Re: [NEW] audio/zynaddsubfx
Ping On Mon, Mar 22, 2021 at 04:00:36PM +, Dimitri Karamazov wrote: > On Mon, Mar 22, 2021 at 11:34:27AM +, Stuart Henderson wrote: > > On 2021/03/21 16:19, Dimitri Karamazov wrote: > > > On Sun, Mar 21, 2021 at 11:46:48AM +, Stuart Henderson wrote: > > > > On 2021/03/21 11:30, Solene Rapenne wrote: > > > > > On Thu, 18 Mar 2021 17:36:55 + > > > > > Dimitri Karamazov : > > > > > > > > > > > Ping > > > > > > > > > > > > > > > > hello, in zyn-fusion I'd add audio/zynaddsubfx as a RUN_DEP because > > > > > it seems mandatory to use zyn-fusion. > > > > > > > > The problem is that the dependency is sort-of circular. zynaddsubfx > > > > requires zyn-fusion or another UI to be useful, and zyn-fusion requires > > > > zynaddsubfx to be useful. > > > > > > > > AUR deals with this by including zynaddsubfx in zyn-fusion (and having > > > > them conflict) but that doesn't seem ideal either. > > > > > > > > Thinking about what happens if other UIs are added though, having > > > > zynaddsubfx depend on zyn-fusion doesn't work because the user might > > > > want > > > > a different UI. > > > > > > > > Maybe the answer is to have zyn-fusion depend on zynaddsubfx and move > > > > the README to zynaddsubfx ("this package provides the backenx > > > > synthesizer, > > > > you must also install a user interface") - or even just including it in > > > > DESCR. > > > > > > > > The other way is to use meta-packages but that's messier .. > > > > > > There is a configure option -DGuiModule in zynaddsubfx which can be set to > > > zest(zyn-fusion), ntk or fltk. So making zyn-fusion depend on zynaddsubfx > > > will not work(restrict our options to zest). Also zyn-fusion has no real > > > dependency on zynaddsubfx it is not needed during build or runtime. It > > > is a hollow ui and is supposed to be so. > > > > > > zyn-fusion(zest) is a runtime dependency, while on the other hand fltk > > > and ntk are linked with zynaddsubfx. Only one of these frontends can be > > > installed at a time. > > > > > > Since files created vary with the value of GuiModule, I will have to use > > > MULTI_PACKAGES, and then set dependencies and CONFIGURE_ARGS for each > > > subpackage(like fltk,ntk and zyn-fusion). I see no other possibility to > > > allow Gui choices but to make zynaddsubfx depend on zyn-fusion, then add > > > other frontends with MULTI_PACKAGES. > > > > If I understand correctly this setting changes existing built files? > > For example bin/zynaddsubfx (or some other file) would be present in > > each build but would be different depending on the build option? > > If so then that would be FLAVOR not MULTI_PACKAGES. > > > > MULTI_PACKAGES is used when you have a core set of files that are > > identical for all build options, and additional/missing files depending > > on the build options. For example where you have a bunch of dlopen()'d > > .so modules which are present or not depending on the build option, > > but the main binary doesn't change. > > So I was mistaken. It needs a mix of both, for different options some > common files are differently linked (zynaddsubfx will either call zyn-fusion > or zynaddsubfx-ext-gui) and an additional bin/zynaddsubfx-ext-gui > file is present for fltk and ntk. Not possible? > > If it is not, there is no real need to fret at it, aside from the rabbit > hole of tabs the rest of the Gui's will lead you into there is nothing > interesting they offer. Only Zyn-Fusion has been good enough for them to > monetize it. > > > > > The result should not be same. Check the images I've attached. > > > `zyn-fusion` - image1.png > > > > > > This interface has no connection with the backend and hence is utterly > > > useless. > > > > How about the attached set of ports: > > > > meta/zyn-fusion > > audio/zynaddsubfx > > x11/mruby-zest > > > > (If you unpack in /usr/ports/mystuff you'll need to copy or symlink > > meta/Makefile.inc from the main ports tree into mystuff/meta/Makefile.inc) > > > > x11/mruby-zest is your x11/zyn-fusion port renamed, and with the binary > > moved into libexec (so it's out of the normal path, which should remove > > most confusion). I used just the date in the version number for this. > > > > meta/zyn-fusion provides a "zyn-fusion" meta-package which only depends > > on both zynaddsubfx+mruby-zest. It's kind-of pointless but gives something > > so that people can pkg_add zyn-fusion. > > > > audio/zynaddsubfx is patched slightly to search in libexec for zyn-fusion. > > (Also fixed a 0/nullptr mismatch that is a warning in clang 10 and I guess > > might possibly break in clang 11). > > > > I also changed from SUBST_CMD to sed (using /usr/X11R6 / /usr/local in the > > patch) so that if you run "make update-patches" after building it doesn't > > undo the subst. > > > > If other guis are added later then we would add FLAVORS in zynaddsubfx and > > tighten/adjust the dependency in meta/zyn-fusion. > > > > This seems a good balance of not changing too
Re: [NEW] audio/zynaddsubfx
On 2021/03/29 16:54, Dimitri Karamazov wrote: > Ping I need an OK before I can commit it. > On Mon, Mar 22, 2021 at 04:00:36PM +, Dimitri Karamazov wrote: > > On Mon, Mar 22, 2021 at 11:34:27AM +, Stuart Henderson wrote: > > > On 2021/03/21 16:19, Dimitri Karamazov wrote: > > > > On Sun, Mar 21, 2021 at 11:46:48AM +, Stuart Henderson wrote: > > > > > On 2021/03/21 11:30, Solene Rapenne wrote: > > > > > > On Thu, 18 Mar 2021 17:36:55 + > > > > > > Dimitri Karamazov : > > > > > > > > > > > > > Ping > > > > > > > > > > > > > > > > > > > hello, in zyn-fusion I'd add audio/zynaddsubfx as a RUN_DEP because > > > > > > it seems mandatory to use zyn-fusion. > > > > > > > > > > The problem is that the dependency is sort-of circular. zynaddsubfx > > > > > requires zyn-fusion or another UI to be useful, and zyn-fusion > > > > > requires > > > > > zynaddsubfx to be useful. > > > > > > > > > > AUR deals with this by including zynaddsubfx in zyn-fusion (and having > > > > > them conflict) but that doesn't seem ideal either. > > > > > > > > > > Thinking about what happens if other UIs are added though, having > > > > > zynaddsubfx depend on zyn-fusion doesn't work because the user might > > > > > want > > > > > a different UI. > > > > > > > > > > Maybe the answer is to have zyn-fusion depend on zynaddsubfx and move > > > > > the README to zynaddsubfx ("this package provides the backenx > > > > > synthesizer, > > > > > you must also install a user interface") - or even just including it > > > > > in > > > > > DESCR. > > > > > > > > > > The other way is to use meta-packages but that's messier .. > > > > > > > > There is a configure option -DGuiModule in zynaddsubfx which can be set > > > > to > > > > zest(zyn-fusion), ntk or fltk. So making zyn-fusion depend on > > > > zynaddsubfx > > > > will not work(restrict our options to zest). Also zyn-fusion has no real > > > > dependency on zynaddsubfx it is not needed during build or runtime. It > > > > is a hollow ui and is supposed to be so. > > > > > > > > zyn-fusion(zest) is a runtime dependency, while on the other hand fltk > > > > and ntk are linked with zynaddsubfx. Only one of these frontends can be > > > > installed at a time. > > > > > > > > Since files created vary with the value of GuiModule, I will have to use > > > > MULTI_PACKAGES, and then set dependencies and CONFIGURE_ARGS for each > > > > subpackage(like fltk,ntk and zyn-fusion). I see no other possibility to > > > > allow Gui choices but to make zynaddsubfx depend on zyn-fusion, then add > > > > other frontends with MULTI_PACKAGES. > > > > > > If I understand correctly this setting changes existing built files? > > > For example bin/zynaddsubfx (or some other file) would be present in > > > each build but would be different depending on the build option? > > > If so then that would be FLAVOR not MULTI_PACKAGES. > > > > > > MULTI_PACKAGES is used when you have a core set of files that are > > > identical for all build options, and additional/missing files depending > > > on the build options. For example where you have a bunch of dlopen()'d > > > .so modules which are present or not depending on the build option, > > > but the main binary doesn't change. > > > > So I was mistaken. It needs a mix of both, for different options some > > common files are differently linked (zynaddsubfx will either call zyn-fusion > > or zynaddsubfx-ext-gui) and an additional bin/zynaddsubfx-ext-gui > > file is present for fltk and ntk. Not possible? > > > > If it is not, there is no real need to fret at it, aside from the rabbit > > hole of tabs the rest of the Gui's will lead you into there is nothing > > interesting they offer. Only Zyn-Fusion has been good enough for them to > > monetize it. > > > > > > > The result should not be same. Check the images I've attached. > > > > `zyn-fusion` - image1.png > > > > > > > > This interface has no connection with the backend and hence is utterly > > > > useless. > > > > > > How about the attached set of ports: > > > > > > meta/zyn-fusion > > > audio/zynaddsubfx > > > x11/mruby-zest > > > > > > (If you unpack in /usr/ports/mystuff you'll need to copy or symlink > > > meta/Makefile.inc from the main ports tree into mystuff/meta/Makefile.inc) > > > > > > x11/mruby-zest is your x11/zyn-fusion port renamed, and with the binary > > > moved into libexec (so it's out of the normal path, which should remove > > > most confusion). I used just the date in the version number for this. > > > > > > meta/zyn-fusion provides a "zyn-fusion" meta-package which only depends > > > on both zynaddsubfx+mruby-zest. It's kind-of pointless but gives something > > > so that people can pkg_add zyn-fusion. > > > > > > audio/zynaddsubfx is patched slightly to search in libexec for zyn-fusion. > > > (Also fixed a 0/nullptr mismatch that is a warning in clang 10 and I guess > > > might possibly break in clang 11). > > > > > > I also changed from SUBST_CMD to sed (using
