simultaneous get plots for a set of
different values of Rx?
Cheers
--
/Jonatan
http://kymatica.com
___
Mailing list: https://launchpad.net/~kicad-developers
Post to : kicad-developers@lists.launchpad.net
Unsubscribe : https://launchpad.net/~kicad-developers
More help : https://help.launchpad.net/ListHelp
Re: [Kicad-developers] Simulator still crashes
Thanks! On Mon, Sep 30, 2019 at 3:19 PM Holger Vogt wrote: > > Please see the discussions at https://bugs.launchpad.net/kicad/+bug/1835296 > > ___ > Mailing list: https://launchpad.net/~kicad-developers > Post to : kicad-developers@lists.launchpad.net > Unsubscribe : https://launchpad.net/~kicad-developers > More help : https://help.launchpad.net/ListHelp -- /Jonatan http://kymatica.com ___ Mailing list: https://launchpad.net/~kicad-developers Post to : kicad-developers@lists.launchpad.net Unsubscribe : https://launchpad.net/~kicad-developers More help : https://help.launchpad.net/ListHelp
[Kicad-developers] Simulator still crashes
I've updated my bug report here: https://bugs.launchpad.net/kicad/+bug/1841752 Can someone else on MacOS please test and see if they can reproduce it? -- /Jonatan http://kymatica.com ___ Mailing list: https://launchpad.net/~kicad-developers Post to : kicad-developers@lists.launchpad.net Unsubscribe : https://launchpad.net/~kicad-developers More help : https://help.launchpad.net/ListHelp
Re: [Kicad-developers] Selection highlighting changes.
On Thu, Sep 5, 2019 at 3:20 PM Jeff Young wrote: > > That was tried, but there were complaints that you then lose the differential > colouring between different items. (And it makes selections look just like > highlighting — and there’s enough confusion there already.) I'm not sure I understand the need for differential coloring inside the selection highlight? Highlighting nets are purple/pink, cross-probing is red, selection could be some other bright color. Another alternative would be to use the new halo but with clean sharp edges instead of the fading-out / glow (which IMHO looks a bit unprofessional). I attach a crude mockup of what I mean.. > Cheers, > Jeff. > > > > On 5 Sep 2019, at 14:05, Jonatan Liljedahl wrote: > > > > Personally, I'd much more prefer it like this (see attached screenshot) :) > > > > On Thu, Sep 5, 2019 at 2:07 PM Wayne Stambaugh wrote: > >> > >> I would prefer that we pick a more obvious color out of the box. Under > >> certain zoom conditions it's difficult to tell if anything is even > >> selected on gtk3. Users could make it more subtle if they find it too > >> obnoxious. I think that would be better than a bug report about > >> selection highlighting not working. > >> > >> Wayne > >> > >> On 9/4/19 7:24 PM, Seth Hillbrand wrote: > >>> Oh, that's right. We had been taking the color from system preferences > >>> but this did not show up correctly on schematic sheets that were not the > >>> system window color. Allowing for customization fixes this but the > >>> coloring is very light. It apparently looks correct on Mac but is very > >>> washed out under GTK3. I've added a conditional for the defaults that > >>> sets GTK/MSW values to the ones that seem appropriate on GTK3. We may > >>> want an additional default value for MSW if it similarly appears off. > >>> > >>> -S > >>> > >>> On 2019-09-04 18:12, Jeff Young wrote: > >>>> The colour is in Preferences now; see if it’s just too washed out on > >>>> your monitor. > >>>> > >>>>> On 4 Sep 2019, at 22:35, Seth Hillbrand wrote: > >>>>> > >>>>> I'm working on one (slowly) but it hasn't been pushed. GTK3 in the > >>>>> master branch looks normal for me. > >>>>> > >>>>> Does it look the same in