Re: Come speak at the Guix Social meetup
On 2024-06-04 10:19, Steve George wrote: Hi, We're looking for speakers - of any level - to give talks at the Guix online meetup. We're a small, friendly group at all levels of experience. Talks can be about anything Guix associated, for example: - Experiences using or learning Guix - Particular commands or capabilities - deploying VM's, building a desktop etc - Using the declarative configuration approach - Packaging or creating system services - Something mad, interesting or curious! Whatever it is, we'd love to hear it! We've had recent talks by Ludo [0] and David Wilson (System crafters) [1]. We have an upcoming talk on June 18th by David Thompson [2] who will be talking about Guile, Guix and his work on Guile Hoot. And, on July 11th Vagrant Cascadian will be talking about reproducibility and Guix. After that ... Just to repeat - talks can be any length (10-20-45 mins), and at any level :-) Thanks, Steve / Futurile [0] https://www.youtube.com/watch?v=339dRDf4c6E [1] https://www.youtube.com/watch?v=oINCgds-8kA [2] https://dthompson.us/ - register here https://libreplanet.org/wiki/Group:Guix/PatchReviewSessions2024 I could talk about the bootstrapping in riscv, maybe it's too technical or idk... If you want, I'm open to do it! Thanks for organizing these talks, they are very interesting. Cheers, Ekaitz
Re: Issues cross-compiling arm with QMK
Hi Kristoffer, On 2024-03-09 18:50, Kristoffer Ström wrote: When trying to compile keyboard firmware using the qmk toolkit, i run into this error: ./lib/pico-sdk/src/rp2_common/hardware_base/include/hardware/address_mapped.h:94:19: error: cast from pointer to integer of different size [-Werror=pointer-to-int-cast] uint32_t rc = (uintptr_t)addr; Which seems to me to indicate that the cross-toolchain is trying to produce 64-bit binaries. i based my build shell manifest off of https://github.com/mrvdb/qmk_atreus which is an avr-based board and replacing (make-avr-toolchain) with (make-arm-none-eabi-nano-toolchain-x), i have tried all available versions. Is there a way to specify a toolchain for cross-compiling into 32bit arm binaries? You should use `(make-arm-none-eabi-toolchain-*whatever*)` but I have bad news. There are *many* issues with this. I reported this a while ago: https://lists.gnu.org/archive/html/help-guix/2023-09/msg00087.html And many other people did more questions after that: https://lists.gnu.org/archive/html/guix-devel/2023-12/msg00180.html But I didn't manage to fix this yet. In fact, I found similar issues in other architectures, mostly because all of them share the same problem: https://issues.guix.gnu.org/69394 Maybe making a Guix package which has your firmware as an output might work, at this very moment, the problem is our modified GCCs are unable to find the proper libraries and they explode. I can't fix this, and it's affecting me severely and I feel a little bit helpless about it... I don't know what we should do. I might organize in the near future and make my own toolchain like the one shown in the second link to fix my personal issue and start helping everyone that reports this... But at the moment there's not much I can do. Maybe someone else can help better. I'll try to keep an eye in this subject, but if you fix it, please let me know. Best, Ekaitz
Help needed: Weird bootstrapping hangs
Hi all,
Really weird question here. Beware.
I'm experiencing weird Guix hangs in the bootstrapping packages.
They happen randomly but frequently in:
- Mes' configure script
- Mes' linking of libtcc
- Some other packages' configure scripts (I had it in Binutils but the
package is not published yet)
- When creating a Makefile or a config.log in several packages. Even
after an error appears, they don't stop, they just hang.
The symptoms are the following:
- The stdout of the configure script is just frozen.
- The CPU usage drops, that specific process goes to exactly 20% of CPU
as seen in `htop`.
- If I connect to the running program using strace I see the following
forever, appearing very fast in the screen:
pselect6(4, [3], NULL, NULL, {tv_sec=0, tv_nsec=16}, NULL) = 0 (Timeout)
wait4(14338, 0x7fffc5a0, WNOHANG, NULL) = 0
pselect6(4, [3], NULL, NULL, {tv_sec=0, tv_nsec=16}, NULL) = 0 (Timeout)
wait4(14338, 0x7fffc5a0, WNOHANG, NULL) = 0
pselect6(4, [3], NULL, NULL, {tv_sec=0, tv_nsec=16}, NULL) = 0 (Timeout)
wait4(14338, 0x7fffc5a0, WNOHANG, NULL) = 0
pselect6(4, [3], NULL, NULL, {tv_sec=0, tv_nsec=16}, NULL) = 0 (Timeout)
wait4(14338, 0x7fffc5a0, WNOHANG, NULL) = 0
I've replicated this in a separate RISC-V machine so it's not Qemu.
Does anyone have an idea of what can this be?
We have two suspects:
- Guix
- Gash
I have no idea of how to dig on this further. Yesterday I had the same
hang in both machines' Mes configure scripts but then I added -x to the
`set -e` on the top of the file and both worked, and they keep working
since. Then one of them exploded later... I went to bed and I woke up
with a "Process unresponsive after 3600 seconds" error and I run it
again and it worked and managed to finish.
Some help would be much appreciated.
Thanks,
Ekaitz
Re: Cross-compilation arm-none-eabi
Hi On 2024-01-15 11:20, Kristoffer Ström wrote: > > Hello! > > I'm trying to use `qmk` to build firmware for my keyboard, but i'm > running into cross complilation issues that i have no idea where to > start on. > > When running in a shell with `qmk` i get > > arm-none-eabi-gcc: command not found > > The only maybe related package i've been able to find is > `raspi-arm-chainloader`, however this fails to build (it seems also in > CI) > > https://data.guix.gnu.org/revision/45a6af1c18b254e2a5bb26c6bab465c3d4476a75/package/raspi-arm-chainloader/0.1?locale=en_US.UTF-8 > > <- link to failing build on data.guix.gnu.org > > Am i missing something? How would i make the cross-compile gcc available > in a shell environment? > > Thx, Kristoffer > Give a try to this in a `manifest.scm` file. (use-modules (gnu packages embedded)) (packages->manifest (list (make-arm-none-eabi-toolchain-7-2018-q2-update))) Since some recent changes in guix we have to make-... the package for these cross-compilers. I don't like it, but it has some reason behind. Cheers, PS: There are other problems with cross compilation I hope you don't get after this, if you do, please keep me posted and I'll try to help.
Re: Question regarding qmk firmware
ElenQ Technology --- Original Message --- On Tuesday, October 10th, 2023 at 14:20, Peter Polidoro wrote: > On Sunday, October 8th, 2023 at 10:28, Ekaitz Zarraga > eka...@elenq.tech wrote: > > > There are other ways to do it, but I find this the easiest. > > You can even make a package for your firmware and let guix > > compile it for you (and maybe flash it too?). > > > Cheers, > > Ekaitz > > > I am also interested in using Guix packages for firmware projects. > > Do you have any links to example firmware projects that package, > compile, and flash firmware to embedded boards? > > This would be a great blog post if you are up for writing it. I > really enjoyed Mitchell Schmeisser's blog post about building > toolchains with Guix. It would be nice to see examples of firmware > that do not use the Zephyr real-time operating system as well, if > anyone is doing that. > > I am tempted to use PlatformIO with Guix as an intermediate > solution until I figure out how to do everything in Guix. I would > like to have a pure Guix solution for Arduino-based projects and > Raspberry Pi RP2040 as well. > > Thanks! Hi Peter, I suggest you to take a look to: https://lists.gnu.org/archive/html/guix-devel/2023-10/msg00014.html Because I'm the ^1 in that email and you can see the problems I've had with ARM (it was a problem with RP4020). The current toolchain is not able to find the stdlib.h and this is a huge inconvenience. I didn't manage to solve it but the email I'm linking there is really promising (I didn't have the time to test, but the author told me he was able to solve my problem using his toolchain). We should fix this in Guix, but right now it just doesn't work. Cheers, Ekaitz
Re: Question regarding qmk firmware
--- Original Message --- On Saturday, October 7th, 2023 at 21:23, Fredrik Salomonsson wrote: > Hi, > > Today I was tweaking a keymap for one of my qmk based keyboards but some > of the packages I used when building the firmware has been removed. My > commad was as follows: > > `sh guix shell avr-toolchain dfu-programmer qmk -- qmk flash -kb > ergodox_infinity -km plattfot -bl dfu-split-left` > > But `avr-toolchain` is gone. When I tried to just drop it and see if it > worked I get > > `Ψ Compiling keymap with make --jobs=1 > ergodox_infinity:plattfot:dfu-split-left QMK Firmware 0.14.19 Making > ergodox_infinity with keymap plattfot and target dfu-split-left > /gnu/store/rib9g2ig1xf3kclyl076w28parmncg4k-bash-minimal-5.1.16/bin/sh: line > 1: arm-none-eabi-gcc: command not found` > > It seems `arm-none-eabi-toolchain` is also removed. Looking at the > commit history for guix it looks like they got replaced by > [proceduers][0] instead. > > [0] > https://git.savannah.gnu.org/cgit/guix.git/commit/?id=35c1df5bd6317b1cd038c1a4aca1c7e4a52d4d93 > > My question is how do I get access to the arm-none-eabi-toolchain from > the commandline with guix shell? > > Thanks Hi Fredrik, Now what you need to do is call to the `make-avr-toolchain` function from `(gnu packages avr)` to get that. Maybe the easiest way is to create a `manifest.scm` where you do something like this (i'm doing it by memory): ``` scheme (use-modules (gnu packages avr) (gnu packages flashing-tools) (gnu packages firmware)) (packages->manifest (list (make-avr-toolchain) dfu-programmer qmk)) ``` Then use `guix shell -m manifest.scm`. There are other ways to do it, but I find this the easiest. You can even make a package for your firmware and let guix compile it for you (and maybe flash it too?). Cheers, Ekaitz
Re: Development shell for diffutils does not appear to work - what am I doing wrong?
ElenQ Technology --- Original Message --- On Tuesday, September 26th, 2023 at 08:08, Daniel Littlewood wrote: > Hi guix help, > > I want to try out making a simple change to the program `diff`, which > is part of GNU diffutils: > https://packages.guix.gnu.org/packages/diffutils/3.8/ > I'd like to set up a dev environment, patch diff.c, rebuild it and try > out the new binary. Maybe install it globally later, but I'm not there > yet. > I cloned the diffutils repo from > https://git.savannah.gnu.org/git/diffutils.git, and in that directory > ran > `guix shell git vim nnn -D diffutils` (but I think it's just the -D > diffutils I'm having trouble with). I believe that the > -D/--development argument should produce a shell within which I can > build `diff`. > The README says that the first step is to run `./bootstrap`, but that > fails because of several missing packages. I don't have the full list > right now, but I think autoconf was one, and texi2pdf was another. > > The packaging for diffutils clearly works (since I can install it), so > I wonder if it does something different from what I'm attempting? I > couldn't find the scheme file that defines diffutils, but I'm not sure > I'd be able to read it anyway (I'm really trying out guix for the > first time). > > Thanks for reading, please let me know if I can provide more info. > Dan Dan, I don't think you are doing anything wrong. I don't know why but when doing `-D package` guix is often not adding all the development dependencies as it doesn't load autotools and related things to the shell. You have to add them by hand. Also, diffutils downloads a tar.xz which probably has the bootstrap step already done because it is considered a release source code. You are working from development code I expect, which probably needs some extra tools. What I would do: go adding them to the shell one by one as the build system complains until it doesn't complain anymore. They will probably be `texinfo`, `automake`, `autoconf`, `libtool` and maybe I'm missing something... If you add them as you go you shouldn't leave anything out. Also, this is an interesting call. It might be cool to have a way to add those directly... I don't know what is best but probably with some kind of flag we should add all the deps from the build-system too. And also, have all the `autotools` in just one package because I always forget some of them. I may start another thread with that... Thanks for your question, it is a very valid one! It happened to me before, too, and it's pretty annoying. Hope this helps, Ekaitz
Re: arm-none-eabi-toolchain: stdlib.h not found
ElenQ Technology
--- Original Message ---
On Tuesday, September 26th, 2023 at 22:28, Ekaitz Zarraga
wrote:
>
> --- Original Message ---
> On Tuesday, September 5th, 2023 at 18:09, Ekaitz Zarraga eka...@elenq.tech
> wrote:
>
>
>
> > Hi,
> >
> > If I guix shell to arm-none-eabi-toolchain:
> >
> > `Ekaitz@tuxedo ~/projects/TESTS$ guix shell arm-none-eabi-toolchain`
> >
> > arm-none-eabi-toolchain can't use cstdlib because:
> >
> > ```
> > Ekaitz@tuxedo ~/projects/TESTS [env]$ cat a.c
> > #include
> >
> > void operator delete(void *p) { std::free(p); }
> >
> > Ekaitz@tuxedo ~/projects/TESTS [env]$ arm-none-eabi-g++ a.c
> > In file included from a.c:1:0:
> > /gnu/store/6bq6svar4b2mkp2k3laxqqs1nsxykrvf-profile/arm-none-eabi/include/c++/cstdlib:75:15:
> > fatal error: stdlib.h: No such file or directory
> > #include_next
> >
> > ^~
> > compilation terminated.