Re: [NEW] audio/zynaddsubfx
On Mon, Mar 22, 2021 at 11:34:27AM +, Stuart Henderson wrote: > On 2021/03/21 16:19, Dimitri Karamazov wrote: > > On Sun, Mar 21, 2021 at 11:46:48AM +, Stuart Henderson wrote: > > > On 2021/03/21 11:30, Solene Rapenne wrote: > > > > On Thu, 18 Mar 2021 17:36:55 + > > > > Dimitri Karamazov : > > > > > > > > > Ping > > > > > > > > > > > > > hello, in zyn-fusion I'd add audio/zynaddsubfx as a RUN_DEP because > > > > it seems mandatory to use zyn-fusion. > > > > > > The problem is that the dependency is sort-of circular. zynaddsubfx > > > requires zyn-fusion or another UI to be useful, and zyn-fusion requires > > > zynaddsubfx to be useful. > > > > > > AUR deals with this by including zynaddsubfx in zyn-fusion (and having > > > them conflict) but that doesn't seem ideal either. > > > > > > Thinking about what happens if other UIs are added though, having > > > zynaddsubfx depend on zyn-fusion doesn't work because the user might want > > > a different UI. > > > > > > Maybe the answer is to have zyn-fusion depend on zynaddsubfx and move > > > the README to zynaddsubfx ("this package provides the backenx synthesizer, > > > you must also install a user interface") - or even just including it in > > > DESCR. > > > > > > The other way is to use meta-packages but that's messier .. > > > > There is a configure option -DGuiModule in zynaddsubfx which can be set to > > zest(zyn-fusion), ntk or fltk. So making zyn-fusion depend on zynaddsubfx > > will not work(restrict our options to zest). Also zyn-fusion has no real > > dependency on zynaddsubfx it is not needed during build or runtime. It > > is a hollow ui and is supposed to be so. > > > > zyn-fusion(zest) is a runtime dependency, while on the other hand fltk > > and ntk are linked with zynaddsubfx. Only one of these frontends can be > > installed at a time. > > > > Since files created vary with the value of GuiModule, I will have to use > > MULTI_PACKAGES, and then set dependencies and CONFIGURE_ARGS for each > > subpackage(like fltk,ntk and zyn-fusion). I see no other possibility to > > allow Gui choices but to make zynaddsubfx depend on zyn-fusion, then add > > other frontends with MULTI_PACKAGES. > > If I understand correctly this setting changes existing built files? > For example bin/zynaddsubfx (or some other file) would be present in > each build but would be different depending on the build option? > If so then that would be FLAVOR not MULTI_PACKAGES. > > MULTI_PACKAGES is used when you have a core set of files that are > identical for all build options, and additional/missing files depending > on the build options. For example where you have a bunch of dlopen()'d > .so modules which are present or not depending on the build option, > but the main binary doesn't change. So I was mistaken. It needs a mix of both, for different options some common files are differently linked (zynaddsubfx will either call zyn-fusion or zynaddsubfx-ext-gui) and an additional bin/zynaddsubfx-ext-gui file is present for fltk and ntk. Not possible? If it is not, there is no real need to fret at it, aside from the rabbit hole of tabs the rest of the Gui's will lead you into there is nothing interesting they offer. Only Zyn-Fusion has been good enough for them to monetize it. > > > The result should not be same. Check the images I've attached. > > `zyn-fusion` - image1.png > > > > This interface has no connection with the backend and hence is utterly > > useless. > > How about the attached set of ports: > > meta/zyn-fusion > audio/zynaddsubfx > x11/mruby-zest > > (If you unpack in /usr/ports/mystuff you'll need to copy or symlink > meta/Makefile.inc from the main ports tree into mystuff/meta/Makefile.inc) > > x11/mruby-zest is your x11/zyn-fusion port renamed, and with the binary > moved into libexec (so it's out of the normal path, which should remove > most confusion). I used just the date in the version number for this. > > meta/zyn-fusion provides a "zyn-fusion" meta-package which only depends > on both zynaddsubfx+mruby-zest. It's kind-of pointless but gives something > so that people can pkg_add zyn-fusion. > > audio/zynaddsubfx is patched slightly to search in libexec for zyn-fusion. > (Also fixed a 0/nullptr mismatch that is a warning in clang 10 and I guess > might possibly break in clang 11). > > I also changed from SUBST_CMD to sed (using /usr/X11R6 / /usr/local in the > patch) so that if you run "make update-patches" after building it doesn't > undo the subst. > > If other guis are added later then we would add FLAVORS in zynaddsubfx and > tighten/adjust the dependency in meta/zyn-fusion. > > This seems a good balance of not changing too much, and reducing user > confusion (seems better to head off problems by making it so people don't > run into them if possible, rather than hope they will read pkg-readme, > some will but I know many won't and will just think it's broken when they > don't get the bank lists etc showing up..) What
Re: [NEW] audio/zynaddsubfx
On 2021/03/21 16:19, Dimitri Karamazov wrote: > On Sun, Mar 21, 2021 at 11:46:48AM +, Stuart Henderson wrote: > > On 2021/03/21 11:30, Solene Rapenne wrote: > > > On Thu, 18 Mar 2021 17:36:55 + > > > Dimitri Karamazov : > > > > > > > Ping > > > > > > > > > > hello, in zyn-fusion I'd add audio/zynaddsubfx as a RUN_DEP because > > > it seems mandatory to use zyn-fusion. > > > > The problem is that the dependency is sort-of circular. zynaddsubfx > > requires zyn-fusion or another UI to be useful, and zyn-fusion requires > > zynaddsubfx to be useful. > > > > AUR deals with this by including zynaddsubfx in zyn-fusion (and having > > them conflict) but that doesn't seem ideal either. > > > > Thinking about what happens if other UIs are added though, having > > zynaddsubfx depend on zyn-fusion doesn't work because the user might want > > a different UI. > > > > Maybe the answer is to have zyn-fusion depend on zynaddsubfx and move > > the README to zynaddsubfx ("this package provides the backenx synthesizer, > > you must also install a user interface") - or even just including it in > > DESCR. > > > > The other way is to use meta-packages but that's messier .. > > There is a configure option -DGuiModule in zynaddsubfx which can be set to > zest(zyn-fusion), ntk or fltk. So making zyn-fusion depend on zynaddsubfx > will not work(restrict our options to zest). Also zyn-fusion has no real > dependency on zynaddsubfx it is not needed during build or runtime. It > is a hollow ui and is supposed to be so. > > zyn-fusion(zest) is a runtime dependency, while on the other hand fltk > and ntk are linked with zynaddsubfx. Only one of these frontends can be > installed at a time. > > Since files created vary with the value of GuiModule, I will have to use > MULTI_PACKAGES, and then set dependencies and CONFIGURE_ARGS for each > subpackage(like fltk,ntk and zyn-fusion). I see no other possibility to > allow Gui choices but to make zynaddsubfx depend on zyn-fusion, then add > other frontends with MULTI_PACKAGES. If I understand correctly this setting changes existing built files? For example bin/zynaddsubfx (or some other file) would be present in each build but would be different depending on the build option? If so then that would be FLAVOR not MULTI_PACKAGES. MULTI_PACKAGES is used when you have a core set of files that are identical for all build options, and additional/missing files depending on the build options. For example where you have a bunch of dlopen()'d .so modules which are present or not depending on the build option, but the main binary doesn't change. > The result should not be same. Check the images I've attached. > `zyn-fusion` - image1.png > > This interface has no connection with the backend and hence is utterly > useless. How about the attached set of ports: meta/zyn-fusion audio/zynaddsubfx x11/mruby-zest (If you unpack in /usr/ports/mystuff you'll need to copy or symlink meta/Makefile.inc from the main ports tree into mystuff/meta/Makefile.inc) x11/mruby-zest is your x11/zyn-fusion port renamed, and with the binary moved into libexec (so it's out of the normal path, which should remove most confusion). I used just the date in the version number for this. meta/zyn-fusion provides a "zyn-fusion" meta-package which only depends on both zynaddsubfx+mruby-zest. It's kind-of pointless but gives something so that people can pkg_add zyn-fusion. audio/zynaddsubfx is patched slightly to search in libexec for zyn-fusion. (Also fixed a 0/nullptr mismatch that is a warning in clang 10 and I guess might possibly break in clang 11). I also changed from SUBST_CMD to sed (using /usr/X11R6 / /usr/local in the patch) so that if you run "make update-patches" after building it doesn't undo the subst. If other guis are added later then we would add FLAVORS in zynaddsubfx and tighten/adjust the dependency in meta/zyn-fusion. This seems a good balance of not changing too much, and reducing user confusion (seems better to head off problems by making it so people don't run into them if possible, rather than hope they will read pkg-readme, some will but I know many won't and will just think it's broken when they don't get the bank lists etc showing up..) What do you think? I'm OK with importing like this. zyn.tgz Description: application/tar-gz
Re: [NEW] audio/zynaddsubfx