OpenGL and Cairo? > >>>>> > >>>>> _Seth > >>>>> > >>>>> On 2019-09-04 15:36, Wayne Stambaugh wrote: > >>>>>> Was there a change recently to the selection highlighting in Eeschema? > >>>>>> I can barely see any difference between the selected and unselected > >>>>>> objects. If there has not been changes then something needs to be done > >>>>>> about the highlighting on gtk3 builds on linux. I thought I would > >>>>>> check > >>>>>> before I filed a bug report. > >>>>>> Cheers, > >>>>>> Wayne > >>>>>> ___ > >>>>>> Mailing list: https://launchpad.net/~kicad-developers > >>>>>> Post to : kicad-developers@lists.launchpad.net > >>>>>> Unsubscribe : https://launchpad.net/~kicad-developers > >>>>>> More help : https://help.launchpad.net/ListHelp > >>>>> > >>>>> ___ > >>>>> Mailing list: https://launchpad.net/~kicad-developers > >>>>> Post to : kicad-developers@lists.launchpad.net > >>>>> Unsubscribe : https://launchpad.net/~kicad-developers > >>>>> More help : https://help.launchpad.net/ListHelp > >> > >> ___ > >> Mailing list: https://launchpad.net/~kicad-developers > >> Post to : kicad-developers@lists.launchpad.net > >> Unsubscribe : https://launchpad.net/~kicad-developers > >> More help : https://help.launchpad.net/ListHelp > > > > > > > > -- > > /Jonatan > > http://kymatica.com > > > -- /Jonatan http://kymatica.com ___ Mailing list: https://launchpad.net/~kicad-developers Post to : kicad-developers@lists.launchpad.net Unsubscribe : https://launchpad.net/~kicad-developers More help : https://help.launchpad.net/ListHelp
Re: [Kicad-developers] knowing last commit included in a nightly build?
On Thu, Sep 5, 2019 at 1:59 PM Nick Østergaard wrote: > > Does the sha listed in the version info from the runtime app not match? Unfortunately not. For example (5.1.4-60-ga2950354a) lijon-mbp:kicad lijon$ git branch * 5.1 master lijon-mbp:kicad lijon$ git log --oneline | grep ga2950354a lijon-mbp:kicad lijon$ -- /Jonatan http://kymatica.com ___ Mailing list: https://launchpad.net/~kicad-developers Post to : kicad-developers@lists.launchpad.net Unsubscribe : https://launchpad.net/~kicad-developers More help : https://help.launchpad.net/ListHelp
Re: [Kicad-developers] Selection highlighting changes.
Personally, I'd much more prefer it like this (see attached screenshot) :) On Thu, Sep 5, 2019 at 2:07 PM Wayne Stambaugh wrote: > > I would prefer that we pick a more obvious color out of the box. Under > certain zoom conditions it's difficult to tell if anything is even > selected on gtk3. Users could make it more subtle if they find it too > obnoxious. I think that would be better than a bug report about > selection highlighting not working. > > Wayne > > On 9/4/19 7:24 PM, Seth Hillbrand wrote: > > Oh, that's right. We had been taking the color from system preferences > > but this did not show up correctly on schematic sheets that were not the > > system window color. Allowing for customization fixes this but the > > coloring is very light. It apparently looks correct on Mac but is very > > washed out under GTK3. I've added a conditional for the defaults that > > sets GTK/MSW values to the ones that seem appropriate on GTK3. We may > > want an additional default value for MSW if it similarly appears off. > > > > -S > > > > On 2019-09-04 18:12, Jeff Young wrote: > >> The colour is in Preferences now; see if it’s just too washed out on > >> your monitor. > >> > >>> On 4 Sep 2019, at 22:35, Seth Hillbrand wrote: > >>> > >>> I'm working on one (slowly) but it hasn't been pushed. GTK3 in the > >>> master branch looks normal for me. > >>> > >>> Does it look the same in OpenGL and Cairo? > >>> > >>> _Seth > >>> > >>> On 2019-09-04 15:36, Wayne Stambaugh wrote: > Was there a change recently to the selection highlighting in Eeschema? > I can barely see any difference between the selected and unselected > objects. If there has not been changes then something needs to be done > about the highlighting on gtk3 builds on linux. I thought I would > check > before I filed a bug report. > Cheers, > Wayne > ___ > Mailing list: https://launchpad.net/~kicad-developers > Post to : kicad-developers@lists.launchpad.net > Unsubscribe : https://launchpad.net/~kicad-developers > More help : https://help.launchpad.net/ListHelp > >>> > >>> ___ > >>> Mailing list: https://launchpad.net/~kicad-developers > >>> Post to : kicad-developers@lists.launchpad.net > >>> Unsubscribe : https://launchpad.net/~kicad-developers > >>> More help : https://help.launchpad.net/ListHelp > > ___ > Mailing list: https://launchpad.net/~kicad-developers > Post to : kicad-developers@lists.launchpad.net > Unsubscribe : https://launchpad.net/~kicad-developers > More help : https://help.launchpad.net/ListHelp -- /Jonatan http://kymatica.com ___ Mailing list: https://launchpad.net/~kicad-developers Post to : kicad-developers@lists.launchpad.net Unsubscribe : https://launchpad.net/~kicad-developers More help : https://help.launchpad.net/ListHelp
Re: [Kicad-developers] knowing last commit included in a nightly build?
Oh, I see. So then how can I check if a commit is included or not in a specific nightly 5.1 download? Cheers On Thu, Sep 5, 2019 at 10:47 AM Ian McInerney wrote: > > I believe that OSX adds several patches to the build process still, so the > last few commits are OSX specific and the one being picked for the version > string is probably one of this. Adam should be able to confirm this. > > -Ian > > On Thu, Sep 5, 2019 at 10:12 AM Jonatan Liljedahl wrote: >> >> Hmm, weird. >> I'm looking at nightlies of the 5.1 branch for OSX. For example the >> latest: >> https://kicad-downloads.s3.cern.ch/osx/testing/5.1/kicad-5.1-nightly-20190904-112458-3d75100e0-10_14.dmg >> >> Looking at the hash (3d75100e0), this yields no results: >> >> $ git log --oneline | grep 3d75100e0 >> >> On Thu, Sep 5, 2019 at 9:49 AM Andrew Lutsenko wrote: >> > >> > Hello, >> > What do you mean? The hex string in the file name is the prefix of the >> > commit hash, common way to identify commits in git. >> > E.g. latest nightly kicad-r14168.6688e8013-x86_64.exe is built from commit >> > 6688e80131e8283ecd2b45e0e2c04a4f3d6fcef2. >> > >> > Regards, >> > Andrew >> > >> > On Thu, Sep 5, 2019 at 12:25 AM Jonatan Liljedahl >> > wrote: >> >> >> >> Hi list, >> >> Is there any way to see from which commit a nightly was built? The hex >> >> strings in the download files doesn't seem to be git refs. >> >> >> >> -- >> >> /Jonatan >> >> http://kymatica.com >> >> >> >> ___ >> >> Mailing list: https://launchpad.net/~kicad-developers >> >> Post to : kicad-developers@lists.launchpad.net >> >> Unsubscribe : https://launchpad.net/~kicad-developers >> >> More help : https://help.launchpad.net/ListHelp >> >> >> >> -- >> /Jonatan >> http://kymatica.com >> >> ___ >> Mailing list: https://launchpad.net/~kicad-developers >> Post to : kicad-developers@lists.launchpad.net >> Unsubscribe : https://launchpad.net/~kicad-developers >> More help : https://help.launchpad.net/ListHelp -- /Jonatan http://kymatica.com ___ Mailing list: https://launchpad.net/~kicad-developers Post to : kicad-developers@lists.launchpad.net Unsubscribe : https://launchpad.net/~kicad-developers More help : https://help.launchpad.net/ListHelp
Re: [Kicad-developers] knowing last commit included in a nightly build?