> > Ekaitz@tuxedo ~/projects/TESTS [env]$
> > `But:`
> > Ekaitz@tuxedo ~/projects/TESTS [env]$ ls $CROSS_C_INCLUDE_PATH | grep stdlib
> > stdlib.h
> > ```
> >
> > Ideas?
>
>
>
> Hi,
>
> I dug a little bit further and I found a very interesting behavior:
>
> Preprocessor doesn't work vanilla:
>
> ```
> Ekaitz@tuxedo ~/projects/rp2040/TESTS/RANDOM [env]$ arm-none-eabi-cpp a.c
> # 1 "a.c"
> # 1 ""
>
> # 1 ""
>
> # 1 "a.c"
> a.c:1:10: fatal error: cstdlib: No such file or directory
> #include
>
> ^
> compilation terminated.
> `We have CROSS_CPLUS_INCLUDE_PATH correctly set:`
> Ekaitz@tuxedo ~/projects/rp2040/TESTS/RANDOM [env]$ echo
> $CROSS_CPLUS_INCLUDE_PATH
> /gnu/store/6zcm3xq6d3a0j6pnjd09b3zj1c9qbmrl-profile/arm-none-eabi/include:/gnu/store/6zcm3xq6d3a0j6pnjd09b3zj1c9qbmrl-profile/arm-none-eabi/include/c++:/gnu/store/6zcm3xq6d3a0j6pnjd09b3zj1c9qbmrl-profile/arm-none-eabi/include/c++/arm-none-eabi
> `If we insert exactly the same directories pointed by
> CROSS_CPLUS_INCLUDE_PATH manually to the preprocessor it works!`
> Ekaitz@tuxedo ~/projects/rp2040/TESTS/RANDOM [env]$ arm-none-eabi-cpp a.c
> -I/gnu/store/6zcm3xq6d3a0j6pnjd09b3zj1c9qbmrl-profile/arm-none-eabi/include
> -I/gnu/store/6zcm3xq6d3a0j6pnjd09b3zj1c9qbmrl-profile/arm-none-eabi/include/c++
>
> -I/gnu/store/6zcm3xq6d3a0j6pnjd09b3zj1c9qbmrl-profile/arm-none-eabi/include/c++/arm-none-eabi
> # 1 "a.c"
> # 1 ""
>
> # 1 ""
>
> # 1 "a.c"
> # 1
> "/gnu/store/6zcm3xq6d3a0j6pnjd09b3zj1c9qbmrl-profile/arm-none-eabi/include/c++/cstdlib"
> 1
> # 39
> "/gnu/store/6zcm3xq6d3a0j6pnjd09b3zj1c9qbmrl-profile/arm-none-eabi/include/c++/cstdlib"
>
> # 40
> "/gnu/store/6zcm3xq6d3a0j6pnjd09b3zj1c9qbmrl-profile/arm-none-eabi/include/c++/cstdlib"
> 3
>
> # 1
> "/gnu/store/6zcm3xq6d3a0j6pnjd09b3zj1c9qbmrl-profile/arm-none-eabi/include/c++/arm-none-eabi/bits/c++config.h"
> 1 3
> # 533
> "/gnu/store/6zcm3xq6d3a0j6pnjd09b3zj1c9qbmrl-profile/arm-none-eabi/include/c++/arm-none-eabi/bits/c++config.h"
> 3
> # 1
> "/gnu/store/6zcm3xq6d3a0j6pnjd09b3zj1c9qbmrl-profile/arm-none-eabi/include/c++/arm-none-eabi/bits/os_defines.h"
> 1 3
> # 534
> "/gnu/store/6zcm3xq6d3a0j6pnjd09b3zj1c9qbmrl-profile/arm-none-eabi/include/c++/arm-none-eabi/bits/c++config.h"
> 2 3
>
>
> # 1
> "/gnu/store/6zcm3xq6d3a0j6pnjd09b3zj1c9qbmrl-profile/arm-none-eabi/include/c++/arm-none-eabi/bits/cpu_defines.h"
> 1 3
> # 537
> "/gnu/store/6zcm3xq6d3a0j6pnjd09b3zj1c9qbmrl-profile/arm-none-eabi/include/c++/arm-none-eabi/bits/c++config.h"
> 2 3
> # 42
> "/gnu/store/6zcm3xq6d3a0j6pnjd09b3zj1c9qbmrl-profile/arm-none-eabi/include/c++/cstdlib"
> 2 3
> # 75
> "/gnu/store/6zcm3xq6d3a0j6pnjd09b3zj1c9qbmrl-profile/arm-none-eabi/include/c++/cstdlib"
> 3
> ...
> ```
>
> Anyone has an idea of what I'm missing here and why isn't it just automagic?
>
> It should just work, or am I superwrong?
>
> Thanks!
Oh! Too fast!
It was a different problem!
```
Ekaitz@tuxedo ~/projects/rp2040/TESTS/RANDOM [env]$ arm-none-eabi-g++
-I/gnu/store/6zcm3xq6d3a0j6pnjd09b3zj1c9qbmrl-profile/arm-none-eabi/include
-I/gnu/store/6zcm3xq6d3a0j6pnjd09b3zj1c9qbmrl-profile/arm-none-eabi/include/c++
-I/gnu/store/6zcm3xq6d3a0j6pnjd09b3zj1c9qbmrl-profile/arm-none-eabi/include/c++/arm-none-eabi
a.c
In file included from a.c:1:0:
/gnu/store/6zcm3xq6d3a0j6pnjd09b3zj1c9qbmrl-profile/arm-none-eabi/include/c++/cstdlib:75:15:
fatal error: stdlib.h: No such file or directory
#include_next
^~
compilation terminated.
```
The problem is with C++ only.
With C, as per the previous email, I managed to pass the preprocessing step,
but in C++ it doesn't!
It doesn't really matter if I play with the -I flag.
It just doesn't work :(
Re: arm-none-eabi-toolchain: stdlib.h not found
--- Original Message ---
On Tuesday, September 5th, 2023 at 18:09, Ekaitz Zarraga
wrote:
> Hi,
>
> If I guix shell to arm-none-eabi-toolchain:
>
> `Ekaitz@tuxedo ~/projects/TESTS$ guix shell arm-none-eabi-toolchain`
>
> arm-none-eabi-toolchain can't use cstdlib because:
>
> ```
> Ekaitz@tuxedo ~/projects/TESTS [env]$ cat a.c
> #include
>
> void operator delete(void *p) { std::free(p); }
>
> Ekaitz@tuxedo ~/projects/TESTS [env]$ arm-none-eabi-g++ a.c
> In file included from a.c:1:0:
> /gnu/store/6bq6svar4b2mkp2k3laxqqs1nsxykrvf-profile/arm-none-eabi/include/c++/cstdlib:75:15:
> fatal error: stdlib.h: No such file or directory
> #include_next
>
> ^~
> compilation terminated.
> Ekaitz@tuxedo ~/projects/TESTS [env]$
> `But:`
> Ekaitz@tuxedo ~/projects/TESTS [env]$ ls $CROSS_C_INCLUDE_PATH | grep stdlib
> stdlib.h
> ```
>
> Ideas?
Hi,
I dug a little bit further and I found a very interesting behavior:
Preprocessor doesn't work vanilla:
```
Ekaitz@tuxedo ~/projects/rp2040/TESTS/RANDOM [env]$ arm-none-eabi-cpp a.c
# 1 "a.c"
# 1 ""
# 1 ""
# 1 "a.c"
a.c:1:10: fatal error: cstdlib: No such file or directory
#include
^
compilation terminated.
```
We have CROSS_CPLUS_INCLUDE_PATH correctly set:
```
Ekaitz@tuxedo ~/projects/rp2040/TESTS/RANDOM [env]$ echo
$CROSS_CPLUS_INCLUDE_PATH
/gnu/store/6zcm3xq6d3a0j6pnjd09b3zj1c9qbmrl-profile/arm-none-eabi/include:/gnu/store/6zcm3xq6d3a0j6pnjd09b3zj1c9qbmrl-profile/arm-none-eabi/include/c++:/gnu/store/6zcm3xq6d3a0j6pnjd09b3zj1c9qbmrl-profile/arm-none-eabi/include/c++/arm-none-eabi
```
If we insert exactly the same directories pointed by CROSS_CPLUS_INCLUDE_PATH
manually to the preprocessor it works!
```
Ekaitz@tuxedo ~/projects/rp2040/TESTS/RANDOM [env]$ arm-none-eabi-cpp a.c
-I/gnu/store/6zcm3xq6d3a0j6pnjd09b3zj1c9qbmrl-profile/arm-none-eabi/include
-I/gnu/store/6zcm3xq6d3a0j6pnjd09b3zj1c9qbmrl-profile/arm-none-eabi/include/c++
-I/gnu/store/6zcm3xq6d3a0j6pnjd09b3zj1c9qbmrl-profile/arm-none-eabi/include/c++/arm-none-eabi
# 1 "a.c"
# 1 ""
# 1 ""
# 1 "a.c"
# 1
"/gnu/store/6zcm3xq6d3a0j6pnjd09b3zj1c9qbmrl-profile/arm-none-eabi/include/c++/cstdlib"
1
# 39
"/gnu/store/6zcm3xq6d3a0j6pnjd09b3zj1c9qbmrl-profile/arm-none-eabi/include/c++/cstdlib"
# 40
"/gnu/store/6zcm3xq6d3a0j6pnjd09b3zj1c9qbmrl-profile/arm-none-eabi/include/c++/cstdlib"
3
# 1
"/gnu/store/6zcm3xq6d3a0j6pnjd09b3zj1c9qbmrl-profile/arm-none-eabi/include/c++/arm-none-eabi/bits/c++config.h"
1 3
# 533
"/gnu/store/6zcm3xq6d3a0j6pnjd09b3zj1c9qbmrl-profile/arm-none-eabi/include/c++/arm-none-eabi/bits/c++config.h"
3
# 1
"/gnu/store/6zcm3xq6d3a0j6pnjd09b3zj1c9qbmrl-profile/arm-none-eabi/include/c++/arm-none-eabi/bits/os_defines.h"
1 3
# 534
"/gnu/store/6zcm3xq6d3a0j6pnjd09b3zj1c9qbmrl-profile/arm-none-eabi/include/c++/arm-none-eabi/bits/c++config.h"
2 3
# 1
"/gnu/store/6zcm3xq6d3a0j6pnjd09b3zj1c9qbmrl-profile/arm-none-eabi/include/c++/arm-none-eabi/bits/cpu_defines.h"
1 3
# 537
"/gnu/store/6zcm3xq6d3a0j6pnjd09b3zj1c9qbmrl-profile/arm-none-eabi/include/c++/arm-none-eabi/bits/c++config.h"
2 3
# 42
"/gnu/store/6zcm3xq6d3a0j6pnjd09b3zj1c9qbmrl-profile/arm-none-eabi/include/c++/cstdlib"
2 3
# 75
"/gnu/store/6zcm3xq6d3a0j6pnjd09b3zj1c9qbmrl-profile/arm-none-eabi/include/c++/cstdlib"
3
...
```
Anyone has an idea of what I'm missing here and why isn't it just automagic?
It should just work, or am I superwrong?
Thanks!
arm-none-eabi-toolchain: stdlib.h not found
Hi,
If I guix shell to arm-none-eabi-toolchain:
```
Ekaitz@tuxedo ~/projects/TESTS$ guix shell arm-none-eabi-toolchain
```
arm-none-eabi-toolchain can't use cstdlib because:
```
Ekaitz@tuxedo ~/projects/TESTS [env]$ cat a.c
#include
void operator delete(void *p) { std::free(p); }
Ekaitz@tuxedo ~/projects/TESTS [env]$ arm-none-eabi-g++ a.c
In file included from a.c:1:0:
/gnu/store/6bq6svar4b2mkp2k3laxqqs1nsxykrvf-profile/arm-none-eabi/include/c++/cstdlib:75:15:
fatal error: stdlib.h: No such file or directory
#include_next
^~
compilation terminated.
Ekaitz@tuxedo ~/projects/TESTS [env]$
```
But:
```
Ekaitz@tuxedo ~/projects/TESTS [env]$ ls $CROSS_C_INCLUDE_PATH | grep stdlib
stdlib.h
```
Ideas?