On 2021/03/21 11:30, Solene Rapenne wrote: > On Thu, 18 Mar 2021 17:36:55 + > Dimitri Karamazov : > > > Ping > > > > hello, in zyn-fusion I'd add audio/zynaddsubfx as a RUN_DEP because > it seems mandatory to use zyn-fusion. The problem is that the dependency is sort-of circular. zynaddsubfx requires zyn-fusion or another UI to be useful, and zyn-fusion requires zynaddsubfx to be useful. AUR deals with this by including zynaddsubfx in zyn-fusion (and having them conflict) but that doesn't seem ideal either. Thinking about what happens if other UIs are added though, having zynaddsubfx depend on zyn-fusion doesn't work because the user might want a different UI. Maybe the answer is to have zyn-fusion depend on zynaddsubfx and move the README to zynaddsubfx ("this package provides the backenx synthesizer, you must also install a user interface") - or even just including it in DESCR. The other way is to use meta-packages but that's messier .. > When I run zynaddsubfx or zyn-fusion, I see no difference, is it > normal? Both open a X window with a keyboard at the bottom of the > window. If both binaries could be used for the same result, I would > remove the README file and just add a RUN_DEP. > > zynaddsubfx has some -O3 in the build log, this is something that must > be fixed before we can import it. > > zynaddsubfx HOMEPAGE can be swiched to https >
Re: [NEW] audio/zynaddsubfx
On Thu, 18 Mar 2021 17:36:55 + Dimitri Karamazov : > Ping > hello, in zyn-fusion I'd add audio/zynaddsubfx as a RUN_DEP because it seems mandatory to use zyn-fusion. When I run zynaddsubfx or zyn-fusion, I see no difference, is it normal? Both open a X window with a keyboard at the bottom of the window. If both binaries could be used for the same result, I would remove the README file and just add a RUN_DEP. zynaddsubfx has some -O3 in the build log, this is something that must be fixed before we can import it. zynaddsubfx HOMEPAGE can be swiched to https
Re: [NEW] audio/zynaddsubfx
Ping On Thu, Mar 11, 2021 at 02:19:01PM +, Dimitri Karamazov wrote: > On Thu, Mar 11, 2021 at 11:34:47AM +, Stuart Henderson wrote: > > On 2021/03/10 14:04, Dimitri Karamazov wrote: > > > On Tue, Mar 09, 2021 at 07:51:48PM +, Stuart Henderson wrote: > > > > zynaddsubfx picks up bash-completion if present at build time, > > > > > > > > $ diff pkg/PLIST.orig pkg/PLIST > > > > --- pkg/PLIST.orig Wed Feb 10 16:31:49 2021 > > > > +++ pkg/PLIST Tue Mar 9 19:39:30 2021 > > > > @@ -47,6 +47,9 @@ share/applications/zynaddsubfx-alsa.desktop > > > > share/applications/zynaddsubfx-jack-multi.desktop > > > > share/applications/zynaddsubfx-jack.desktop > > > > share/applications/zynaddsubfx-oss.desktop > > > > +share/bash_completion/ > > > > +share/bash_completion/completions/ > > > > +share/bash_completion/completions/zynaddsubfx > > > > share/doc/zynaddsubfx/ > > > > share/doc/zynaddsubfx/AUTHORS.txt > > > > share/doc/zynaddsubfx/COPYING > > > > > > > > can you force that either on or off please? either add a build dep on > > > > shells/bash-completion or set BASHCOMP_PKG_PATH. > > > > > > The install script uses bash which I've added to BUILD_DEPENDS. > > > > I don't see it using bash anywhere? > > > > It only looks for the pkg-config file from shells/bash-completion, > > installing bash won't have it create the completion files, only installing > > bash-completion (or passing the configure arg, but it looks like the > > directory may need to exist too, so it's less messy to just build dep > > on bash-completion). > > I've added bash-completion to BUILD_DEPENDS and this patch below to remove > the bash dependency. The script uses mkdir if the directory doesn't exist. > > $OpenBSD$ > > fix bash_completion install > > --- cmake/BashCompletion.cmake.orig Sat Feb 23 21:13:37 2019 > +++ cmake/BashCompletion.cmakeThu Mar 11 18:06:57 2021 > @@ -71,9 +71,9 @@ ELSE() > SET(BASHCOMP_SCRIPT > "${CMAKE_CURRENT_BINARY_DIR}/install_${SCRIPT_NAME}_completion.sh") > > FILE(WRITE ${BASHCOMP_SCRIPT} "\ > -#!${BASH}\n\ > +#!/bin/sh\n\ > set -e\n\ > -BASHCOMP_PKG_PATH=\"${BASHCOMP_USER_PATH}\"\n\ > +BASHCOMP_PKG_PATH=\"\$DESTDIR${BASHCOMP_USER_PATH}\"\n\ > if [ -w \"${BASHCOMP_PKG_PATH}\" ]; then\n\ >BASHCOMP_PKG_PATH=\"${BASHCOMP_PKG_PATH}\"\n\ > fi\n\ > > > > > Nice work on the zyn-fusion Makefile with the distfiles. It hardcodes > > > > gcc somewhere though (and I just noticed it forces -O3 as well), > > > > > > I couldn't recreate the gcc issue(after removing ports-gcc) so I've set > > > CC and LD in MAKE_FLAGS but why will the command fail if gcc is > > > available in the base? > > > > gcc is only available in base on alpha amd64 hppa sh sparc64, and it > > should not be used by anything in base/ports on amd64 (it's only there to > > make it easier for people to test building things with old-gcc but really > > it causes more problems than it solves now). > > I've tested after removing both ports-gcc and base-gcc(actually renaming the > binary for base-gcc) and I cannot encounter that issue, you can test it > again now. I can think it was an issue with CC CXX and LD not being set. > > > > > Looks like you maybe able to steal some bits from ports/lang/mruby. > > > > (Guessing it needs some specific version and can't just use the mruby > > > > port directly?) > > > > > > In the present case the mruby interpretor is a internal part of the > > > program > > > and thus requires the source code to be available during the build. > > > Developers of zyn-fusion didn't intend this program to be open-source so > > > everything is static-linked by default the refractoring needed here is > > > immense. > > > > ack. In which case try borrowing patches from the mruby port which > > already solved these problems. > > I've taken the necessary patches. > > > > > The dependencies seem the wrong way round, shouldn't it be zyn-fusion > > > > that depends on zynaddsubfx instead? (For discoverability I'd probably > > > > but zyn-fusion in audio/ instead of x11/). > > > > > > zyn-fusion unlike other frontends cannot interact with the backend. > > > You can check this yourself, executing zyn-fusion gives you an empty > > > interface and this is by design. So moving zyn-fusion into audio > > > wouldn't make sense since it is only an x11-interface(like ntk/fltk) > > > and more like a dependency for zynaddsubfx than the other way around. > > > It is by executing the zynaddsubfx binary that a user will actually > > > trigger > > > the usable interface. Ofcourse this goes contrary to how frontends are > > > actually developed but that is what we are dealing with. > > > So I think the categories should be kept as is. > > > > In that case I think there needs to be some user-facing information > > about this, maybe a short pkg-readme, or at least in DESCR, otherwise > > user experience for someone that has read about zyn-fusion and wants to > > try it with "pkg_add
Re: [NEW] audio/zynaddsubfx
On Thu, Mar 11, 2021 at 11:34:47AM +, Stuart Henderson wrote: > On 2021/03/10 14:04, Dimitri Karamazov wrote: > > On Tue, Mar 09, 2021 at 07:51:48PM +, Stuart Henderson wrote: > > > zynaddsubfx picks up bash-completion if present at build time, > > > > > > $ diff pkg/PLIST.orig pkg/PLIST > > > --- pkg/PLIST.origWed Feb 10 16:31:49 2021 > > > +++ pkg/PLIST Tue Mar 9 19:39:30 2021 > > > @@ -47,6 +47,9 @@ share/applications/zynaddsubfx-alsa.desktop > > > share/applications/zynaddsubfx-jack-multi.desktop > > > share/applications/zynaddsubfx-jack.desktop > > > share/applications/zynaddsubfx-oss.desktop > > > +share/bash_completion/ > > > +share/bash_completion/completions/ > > > +share/bash_completion/completions/zynaddsubfx > > > share/doc/zynaddsubfx/ > > > share/doc/zynaddsubfx/AUTHORS.txt > > > share/doc/zynaddsubfx/COPYING > > > > > > can you force that either on or off please? either add a build dep on > > > shells/bash-completion or set BASHCOMP_PKG_PATH. > > > > The install script uses bash which I've added to BUILD_DEPENDS. > > I don't see it using bash anywhere? > > It only looks for the pkg-config file from shells/bash-completion, > installing bash won't have it create the completion files, only installing > bash-completion (or passing the configure arg, but it looks like the > directory may need to exist too, so it's less messy to just build dep > on bash-completion). I've added bash-completion to BUILD_DEPENDS and this patch below to remove the bash dependency. The script uses mkdir if the directory doesn't exist. $OpenBSD$ fix bash_completion install --- cmake/BashCompletion.cmake.orig Sat Feb 23 21:13:37 2019 +++ cmake/BashCompletion.cmake Thu Mar 11 18:06:57 2021 @@ -71,9 +71,9 @@ ELSE() SET(BASHCOMP_SCRIPT "${CMAKE_CURRENT_BINARY_DIR}/install_${SCRIPT_NAME}_completion.sh") FILE(WRITE ${BASHCOMP_SCRIPT} "\ -#!${BASH}\n\ +#!/bin/sh\n\ set -e\n\ -BASHCOMP_PKG_PATH=\"${BASHCOMP_USER_PATH}\"\n\ +BASHCOMP_PKG_PATH=\"\$DESTDIR${BASHCOMP_USER_PATH}\"\n\ if [ -w \"${BASHCOMP_PKG_PATH}\" ]; then\n\ BASHCOMP_PKG_PATH=\"${BASHCOMP_PKG_PATH}\"\n\ fi\n\ > > > Nice work on the zyn-fusion Makefile with the distfiles. It hardcodes > > > gcc somewhere though (and I just noticed it forces -O3 as well), > > > > I couldn't recreate the gcc issue(after removing ports-gcc) so I've set > > CC and LD in MAKE_FLAGS but why will the command fail if gcc is > > available in the base? > > gcc is only available in base on alpha amd64 hppa sh sparc64, and it > should not be used by anything in base/ports on amd64 (it's only there to > make it easier for people to test building things with old-gcc but really > it causes more problems than it solves now). I've tested after removing both ports-gcc and base-gcc(actually renaming the binary for base-gcc) and I cannot encounter that issue, you can test it again now. I can think it was an issue with CC CXX and LD not being set. > > > Looks like you maybe able to steal some bits from ports/lang/mruby. > > > (Guessing it needs some specific version and can't just use the mruby > > > port directly?) > > > > In the present case the mruby interpretor is a internal part of the program > > and thus requires the source code to be available during the build. > > Developers of zyn-fusion didn't intend this program to be open-source so > > everything is static-linked by default the refractoring needed here is > > immense. > > ack. In which case try borrowing patches from the mruby port which > already solved these problems. I've taken the necessary patches. > > > The dependencies seem the wrong way round, shouldn't it be zyn-fusion > > > that depends on zynaddsubfx instead? (For discoverability I'd probably > > > but zyn-fusion in audio/ instead of x11/). > > > > zyn-fusion unlike other frontends cannot interact with the backend. > > You can check this yourself, executing zyn-fusion gives you an empty > > interface and this is by design. So moving zyn-fusion into audio > > wouldn't make sense since it is only an x11-interface(like ntk/fltk) > > and more like a dependency for zynaddsubfx than the other way around. > > It is by executing the zynaddsubfx binary that a user will actually trigger > > the usable interface. Ofcourse this goes contrary to how frontends are > > actually developed but that is what we are dealing with. > > So I think the categories should be kept as is. > > In that case I think there needs to be some user-facing information > about this, maybe a short pkg-readme, or at least in DESCR, otherwise > user experience for someone that has read about zyn-fusion and wants to > try it with "pkg_add zyn-fusion" won't be good. This is part of zyn-fusion port now. $OpenBSD: README.template,v 1.7 2019/05/09 17:45:05 ajacoutot Exp $ +--- | Running ${PKGSTEM} on OpenBSD