Hmm, weird. I'm looking at nightlies of the 5.1 branch for OSX. For example the latest: https://kicad-downloads.s3.cern.ch/osx/testing/5.1/kicad-5.1-nightly-20190904-112458-3d75100e0-10_14.dmg Looking at the hash (3d75100e0), this yields no results: $ git log --oneline | grep 3d75100e0 On Thu, Sep 5, 2019 at 9:49 AM Andrew Lutsenko wrote: > > Hello, > What do you mean? The hex string in the file name is the prefix of the commit > hash, common way to identify commits in git. > E.g. latest nightly kicad-r14168.6688e8013-x86_64.exe is built from commit > 6688e80131e8283ecd2b45e0e2c04a4f3d6fcef2. > > Regards, > Andrew > > On Thu, Sep 5, 2019 at 12:25 AM Jonatan Liljedahl wrote: >> >> Hi list, >> Is there any way to see from which commit a nightly was built? The hex >> strings in the download files doesn't seem to be git refs. >> >> -- >> /Jonatan >> http://kymatica.com >> >> ___ >> Mailing list: https://launchpad.net/~kicad-developers >> Post to : kicad-developers@lists.launchpad.net >> Unsubscribe : https://launchpad.net/~kicad-developers >> More help : https://help.launchpad.net/ListHelp -- /Jonatan http://kymatica.com ___ Mailing list: https://launchpad.net/~kicad-developers Post to : kicad-developers@lists.launchpad.net Unsubscribe : https://launchpad.net/~kicad-developers More help : https://help.launchpad.net/ListHelp
[Kicad-developers] knowing last commit included in a nightly build?
Hi list, Is there any way to see from which commit a nightly was built? The hex strings in the download files doesn't seem to be git refs. -- /Jonatan http://kymatica.com ___ Mailing list: https://launchpad.net/~kicad-developers Post to : kicad-developers@lists.launchpad.net Unsubscribe : https://launchpad.net/~kicad-developers More help : https://help.launchpad.net/ListHelp
Re: [Kicad-developers] 5.1.5 RC?
On Fri, Aug 30, 2019 at 7:57 PM Nick Østergaard wrote: > > Att. Wayne and Adam > > According to Jonatans testing in > https://bugs.launchpad.net/kicad/+bug/1841752/comments/9 > > It looks like ngspice-30 do work for him on the latest nightlies... > > Next step, checking if the 5.1 branch testing builds are ok with ngspice-30. > > If that works it may be because of some kicad-mac-builder changes that > fixed something. (I didn't try to correlate this) Seems something was fixed! Because kicad-5.1-unified-20190830-121212-c1f9abc98-10_14.dmg works fine here, and it's running libngspice-30. Cheers -- /Jonatan http://kymatica.com ___ Mailing list: https://launchpad.net/~kicad-developers Post to : kicad-developers@lists.launchpad.net Unsubscribe : https://launchpad.net/~kicad-developers More help : https://help.launchpad.net/ListHelp
Re: [Kicad-developers] 5.1.5 RC?