Re: python package managers: which works on Guix
Hi, --- Original Message --- On Thursday, July 27th, 2023 at 5:57 PM, Andy Tai wrote: > Hi, a general question: Python-native package managers like conda and > poetry both do not work (or build) on Guix now (I mean the packaged > versions of these in Guix). Curious if any Python package manager > works on top of Guix at this time? Thanks. Guix itself is a pretty reasonable package manager for python too. I've been using Pipenv, but pipenv shell has issues. I decided to use Guix for it directly like: guix shell -m python-wrapper python-requests python-... Or even making a proper manifest.scm or a guix package. It's the best, but that doesn't mean we shouldn't try to make the python package managers work. Hope this helps, Ekaitz
Re: VSCode DevContainer for Guix? (Yasuaki Kudo)
> Careful about this assumption, Guix-generated container images are pretty > large > and include a lot of things you might not want in there (man, info pages, > bash, > ...). I am not saying "do not use it", but if you try to convince people to > use > Guix for container builds, size should not be your argument (especially since > Alpine exists as an alternative). That is very true. This week in System Crafters they have been workin on trying to reduce this. But I didn't have the time to watch the results deeply. Be careful, youtube link: https://www.youtube.com/watch?v=ItGRcOLT-BM
Building JavaFX applications
Hi, I need help building javafx applications. I don't know how to build a simple app: This is my environment: ``` Ekaitz@tuxedo ~/projects/window$$ guix shell openjdk openjdk:jdk java-openjfx-graphics java-openjfx-controls java-openjfx-media java-openjfx-base java-openjfx-build ``` I look for the javafx module paths: ``` Ekaitz@tuxedo ~/projects/window$ [env]$ guix build java-openjfx-graphics java-openjfx-controls java-openjfx-media java-openjfx-base java-openjfx-build /gnu/store/rynjd6sbgqhz7zpgzrdzl8fcpbykwxri-java-openjfx-build-8.202 /gnu/store/13l0qiwy3z88hiq49g9zdpcj77i16fx9-java-openjfx-base-8.202 /gnu/store/bwpkcbcm8cwkdcq7sviyya6pjb4ma3zk-java-openjfx-media-8.202 /gnu/store/f4bbx3524avpw1yn9b0z0fax9rmx4ayd-java-openjfx-controls-8.202 /gnu/store/hpcja76y3y0hpgkxvmv68jhxjlny2316-java-openjfx-graphics-8.202 ``` If I don't add them to module paths they don't appear: ``` Ekaitz@tuxedo ~/projects/window$ [env]$ java --list-modules java.base@19.0.1 java.compiler@19.0.1 java.datatransfer@19.0.1 java.desktop@19.0.1 java.instrument@19.0.1 java.logging@19.0.1 java.management@19.0.1 java.management.rmi@19.0.1 java.naming@19.0.1 java.net.http@19.0.1 java.prefs@19.0.1 java.rmi@19.0.1 java.scripting@19.0.1 java.se@19.0.1 java.security.jgss@19.0.1 java.security.sasl@19.0.1 java.smartcardio@19.0.1 java.sql@19.0.1 java.sql.rowset@19.0.1 java.transaction.xa@19.0.1 java.xml@19.0.1 java.xml.crypto@19.0.1 jdk.accessibility@19.0.1 jdk.attach@19.0.1 jdk.charsets@19.0.1 jdk.compiler@19.0.1 jdk.crypto.cryptoki@19.0.1 jdk.crypto.ec@19.0.1 jdk.dynalink@19.0.1 jdk.editpad@19.0.1 jdk.hotspot.agent@19.0.1 jdk.httpserver@19.0.1 jdk.incubator.concurrent@19.0.1 jdk.incubator.vector@19.0.1 jdk.internal.ed@19.0.1 jdk.internal.jvmstat@19.0.1 jdk.internal.le@19.0.1 jdk.internal.opt@19.0.1 jdk.internal.vm.ci@19.0.1 jdk.internal.vm.compiler@19.0.1 jdk.internal.vm.compiler.management@19.0.1 jdk.jartool@19.0.1 jdk.javadoc@19.0.1 jdk.jcmd@19.0.1 jdk.jconsole@19.0.1 jdk.jdeps@19.0.1 jdk.jdi@19.0.1 jdk.jdwp.agent@19.0.1 jdk.jfr@19.0.1 jdk.jlink@19.0.1 jdk.jpackage@19.0.1 jdk.jshell@19.0.1 jdk.jsobject@19.0.1 jdk.jstatd@19.0.1 jdk.localedata@19.0.1 jdk.management@19.0.1 jdk.management.agent@19.0.1 jdk.management.jfr@19.0.1 jdk.naming.dns@19.0.1 jdk.naming.rmi@19.0.1 jdk.net@19.0.1 jdk.nio.mapmode@19.0.1 jdk.random@19.0.1 jdk.sctp@19.0.1 jdk.security.auth@19.0.1 jdk.security.jgss@19.0.1 jdk.unsupported@19.0.1 jdk.unsupported.desktop@19.0.1 jdk.xml.dom@19.0.1 jdk.zipfs@19.0.1 ``` But If I add them they appear: ``` Ekaitz@tuxedo ~/projects/window$ [env]$ java --list-modules --module-path /gnu/store/rynjd6sbgqhz7zpgzrdzl8fcpbykwxri-java-openjfx-build-8.202/share/java:/gnu/store/13l0qiwy3z88hiq49g9zdpcj77i16fx9-java-openjfx-base-8.202/share/java:/gnu/store/bwpkcbcm8cwkdcq7sviyya6pjb4ma3zk-java-openjfx-media-8.202/share/java:/gnu/store/f4bbx3524avpw1yn9b0z0fax9rmx4ayd-java-openjfx-controls-8.202/share/java:/gnu/store/hpcja76y3y0hpgkxvmv68jhxjlny2316-java-openjfx-graphics-8.202/share/java java.base@19.0.1 java.compiler@19.0.1 java.datatransfer@19.0.1 java.desktop@19.0.1 java.instrument@19.0.1 java.logging@19.0.1 java.management@19.0.1 java.management.rmi@19.0.1 java.naming@19.0.1 java.net.http@19.0.1 java.prefs@19.0.1 java.rmi@19.0.1 java.scripting@19.0.1 java.se@19.0.1 java.security.jgss@19.0.1 java.security.sasl@19.0.1 java.smartcardio@19.0.1 java.sql@19.0.1 java.sql.rowset@19.0.1 java.transaction.xa@19.0.1 java.xml@19.0.1 java.xml.crypto@19.0.1 jdk.accessibility@19.0.1 jdk.attach@19.0.1 jdk.charsets@19.0.1 jdk.compiler@19.0.1 jdk.crypto.cryptoki@19.0.1 jdk.crypto.ec@19.0.1 jdk.dynalink@19.0.1 jdk.editpad@19.0.1 jdk.hotspot.agent@19.0.1 jdk.httpserver@19.0.1 jdk.incubator.concurrent@19.0.1 jdk.incubator.vector@19.0.1 jdk.internal.ed@19.0.1 jdk.internal.jvmstat@19.0.1 jdk.internal.le@19.0.1 jdk.internal.opt@19.0.1 jdk.internal.vm.ci@19.0.1 jdk.internal.vm.compiler@19.0.1 jdk.internal.vm.compiler.management@19.0.1 jdk.jartool@19.0.1 jdk.javadoc@19.0.1 jdk.jcmd@19.0.1 jdk.jconsole@19.0.1 jdk.jdeps@19.0.1 jdk.jdi@19.0.1 jdk.jdwp.agent@19.0.1 jdk.jfr@19.0.1 jdk.jlink@19.0.1 jdk.jpackage@19.0.1 jdk.jshell@19.0.1 jdk.jsobject@19.0.1 jdk.jstatd@19.0.1 jdk.localedata@19.0.1 jdk.management@19.0.1 jdk.management.agent@19.0.1 jdk.management.jfr@19.0.1 jdk.naming.dns@19.0.1 jdk.naming.rmi@19.0.1 jdk.net@19.0.1 jdk.nio.mapmode@19.0.1 jdk.random@19.0.1 jdk.sctp@19.0.1 jdk.security.auth@19.0.1 jdk.security.jgss@19.0.1 jdk.unsupported@19.0.1 jdk.unsupported.desktop@19.0.1 jdk.xml.dom@19.0.1 jdk.zipfs@19.0.1 java.openjfx file:///gnu/store/rynjd6sbgqhz7zpgzrdzl8fcpbykwxri-java-openjfx-build-8.202/share/java/java-openjfx.jar automatic java.openjfx.base file:///gnu/store/13l0qiwy3z88hiq49g9zdpcj77i16fx9-java-openjfx-base-8.202/share/java/java-openjfx-base.jar automatic java.openjfx.controls file:///gnu/store/f4bbx3524avpw1yn9b0z0fax9rmx4ayd-java-openjfx-controls-8.202/share/java/java-openjfx-controls.jar automatic java.openjfx.graphics
Re: Need help packaging IUP (part 1: IM)
> > Have you tried something like
> >
> > (add-before 'configure 'set-ldflags
> > (lambda _
> > (substitute* "tecmake.mak"
> > (("STDLDFLAGS := -shared" all) (string-append all "
> > -Wl,-rpath="#$output "/lib")
> >
> > not tested though.
>
>
> I'll give a try to this. I didn't try it!
> Thanks!
>
Worked! Thank you!!
Re: Need help packaging IUP (part 1: IM)
Hi Reza,
Thanks for the help!
I'll answer inline:
--- Original Message ---
On Tuesday, March 21st, 2023 at 13:25, Reza Housseini
wrote:
> Have you tried something like
>
> (add-before 'configure 'set-ldflags
> (lambda _
> (substitute* "tecmake.mak"
> (("STDLDFLAGS := -shared" all) (string-append all "
> -Wl,-rpath="#$output "/lib")
>
> not tested though.
I'll give a try to this. I didn't try it!
Thanks!
>
> Also it looks like a cmake project, why are you not using the
> cmake-build-system?
The CMake they use gives syntax errors and the official docs describe to use
the Makefile and they don't mention the CMake.
It's probably poorly maintained.
Thanks for the help! I'll answer back when I try the STDLDFLAGS!
Need help packaging IUP (part 1: IM)
Hi,
I'm trying to package IUP (https://www.tecgraf.puc-rio.br/iup/) and I'm already
stuck on the first package :)
They have a pretty wild general Makefile but I think I'm managing to deal with
it.
The main problem I have now is the RUNPATH is not set properly. If I keep the
RUNPATH validation phase it fails.
I tried several ways to add it but I didn't manage to fix it.
I leave my current package description below.
Thanks for your help!
Cheers,
Ekaitz
PS: Also, anyone has a suggestion on how can I transform the phases to turn
then something like:
(add-build-phase "im_jp2") => (add-before build' build-im_jp2 (lambda _ (invoke
...)))
I tried with a simple function but the the quoting killed all my hopes and I'm
not good with macros (yet).
---
(define-module (iup)
#:use-module (guix packages)
#:use-module (guix gexp)
#:use-module (gnu packages)
#:use-module ((gnu packages base) #:prefix base:)
#:use-module (gnu packages algebra)
#:use-module (gnu packages commencement)
#:use-module (gnu packages gl)
#:use-module (gnu packages pkg-config)
#:use-module (gnu packages fontutils)
#:use-module (gnu packages glib)
#:use-module (gnu packages gtk)
#:use-module (gnu packages image)
#:use-module (gnu packages lua)
#:use-module (gnu packages xorg)
#:use-module (guix download)
#:use-module (guix build utils)
#:use-module (guix build-system gnu)
#:use-module (guix build-system cmake)
#:use-module ((guix licenses) #:prefix license:))
(define-public im
(package
(name "im")
(version "3.15")
(source (origin
(method url-fetch)
(uri (string-append
"https://sourceforge.net/projects/imtoolkit/files/;
version "/Docs%20and%20Sources/im-" version
"_Sources.tar.gz"))
(sha256
(base32
"1c5p8j7sw3r7kmkd4qa49zmh3dlp5zszp2sagi71aqjq84pngp40"
(inputs (list base:which
libpng
fftw
fftwf))
(build-system gnu-build-system)
(arguments
(list
#:tests? #f
#:validate-runpath? #f ;; FIXME: Fix the runpath!!
#:phases
#~(modify-phases %standard-phases
(add-before 'configure 'set-ldflags
;; FIXME: LDFLAGS don't work, so the runpath is broken
(lambda _
(setenv "LDFLAGS" (string-append "-Wl,-rpath=" #$output
"/lib"
(delete 'configure)
;; Left lua-related libs out of the mix for the moment as they
;; have some build errors and I don't need them yet.