Re: [NEW] audio/zynaddsubfx
On 2021/03/10 14:04, Dimitri Karamazov wrote: > On Tue, Mar 09, 2021 at 07:51:48PM +, Stuart Henderson wrote: > > zynaddsubfx picks up bash-completion if present at build time, > > > > $ diff pkg/PLIST.orig pkg/PLIST > > --- pkg/PLIST.orig Wed Feb 10 16:31:49 2021 > > +++ pkg/PLIST Tue Mar 9 19:39:30 2021 > > @@ -47,6 +47,9 @@ share/applications/zynaddsubfx-alsa.desktop > > share/applications/zynaddsubfx-jack-multi.desktop > > share/applications/zynaddsubfx-jack.desktop > > share/applications/zynaddsubfx-oss.desktop > > +share/bash_completion/ > > +share/bash_completion/completions/ > > +share/bash_completion/completions/zynaddsubfx > > share/doc/zynaddsubfx/ > > share/doc/zynaddsubfx/AUTHORS.txt > > share/doc/zynaddsubfx/COPYING > > > > can you force that either on or off please? either add a build dep on > > shells/bash-completion or set BASHCOMP_PKG_PATH. > > The install script uses bash which I've added to BUILD_DEPENDS. I don't see it using bash anywhere? It only looks for the pkg-config file from shells/bash-completion, installing bash won't have it create the completion files, only installing bash-completion (or passing the configure arg, but it looks like the directory may need to exist too, so it's less messy to just build dep on bash-completion). > > Nice work on the zyn-fusion Makefile with the distfiles. It hardcodes > > gcc somewhere though (and I just noticed it forces -O3 as well), > > I couldn't recreate the gcc issue(after removing ports-gcc) so I've set > CC and LD in MAKE_FLAGS but why will the command fail if gcc is > available in the base? gcc is only available in base on alpha amd64 hppa sh sparc64, and it should not be used by anything in base/ports on amd64 (it's only there to make it easier for people to test building things with old-gcc but really it causes more problems than it solves now). > > Looks like you maybe able to steal some bits from ports/lang/mruby. > > (Guessing it needs some specific version and can't just use the mruby > > port directly?) > > In the present case the mruby interpretor is a internal part of the program > and thus requires the source code to be available during the build. > Developers of zyn-fusion didn't intend this program to be open-source so > everything is static-linked by default the refractoring needed here is > immense. ack. In which case try borrowing patches from the mruby port which already solved these problems. > > The dependencies seem the wrong way round, shouldn't it be zyn-fusion > > that depends on zynaddsubfx instead? (For discoverability I'd probably > > but zyn-fusion in audio/ instead of x11/). > > zyn-fusion unlike other frontends cannot interact with the backend. > You can check this yourself, executing zyn-fusion gives you an empty > interface and this is by design. So moving zyn-fusion into audio > wouldn't make sense since it is only an x11-interface(like ntk/fltk) > and more like a dependency for zynaddsubfx than the other way around. > It is by executing the zynaddsubfx binary that a user will actually trigger > the usable interface. Ofcourse this goes contrary to how frontends are > actually developed but that is what we are dealing with. > So I think the categories should be kept as is. In that case I think there needs to be some user-facing information about this, maybe a short pkg-readme, or at least in DESCR, otherwise user experience for someone that has read about zyn-fusion and wants to try it with "pkg_add zyn-fusion" won't be good. > I will be introducing MULTI_PACKAGES in zynaddsubfx for the ntk and fltk > interfaces later on so distinctions in the frontend available will be more > clear while installing. > > > > > On 2021/03/07 15:03, Dimitri Karamazov wrote: > > > Ping > > > > > > On Sat, Feb 27, 2021 at 05:32:23PM +, Dimitri Karamazov wrote: > > > > Ping > > > > > > > > On Wed, Feb 17, 2021 at 05:39:57PM +, Dimitri Karamazov wrote: > > > > > This here is a terrific synthesizer and can be put to good use > > > > > with some DAW with lv2 support. Can be used with lmms if > > > > > carla is ported or just wait until lv2 is supported(ongoing). > > > > > https://github.com/LMMS/lmms/issues/562 > > > > > > > > > > I've a muse(https://vstwarehouse.com/d/muse/) port with me which > > > > > can run this as plugin(both vst & lv2) but it brings along with > > > > > it 4-5 dependencies so might release it some other time. > > > > > > > > > > Information for inst:zynaddsubfx-3.0.5 > > > > > > > > > > Comment: > > > > > synth capable of making a countless number of instruments > > > > > > > > > > Description: > > > > > ZynAddSubFX is a opensource software synthesizer capable of making > > > > > a countless number of instruments, from some common heard from > > > > > expensive hardware to interesting sounds that you'll boost to an > > > > > amazing universe of sounds. > > > > > > *Polyphonic with support for legato and mono playing modes > > > *Three synthesizer
Re: [NEW] audio/zynaddsubfx
On Tue, Mar 09, 2021 at 07:51:48PM +, Stuart Henderson wrote: > zynaddsubfx picks up bash-completion if present at build time, > > $ diff pkg/PLIST.orig pkg/PLIST > --- pkg/PLIST.origWed Feb 10 16:31:49 2021 > +++ pkg/PLIST Tue Mar 9 19:39:30 2021 > @@ -47,6 +47,9 @@ share/applications/zynaddsubfx-alsa.desktop > share/applications/zynaddsubfx-jack-multi.desktop > share/applications/zynaddsubfx-jack.desktop > share/applications/zynaddsubfx-oss.desktop > +share/bash_completion/ > +share/bash_completion/completions/ > +share/bash_completion/completions/zynaddsubfx > share/doc/zynaddsubfx/ > share/doc/zynaddsubfx/AUTHORS.txt > share/doc/zynaddsubfx/COPYING > > can you force that either on or off please? either add a build dep on > shells/bash-completion or set BASHCOMP_PKG_PATH. The install script uses bash which I've added to BUILD_DEPENDS. > Nice work on the zyn-fusion Makefile with the distfiles. It hardcodes > gcc somewhere though (and I just noticed it forces -O3 as well), I couldn't recreate the gcc issue(after removing ports-gcc) so I've set CC and LD in MAKE_FLAGS but why will the command fail if gcc is available in the base? > [...] > configure: creating ./config.status > config.status: creating config.h > cd src/osc-bridge&& make lib > cc -O2 -pipe -I/usr/local/include -I/usr/X11R6/include -std=gnu99 -Wall > -Wextra -I . -O3 -g -fPIC -c src/bridge.c src/cache.c src/parse-schema.c > src/schema.c rtosc/rtosc.c > src/parse-schema.c:11:9: warning: 'putchar' macro redefined > [-Wmacro-redefined] > #define putchar(x) (void)(x) > ^ > /usr/include/stdio.h:435:9: note: previous definition is here > #define putchar(x) putc(x, stdout) > ^ > 1 warning generated. > ar rcs libosc-bridge.a bridge.o cache.o parse-schema.o schema.o rtosc.o > cd mruby && MRUBY_CONFIG=../build_config.rb /usr/local/bin/rake27 > Environment is: > > CPP src/array.c -> build/host/src/array.i > sh: gcc: not found > rake aborted! > Command failed with status (127): [gcc -E -P -std=gnu99 -O2 -pipe > -I/usr/lo...] > [...] > > Looks like you maybe able to steal some bits from ports/lang/mruby. > (Guessing it needs some specific version and can't just use the mruby > port directly?) In the present case the mruby interpretor is a internal part of the program and thus requires the source code to be available during the build. Developers of zyn-fusion didn't intend this program to be open-source so everything is static-linked by default the refractoring needed here is immense. > The dependencies seem the wrong way round, shouldn't it be zyn-fusion > that depends on zynaddsubfx instead? (For discoverability I'd probably > but zyn-fusion in audio/ instead of x11/). zyn-fusion unlike other frontends cannot interact with the backend. You can check this yourself, executing zyn-fusion gives you an empty interface and this is by design. So moving zyn-fusion into audio wouldn't make sense since it is only an x11-interface(like ntk/fltk) and more like a dependency for zynaddsubfx than the other way around. It is by executing the zynaddsubfx binary that a user will actually trigger the usable interface. Ofcourse this goes contrary to how frontends are actually developed but that is what we are dealing with. So I think the categories should be kept as is. I will be introducing MULTI_PACKAGES in zynaddsubfx for the ntk and fltk interfaces later on so distinctions in the frontend available will be more clear while installing. > > On 2021/03/07 15:03, Dimitri Karamazov wrote: > > Ping > > > > On Sat, Feb 27, 2021 at 05:32:23PM +, Dimitri Karamazov wrote: > > > Ping > > > > > > On Wed, Feb 17, 2021 at 05:39:57PM +, Dimitri Karamazov wrote: > > > > This here is a terrific synthesizer and can be put to good use > > > > with some DAW with lv2 support. Can be used with lmms if > > > > carla is ported or just wait until lv2 is supported(ongoing). > > > > https://github.com/LMMS/lmms/issues/562 > > > > > > > > I've a muse(https://vstwarehouse.com/d/muse/) port with me which > > > > can run this as plugin(both vst & lv2) but it brings along with > > > > it 4-5 dependencies so might release it some other time. > > > > > > > > Information for inst:zynaddsubfx-3.0.5 > > > > > > > > Comment: > > > > synth capable of making a countless number of instruments > > > > > > > > Description: > > > > ZynAddSubFX is a opensource software synthesizer capable of making > > > > a countless number of instruments, from some common heard from > > > > expensive hardware to interesting sounds that you'll boost to an > > > > amazing universe of sounds. > > > > *Polyphonic with support for legato and mono playing modes > > *Three synthesizer engines: > > -Additive Synthesis for classic synth sounds composed of a variety of > > voices with powerful modulation. > > This exposes modulators ranging from LFOs and envelopes to oscillator > > modulators for FM, PM, and AM. > > -Subtractive