Thanks. I've updated with a comment on my launchpad report: https://bugs.launchpad.net/kicad/+bug/1841752 Cheers On Thu, Aug 29, 2019 at 6:12 PM Wayne Stambaugh wrote: > > If ngspice-30 is completely broken on macos, we may have no choice but > to fall back to ngspice-26. The brings the bugs that have been fixed > since ngspice-26 back into play which is less than ideal but I don't see > a solution unless there is a new release of ngspice that resolves the > issues. We should try to work with Holger to try to resolve the issues > so we can provide a solution as soon as it's available. > > Cheers, > > Wayne > > On 8/28/19 10:26 PM, Adam Wolf wrote: > > Jonathan, I hear you. I rolled back to ngspice-26 for the 5.1.4 > > release, and I think it probably makes sense to roll back to > > ngspice-26 for 5.1.5 as well. > > > > I am not really comfortable with the state of this situation. We're > > running nightlies with 30, and release with 26, but how long until > > there's a bug caused by rolling back for stable releases that we never > > see since all the nightlies are on ngspice-30? > > > > Do I simply need to make another set of builds with ngspice-30 and a > > set with ngspice-26? Is this fine? Maybe? Am I overthinking this? > > > > Wayne, dev team, packagers, everyone--what do you think? I don't have > > any skin in the game here. > > > > Adam > > > > On Wed, Aug 28, 2019 at 1:21 PM Jonatan Liljedahl > > wrote: > >> > >> It would be great if you could roll back to ngspice-26 in 5.1.5, at > >> least for macOS, since it seems completely broken (can't simulate > >> op-amps, etc). > >> > >> Cheers > >> /Jonatan > >> > >> On Wed, Aug 28, 2019 at 1:58 PM Wayne Stambaugh > >> wrote: > >>> > >>> On 8/27/19 3:56 PM, Seth Hillbrand wrote: > >>>> Hi All- > >>>> > >>>> I know 5.1.4 is still in its infancy but I am hoping that we can plan > >>>> for a 5.1.5 during September (fixing at least 2 critical bugs). We've > >>>> been a bit rocky with the point updates, so I'd like to propose a > >>>> sequence: > >>>> > >>>> 1) Sometime around September 15: Tag 5.1.5-rc1 and announce string > >>>> freeze to give give translations a chance to update and tag 5.1.5 > >>>> 2) At this point, we don't push anything to 5.1 unless it is fix for a > >>>> critical bug. No other fixes until after 5.1.5. > >>>> 3) If there is a fix for a critical bug, we tag 5.1.5-rc2 after. > >>>> 4) 7 days after the last rc2 is tagged, we tag 5.1.5 > >>>> > >>>> What do folks think about this scheme? Maybe a way to avoid the 5.1.3 > >>>> (and 5.1.1 and 5.0.1 before it) situation? > >>>> > >>>> -Seth > >>> > >>> Absolutely. I was planning on this after the last debacle. Hopefully > >>> the 5.1.5 release will go a bit more smoothly. > >>> > >>> Cheers, > >>> > >>> Wayne > >>> > >>> ___ > >>> Mailing list: https://launchpad.net/~kicad-developers > >>> Post to : kicad-developers@lists.launchpad.net > >>> Unsubscribe : https://launchpad.net/~kicad-developers > >>> More help : https://help.launchpad.net/ListHelp > >> > >> > >> > >> -- > >> /Jonatan > >> http://kymatica.com > >> > >> ___ > >> Mailing list: https://launchpad.net/~kicad-developers > >> Post to : kicad-developers@lists.launchpad.net > >> Unsubscribe : https://launchpad.net/~kicad-developers > >> More help : https://help.launchpad.net/ListHelp > > > > ___ > > Mailing list: https://launchpad.net/~kicad-developers > > Post to : kicad-developers@lists.launchpad.net > > Unsubscribe : https://launchpad.net/~kicad-developers > > More help : https://help.launchpad.net/ListHelp > > > > ___ > Mailing list: https://launchpad.net/~kicad-developers > Post to : kicad-developers@lists.launchpad.net > Unsubscribe : https://launchpad.net/~kicad-developers > More help : https://help.launchpad.net/ListHelp -- /Jonatan http://kymatica.com ___ Mailing list: https://launchpad.net/~kicad-developers Post to : kicad-developers@lists.launchpad.net Unsubscribe : https://launchpad.net/~kicad-developers More help : https://help.launchpad.net/ListHelp
Re: [Kicad-developers] 5.1.5 RC?