(add-before 'build 'cd
(lambda _
(chdir "src")))
(add-before 'build 'build-im
(lambda _
(invoke "make" "im")))
(add-before 'build 'build-im-jp2
(lambda _
(invoke "make" "im_jp2")))
(add-before 'build 'build-im-process
(lambda _
(invoke "make" "im_process")))
(add-before 'build 'build-im-process-omp
(lambda _
(invoke "make" "im_process_omp")))
(add-before 'build 'build-im-fftw3
(lambda _
(invoke "make" "im_fftw3")))
(delete 'build)
(replace 'install
(lambda* (#:key outputs #:allow-other-keys)
(let* ((include (string-append #$output "/include"))
(lib (string-append #$output "/lib")))
(mkdir-p include)
(mkdir-p lib)
(copy-recursively "../include" include)
(invoke "find" "../lib" "-exec" "cp" "{}" lib ";")))
(home-page "https://www.tecgraf.puc-rio.br/im;)
(synopsis "IM is a toolkit for image representation, storage, capture and
processing")
(description "The main goal of the library is to provide a simple API and
abstraction of imaging for scientific applications. The most popular file
formats are supported: TIFF, BMP, PNG, JPEG, GIF and AVI. Image representation
includes scientific data types, and about a hundred Image Processing operations
are available.")
(license license:expat)))
Re: Intel i7-1165G7 vulnerable to Spectre v2
Hi, > Does > > $ echo 1 | sudo tee /proc/sys/kernel/unprivileged_bpf_disabled > > change this? > > What does Debian's kconfig list for CONFIG_BPF_UNPRIV_DEFAULT_OFF? > > Guix has it unset (which means default on) which means that > unprivileged_bpf_disabled is 0 (which means enabled) because > Linux is a hot mess and nobody cares. > > Kind regards, > > T G-R In my CPU (i7-10510U) I had the same problem and that fixes it. Cheers, Ekaitz
Re: bash scripts in Guix question
Hi, > Hi guixers, > > i'm wondering...for my scripts to work in Guix, instead of > '#! /bin/bash' , > > what do i have to start them with ? > > Thanks,thanks, thanks > > F We have a link to bash located in /bin/sh so you can just leave /bin/sh. Or you can also use `#!/usr/bin/env bash` Both should work. Cheers, Ekaitz
Re: Trivia question: What does the "G" in G-expressions stand for?
> Hi! > > jgart jg...@dismail.de skribis: > > > Does the G in G-Expressions stand for Guix? > > > No, it stands for “G”, just like the “s” in “s-expression” stands for > nothing (AFAIK). :-) > > Ludo’. In s-expression it doesn't stand for symbolic expression?
Re: How does a gexp relate to a derivation?
> > How does a gexp relate to a derivation? > > > The module comment of (guix gexp) says this: > > --8<---cut here---start->8--- > > ;;; This module implements "G-expressions", or "gexps". Gexps are like > ;;; S-expressions (sexps), with two differences: > ;;; > ;;; 1. References (un-quotations) to derivations or packages in a gexp are > ;;; replaced by the corresponding output file name; in addition, the > ;;; 'ungexp-native' unquote-like form allows code to explicitly refer to > ;;; the native code of a given package, in case of cross-compilation; > ;;; > ;;; 2. Gexps embed information about the derivations they refer to. > ;;; > ;;; Gexps make it easy to write to files Scheme code that refers to store > ;;; items, or to write Scheme code to build derivations. > --8<---cut here---end--->8--- > > > > > -- > Ricardo So, in other words, a Gexp is a quotation system that has a context: the current derivation. So all the expansions happen with that in mind. Correct?
Re: How to install OBS plugins and scripts?
Thanks a lot Andrew! Just tested the scripts and they work well. Also, mention that you can keep them anywhere you want, which is also a plus. --- Original Message --- On Friday, August 26th, 2022 at 7:48 AM, Andrew Tropin wrote: > OBS native plugins should install their .so files to lib/obs-plugins/ > and data files to share/obs/obs-plugins/ and they will be automatically > loaded, when obs installed in the same profile (directories for finding > plugins defined in native-search-paths of obs package). Also, take a > look at obs-* packages for examples on how to package plugins. > > There was no scripts support, updated obs package to 27.2.4 and enabled > python and lua scripting support. > > https://git.savannah.gnu.org/cgit/guix.git/commit/?id=4ff0ceb26462d27281eed5cf789ca4158f94409a > > -- > Best regards, > Andrew Tropin
Re: How to install OBS plugins and scripts?
Hi, --- Original Message --- On Thursday, August 18th, 2022 at 6:12 PM, pelzflorian (Florian Pelz) wrote: > Ekaitz Zarraga eka...@elenq.tech writes: > > > Does anyone have OBS plugins? > > > Not me, but maybe it helps that Andrew Tropin wrote in > https://lists.gnu.org/archive/html/guix-patches/2021-01/msg01022.html > > that you can use OBS_PLUGINS_DATA_PATH and OBS_PLUGINS_PATH and made obs > plugins that are shipped with guix use those. > > > Also our OBS does not depend on Python or Lua so I don't think it's > > able to run scripts... Anyone has info about this? > > > I suppose Python or Lua just need to be installed along with obs? > > Sorry I did not try any of this. > > Regards, > Florian Thanks a lot for your answer Florian. I wanted to dig on this further in order to answer you but I didn't have the time to do my proper research. I just want to thank you for your help. I'll dig on this further and answer back if I find any solution. I'll contact Andrew directly too. Thanks again, Ekaitz
How to install OBS plugins and scripts?
Hi, Does anyone have OBS plugins? How should I configure them? There's no `scripts` button under `Tools` as some tutorials suggest. Also our OBS does not depend on Python or Lua so I don't think it's able to run scripts... Anyone has info about this? Thanks! Ekaitz
Re: Gnome Tracker (tracker-miner-fs-3) behaving badly
Hi, --- Original Message --- On Thursday, August 18th, 2022 at 2:50 AM, Ben Sturmfels via wrote: > Hi Folks! > > I really like the idea of Gnome Tracker - index all your files, > contacts, etc. so you can quickly search them. It just doesn't seem to > work properly, it uses a lot of resources and it's behaviour is > opaque. I'm not sure whether this is the configuration in Guix System, > or whether Tracker just isn't well designed. My hunch is that it's a > design problem. Does anyone have good/bad experience with Tracker > outside of Guix? > > Specific issues: > > * the tracker-miner-fs-3 process runs spontaneously consuming at least > one CPU for a few minutes (I haven't kept track of how long or how > often) > > * after all this scanning, searching in Gnome Activities doesn't > display any matches for the contents of files, only the filenames - > I thought indexing the contents was the whole point > > * when it's busy, tracker-miner-fs-3 seems to ignore `kill` (SIGTERM) > and needs `kill -9` (SIGKILL) to kill it > > * it's not clear exactly what files Tracker is indexing - there doesn't > appear to be a log - I'd really like to be able to just `tail -f` to > see what's going on > > I'd like to also understand how tracker-miner-fs-3 is started by > Gnome. It seems to start automatically and restart after some time if > killed. It doesn't seem to be listed under "Startup Applications" in > Gnome Tweak Tool. > > Debugging Tracker: > > After killing it with `pkill --full tracker-miner-fs-3 --echo`, I can > restart it with some debugging information using: > > env TRACKER_DEBUG=config,miner-fs-events,monitors,statistics,status > /gnu/store/[...]-tracker-miners-3.3.1/libexec/tracker-miner-fs-3 > > This seems to show some scheduling information about the indexing, but > not exactly what it's indexing. > > Regards, > Ben I experienced a very similar situation. I hate hen it happens. Every time I move a file or I rename something the thing starts to run and my cpu goes crazy. On the other hand, this is a recent problem, in previous versions it didn't happen and I didn't really have more effort searching for files in nautilus or similar applications. For me, it's just bad. Also, I don't use GNOME directly, even if I have it installed and use most of its apps, I run an i3 session, but tracker-miner is still running. I don't know if that helps. If I could, I'd like to remove it. Is it possible? Thanks, Ekaitz
Re: Web Browsers Report being "Unable to Connect" after Wi-Fi Connection Drop Period
Hi --- Original Message --- On Monday, May 23rd, 2022 at 9:34 PM, Wamm K. D. wrote: > I think it's happened most often with IceCat but pretty much every > browser I've tried has done this. > > I'm not sure the cause but my Wi-Fi card drops the connection, /really/ > often; it sucks but it's what I've got. > > The consequence is that I've noticed a particular behavior relatively > frequently: when the connection drops, the browser I'm using will get > stuck thinking that it cannot connect to the internet once the internet > connection comes back, sometimes (especially if I tried to load a page > while the internet connection was down). Trying to load any other page > results in the "Unable to Connect" (or equivalent) page. > > I've tried IceCat, Ungoogled Chromium, Qutebrowser, Luakit, and Nyxt and > had the same experience, in some capacity, with all of them. > > Loading a local HTML file works fine, during these spells, and I either > have to wait a variable length of time before the spell passes or > restart the browser. > > It doesn't seem to be something which impacts all browsers, at the same > time: opening a new browser after one gets affected works fine or trying > to use one of the other browsers I had open (especially if I hadn't been > using this other browser while the internet connection had dropped). That's have been my experience with Icecat in the past and I think others have mentioned this in the past too. I don't know where it comes from, or if I still have the same issue, as my wifi is pretty stable nowadays. I'm sure more people have experienced this. It's something we should fix soonish. Ideas anyone?
Re: IceWM for Guix System
Hi, --- Original Message --- On Saturday, May 14th, 2022 at 8:38 PM, 白い熊＠相撲道 wrote: > On 2022-05-13 10:22, zimoun wrote: > > > guix build -L /tmp/my-pkgs icewm > > > > where /tmp/my-pkgs contains this file icewm.scm: > > ... > > > You can address these issues, adds the list of inputs, etc. Here, > > icewm > > is bootstrapped when you could also use the release version and avoid > > automake and friend. Be careful, the ’bootstrap’ phase is triggered by > > the check of the ’configure’ file, and here it is probably named > > ’configure.sh’. > > ... > > > Feel free to share your progress and ask more help on specific points > > if > > > Many thanks Simon for the guidance - I'm learning many things working > through this. Now I need more help please :@) > > I've been able to get it built with the following package definition: > --8<---cut here---start->8--- > > (define-module (icewm) > #:use-module (guix packages) > #:use-module (gnu packages) > #:use-module (guix build-system gnu) > #:use-module (guix download) > #:use-module ((guix licenses) #:prefix license:) > #:use-module (gnu packages autogen) > #:use-module (gnu packages autotools) > #:use-module (gnu packages compression) > #:use-module (gnu packages fontutils) > #:use-module (gnu packages fribidi) > #:use-module (gnu packages gettext) > #:use-module (gnu packages image) > #:use-module (gnu packages perl) > #:use-module (gnu packages pkg-config) > #:use-module (gnu packages xorg) > ) > > (define-public icewm > (package > (name "icewm") > (version "2.9.7") > (source > (origin > (method url-fetch) > (uri (string-append > > "https://github.com/ice-wm/icewm/releases/download/2.9.7/icewm-; > version > ".tar.lz")) > (sha256 (base32 > "0an7h55sacikkvb7llz2n93dl7xds24ics8mqzmznjddygiphpy2" > (build-system gnu-build-system) > (inputs > (list fontconfig fribidi imlib2 libice libjpeg-turbo libsm > libxcomposite libxdamage libxext libxfixes libxft libxinerama libxpm > libxrandr libxrender libx11 lzip perl pkg-config)) > (home-page "https://ice-wm.org/;) > (synopsis " window manager for the X Window System") > (description "IceWM is a window manager for the X Window System. The > goal > of IceWM is speed, simplicity, and not getting in the user’s way. It > comes > with a taskbar with pager, global and per-window keybindings and a > dynamic > menu system. Application windows can be managed by keyboard and mouse. > Windows > can be iconified to the taskbar, to the tray, to the desktop or be made > hidden. They are controllable by a quick switch window (Alt+Tab) and in > a > window list. A handful of configurable focus models are > menu-selectable. Setups with multiple monitors are supported by RandR > and > Xinerama. IceWM is very configurable, themeable and well documented. It > includes an optional external background wallpaper manager with > transparency > support, a simple session manager and a system tray.") > (license license:gpl2))) > --8<---cut here---end--->8--- > > > However it fails on running 'make check' with one test failed: > FAIL: strtest > > I do not know how to solve this - configure doesn't have options for > skip check, don't know how to overcome this. Could you help me get over > this hurdle and then indicate how to proceed further? > > --- > Best regards / 宜しく御願い致します / S pozdravem / C уважением / Z poważaniem / > Mit freundlichen Grüßen > > 白い熊 > ShiroiKuma We try to avoid disabling tests, but for giving it a try you can set the #:tests? argument to #f. You can see how that is done in sc-im package for instance (see gnu/packages/spreadsheet.scm:62 or run `guix edit sc-im`) Hope this helps, Ekaitz
Re: Blender has no denoiser
Hi, --- Original Message --- On Saturday, April 16th, 2022 at 6:29 PM, Haider Mirza wrote: > The Guix installation of Blender has no denoiser. > Blender's compositor says that Blender is compiled without > "OpenImageDenoise" > > Last time I checked OIDN (OpenImageDenoise) is free software and > all it's dependancies are also. > I tried to packaged OIDN but it just ended up being a massive pain > and It's depenadancy (ispc) just refused to compile. > Can you suggest some sort of solution (The IRC has no answers) > -- > Haider Mirza via Emacs on a GNU/Linux system I had a similar issue in the past with blender, it was missing a dependency and I had to package it myself. It is what it is. Sometimes you need to package this kind of software, it's full of compilation errors and so on. The only thing we can do is to help you package it. Can you share the package you developed?