Re: [NEW] audio/zynaddsubfx
zynaddsubfx picks up bash-completion if present at build time, $ diff pkg/PLIST.orig pkg/PLIST --- pkg/PLIST.orig Wed Feb 10 16:31:49 2021 +++ pkg/PLIST Tue Mar 9 19:39:30 2021 @@ -47,6 +47,9 @@ share/applications/zynaddsubfx-alsa.desktop share/applications/zynaddsubfx-jack-multi.desktop share/applications/zynaddsubfx-jack.desktop share/applications/zynaddsubfx-oss.desktop +share/bash_completion/ +share/bash_completion/completions/ +share/bash_completion/completions/zynaddsubfx share/doc/zynaddsubfx/ share/doc/zynaddsubfx/AUTHORS.txt share/doc/zynaddsubfx/COPYING can you force that either on or off please? either add a build dep on shells/bash-completion or set BASHCOMP_PKG_PATH. Nice work on the zyn-fusion Makefile with the distfiles. It hardcodes gcc somewhere though (and I just noticed it forces -O3 as well), [...] configure: creating ./config.status config.status: creating config.h cd src/osc-bridge&& make lib cc -O2 -pipe -I/usr/local/include -I/usr/X11R6/include -std=gnu99 -Wall -Wextra -I . -O3 -g -fPIC -c src/bridge.c src/cache.c src/parse-schema.c src/schema.c rtosc/rtosc.c src/parse-schema.c:11:9: warning: 'putchar' macro redefined [-Wmacro-redefined] #define putchar(x) (void)(x) ^ /usr/include/stdio.h:435:9: note: previous definition is here #define putchar(x) putc(x, stdout) ^ 1 warning generated. ar rcs libosc-bridge.a bridge.o cache.o parse-schema.o schema.o rtosc.o cd mruby && MRUBY_CONFIG=../build_config.rb /usr/local/bin/rake27 Environment is: CPP src/array.c -> build/host/src/array.i sh: gcc: not found rake aborted! Command failed with status (127): [gcc -E -P -std=gnu99 -O2 -pipe -I/usr/lo...] [...] Looks like you maybe able to steal some bits from ports/lang/mruby. (Guessing it needs some specific version and can't just use the mruby port directly?) The dependencies seem the wrong way round, shouldn't it be zyn-fusion that depends on zynaddsubfx instead? (For discoverability I'd probably but zyn-fusion in audio/ instead of x11/). On 2021/03/07 15:03, Dimitri Karamazov wrote: > Ping > > On Sat, Feb 27, 2021 at 05:32:23PM +, Dimitri Karamazov wrote: > > Ping > > > > On Wed, Feb 17, 2021 at 05:39:57PM +, Dimitri Karamazov wrote: > > > This here is a terrific synthesizer and can be put to good use > > > with some DAW with lv2 support. Can be used with lmms if > > > carla is ported or just wait until lv2 is supported(ongoing). > > > https://github.com/LMMS/lmms/issues/562 > > > > > > I've a muse(https://vstwarehouse.com/d/muse/) port with me which > > > can run this as plugin(both vst & lv2) but it brings along with > > > it 4-5 dependencies so might release it some other time. > > > > > > Information for inst:zynaddsubfx-3.0.5 > > > > > > Comment: > > > synth capable of making a countless number of instruments > > > > > > Description: > > > ZynAddSubFX is a opensource software synthesizer capable of making > > > a countless number of instruments, from some common heard from > > > expensive hardware to interesting sounds that you'll boost to an > > > amazing universe of sounds. > > *Polyphonic with support for legato and mono playing modes > *Three synthesizer engines: > -Additive Synthesis for classic synth sounds composed of a variety of voices > with powerful modulation. > This exposes modulators ranging from LFOs and envelopes to oscillator > modulators for FM, PM, and AM. > -Subtractive Synthesis for creating variable bandwidth harmonics from > filtered white noise > -PAD synthesis for creating beautiful pads and other instruments > *Powerful waveform generator with up to 128 sine/non-sine harmonics > *A variety of filters including analogue modeled filters, formant filters, > and state variable filters. > *Envelopes can have ADSR (or ASR, etc..) modes or can be free modes (with any > shape) > *Effects for Reverb, Echo, Chorus/Flange, Phasing, Wave-shaping, Equalizing, > Dynamic Filtering with > flexible signal routing > *Instruments can be organized in kits, which allows you to make drum kits or > layered instruments; this > makes possible to use more than one instrument for a single part. It is > possible to choose what items > from the kit should be processed by the Part's effects. > *Randomness settings to create subtle differences in each sound to help > create that familiar analogue warmth. > *Microtonal capabilities with any scale, up to 128 notes per octave, and key > mapping > *Extensive MIDI/Audio driver support including JACK, ALSA, OSS, and PortAudio > *A built-in Virtual Keyboard which can be used even if you don't have a > physical midi keyboard > *The graphical user interface can be completely disabled if the user need so > *Session Management Support via LASH/NSM > *Plugin Support via DSSI/LV2/VST > *Over 1100 high quality instruments included > > > > > > > Maintainer: Dimitri Karamazov > > > > > > WWW: http://zynaddsubfx.sourceforge.net > > > > > > Port below
Re: [NEW] audio/zynaddsubfx
Ping On Sat, Feb 27, 2021 at 05:32:23PM +, Dimitri Karamazov wrote: > Ping > > On Wed, Feb 17, 2021 at 05:39:57PM +, Dimitri Karamazov wrote: > > This here is a terrific synthesizer and can be put to good use > > with some DAW with lv2 support. Can be used with lmms if > > carla is ported or just wait until lv2 is supported(ongoing). > > https://github.com/LMMS/lmms/issues/562 > > > > I've a muse(https://vstwarehouse.com/d/muse/) port with me which > > can run this as plugin(both vst & lv2) but it brings along with > > it 4-5 dependencies so might release it some other time. > > > > Information for inst:zynaddsubfx-3.0.5 > > > > Comment: > > synth capable of making a countless number of instruments > > > > Description: > > ZynAddSubFX is a opensource software synthesizer capable of making > > a countless number of instruments, from some common heard from > > expensive hardware to interesting sounds that you'll boost to an > > amazing universe of sounds. *Polyphonic with support for legato and mono playing modes *Three synthesizer engines: -Additive Synthesis for classic synth sounds composed of a variety of voices with powerful modulation. This exposes modulators ranging from LFOs and envelopes to oscillator modulators for FM, PM, and AM. -Subtractive Synthesis for creating variable bandwidth harmonics from filtered white noise -PAD synthesis for creating beautiful pads and other instruments *Powerful waveform generator with up to 128 sine/non-sine harmonics *A variety of filters including analogue modeled filters, formant filters, and state variable filters. *Envelopes can have ADSR (or ASR, etc..) modes or can be free modes (with any shape) *Effects for Reverb, Echo, Chorus/Flange, Phasing, Wave-shaping, Equalizing, Dynamic Filtering with flexible signal routing *Instruments can be organized in kits, which allows you to make drum kits or layered instruments; this makes possible to use more than one instrument for a single part. It is possible to choose what items from the kit should be processed by the Part's effects. *Randomness settings to create subtle differences in each sound to help create that familiar analogue warmth. *Microtonal capabilities with any scale, up to 128 notes per octave, and key mapping *Extensive MIDI/Audio driver support including JACK, ALSA, OSS, and PortAudio *A built-in Virtual Keyboard which can be used even if you don't have a physical midi keyboard *The graphical user interface can be completely disabled if the user need so *Session Management Support via LASH/NSM *Plugin Support via DSSI/LV2/VST *Over 1100 high quality instruments included > > > > Maintainer: Dimitri Karamazov > > > > WWW: http://zynaddsubfx.sourceforge.net > > > > Port below adds a nice graphical interface to the synthesizer. > > Looks like this:https://zynaddsubfx.sourceforge.io/zyn-fusion.html > > Although there other interfaces available like fltk and ntk, both > > are quite cumbersome to use, and very ugly in comparison. > > > > Information for inst:zyn-fusion-3.0.5pre > > > > Comment: > > single window & tabbed version of the ZynAddSubFX interface > > > > Required by: > > zynaddsubfx-3.0.5 > > > > Description: > > Zyn was plagued with a difficult to use interface and unstable, unofficial > > plugin versions. Fusion blows those issues away with a complete redesign and > > rewrite. Powered by a new custom window toolkit: zest, Zyn is now ready for > > any platform. With total separation of the UI from the synthesis engine. > > > > Zyn-Fusion is enabling powerful features such as the new, > > officially-supported > > VST plugin version, remote control of a headless CPU with the UI on > > touchscreen, even multiple UI windows controlling a single sound engine. By > > fusing 28 windows of fine grained control into one, Zyn is now controlled > > with > > a single window design of tabbed views for a consistent, navigable > > interface. > > You think you've heard the sounds before, but you've never had the precision > > control of Fusion. > > > > Maintainer: Dimitri Karamazov > > > > WWW: https://github.com/mruby-zest/mruby-zest-build > > > > Both Build & Run tested on amd64 zynaddsubfx.tar.gz Description: Binary data zyn-fusion.tar.gz Description: Binary data