Hi, On Thu, Aug 29, 2019 at 4:27 AM Adam Wolf wrote: > > Jonathan, I hear you. I rolled back to ngspice-26 for the 5.1.4 > release, and I think it probably makes sense to roll back to > ngspice-26 for 5.1.5 as well. Are you saying you meant to roll back in the official 5.1.4 release, or that you did it locally for yourself? I downloaded and installed the official 5.1.4 release for macOS, and it was bundled with ngspice-30, not 26. > I am not really comfortable with the state of this situation. We're > running nightlies with 30, and release with 26, but how long until > there's a bug caused by rolling back for stable releases that we never > see since all the nightlies are on ngspice-30? > > Do I simply need to make another set of builds with ngspice-30 and a > set with ngspice-26? Is this fine? Maybe? Am I overthinking this? Why include ngspice-30 if it's broken? I don't know if the bug is in ngspice or KiCad, but as long as it's not fixed there's no reason to use ngspice-30 since it's unusable for macOS users, IMHO. Cheers -- /Jonatan http://kymatica.com ___ Mailing list: https://launchpad.net/~kicad-developers Post to : kicad-developers@lists.launchpad.net Unsubscribe : https://launchpad.net/~kicad-developers More help : https://help.launchpad.net/ListHelp
Re: [Kicad-developers] MacOS ngspice version
Thanks! I reported the same issue earlier today[1] so I guess it could be marked as duplicate. [1] https://bugs.launchpad.net/kicad/+bug/1841752 Cheers On Wed, Aug 28, 2019 at 3:24 PM Seth Hillbrand wrote: > > Note: Changing thread to avoid hijacking. > > On 2019-08-28 08:05, Jonatan Liljedahl wrote: > > It would be great if you could roll back to ngspice-26 in 5.1.5, at > > least for macOS, since it seems completely broken (can't simulate > > op-amps, etc). > > > > Cheers > > /Jonatan > > Hi Jonatan- > > Please see bug tracker[1] for information on this. > > [1] https://bugs.launchpad.net/kicad/+bug/1836104 -- /Jonatan http://kymatica.com ___ Mailing list: https://launchpad.net/~kicad-developers Post to : kicad-developers@lists.launchpad.net Unsubscribe : https://launchpad.net/~kicad-developers More help : https://help.launchpad.net/ListHelp
Re: [Kicad-developers] 5.1.5 RC?
It would be great if you could roll back to ngspice-26 in 5.1.5, at least for macOS, since it seems completely broken (can't simulate op-amps, etc). Cheers /Jonatan On Wed, Aug 28, 2019 at 1:58 PM Wayne Stambaugh wrote: > > On 8/27/19 3:56 PM, Seth Hillbrand wrote: > > Hi All- > > > > I know 5.1.4 is still in its infancy but I am hoping that we can plan > > for a 5.1.5 during September (fixing at least 2 critical bugs). We've > > been a bit rocky with the point updates, so I'd like to propose a sequence: > > > > 1) Sometime around September 15: Tag 5.1.5-rc1 and announce string > > freeze to give give translations a chance to update and tag 5.1.5 > > 2) At this point, we don't push anything to 5.1 unless it is fix for a > > critical bug. No other fixes until after 5.1.5. > > 3) If there is a fix for a critical bug, we tag 5.1.5-rc2 after. > > 4) 7 days after the last rc2 is tagged, we tag 5.1.5 > > > > What do folks think about this scheme? Maybe a way to avoid the 5.1.3 > > (and 5.1.1 and 5.0.1 before it) situation? > > > > -Seth > > Absolutely. I was planning on this after the last debacle. Hopefully > the 5.1.5 release will go a bit more smoothly. > > Cheers, > > Wayne > > ___ > Mailing list: https://launchpad.net/~kicad-developers > Post to : kicad-developers@lists.launchpad.net > Unsubscribe : https://launchpad.net/~kicad-developers > More help : https://help.launchpad.net/ListHelp -- /Jonatan http://kymatica.com ___ Mailing list: https://launchpad.net/~kicad-developers Post to : kicad-developers@lists.launchpad.net Unsubscribe : https://launchpad.net/~kicad-developers More help : https://help.launchpad.net/ListHelp
Re: [Kicad-developers] [RFC] Set spice parameters in component library editor
I haven't tried your new revision, but I have simply added parameters
after the model name in the Spice_Model field, like "pot r=20k k=0.75"
for a subckt like this:
.SUBCKT pot 1 2 3 r=100k k=0.5
R1 2 3 {r*(1-k)}
R2 1 2 {r*(k)}
.ENDS
On Sat, Oct 15, 2016 at 8:15 PM, jp charras wrote:
> Hi All,
>
> In rev "2016-10-15 revision d1e4399" I added a button to add/edit spice
> parameters in component
> library editor (Field Properties dialog editor).