Re: How to work with npm from guix shell
Hi > Hello, > > I try to install the community server of the solid project using the > following documentation in a guix shell session. > > https://solidproject.org//self-hosting/css > > Unfortunatelly I got the following error during the execution of 'npm install > -g @solid/community-server' This command installs stuff globally, which tries to access system's directories and that's what's giving you the error (the store is immutable). What I suggest is to remove the -g so the packages are going to be installed in the current folder under node_modules, and you'll be able to run them by their path: ./node_modules/community-server/bin/whatever You can also add that to the PATH and simplify some things. That's what I do. But there might be other better options, like setting where should npm install the packages globally? I don't know about that. Hope that helps
Re: LuaJIT ffi issue with dynamic libraries
Hi,
I'm pushing this again just in case anyone has an answer.
Thanks!
‐‐‐ Original Message ‐‐‐
On Thursday, November 25th, 2021 at 5:46 PM, Ekaitz Zarraga
wrote:
> Hi all,
>
> I've been having some issues with LuaJIT's FFI. More specifically with the
> `load` function, that loads a library in the "search path for dynamic
> libraries".
>
> The LuaJIT documentation says the following:
>
> > On POSIX systems, if the name contains no dot, the extension .so is
> > appended.
> > Also, the lib prefix is prepended if necessary. So ffi.load("z") looks for
> > "libz.so" in the default shared library search path.
>
>
> In my case I'm unable to load a library, in this case `librec.so` simply by
> calling `ffi.load` even if both are part of the same environment. I read a
> little bit of the LuaJIT code and I found that it is doing a `dlopen` inside
> so
> I made a test:
>
> First I made a simple C file with a call to `dlopen` and straced its result.
> It
> simply finds the library after proving some folders. I also noted that the
> environment automatically set the `LIBRARY_PATH` variable for me (even if the
> value of the variable does not match the patch where librec.so is being
> searched):
>
> --8<---cut here---start->8---
> $ guix shell gcc-toolchain strace recutils
> $ cat < dl.c
> > #include
> >
> > int main(int argc, char *argv[]){
> > void * i = dlopen("librec.so", RTLD_LAZY);
> > return (int)i;
> > }
> EOF
> $ gcc dl.c -ldl
> $ strace ./a.out 2>&1 | grep librec.so
> openat(AT_FDCWD,
> "/gnu/store/fa6wj5bxkj5ll1d7292a70knmyl7a0cr-glibc-2.31/lib/librec.so",
> O_RDONLY|O_CLOEXEC) = -1 ENOENT (No such file or directory)
> openat(AT_FDCWD,
> "/gnu/store/28bcmy08ki5krvr2g9hbm3bhys822fvn-gcc-11.2.0-lib/lib/librec.so",
> O_RDONLY|O_CLOEXEC) = -1 ENOENT (No such file or directory)
> openat(AT_FDCWD,
> "/gnu/store/fz52krs37spvca6q0g0zmx6jmc1n388g-profile/lib/tls/haswell/x86_64/librec.so",
> O_RDONLY|O_CLOEXEC) = -1 ENOENT (No such file or directory)
> openat(AT_FDCWD,
> "/gnu/store/fz52krs37spvca6q0g0zmx6jmc1n388g-profile/lib/tls/haswell/librec.so",
> O_RDONLY|O_CLOEXEC) = -1 ENOENT (No such file or directory)
> openat(AT_FDCWD,
> "/gnu/store/fz52krs37spvca6q0g0zmx6jmc1n388g-profile/lib/tls/x86_64/librec.so",
> O_RDONLY|O_CLOEXEC) = -1 ENOENT (No such file or directory)
> openat(AT_FDCWD,
> "/gnu/store/fz52krs37spvca6q0g0zmx6jmc1n388g-profile/lib/tls/librec.so",
> O_RDONLY|O_CLOEXEC) = -1 ENOENT (No such file or directory)
> openat(AT_FDCWD,
> "/gnu/store/fz52krs37spvca6q0g0zmx6jmc1n388g-profile/lib/haswell/x86_64/librec.so",
> O_RDONLY|O_CLOEXEC) = -1 ENOENT (No such file or directory)
> openat(AT_FDCWD,
> "/gnu/store/fz52krs37spvca6q0g0zmx6jmc1n388g-profile/lib/haswell/librec.so",
> O_RDONLY|O_CLOEXEC) = -1 ENOENT (No such file or directory)
> openat(AT_FDCWD,
> "/gnu/store/fz52krs37spvca6q0g0zmx6jmc1n388g-profile/lib/x86_64/librec.so",
> O_RDONLY|O_CLOEXEC) = -1 ENOENT (No such file or directory)
> openat(AT_FDCWD,
> "/gnu/store/fz52krs37spvca6q0g0zmx6jmc1n388g-profile/lib/librec.so",
> O_RDONLY|O_CLOEXEC) = 3
> $ echo $LIBRARY_PATH
> /gnu/store/hiiyy2b8pw2cq9mw8iik9xi9lpclcv7b-profile/lib
> --8<---cut here---end--->8---
>
> When I do a similar approach with LuaJIT I get the following:
>
> - The paths the `dlopen` call is proving are different to the previous case
> and
> they are pretty weird too.
> - There's no `LIBRARY_PATH` set.
> - Even if I set `LIBRARY_PATH` myself it's ignored.
> - It works correctly when I set `LD_LIBRARY_PATH`, but should I?
>
>
> --8<---cut here---start->8---
> $ guix shell luajit strace recutils
> $ cat < dl.lua
> > ffi = require "ffi"
> > ffi.load "rec"
> > EOF
> $ strace luajit dl.lua 2>&1 | grep librec.so
> openat(AT_FDCWD,
> "/gnu/store/fa6wj5bxkj5ll1d7292a70knmyl7a0cr-glibc-2.31/lib/librec.so",
> O_RDONLY|O_CLOEXEC) = -1 ENOENT (No such file or directory)
> openat(AT_FDCWD,
> "/gnu/store/01b4w3m6mp55y531kyi1g8shh722kwqm-gcc-7.5.0-lib/lib/librec.so",
> O_RDONLY|O_CLOEXEC) = -1 ENOENT (No such file or directory)
> openat(AT_FDCWD,
> "/gnu/store/01b4w3m6mp55y531kyi1g8shh722kwqm-gcc-7.5.0-lib/lib/gcc/x86_64-unknown-linux-gnu/7.5.0/../../../tls/haswell/x86_64/librec.so",
> O_RDONLY|O_CLOEXEC) = -1 ENOENT (No such file or directory)
> openat(AT_FDCWD,
> "/gnu/
LuaJIT ffi issue with dynamic libraries
Hi all,
I've been having some issues with LuaJIT's FFI. More specifically with the
`load` function, that loads a library in the "search path for dynamic
libraries".
The LuaJIT documentation says the following:
> On POSIX systems, if the name contains no dot, the extension .so is appended.
> Also, the lib prefix is prepended if necessary. So ffi.load("z") looks for
> "libz.so" in the default shared library search path.
In my case I'm unable to load a library, in this case `librec.so` simply by
calling `ffi.load` even if both are part of the same environment. I read a
little bit of the LuaJIT code and I found that it is doing a `dlopen` inside so
I made a test:
First I made a simple C file with a call to `dlopen` and straced its result. It
simply finds the library after proving some folders. I also noted that the
environment automatically set the `LIBRARY_PATH` variable for me (even if the
value of the variable does not match the patch where librec.so is being
searched):
--8<---cut here---start->8---
$ guix shell gcc-toolchain strace recutils
$ cat < dl.c
> #include
>
> int main(int argc, char *argv[]){
> void * i = dlopen("librec.so", RTLD_LAZY);
> return (int)i;
> }
EOF
$ gcc dl.c -ldl
$ strace ./a.out 2>&1 | grep librec.so
openat(AT_FDCWD,
"/gnu/store/fa6wj5bxkj5ll1d7292a70knmyl7a0cr-glibc-2.31/lib/librec.so",
O_RDONLY|O_CLOEXEC) = -1 ENOENT (No such file or directory)
openat(AT_FDCWD,
"/gnu/store/28bcmy08ki5krvr2g9hbm3bhys822fvn-gcc-11.2.0-lib/lib/librec.so",
O_RDONLY|O_CLOEXEC) = -1 ENOENT (No such file or directory)
openat(AT_FDCWD,
"/gnu/store/fz52krs37spvca6q0g0zmx6jmc1n388g-profile/lib/tls/haswell/x86_64/librec.so",
O_RDONLY|O_CLOEXEC) = -1 ENOENT (No such file or directory)
openat(AT_FDCWD,
"/gnu/store/fz52krs37spvca6q0g0zmx6jmc1n388g-profile/lib/tls/haswell/librec.so",
O_RDONLY|O_CLOEXEC) = -1 ENOENT (No such file or directory)
openat(AT_FDCWD,
"/gnu/store/fz52krs37spvca6q0g0zmx6jmc1n388g-profile/lib/tls/x86_64/librec.so",
O_RDONLY|O_CLOEXEC) = -1 ENOENT (No such file or directory)
openat(AT_FDCWD,
"/gnu/store/fz52krs37spvca6q0g0zmx6jmc1n388g-profile/lib/tls/librec.so",
O_RDONLY|O_CLOEXEC) = -1 ENOENT (No such file or directory)
openat(AT_FDCWD,
"/gnu/store/fz52krs37spvca6q0g0zmx6jmc1n388g-profile/lib/haswell/x86_64/librec.so",
O_RDONLY|O_CLOEXEC) = -1 ENOENT (No such file or directory)
openat(AT_FDCWD,
"/gnu/store/fz52krs37spvca6q0g0zmx6jmc1n388g-profile/lib/haswell/librec.so",
O_RDONLY|O_CLOEXEC) = -1 ENOENT (No such file or directory)
openat(AT_FDCWD,
"/gnu/store/fz52krs37spvca6q0g0zmx6jmc1n388g-profile/lib/x86_64/librec.so",
O_RDONLY|O_CLOEXEC) = -1 ENOENT (No such file or directory)
openat(AT_FDCWD,
"/gnu/store/fz52krs37spvca6q0g0zmx6jmc1n388g-profile/lib/librec.so",
O_RDONLY|O_CLOEXEC) = 3
$ echo $LIBRARY_PATH
/gnu/store/hiiyy2b8pw2cq9mw8iik9xi9lpclcv7b-profile/lib
--8<---cut here---end--->8---
When I do a similar approach with LuaJIT I get the following:
- The paths the `dlopen` call is proving are different to the previous case and
they are pretty weird too.
- There's no `LIBRARY_PATH` set.
- Even if I set `LIBRARY_PATH` myself it's ignored.
- It works correctly when I set `LD_LIBRARY_PATH`, but should I?
--8<---cut here---start->8---
$ guix shell luajit strace recutils
$ cat < dl.lua
> ffi = require "ffi"
> ffi.load "rec"
> EOF
$ strace luajit dl.lua 2>&1 | grep librec.so
openat(AT_FDCWD,
"/gnu/store/fa6wj5bxkj5ll1d7292a70knmyl7a0cr-glibc-2.31/lib/librec.so",
O_RDONLY|O_CLOEXEC) = -1 ENOENT (No such file or directory)
openat(AT_FDCWD,
"/gnu/store/01b4w3m6mp55y531kyi1g8shh722kwqm-gcc-7.5.0-lib/lib/librec.so",
O_RDONLY|O_CLOEXEC) = -1 ENOENT (No such file or directory)
openat(AT_FDCWD,
"/gnu/store/01b4w3m6mp55y531kyi1g8shh722kwqm-gcc-7.5.0-lib/lib/gcc/x86_64-unknown-linux-gnu/7.5.0/../../../tls/haswell/x86_64/librec.so",
O_RDONLY|O_CLOEXEC) = -1 ENOENT (No such file or directory)
openat(AT_FDCWD,
"/gnu/store/01b4w3m6mp55y531kyi1g8shh722kwqm-gcc-7.5.0-lib/lib/gcc/x86_64-unknown-linux-gnu/7.5.0/../../../tls/haswell/librec.so",
O_RDONLY|O_CLOEXEC) = -1 ENOENT (No such file or directory)
openat(AT_FDCWD,
"/gnu/store/01b4w3m6mp55y531kyi1g8shh722kwqm-gcc-7.5.0-lib/lib/gcc/x86_64-unknown-linux-gnu/7.5.0/../../../tls/x86_64/librec.so",
O_RDONLY|O_CLOEXEC) = -1 ENOENT (No such file or directory)
openat(AT_FDCWD,
"/gnu/store/01b4w3m6mp55y531kyi1g8shh722kwqm-gcc-7.5.0-lib/lib/gcc/x86_64-unknown-linux-gnu/7.5.0/../../../tls/librec.so",
O_RDONLY|O_CLOEXEC) = -1 ENOENT (No such file or directory)
openat(AT_FDCWD,
"/gnu/store/01b4w3m6mp55y531kyi1g8shh722kwqm-gcc-7.5.0-lib/lib/gcc/x86_64-unknown-linux-gnu/7.5.0/../../../haswell/x86_64/librec.so",
O_RDONLY|O_CLOEXEC) = -1 ENOENT (No such file or directory)
openat(AT_FDCWD,
Re: Help with a GraalVM package?