Re: [NEW] audio/zynaddsubfx
Ping On Wed, Feb 17, 2021 at 05:39:57PM +, Dimitri Karamazov wrote: > This here is a terrific synthesizer and can be put to good use > with some DAW with lv2 support. Can be used with lmms if > carla is ported or just wait until lv2 is supported(ongoing). > https://github.com/LMMS/lmms/issues/562 > > I've a muse(https://vstwarehouse.com/d/muse/) port with me which > can run this as plugin(both vst & lv2) but it brings along with > it 4-5 dependencies so might release it some other time. > > Information for inst:zynaddsubfx-3.0.5 > > Comment: > synth capable of making a countless number of instruments > > Description: > ZynAddSubFX is a opensource software synthesizer capable of making > a countless number of instruments, from some common heard from > expensive hardware to interesting sounds that you'll boost to an > amazing universe of sounds. > > Maintainer: Dimitri Karamazov > > WWW: http://zynaddsubfx.sourceforge.net > > Port below adds a nice graphical interface to the synthesizer. > Looks like this:https://zynaddsubfx.sourceforge.io/zyn-fusion.html > Although there other interfaces available like fltk and ntk, both > are quite cumbersome to use, and very ugly in comparison. > > Information for inst:zyn-fusion-3.0.5pre > > Comment: > single window & tabbed version of the ZynAddSubFX interface > > Required by: > zynaddsubfx-3.0.5 > > Description: > Zyn was plagued with a difficult to use interface and unstable, unofficial > plugin versions. Fusion blows those issues away with a complete redesign and > rewrite. Powered by a new custom window toolkit: zest, Zyn is now ready for > any platform. With total separation of the UI from the synthesis engine. > > Zyn-Fusion is enabling powerful features such as the new, officially-supported > VST plugin version, remote control of a headless CPU with the UI on > touchscreen, even multiple UI windows controlling a single sound engine. By > fusing 28 windows of fine grained control into one, Zyn is now controlled with > a single window design of tabbed views for a consistent, navigable interface. > You think you've heard the sounds before, but you've never had the precision > control of Fusion. > > Maintainer: Dimitri Karamazov > > WWW: https://github.com/mruby-zest/mruby-zest-build > > Both Build & Run tested on amd64 zynaddsubfx.tar.gz Description: Binary data zyn-fusion.tar.gz Description: Binary data
[NEW] audio/zynaddsubfx
This here is a terrific synthesizer and can be put to good use with some DAW with lv2 support. Can be used with lmms if carla is ported or just wait until lv2 is supported(ongoing). https://github.com/LMMS/lmms/issues/562 I've a muse(https://vstwarehouse.com/d/muse/) port with me which can run this as plugin(both vst & lv2) but it brings along with it 4-5 dependencies so might release it some other time. Information for inst:zynaddsubfx-3.0.5 Comment: synth capable of making a countless number of instruments Description: ZynAddSubFX is a opensource software synthesizer capable of making a countless number of instruments, from some common heard from expensive hardware to interesting sounds that you'll boost to an amazing universe of sounds. Maintainer: Dimitri Karamazov WWW: http://zynaddsubfx.sourceforge.net Port below adds a nice graphical interface to the synthesizer. Looks like this:https://zynaddsubfx.sourceforge.io/zyn-fusion.html Although there other interfaces available like fltk and ntk, both are quite cumbersome to use, and very ugly in comparison. Information for inst:zyn-fusion-3.0.5pre Comment: single window & tabbed version of the ZynAddSubFX interface Required by: zynaddsubfx-3.0.5 Description: Zyn was plagued with a difficult to use interface and unstable, unofficial plugin versions. Fusion blows those issues away with a complete redesign and rewrite. Powered by a new custom window toolkit: zest, Zyn is now ready for any platform. With total separation of the UI from the synthesis engine. Zyn-Fusion is enabling powerful features such as the new, officially-supported VST plugin version, remote control of a headless CPU with the UI on touchscreen, even multiple UI windows controlling a single sound engine. By fusing 28 windows of fine grained control into one, Zyn is now controlled with a single window design of tabbed views for a consistent, navigable interface. You think you've heard the sounds before, but you've never had the precision control of Fusion. Maintainer: Dimitri Karamazov WWW: https://github.com/mruby-zest/mruby-zest-build Both Build & Run tested on amd64 zynaddsubfx.tar.gz Description: Binary data zyn-fusion.tar.gz Description: Binary data
Re: NEW: audio/zynaddsubfx
Ping On Tue, August 4, 2020 06:59, Dimitri Karamazov wrote: > On Sun, July 26, 2020 03:32, Dimitri Karamazov wrote: > >> On Sat, July 25, 2020 20:44, Stuart Henderson wrote: >> >> >>> On 2020/07/25 18:30, Dimitri Karamazov wrote: >>> >>> >>> Any comments? >>> >>> >>> just use normal patches (plus SUBST_CMD if necessary e.g. where you are >>> adding X11BASE) rather than seds please. >>> >>> patch the input cmake file not the generated build.ninja file (don't assume >>> that it's using ninja to build, >>> sometimes there are reasons to not use this) >>> >>> >> Hi, and thanks for the comments! >> >> >> >> I've made all the appropriate changes, let me know if there are any more. >> Also corrected the patch file names. >> I hope this gets in, it's an important synthesizer, probably the most >> exhaustive imho. >> >> I'll try Zyn-fusion sometime later, it has a more in place Gui. >> Compare that to the default which is really a rabbit hole of tabs, >> and difficult to deal with on some wm's. >> >> OK? >> >> >> >> regards, Dimitri > Nobody interested in taking this in? There have been a couple of requests > for this port in the mailing list. >
Re: NEW: audio/zynaddsubfx
On Sun, July 26, 2020 03:32, Dimitri Karamazov wrote: > On Sat, July 25, 2020 20:44, Stuart Henderson wrote: > >> On 2020/07/25 18:30, Dimitri Karamazov wrote: >> >> >>> Any comments? >>> >>> >> >> >> just use normal patches (plus SUBST_CMD if necessary e.g. where you are >> adding X11BASE) rather than seds please. >> >> patch the input cmake file not the generated build.ninja file (don't assume >> that it's using ninja to build, >> sometimes there are reasons to not use this) >> >> > Hi, and thanks for the comments! > > > I've made all the appropriate changes, let me know if there are any more. > Also corrected the patch file names. > I hope this gets in, it's an important synthesizer, probably the most > exhaustive imho. > > I'll try Zyn-fusion sometime later, it has a more in place Gui. > Compare that to the default which is really a rabbit hole of tabs, > and difficult to deal with on some wm's. > > OK? > > > regards, Dimitri Nobody interested in taking this in? There have been a couple of requests for this port in the mailing list.
Re: NEW: audio/zynaddsubfx
On Sat, July 25, 2020 20:44, Stuart Henderson wrote: > On 2020/07/25 18:30, Dimitri Karamazov wrote: > >> Any comments? >> > > > just use normal patches (plus SUBST_CMD if necessary e.g. where you are > adding X11BASE) rather than seds please. > > patch the input cmake file not the generated build.ninja file (don't assume > that it's using ninja to build, sometimes > there are reasons to not use this) > > Hi, and thanks for the comments! I've made all the appropriate changes, let me know if there are any more. Also corrected the patch file names. I hope this gets in, it's an important synthesizer, probably the most exhaustive imho. I'll try Zyn-fusion sometime later, it has a more in place Gui. Compare that to the default which is really a rabbit hole of tabs, and difficult to deal with on some wm's. OK? regards, Dimitri zynaddsubfx.tar.gz Description: application/gzip
NEW: audio/zynaddsubfx
ZynAddSubFX ZynAddSubFX is a fully featured musical software synthesizer for Linux, MacOS, BSD, and Windows. ZynAddSubFX exposes a wide array of synthesis parameters to make it flexible tool for sound design and a fun experience for playing instruments. Features * Polyphonic with support for legato and mono playing modes * Three synthesizer engines: ** Additive Synthesis for classic synth sounds composed of a variety of voices with powerful modulation. This exposes modulators ranging from LFOs and envelopes to oscillator modulators for FM, PM, and AM. ** Subtractive Synthesis for creating variable bandwidth harmonics from filtered white noise ** PAD synthesis for creating beautiful pads and other instruments * Powerful waveform generator with up to 128 sine/non-sine harmonics * A variety of filters including analogue modeled filters, formant filters, and state variable filters. * Envelopes can have ADSR (or ASR, etc..) modes or can be free modes (with any shape) * Effects for Reverb, Echo, Chorus/Flange, Phasing, Wave-shaping, Equalizing, Dynamic Filtering with flexible signal routing * Instruments can be organized in kits, which allows you to make drum kits or layered instruments; this makes possible to use more than one instrument for a single part. It is possible to choose what items from the kit should be processed by the Part's effects. * Randomness settings to create subtle differences in each sound to help create that familiar analogue warmth. * Microtonal capabilities with any scale, up to 128 notes per octave, and key mapping * Extensive MIDI/Audio driver support including JACK, ALSA, OSS, and PortAudio * Session Management Support via LASH/NSM * Plugin Support via DSSI/LV2/VST * Over 1100 high quality instruments included Build and run-tested on amd64,this port depends on the x11/ntk, which I've already sent. The port works fine, I tested pretty much every instrument, but there are more knobs than days in a year, so not so exhaustive in that regard. This also serves as a pretty good test for x11/ntk. Any comments? regards, Dimitri zynaddsubfx.tar.gz Description: application/gzip