>
> Could you try this feature and see if it is OK or if some enhancement is
> needed (I am not a spice
> specialist)?
>
> Thanks.
>
> --
> Jean-Pierre CHARRAS
>
> ___
> Mailing list: https://launchpad.net/~kicad-developers
> Post to : kicad-developers@lists.launchpad.net
> Unsubscribe : https://launchpad.net/~kicad-developers
> More help : https://help.launchpad.net/ListHelp
--
/Jonatan
http://kymatica.com
___
Mailing list: https://launchpad.net/~kicad-developers
Post to : kicad-developers@lists.launchpad.net
Unsubscribe : https://launchpad.net/~kicad-developers
More help : https://help.launchpad.net/ListHelp
Re: [Kicad-developers] Exclude components from PCB
Hi Wayne, Thanks, good to know! For this issue, it's important to have separate flags for inclusion in PCB netlist and Spice netlist, but maybe that's already the plan? On Wed, Oct 12, 2016 at 4:00 PM, Wayne Stambaugh <stambau...@gmail.com> wrote: > Jonaton, > > Since you are new to the list, you have missed the discussion about the > new schematic file format. New flags are going to bet added for taggind > schematic symbols as not in netlist and/or not in BOM. Adding this as a > field will get lost in the new file format. The new file format is > currently under development so adding this now doesn't make a lot of sense. > > Cheers, > > Wayne > > On 10/12/2016 9:57 AM, Jonatan Liljedahl wrote: >> Hi list, I'm new here and quite new to Kicad. I'm actually a software >> developer but doing hobby DIY electronics as well. That's my little >> presentation, now on to the topic: >> >> See https://forum.kicad.info/t/simulator-vs-pcb/4219/8 >> >> What do you think? Perhaps it would make sense to add a >> "PCB_Netlist_Enabled" field (defaults to Y if no such field), similar >> to the now existing "Spice_Netlist_Enabled"? Spice-specific symbols >> like voltage sources can then have this field pre-filled in the symbol >> library. And users can add it themselves for other components specific >> to simulation, like dummy load resistors etc. >> >> Best regards >> > > ___ > Mailing list: https://launchpad.net/~kicad-developers > Post to : kicad-developers@lists.launchpad.net > Unsubscribe : https://launchpad.net/~kicad-developers > More help : https://help.launchpad.net/ListHelp -- /Jonatan http://kymatica.com ___ Mailing list: https://launchpad.net/~kicad-developers Post to : kicad-developers@lists.launchpad.net Unsubscribe : https://launchpad.net/~kicad-developers More help : https://help.launchpad.net/ListHelp
[Kicad-developers] Exclude components from PCB
Hi list, I'm new here and quite new to Kicad. I'm actually a software developer but doing hobby DIY electronics as well. That's my little presentation, now on to the topic: See https://forum.kicad.info/t/simulator-vs-pcb/4219/8 What do you think? Perhaps it would make sense to add a "PCB_Netlist_Enabled" field (defaults to Y if no such field), similar to the now existing "Spice_Netlist_Enabled"? Spice-specific symbols like voltage sources can then have this field pre-filled in the symbol library. And users can add it themselves for other components specific to simulation, like dummy load resistors etc. Best regards -- /Jonatan http://kymatica.com ___ Mailing list: https://launchpad.net/~kicad-developers Post to : kicad-developers@lists.launchpad.net Unsubscribe : https://launchpad.net/~kicad-developers More help : https://help.launchpad.net/ListHelp