Hi Gary, > Hi Guix, > > Has anyone had any luck in building a package definition for GraalVM? > > Https://www.graalvm.org/ Probably not, but it can be you, or maybe me in the near future. > I'm a Clojure programmer in my day job, and one of the areas I work in > Is (open source) high performance wildland fire modeling. GraalVM could > Really help me speed up my existing OpenJDK-based models as well as help > Me build other GraalVM-dependent Guix packages like the Clojure linter, > Clj-kondo. > > Any help with this task (or a pointer to a channel with a working > Package definition) would be much appreciated. > The guix documentation has a good explanation about how to make packages, but this one might be complex. I searched a little bit and I didn't find any guide on how to build GraalVM itself, if you can find me one I can try to guide you in making the package or at least we could evaluate how hard is it. Cheers, Ekaitz
Re: help: screen tearing in guix
Hi Andre
> I've noticed that on my laptop's built-in screen there's some kind of
> weird "delay" going on. For instance, the cursor lags behind after
> hitting Space. On my external monitor, there are no issues at all. If
> that matters, the built-in screen native resolution is almost 4K. I've
> tried to play with the ModeLine settings in the past, but I wasn't able
> to solve it.
>
> I'm wondering if this "TearFree" option sounds like it could solve my
> issue. Regardless, I'll give it a try.
In my laptop it didn't resolve tearing totally but it reduced it (I think,
I made the change recently and I was trained to ignore it :S)
> I also took a look at your config, and I noticed the last section
> relative to ModemManager.
>
> I've been playing with DIY keyboards (as you know) and I'm just curious
> why exactly ModemManager interferes with USB devices.
>
> For now I've solved the issue by adding the following udev rule:
>
> --8<---cut here---start->8---
> SUBSYSTEMS=="usb", ATTRS{idVendor}=="1b4f", ATTRS{idProduct}=="9205",
> TAG+="uaccess", ENV{ID_MM_DEVICE_IGNORE}="1"
> --8<---cut here---end--->8---
>
> My understand is that MM from "ID_MM_DEVICE_IGNORE" stands for
> ModemManager. I'd like to get a deeper understanding of what's going
> on. If you can share some pointers, I'd appreciate it.
The modem manager basically tries to know if anything you connected
is a modem so it sends AT commands to the serial devices right when
they are connected. That may interfere with your work.
The udev rule you added id more than enough I think. I'm a radical guy
so I just removed the modem manager service :)
Re: help: screen tearing in guix
> > I don't know if Guix's documentation is designed to include this kind of > > specific things, but we could try to add it. > > I think that's what the cookbook is for. > Thanks! Any suggestion on where to include it or how?
Re: help: screen tearing in guix
> Thank you very much @ekaitz, problem solved. > > for the future maybe a screen tearing section in guix wiki would be a great > help and quick access for starter :) It's a shame Guix doesn't have a wiki :P I don't know if Guix's documentation is designed to include this kind of specific things, but we could try to add it. The change is basically the same that comes in the ArchWiki but it must be set through the Guix configuration system that fills xorg.conf. :)
Re: help: screen tearing in guix
I think this is what you are looking for: https://gitlab.com/ekaitz-zarraga/guix-configuration/-/blob/master/desktop.scm#L115 HTH!! ElenQ Technology Ethical Innovation
Re: Blender freezes i3wm
> On Sun, Oct 3, 2021, at 15:59, Ekaitz Zarraga wrote: > > > Is there any chance that we can fix this? > > There's not a fix available upstream yet, right? There's a mesa suggestion: https://gitlab.freedesktop.org/mesa/mesa/-/issues/3751 But I didn't try it yet. > > Any ideas? > > I think we have two different versions of Blender. Does the old one work? It surely will but it lacks most of the features I need :(
Re: Blender freezes i3wm
> Well I was wrong about the patch applying cleanly too. Not paying close > Attention, apparently. And that patch was merged already in the current > Version of Blender we ship. > > It looks like my GPU, which is (according to lspci): > > 00:02.0 VGA compatible controller: Intel Corporation UHD Graphics 620 > (Whiskey Lake) > > Has particular trouble: > > https://developer.blender.org/T80458#1182395 > > Frustrating. Hi all, I found this error again and I don't know what to do. Honestly this sucks really hard. I checked the error report at blender.org and I find the same error even with the same file that is added in the error report: ``` Oct 3 21:52:44 localhost vmunix: [35774.167860] i915 :00:02.0: [drm] Resetting rcs0 for preemption time out Oct 3 21:52:44 localhost vmunix: [35774.167877] i915 :00:02.0: [drm] blender[12755] context reset due to GPU hang Oct 3 21:52:44 localhost vmunix: [35774.197793] i915 :00:02.0: [drm] GPU HANG: ecode 9:1:85df9ebf, in blender [12755] Oct 3 21:52:45 localhost vmunix: [35774.871808] i915 :00:02.0: [drm] Resetting rcs0 for preemption time out Oct 3 21:52:45 localhost vmunix: [35774.871827] i915 :00:02.0: [drm] blender[12755] context reset due to GPU hang Oct 3 21:52:45 localhost vmunix: [35774.879501] i915 :00:02.0: [drm] GPU HANG: ecode 9:1:85dfbfff, in blender [12755] Oct 3 21:52:54 localhost vmunix: [35784.215617] i915 :00:02.0: [drm] Resetting rcs0 for preemption time out Oct 3 21:52:54 localhost vmunix: [35784.215643] i915 :00:02.0: [drm] blender[12755] context reset due to GPU hang Oct 3 21:52:54 localhost vmunix: [35784.224865] i915 :00:02.0: [drm] GPU HANG: ecode 9:1:85dfbfff, in blender [12755] ``` Is there any chance that we can fix this? Any ideas?
Re: System doesn't boot after recent pull and system reconfigure.
> At commit 189003c83ee04e1819245c861c1ba31597db537d, system reconfigure gives > that warning and reboot fails. I checked the commit and it's superweird that it fails there, the only changes are removals of urls that are dead. I tested the urls and they are actually dead so it shouldn't change anything...
Re: System doesn't boot after recent pull and system reconfigure.
Hi, ‐‐‐ Original Message ‐‐‐ On Tuesday, August 10th, 2021 at 9:36 PM, Raghav Gururajan wrote: > Hello Folks! > > After a recent `guix pull` and `guix system reconfigure`, the boot process > gets > > stuck at the line "Error in finalization thread: success". I also got these > > warnings (https://paste.debian.net/plain/1207184) at the end of system > > reconfigure. > > Guix Describe: https://paste.debian.net/plain/1207192 > > System Configuration: https://paste.debian.net/plain/1207192 You pasted the same link twice so there's no way to see the config. :) Also, did you change anything in the config since the previous reconfigure? Cheers, Ekaitz
Re: Auto-start service
‐‐‐ Original Message ‐‐‐ On Friday, July 9th, 2021 at 12:28 AM, excalamus--- via wrote: > I'm setting up the Emacs daemon following a thread on the mailing list: > https://lists.gnu.org/archive/html/help-guix/2019-11/msg00148.html. I have > services.scm and init.scm defined. I can call `shepherd -c > ~/.config/shepherd/init.scm` and the service starts. I can kill the terminal > and connect a client with `emacsclient -c`. Trouble is, I can't figure out > how to start the service on boot. > > The config files I'm using are: > > ;; services.scm > > (define emacsd > > (make > > #:provides '(emacsd) > > #:start (make-system-constructor "emacs --daemon") > > #:stop (make-system-destructor "emacsclient --eval \"(kill-emacs)\""))) > > ;; init.scm > > ;;; Shepherd User Services > > (load "/home/ahab/.config/shepherd/services.scm") > > (register-services > > emacsd) > > ;; Send shepherd into the background. > > (action 'shepherd 'daemonize) > > ;; Services to start when shepherd starts: > > (for-each start '(emacsd)) > > Based on what (shepherd) Jump Start paragraph 3 says, I'm confused about why > Shepard doesn't find the config files. Is it because Shepherd starts as > superuser on boot? > > Meanwhile, I was advised to define a service in my config. I'm in the > process of reading the documentation (really, figuring out how to set Geiser > up to read the documentation) so that I can understand the relevant parts. > Is the config the generally recommended approach for Guix SD versus the > $HOME/.config/shepherd directory? > > Surely someone else has set up an Emacs daemon service? The linked thread > mentioned this being a good idea for a cookbook recipe. I'd be happy to > contribute to that as I'm confident I'm not alone in wanting an Emacs daemon > service defined. There's a blog post in Guix's blog about this configuration: https://guix.gnu.org/en/blog/2020/gnu-shepherd-user-services/ I remember to have followed that and I have it working. I had to activate the user-level Shepherd by hand in any file that is executed when the system is started. In my case I have it in my i3 config. You can put it in .xsession like the thread you shared suggests. Hope this helps. Ekaitz
Re: Packaging Golang programs
‐‐‐ Original Message ‐‐‐ On Friday, April 30, 2021 3:24 PM, Formbi wrote: > Hi > > I'm trying to package Protonmail's proton-bridge, which is written in Go. In > most cases, it's quite easy to import a package or write the definition > manually. However, some packages are made of sub-packages and the build > system says, for example: > > can't load package: package golang.org/x/mobile: no Go files in (…) > > Those packages are referenced as whole, though. If the sub-packages (with > #:unpack-path) are needed, they don't really work either. For example > github.com/shurcooL/httpfs/vfsutil in the test suite can't find itself: > > package github.com/shurcool/httpfs/vfsutil_test > imports github.com/shurcooL/httpfs/vfsutil: cannot find package > "github.com/shurcooL/httpfs/vfsutil" in any of: > /gnu/store/y5rwacd5l4q26pxis28wsmswj2603hkw-go-1.14.15/src/github.com/shurcooL/httpfs/vfsutil > (from $GOROOT) > /tmp/guix-build-go-github-com-shurcool-httpfs-vfsutil-0.0.0-20190707220628-8d4bc4ba7749.drv-0/src/github.com/shurcooL/httpfs/vfsutil > (from $GOPATH) > FAIL github.com/shurcool/httpfs/vfsutil [setup failed] > FAIL > > Disabling the tests doesn't suffice here, because after it «builds», it can't > be found by a package which needs it. Could you please give me some > directions? > > Kind regards. Hi, I've been waiting until you got an answer here but seems you're out of luck. I'm writing just to tell you that I had the same issue in the past, trying to package the same exact thing and I was unable to solve it. You have my solidarity at least. That being said, we can join forces and make this package. I'll ping some people in Guix's IRC to see if someone can help. It is true that Go programs are pretty crazy with dependencies... Sorry for the late response. Best, Ekaitz
Re: Import package and generate use-module
Hi, ‐‐‐ Original Message ‐‐‐ On Saturday, April 24, 2021 11:26 AM, phodina via wrote: > I'm making a package for octoprint which is imported from PyPI. guix import > -r octoprint generates the template to build the package. > > However, I'm missing the :#use-module ... in order to build the package when > I add (define-module (python-octoprint). Given the fact there are tens of > inputs doing it manually seems tedious and writing a script to parse the > package definition seems like reinventing the wheel as I'm certainly not the > first person to run into this issue. > > So is there a way to automate also this part? > > Let me give you context:If I look at Guix package repository > gnu/packages/python-xyz.scm I see there definitions of packages: > > (define-public python-numpy > (package > (name "python-numpy") > ... > > If I have another package that depends on python-numpy I just provide the > name in e.g. inputs and I don't have to import it with #:use-module within > this file. > > However, when I run guix import pypi PKG I get package definition and when I > run guix build -f python-octoprint.scm I get this error: > > /tmp/python-octoprint.scm:275:6: In procedure propagated-inputs: > error: python-flask: unbound variable > hint: Did you forget `(use-modules (gnu packages python-web))'? > > Of course writing down all the unbound variables by hand is tedious and > error-prone. And since guix gives a hint is there a way to write the > dependency into the file? > > And the question in general is: When I find out that the program is not > packaged and I want to write the package definition and then verify it > installs and runs correctly, what steps should I take? > > guix pull pypi PKG > PKG.scm > > Add 'define-module' > > > > ?? > guix build -f PKG.scm > > Note: Doesn't matter if it's python or rust program. I'm interested in the > workflow > > Kind regards > Petr Hodina > > Sent with ProtonMail Secure Email. I'm not sure if I understand it correctly. What I understand is you need to add the define-module like in any guile file, but you don't need to import every single variable in the module you are importing. You just import the whole module and every `define-public` from the module will be available. See for instance this: https://gitlab.com/ekaitz-zarraga/guix-packages/-/blob/master/electronics.scm#L7 That import provides *all* the public variables from the `python` module, not only one. The case of `python-numpy` you describe works because variables share module scope, so they don't need to import other vars in the same file: they can already see them. In your case you are pretty much fine importing a couple of modules: python-web for flask, and probably python-science or so. The lazy approach is: I try to `guix build` if there's a missing module add it with #:use-modules, as shown in the example. Rinse and repeat. The direct approach is to use tools like `guix show` to check the module where your inputs are described and import them. For example: ``` $ guix show python-flask name: python-flask version: 1.1.2 outputs: out systems: x86_64-linux i686-linux dependencies: python-click@7.1.2 [...] location: gnu/packages/python-web.scm:2946:2 homepage: https://www.palletsprojects.com/p/flask/ license: Modified BSD [...] ``` So now you know it is defined in `python-web`, as it says in the `location` field. Now, you can add the `#:use-modules` form that includes `python-web` to your module definition. If other of your inputs is located in the same module you *don't* need to import it again, because it will be already known by that import. So the point here to clarify is: Use modules don't import PACKAGES, they import THE WHOLE FILE, with all the (public) packages it contains. So it's not as tedious as you suggested.[^1] I hope this helps. Best, Ekaitz [^1]: you can actually choose specific variables to import, exclude some or use a prefix for the imported ones. But our normal usage is just to import the whole file. See: https://www.gnu.org/software/guile/manual/html_node/Creating-Guile-Modules.html and: https://www.gnu.org/software/guile/manual/html_node/Using-Guile-Modules.html
Re: How to debug hanging "check" phase when building Kakoune
Hey! Pretty cool question, let's see if my Bash skills are not lost like tears in the rain. ‐‐‐ Original Message ‐‐‐ On Sunday, April 18, 2021 8:26 AM, Dmitry Matveyev wrote: > > I have assumption that this is somehow connected with read-write > permissions because when I keep failed builds and later try to delete > them, it asks whether I really want to delete some read-only files. > > I have no idea how exactly that `test/run` script works with all the > pipes and output redirection. And I have no idea how to check why this > build phase hangs. Any tips would be really helpful. I agree, it looks like it's related with that. That `cat` you marked there is reading from the file descriptor 4 and discarding its output. That's not a reasonable thing to do in general, *unless* you are reading from a file that is being written to as a pipe (this case is a named pipe, done with `mkfifo`). The `exec 4ui-in` is opening those files and binding them to the file descriptors. So any time you see reading from the fd 4 they are reading from ui-out Pipes are special files that can be used for inter-process communication in this case I think this cat is reading from the pipe until it's empty, in order to consume the data that is being sent by a process. My guess is that the ui-out and ui-in fifos don't have write permission so the process is not able to run, and it hangs in the write, or it's the `cat` who dies in the reading from an empty fifo. It's going to wait until an EOF comes, and that's not happening, because none can write the EOF there. Does this make any sense? HTH, Ekaitz
Re: freecad 19.1 / flatpak
Hi all, This was already merged the other day so you'll have access to the 0.19.1 version of freecad if you `guix pull`. Best, Ekaitz
Re: freecad 19.1 / flatpak
Hi! > Hi Ekaitz, > > Great news: I was able to build the package and use freecad without problems. > I tried several features, addons, with old complex models and new tests for > the occasion. > I could do more tests in the next days if needed. > > Many thanks for your work! > > Christophe That's great! When you are done testing it, please, ping me and I'll prepare the patch for guix. Thanks for your help!