Re: NEW: audio/zynaddsubfx
On 2020/07/25 18:30, Dimitri Karamazov wrote: > Any comments? V= 3.0.5 DISTNAME= zynaddsubfx-${V} V isn't used elsewhere -> DISTNAME= zynaddsubfx-3.0.5 MASTER_SITES= ${MASTER_SITE_SOURCEFORGE:=zynaddsubfx/3.0.5/} version not needed here -> MASTER_SITES= ${MASTER_SITE_SOURCEFORGE:=zynaddsubfx/} BASH_COMPLETION=${DESTDIR}${PREFIX}/share/bash_completion/completions CONFIGURE_ARGS= -DPaEnable=ON \ -DGuiModule=ntk \ -DBASHCOMP_PKG_PATH=${BASH_COMPLETION} \ no need for the separate variable -> CONFIGURE_ARGS= -DPaEnable=ON \ -DGuiModule=ntk \ -DBASHCOMP_PKG_PATH=${DESTDIR}${PREFIX}/share/bash_completion/completions post-patch: sed -ie "s@-w@! -d@" ${WRKSRC}/cmake/BashCompletion.cmake post-configure: sed -ie "s@-lrt@@" ${WRKBUILD}/build.ninja sed -ie "s@\(-L${LOCALBASE}/lib\)@\1 -L${X11BASE}/lib@" \ ${WRKBUILD}/build.ninja just use normal patches (plus SUBST_CMD if necessary e.g. where you are adding X11BASE) rather than seds please. patch the input cmake file not the generated build.ninja file (don't assume that it's using ninja to build, sometimes there are reasons to not use this)
new: audio/zynaddsubfx
this is an update to the ports of this I've sent before as there was actually a new release a few months ago. mostly bugfixes and some new instruments in the new release. -- jake...@sdf.lonestar.org SDF Public Access UNIX System - http://sdf.lonestar.org zynaddsubfx-2.4.0-port.tgz Description: application/tar-gz
Re: new: audio/zynaddsubfx
On Tue, Mar 10, 2009 at 05:15:06AM +, Jacob Meuser wrote: On Fri, Jan 16, 2009 at 10:36:52PM -0800, patrick keshishian wrote: I don't really know what to do with it really, but seems to work just fine on this fairly old 800MHz iBook G4. here's a slightly updated version. defined PORTHOME to quite the systrace warnings about fluidtrying to write to HOME; fixed a couple warnings. tested on i386, amd64 and powerpc so far. would be nice to know it at least builds on other archs. with the port this time -- jake...@sdf.lonestar.org SDF Public Access UNIX System - http://sdf.lonestar.org zynaddsubfx-2.2.1-port.tgz Description: application/tar-gz
Re: new: audio/zynaddsubfx
On Fri, Jan 16, 2009 at 10:36:52PM -0800, patrick keshishian wrote: I don't really know what to do with it really, but seems to work just fine on this fairly old 800MHz iBook G4. here's a slightly updated version. defined PORTHOME to quite the systrace warnings about fluidtrying to write to HOME; fixed a couple warnings. tested on i386, amd64 and powerpc so far. would be nice to know it at least builds on other archs. -- jake...@sdf.lonestar.org SDF Public Access UNIX System - http://sdf.lonestar.org
Re: new: audio/zynaddsubfx
On Fri, Jan 16, 2009 at 11:39 PM, Jacob Meuser jake...@sdf.lonestar.org wrote: On Fri, Jan 16, 2009 at 10:20:50PM -0800, patrick keshishian wrote: Quick fix is to surround '#include sndio.h' with #ifdef __cplusplus extern C { #endif This was silly of me. The #ifdef __cplusplus bits are superfluous, it obviously is a C++ header file. But shouldn't this be done in sndio.h header file itself with __{BEGIN,END}_DECLS pairs? well, maybe. I think it should. I'd like to know why it's needed on ppc and not i386 or amd64. That seems odd that it compiled and linked alright on those. seems like the problem is in ppc, or at least there's an (unnecessary?) inconsistency between archs. The compiler does the name mangling. I imagine both arch are using the same version of gcc 3.3.5, granted the spec file would be different. What is the output of nm on your SNDIOoutput.o when compiled without the 'extern C' bit? --patrick
Re: new: audio/zynaddsubfx
On Sat, Jan 17, 2009 at 07:42:48AM +, Jacob Meuser wrote: On Fri, Jan 16, 2009 at 10:36:52PM -0800, patrick keshishian wrote: I don't really know what to do with it you're supposed to have fun making choose an adjective sounds :) heh, maybe like this: http://jakemsr.trancell.org/mary3.mp3 ... or maybe not :P ok, that's not just from zynaddsubfx. I used rubberband to create a harmonic, aucat -n to mix them together and audacity to do the fade in, a little cutting, and export to mp3. I need some sheet music ... -- jake...@sdf.lonestar.org SDF Public Access UNIX System - http://sdf.lonestar.org
Re: new: audio/zynaddsubfx
On Sat, Jan 17, 2009 at 02:47:11AM -0800, patrick keshishian wrote: On Fri, Jan 16, 2009 at 11:39 PM, Jacob Meuser jake...@sdf.lonestar.org wrote: On Fri, Jan 16, 2009 at 10:20:50PM -0800, patrick keshishian wrote: Quick fix is to surround '#include sndio.h' with #ifdef __cplusplus extern C { #endif This was silly of me. The #ifdef __cplusplus bits are superfluous, it obviously is a C++ header file. But shouldn't this be done in sndio.h header file itself with __{BEGIN,END}_DECLS pairs? well, maybe. I think it should. I'd like to know why it's needed on ppc and not i386 or amd64. That seems odd that it compiled and linked alright on those. seems like the problem is in ppc, or at least there's an (unnecessary?) inconsistency between archs. The compiler does the name mangling. I imagine both arch are using the same version of gcc 3.3.5, granted the spec file would be different. What is the output of nm on your SNDIOoutput.o when compiled without the 'extern C' bit? U SAMPLE_RATE F SNDIOoutput.C U SOUND_BUFFER_SIZE 03b0 T _ZN11SNDIOoutput8SNDIOoutEPfS0_ 01d8 T _ZN11SNDIOoutputC1Ev T _ZN11SNDIOoutputC2Ev 0518 T _ZN11SNDIOoutputD1Ev 04d0 T _ZN11SNDIOoutputD2Ev U __gxx_personality_v0 U __sF U config U fprintf U free U malloc U memset U sio_close U sio_getpar U sio_initpar U sio_open U sio_setpar U sio_start U sio_write -- jake...@sdf.lonestar.org SDF Public Access UNIX System - http://sdf.lonestar.org
Re: new: audio/zynaddsubfx
On Sat, Jan 17, 2009 at 3:00 AM, Jacob Meuser jake...@sdf.lonestar.org wrote: On Sat, Jan 17, 2009 at 07:42:48AM +, Jacob Meuser wrote: On Fri, Jan 16, 2009 at 10:36:52PM -0800, patrick keshishian wrote: I don't really know what to do with it you're supposed to have fun making choose an adjective sounds :) heh, maybe like this: http://jakemsr.trancell.org/mary3.mp3 ... or maybe not :P ok, that's not just from zynaddsubfx. I used rubberband to create a harmonic, aucat -n to mix them together and audacity to do the fade in, a little cutting, and export to mp3. It is definitely neat. no questions about it. The rubberband port did interest me. If i understood its purpose correctly, it would be a great tool for transcribing music or learning songs by slowing them down w/o changing the pitch. I plan on playing with it. --patrick
Re: new: audio/zynaddsubfx
On Sat, Jan 17, 2009 at 10:32:43AM -0800, patrick keshishian wrote: On Sat, Jan 17, 2009 at 3:00 AM, Jacob Meuser jake...@sdf.lonestar.org wrote: On Sat, Jan 17, 2009 at 07:42:48AM +, Jacob Meuser wrote: On Fri, Jan 16, 2009 at 10:36:52PM -0800, patrick keshishian wrote: I don't really know what to do with it you're supposed to have fun making choose an adjective sounds :) heh, maybe like this: http://jakemsr.trancell.org/mary3.mp3 ... or maybe not :P ok, that's not just from zynaddsubfx. I used rubberband to create a harmonic, aucat -n to mix them together and audacity to do the fade in, a little cutting, and export to mp3. It is definitely neat. no questions about it. The rubberband port did interest me. If i understood its purpose correctly, it would be a great tool for transcribing music or learning songs by slowing them down w/o changing the pitch. I plan on playing with it. yes, it could be used for that purpose. soundstretch from the soundtouch port can do that too. however, soundtouch does not have VAMP or LADSPA plugins, and the algorithms differ a bit. -- jake...@sdf.lonestar.org SDF Public Access UNIX System - http://sdf.lonestar.org
Re: new: audio/zynaddsubfx
On Wed, Jan 14, 2009 at 04:53:51AM +, Jacob Meuser wrote: On Mon, Jan 12, 2009 at 11:46:56AM +, Jacob Meuser wrote: DESCR: ZynAddSubFX is a pure software synthesizer that uses no samples. It includes many beautiful sounding predefined instruments. These instruments can be modified in numerous ways, and/or completely new instruments can be built. this is quite different than audio/fluidsynth which is a SoundFont based (and thus sample based) synthesizer. the instruments in ZynAddSubFX are also much more tweakable. it really does make some nice and/or interesting sounds. it has a virtual keyboard that can be played with a mouse, it can be played with a computer keyboard, or it can use MIDI input. includes only a sndio backend. the jack backend is somewhat broken, so I did not include it. here's a new version with a couple changes. this should work better on big endian machines. also should work better when there is no midi input device attached. can someone check that this builds on !x86? thanks -- jake...@sdf.lonestar.org SDF Public Access UNIX System - http://sdf.lonestar.org
Re: new: audio/zynaddsubfx
I don't really know what to do with it really, but seems to work just fine on this fairly old 800MHz iBook G4.
Re: new: audio/zynaddsubfx
On Fri, Jan 16, 2009 at 10:20:50PM -0800, patrick keshishian wrote: On Fri, Jan 16, 2009 at 9:27 PM, Jacob Meuser jake...@sdf.lonestar.org wrote: On Sat, Jan 17, 2009 at 04:44:35AM +, Jacob Meuser wrote: http://www.cs.cmu.edu/~poladian/arch/project/report_ning_poladian.htm which led me to this, which says to use lrintf: http://mega-nerd.com/FPcast/ The was first on my google results, I skimmed but didn't read through it. so, here's a version thst uses lrintf instead of assembly. I didn't find any other uses of assembly. fails with undefined references to sio_*(): c++ -O2 -pipe -DOS_LINUX -DOSSMIDIIN -DFFTW_VERSION_3 -DASM_F2I_NO `fltk-config --cflags` -DSNDIOAUDIOOUT -DOS_LINUX -DOSSMIDIIN -DFFTW_VERSION_3 -DASM_F2I_NO `fltk-config --cflags` -DSNDIOAUDIOOUT -c -o main.o main.C gmake[1]: Leaving directory `/usr/ports/mystuff/audio/zynaddsubfx/w-zynaddsubfx-2.2.1/ZynAddSubFX-2.2.1/src' rm -f zynaddsubfx zynaddsubfx.exe rm -f Make.deps c++ -o zynaddsubfx */*.o *.o -lm -lmxml -lz `fltk-config --ldflags` -lfftw3 -pthread -lsndio Misc/Bank.o(.text+0xed4): In function `Bank::swapslot(unsigned int, unsigned int)': : warning: strcpy() is almost always misused, please use strlcpy() Misc/Bank.o(.text+0x23c): In function `Bank::~Bank()': : warning: sprintf() is often misused, please use snprintf() Misc/Bank.o(.text+0x7d4): In function `Bank::savetoslot(unsigned int, Part*)': : warning: strcat() is almost always misused, please use strlcat() Output/SNDIOoutput.o(.text+0x68): In function `SNDIOoutput::SNDIOoutput()': : undefined reference to `sio_open(char*, unsigned int, int)' Output/SNDIOoutput.o(.text+0xb8): In function `SNDIOoutput::SNDIOoutput()': : undefined reference to `sio_initpar(sio_par*)' Output/SNDIOoutput.o(.text+0x110): In function `SNDIOoutput::SNDIOoutput()': : undefined reference to `sio_setpar(sio_hdl*, sio_par*)' Output/SNDIOoutput.o(.text+0x138): In function `SNDIOoutput::SNDIOoutput()': : undefined reference to `sio_getpar(sio_hdl*, sio_par*)' Output/SNDIOoutput.o(.text+0x1d0): In function `SNDIOoutput::SNDIOoutput()': : undefined reference to `sio_start(sio_hdl*)' Output/SNDIOoutput.o(.text+0x280): In function `SNDIOoutput::SNDIOoutput()': : undefined reference to `sio_open(char*, unsigned int, int)' Output/SNDIOoutput.o(.text+0x2d0): In function `SNDIOoutput::SNDIOoutput()': : undefined reference to `sio_initpar(sio_par*)' Output/SNDIOoutput.o(.text+0x328): In function `SNDIOoutput::SNDIOoutput()': : undefined reference to `sio_setpar(sio_hdl*, sio_par*)' Output/SNDIOoutput.o(.text+0x350): In function `SNDIOoutput::SNDIOoutput()': : undefined reference to `sio_getpar(sio_hdl*, sio_par*)' Output/SNDIOoutput.o(.text+0x3e8): In function `SNDIOoutput::SNDIOoutput()': : undefined reference to `sio_start(sio_hdl*)' Output/SNDIOoutput.o(.text+0x500): In function `SNDIOoutput::SNDIOout(float*, float*)': : undefined reference to `sio_write(sio_hdl*, void*, unsigned long)' Output/SNDIOoutput.o(.text+0x5e0): In function `SNDIOoutput::~SNDIOoutput()': : undefined reference to `sio_close(sio_hdl*)' Output/SNDIOoutput.o(.text+0x640): In function `SNDIOoutput::~SNDIOoutput()': : undefined reference to `sio_close(sio_hdl*)' gmake: *** [all] Error 1 *** Error code 2 Stop in /usr/ports/mystuff/audio/zynaddsubfx (line 2172 of /usr/ports/infrastructure/mk/bsd.port.mk). I running -current built with cvs update on Jan 10th: OpenBSD 4.4-current (GENERIC) #1: Sat Jan 10 17:26:43 PST 2009 Quick fix is to surround '#include sndio.h' with #ifdef __cplusplus extern C { #endif But shouldn't this be done in sndio.h header file itself with __{BEGIN,END}_DECLS pairs? well, maybe. I'd like to know why it's needed on ppc and not i386 or amd64. seems like the problem is in ppc, or at least there's an (unnecessary?) inconsistency between archs. Also I see quite a lot of systrace denial for fluid trying to write to home directory: systrace: deny user: sidster, prog: /usr/local/bin/fluid, pid: 23911(0)[22438], policy: /usr/bin/env, filters: 195, syscall: native-fswrite(136), filename: /zynaddsubfx-2.2.1_writes_to_HOME systrace: deny user: sidster, prog: /usr/local/bin/fluid, pid: 23911(0)[22438], policy: /usr/bin/env, filters: 195, syscall: native-fswrite(136), filename: /zynaddsubfx-2.2.1_writes_to_HOME/.fltk fluid is part of fltk-1.1.7p0 package. yeah, I see that too. also with other fluid using ports. thanks for your feedback :) -- jake...@sdf.lonestar.org SDF Public Access UNIX System - http://sdf.lonestar.org
Re: new: audio/zynaddsubfx
On Fri, Jan 16, 2009 at 10:36:52PM -0800, patrick keshishian wrote: I don't really know what to do with it you're supposed to have fun making choose an adjective sounds :) -- jake...@sdf.lonestar.org SDF Public Access UNIX System - http://sdf.lonestar.org
Re: new: audio/zynaddsubfx
2009/1/14 Jacob Meuser jake...@sdf.lonestar.org: here's a new version with a couple changes. this should work better on big endian machines. also should work better when there is no midi input device attached. Builds and works fine here on i386 (without a midi input device though.) It did not crash, even though I may have turned every single button I could :)
Re: new: audio/zynaddsubfx
On Mon, Jan 12, 2009 at 11:46:56AM +, Jacob Meuser wrote: DESCR: ZynAddSubFX is a pure software synthesizer that uses no samples. It includes many beautiful sounding predefined instruments. These instruments can be modified in numerous ways, and/or completely new instruments can be built. this is quite different than audio/fluidsynth which is a SoundFont based (and thus sample based) synthesizer. the instruments in ZynAddSubFX are also much more tweakable. it really does make some nice and/or interesting sounds. it has a virtual keyboard that can be played with a mouse, it can be played with a computer keyboard, or it can use MIDI input. includes only a sndio backend. the jack backend is somewhat broken, so I did not include it. here's a new version with a couple changes. this should work better on big endian machines. also should work better when there is no midi input device attached. -- jake...@sdf.lonestar.org SDF Public Access UNIX System - http://sdf.lonestar.org zynaddsubfx-2.2.1-port.tgz Description: application/tar-gz
new: audio/zynaddsubfx
DESCR: ZynAddSubFX is a pure software synthesizer that uses no samples. It includes many beautiful sounding predefined instruments. These instruments can be modified in numerous ways, and/or completely new instruments can be built. this is quite different than audio/fluidsynth which is a SoundFont based (and thus sample based) synthesizer. the instruments in ZynAddSubFX are also much more tweakable. it really does make some nice and/or interesting sounds. it has a virtual keyboard that can be played with a mouse, it can be played with a computer keyboard, or it can use MIDI input. includes only a sndio backend. the jack backend is somewhat broken, so I did not include it. -- jake...@sdf.lonestar.org SDF Public Access UNIX System - http://sdf.lonestar.org zynaddsubfx-2.2.1-port.tgz Description: application/tar-gz
Re: NEW: audio/zynaddsubfx
On Sun, Sep 30, 2007 at 01:59:09AM +, Jacob Meuser wrote: DESCR: ZynAddSubFX is an opensource software synthesizer capable of making a countless number of instruments, from some commonly heard from expensive hardware to interesting sounds that you'll boost to an amazing universe of sounds. this has a midi interface, but unfortunately, it uses /dev/sequencer which does not support read()ing. actually it's using /dev/sequencer as a raw midi device, so the following makes midi input work: --- Input/OSSMidiIn.C.okMon Oct 1 22:07:04 2007 +++ Input/OSSMidiIn.C Mon Oct 1 22:33:52 2007 @@ -56,11 +56,11 @@ unsigned char OSSMidiIn::readbyte(){ unsigned char tmp[4]; -read(midi_handle,tmp[0],1); -while (tmp[0]!=SEQ_MIDIPUTC){ - read(midi_handle,tmp[0],4); -}; -return(tmp[1]); +if (read(midi_handle,tmp[0],1) 0) { + inputok = 0; + return 0xff; +} +return tmp[0]; }; unsigned char OSSMidiIn::getmidibyte(){ --- Misc/Config.C.okMon Oct 1 22:38:05 2007 +++ Misc/Config.C Mon Oct 1 22:38:25 2007 @@ -49,7 +49,7 @@ snprintf(cfg.LinuxOSSWaveOutDev,MAX_STRING_SIZE,/dev/dsp); #endif cfg.LinuxOSSSeqInDev=new char[MAX_STRING_SIZE]; -snprintf(cfg.LinuxOSSSeqInDev,MAX_STRING_SIZE,/dev/sequencer); +snprintf(cfg.LinuxOSSSeqInDev,MAX_STRING_SIZE,/dev/rmidi1); cfg.DumpFile=new char[MAX_STRING_SIZE]; snprintf(cfg.DumpFile,MAX_STRING_SIZE,zynaddsubfx_dump.txt); -- Alexandre
NEW: audio/zynaddsubfx
DESCR: ZynAddSubFX is an opensource software synthesizer capable of making a countless number of instruments, from some commonly heard from expensive hardware to interesting sounds that you'll boost to an amazing universe of sounds. this has a midi interface, but unfortunately, it uses /dev/sequencer which does not support read()ing. the virtual keyboard is still fun to play with though :) -- [EMAIL PROTECTED] SDF Public Access UNIX System - http://sdf.lonestar.org zynaddsubfx-2.2.1-port.tgz Description: application/tar-gz