Re: freecad 19.1 / flatpak
Hi Christophe, Yes, I just sent a new version with the package to the thread. The problem was Luis Felipe was able to compile it, probably because his system wasn't as up-to-date as yours. (That's why it worked in my computer too) So I found the same error you found after doing a guix-pull. The new file I shared should work with the latest guix. As I explain in the previous message, the error was related with the update of some packages. More specifically with some graphics libraries and `vtk`, that changed its API in the 9.0.0 version and it doesn't work. Freecad is pretty tied to Debian-based machines so they use old-ish packages for development. It's not the first time we have issues with the packaging due to new versions not being compatible. > Sorry for the news... Software, mate. It is like that all the time. After all this time on the industry, I think I'm getting used to it :)
Re: freecad 19.1 / flatpak
> I think it was related with vtk being updated to 9.0.0.
>
> I'll package it back to 8.2.0 and see what happens.
>
> https://forum.freecadweb.org/viewtopic.php?style=4=4=46934=10
>
> I'll keep you all informed.
> Thanks for the help!
I think I have it working for the latest Guix version.
Please, try with this file.
WARNING: it's going to take ages to compile.
I had to re-package vtk for this case.
There was a problem with the latest version of Guix, because
it includes vtk9.0.0 and freecad is not compatible with that...
Hope this works as expected.
Please tell me if it works fine and if it does I'll submit a
patch
Cheers!(define-module (freecad)
#:use-module (guix download)
#:use-module (guix utils)
#:use-module (srfi srfi-1)
#:use-module (guix packages)
#:use-module (guix git-download)
#:use-module ((guix licenses) #:prefix license:)
#:use-module (guix build-system qt)
#:use-module (gnu packages)
#:use-module (gnu packages algebra)
#:use-module (gnu packages boost)
#:use-module (gnu packages compression)
#:use-module (gnu packages documentation)
#:use-module (gnu packages engineering)
#:use-module (gnu packages fontutils)
#:use-module (gnu packages geo)
#:use-module (gnu packages gl)
#:use-module (gnu packages graphics)
#:use-module (gnu packages graphviz)
#:use-module (gnu packages image)
#:use-module (gnu packages image-processing)
#:use-module (gnu packages maths)
#:use-module (gnu packages mpi)
#:use-module (gnu packages multiprecision)
#:use-module (gnu packages pcre)
#:use-module (gnu packages pkg-config)
#:use-module (gnu packages python)
#:use-module (gnu packages python-xyz)
#:use-module (gnu packages qt)
#:use-module (gnu packages serialization)
#:use-module (gnu packages sqlite)
#:use-module (gnu packages swig)
#:use-module (gnu packages tbb)
#:use-module (gnu packages version-control)
#:use-module (gnu packages xiph)
#:use-module (gnu packages xml)
#:use-module (gnu packages xorg))
;; itksnap needs an older variant of VTK.
(define-public vtk-8
(package (inherit vtk)
(version "8.2.0")
(source (origin
(method url-fetch)
(uri (string-append "https://vtk.org/files/release/;
(version-major+minor version)
"/VTK-" version ".tar.gz"))
(sha256
(base32
"1fspgp8k0myr6p2a6wkc21ldcswb4bvmb484m12mxgk1a9vxrhrl"
(inputs
`(("jsoncpp" ,jsoncpp-for-tensorflow)
,@(alist-delete "jsoncpp" (package-inputs vtk))
(package
(name "freecad-mine")
(version "0.19.1")
(source
(origin
(method git-fetch)
(uri (git-reference
(url "https://github.com/FreeCAD/FreeCAD;)
(commit version)))
(file-name (git-file-name name version))
(sha256
(base32
"0c53q2iawy4yfp11czyc7lbr9ivp3r7v24x4c20myh11wyplffc0"
(build-system qt-build-system)
(native-inputs
`(("doxygen" ,doxygen)
("graphviz" ,graphviz)
("qttools" ,qttools)
("pkg-config" ,pkg-config)
("python-pyside-2-tools" ,python-pyside-2-tools)
("swig" ,swig)))
(inputs
`(("boost" ,boost)
("coin3D" ,coin3D)
("double-conversion" ,double-conversion)
("eigen" ,eigen)
("freetype" ,freetype)
("gl2ps" ,gl2ps)
("glew" ,glew)
("hdf5" ,hdf5-1.10)
("jsoncpp" ,jsoncpp)
("libarea" ,libarea)
("libjpeg-turbo" ,libjpeg-turbo)
("libmedfile" ,libmedfile)
("libspnav" ,libspnav)
("libtheora" ,libtheora)
("libtiff" ,libtiff)
("libxi" ,libxi)
("libxmlplusplus" ,libxmlplusplus)
("libxmu" ,libxmu)
("lz4" ,lz4)
("netcdf" ,netcdf)
("opencascade-occt" ,opencascade-occt)
("openmpi" ,openmpi)
("proj" ,proj)
("python-gitpython" ,python-gitpython)
("python-matplotlib" ,python-matplotlib)
("python-pivy" ,python-pivy)
("python-pyside-2" ,python-pyside-2)
("python-pyyaml" ,python-pyyaml)
("python-shiboken-2" ,python-shiboken-2)
("python-wrapper" ,python-wrapper)
("qtbase" ,qtbase)
("qtsvg" ,qtsvg)
("qtwebkit" ,qtwebkit)
("qtx11extras" ,qtx11extras)
("qtxmlpatterns" ,qtxmlpatterns)
("sqlite" ,sqlite)
("tbb" ,tbb)
("vtk" ,vtk-8)
("xerces-c" ,xerces-c)
("zlib" ,zlib)))
(arguments
`(#:tests? #f
#:configure-flags
(list
"-DBUILD_QT5=ON"
"-DBUILD_FLAT_MESH:BOOL=ON"
(string-append "-DCMAKE_INSTALL_LIBDIR=" (assoc-ref %outputs "out") "/lib")
(string-append "-DPYSIDE2UICBINARY="
(assoc-ref %build-inputs "python-pyside-2-tools")
"/bin/uic")
(string-append "-DPYSIDE2RCCBINARY="
(assoc-ref %build-inputs
Re: freecad 19.1 / flatpak
> > > > > Many thanks! > > > > Unfortunately, the build does not work and has stopped after a few > > > > minutes. > > > > See attached the build log. > > > > Christophe > > > > > > I'll try to compile it again... I wasn't expecting the package to fail. > > > LOL > > > I'll keep you posted, thanks for the log! > > > > Eh, I just built it successfully, and it starts (0.19). > > I'll follow a couple of tutorials and see how it performs. > > I supposed I had an old system so I`guix pulled` and it > doesn't compile anymore. Looks like we've changed the order of > the libraries and now it fails. > > I re-made the package, but now it doesn't compile with some > weird compilation error related with C++ stuff. I can't really > believe what I see. > > Please, try to guix pull and build it again. > > I attach what I have right now, I had to add some deps on top of > what I sent before and it doesn't work :S > > This is crazy :D I think it was related with vtk being updated to 9.0.0. I'll package it back to 8.2.0 and see what happens. https://forum.freecadweb.org/viewtopic.php?style=4=4=46934=10 I'll keep you all informed. Thanks for the help!
Re: freecad 19.1 / flatpak
> > > Many thanks!
> > > Unfortunately, the build does not work and has stopped after a few
> > > minutes.
> > > See attached the build log.
> > > Christophe
> >
> > I'll try to compile it again... I wasn't expecting the package to fail. LOL
> > I'll keep you posted, thanks for the log!
>
> Eh, I just built it successfully, and it starts (0.19).
>
> I'll follow a couple of tutorials and see how it performs.
I supposed I had an old system so I `guix pulled` and it
doesn't compile anymore. Looks like we've changed the order of
the libraries and now it fails.
I re-made the package, but now it doesn't compile with some
weird compilation error related with C++ stuff. I can't really
believe what I see.
Please, try to guix pull and build it again.
I attach what I have right now, I had to add some deps on top of
what I sent before and it doesn't work :S
This is crazy :D(define-module (freecad)
#:use-module (guix packages)
#:use-module (guix git-download)
#:use-module ((guix licenses) #:prefix license:)
#:use-module (guix build-system qt)
#:use-module (gnu packages)
#:use-module (gnu packages algebra)
#:use-module (gnu packages boost)
#:use-module (gnu packages compression)
#:use-module (gnu packages documentation)
#:use-module (gnu packages engineering)
#:use-module (gnu packages fontutils)
#:use-module (gnu packages geo)
#:use-module (gnu packages gl)
#:use-module (gnu packages graphics)
#:use-module (gnu packages graphviz)
#:use-module (gnu packages image)
#:use-module (gnu packages image-processing)
#:use-module (gnu packages maths)
#:use-module (gnu packages mpi)
#:use-module (gnu packages multiprecision)
#:use-module (gnu packages pcre)
#:use-module (gnu packages pkg-config)
#:use-module (gnu packages python)
#:use-module (gnu packages python-xyz)
#:use-module (gnu packages qt)
#:use-module (gnu packages serialization)
#:use-module (gnu packages sqlite)
#:use-module (gnu packages swig)
#:use-module (gnu packages tbb)
#:use-module (gnu packages version-control)
#:use-module (gnu packages xiph)
#:use-module (gnu packages xml)
#:use-module (gnu packages xorg))
(package
(name "freecad-mine")
(version "0.19.1")
(source
(origin
(method git-fetch)
(uri (git-reference
(url "https://github.com/FreeCAD/FreeCAD;)
(commit version)))
(file-name (git-file-name name version))
(sha256
(base32
"0c53q2iawy4yfp11czyc7lbr9ivp3r7v24x4c20myh11wyplffc0"
(build-system qt-build-system)
(native-inputs
`(("doxygen" ,doxygen)
("graphviz" ,graphviz)
("qttools" ,qttools)
("pkg-config" ,pkg-config)
("python-pyside-2-tools" ,python-pyside-2-tools)
("swig" ,swig)))
(inputs
`(("boost" ,boost)
("coin3D" ,coin3D)
("double-conversion" ,double-conversion)
("eigen" ,eigen)
("freetype" ,freetype)
("gl2ps" ,gl2ps)
("glew" ,glew)
("hdf5" ,hdf5-1.10)
("jsoncpp" ,jsoncpp)
("libarea" ,libarea)
("libjpeg-turbo" ,libjpeg-turbo)
("libmedfile" ,libmedfile)
("libspnav" ,libspnav)
("libtheora" ,libtheora)
("libtiff" ,libtiff)
("libxi" ,libxi)
("libxmlplusplus" ,libxmlplusplus)
("libxmu" ,libxmu)
("lz4" ,lz4)
("netcdf" ,netcdf)
("opencascade-occt" ,opencascade-occt)
("openmpi" ,openmpi)
("proj" ,proj)
("python-gitpython" ,python-gitpython)
("python-matplotlib" ,python-matplotlib)
("python-pivy" ,python-pivy)
("python-pyside-2" ,python-pyside-2)
("python-pyyaml" ,python-pyyaml)
("python-shiboken-2" ,python-shiboken-2)
("python-wrapper" ,python-wrapper)
("qtbase" ,qtbase)
("qtsvg" ,qtsvg)
("qtwebkit" ,qtwebkit)
("qtx11extras" ,qtx11extras)
("qtxmlpatterns" ,qtxmlpatterns)
("sqlite" ,sqlite)
("tbb" ,tbb)
("vtk" ,vtk)
("xerces-c" ,xerces-c)
("zlib" ,zlib)))
(arguments
`(#:tests? #f
#:configure-flags
(list
"-DBUILD_QT5=ON"
"-DBUILD_FLAT_MESH:BOOL=ON"
(string-append "-DCMAKE_INSTALL_LIBDIR=" (assoc-ref %outputs "out") "/lib")
(string-append "-DPYSIDE2UICBINARY="
(assoc-ref %build-inputs "python-pyside-2-tools")
"/bin/uic")
(string-append "-DPYSIDE2RCCBINARY="
(assoc-ref %build-inputs "python-pyside-2-tools")
"/bin/rcc")
"-DPYSIDE_LIBRARY=PySide2::pyside2"
(string-append
"-DPYSIDE_INCLUDE_DIR="
(assoc-ref %build-inputs "python-pyside-2") "/include;"
(assoc-ref %build-inputs "python-pyside-2") "/include/PySide2;"
(assoc-ref %build-inputs "python-pyside-2") "/include/PySide2/QtCore;"
(assoc-ref %build-inputs "python-pyside-2")
Re: freecad 19.1 / flatpak
> Many thanks! > Unfortunately, the build does not work and has stopped after a few minutes. > See attached the build log. > > Christophe I'll try to compile it again... I wasn't expecting the package to fail. LOL I'll keep you posted, thanks for the log!
Re: freecad 19.1 / flatpak
Hi,
You can try with the file attached.
You just need to run the following, considering `freecad.scm` is the attached
file, and that you are in the folder that contains the file.
```
guix build -f freecad.scm
```
Like half an hour later, you'll get an output similar to this:
```
$ guix build -f freecad.scm
>>TONS OF COMPILATION RELATED NONSENSE<<
/gnu/store/74c1nrijb1x29pj9f94265zaq8frqf4f-freecad-0.19.1
```
That's the folder where the result of the build is. You can run that freecad
using the path you got like this:
```
$ /gnu/store/74c1nrijb1x29pj9f94265zaq8frqf4f-freecad-0.19.1/bin/FreeCAD
```
Tell me if it works and how does it look.
If it works fine I'll send a patch to Guix with this package and you'll soon
have it in Guix. Also, if you need it now, you can use the file I gave you and
install it from there with the same -f trick, or even create an environment for
it, that way it will be easier to run.
For any question you have, just ask.
Good luck!
Ekaitz(define-module (freecad)
#:use-module (guix packages)
#:use-module (guix git-download)
#:use-module ((guix licenses) #:prefix license:)
#:use-module (guix build-system qt)
#:use-module (gnu packages)
#:use-module (gnu packages engineering)
#:use-module (gnu packages algebra)
#:use-module (gnu packages boost)
#:use-module (gnu packages fontutils)
#:use-module (gnu packages gl)
#:use-module (gnu packages xorg)
#:use-module (gnu packages tbb)
#:use-module (gnu packages image-processing)
#:use-module (gnu packages xml)
#:use-module (gnu packages documentation)
#:use-module (gnu packages graphviz)
#:use-module (gnu packages pkg-config)
#:use-module (gnu packages swig)
#:use-module (gnu packages compression)
#:use-module (gnu packages graphics)
#:use-module (gnu packages maths)
#:use-module (gnu packages multiprecision)
#:use-module (gnu packages mpi)
#:use-module (gnu packages pcre)
#:use-module (gnu packages python)
#:use-module (gnu packages python-xyz)
#:use-module (gnu packages qt)
#:use-module (gnu packages version-control))
(define-public freecad
(package
(name "freecad")
(version "0.19.1")
(source
(origin
(method git-fetch)
(uri (git-reference
(url "https://github.com/FreeCAD/FreeCAD;)
(commit version)))
(file-name (git-file-name name version))
(sha256
(base32
"0c53q2iawy4yfp11czyc7lbr9ivp3r7v24x4c20myh11wyplffc0"
(build-system qt-build-system)
(native-inputs
`(("doxygen" ,doxygen)
("graphviz" ,graphviz)
("qttools" ,qttools)
("pkg-config" ,pkg-config)
("python-pyside-2-tools" ,python-pyside-2-tools)
("swig" ,swig)))
(inputs
`(("boost" ,boost)
("coin3D" ,coin3D)
("eigen" ,eigen)
("freetype" ,freetype)
("glew" ,glew)
("hdf5" ,hdf5-1.10)
("libarea" ,libarea)
("libmedfile" ,libmedfile)
("libspnav" ,libspnav)
("libxi" ,libxi)
("libxmu" ,libxmu)
("openmpi" ,openmpi)
("opencascade-occt" ,opencascade-occt)
("python-matplotlib" ,python-matplotlib)
("python-pyside-2" ,python-pyside-2)
("python-shiboken-2" ,python-shiboken-2)
("python-pivy" ,python-pivy)
("python-wrapper" ,python-wrapper)
("python-gitpython" ,python-gitpython)
("python-pyyaml" ,python-pyyaml)
("qtbase" ,qtbase)
("qtsvg" ,qtsvg)
("qtx11extras" ,qtx11extras)
("qtxmlpatterns" ,qtxmlpatterns)
("qtwebkit" ,qtwebkit)
("tbb" ,tbb)
("vtk" ,vtk)
("xerces-c" ,xerces-c)
("zlib" ,zlib)))
(arguments
`(#:tests? #f
#:configure-flags
(list
"-DBUILD_QT5=ON"
(string-append "-DCMAKE_INSTALL_LIBDIR=" (assoc-ref %outputs "out") "/lib")
(string-append "-DPYSIDE2UICBINARY="
(assoc-ref %build-inputs "python-pyside-2-tools")
"/bin/uic")
(string-append "-DPYSIDE2RCCBINARY="
(assoc-ref %build-inputs "python-pyside-2-tools")
"/bin/rcc")
"-DPYSIDE_LIBRARY=PySide2::pyside2"
(string-append
"-DPYSIDE_INCLUDE_DIR="
(assoc-ref %build-inputs "python-pyside-2") "/include;"
(assoc-ref %build-inputs "python-pyside-2") "/include/PySide2;"
(assoc-ref %build-inputs "python-pyside-2") "/include/PySide2/QtCore;"
(assoc-ref %build-inputs "python-pyside-2") "/include/PySide2/QtWidgets;"
(assoc-ref %build-inputs "python-pyside-2") "/include/PySide2/QtGui;")
"-DSHIBOKEN_LIBRARY=Shiboken2::libshiboken"
(string-append "-DSHIBOKEN_INCLUDE_DIR="
(assoc-ref %build-inputs "python-shiboken-2")
Re: freecad 19.1 / flatpak
Hi all,
I had some free time and I decided to spend it on updating the
freecad package to the 0.19.1 version.
I managed it to compile with the changes I add below, but I
prefer someone to test it thoroughly because it's a program
I don't use that much so I don't know if the workbenches are
working as expected, or if the addons work.
Pleas Christophe, could you install try it and see if it works?
This is the diff on the current guix repository, I hope that's
enough to let you use it. If its not, I can make a separate
package for you.
Cheers,
Ekaitz
diff --git a/gnu/packages/engineering.scm b/gnu/packages/engineering.scm
index 8d7ba9c1da..8e7974f1f1 100644
--- a/gnu/packages/engineering.scm
+++ b/gnu/packages/engineering.scm
@@ -2351,21 +2351,19 @@ OpenSCAD code. It supports syntax highlighting,
indenting and refilling of
comments.")))
(define-public freecad
- (let ((commit-ref "7616153b3c31ace006169cdc2fdafab484498858")
-(revision "1"))
-(package
+ (package
(name "freecad")
- (version (git-version "0.18.5" revision commit-ref))
+ (version "0.19.1")
(source
(origin
(method git-fetch)
(uri (git-reference
(url "https://github.com/FreeCAD/FreeCAD;)
- (commit commit-ref)))
+ (commit version)))
(file-name (git-file-name name version))
(sha256
(base32
- "16965yxnp2pq7nm8z3p0pjkzjdyq62vfrj8j3nk26bwc898czyn2"
+ "0c53q2iawy4yfp11czyc7lbr9ivp3r7v24x4c20myh11wyplffc0"
(build-system qt-build-system)
(native-inputs
`(("doxygen" ,doxygen)
@@ -2393,6 +2391,8 @@ comments.")))
("python-shiboken-2" ,python-shiboken-2)
("python-pivy" ,python-pivy)
("python-wrapper" ,python-wrapper)
+ ("python-gitpython" ,python-gitpython)
+ ("python-pyyaml" ,python-pyyaml)
("qtbase" ,qtbase)
("qtsvg" ,qtsvg)
("qtx11extras" ,qtx11extras)
@@ -2454,7 +2454,7 @@ customization.")
license:lgpl2.1+
license:lgpl2.0+
license:gpl3+
-license:bsd-3)
+license:bsd-3
(define-public libmedfile
(package
Re: freecad 19.1 / flatpak
hi, > Ok, I will try. > If I understand correctly, we will communicate privately, won't we? > I'll start by reading the document on the definition of package at guix. > > Christophe You can send me the questions privately if you want, yes. If you are not used to package stuff on Guix, you'll lose time reading the package. It's a pretty complex one. You should report to freecad that the flatpak version doesn't provide plugin support. They may fix it fast and you'll have a working freecad until we fix the package definition on the guix side. Best, Ekaitz
Re: freecad 19.1 / flatpak
‐‐‐ Original Message ‐‐‐ On Saturday, April 3, 2021 9:10 PM, Christophe Pisteur wrote: > Yes, there is a reason: ignorance, incompetence, idleness ;-) Don't be so hard on yourself. > More seriously: I have no computer skills, I just know how to do: guix > package -i freecad... > > it seems to me to be very far from what would allow me to start from the > current definition package > > to make a freecad 19.1 package. Isn't it? At the moment we have the 18.5 version of freecad available. Surprising or not, I was the guy who updated it to that version (and fixed the package because it wasn't even working) and I have bad news for you. Freecad is very hard to compile. It has tons of dependencies and all of them require special effort (some are broken, have broken compilation tools...). Maybe they changed that in the new version but... I don't think so. I spent a hella lot of time trying to fix it. I will take a look to the Nix package and try to convert it to Guix but it will take time. If anyone is interested on taking a look... https://github.com/NixOS/nixpkgs/blob/3ab6e84b79abc45855e666fff33078189b4960c3/pkgs/applications/graphics/freecad/default.nix > That said, why not give it a try if someone has the patience to answer the > > countless questions I'm sure to have... with no guarantee of success. > > See if the collective effort is worth it. I can (try to) answer every single question you have so please, don't be shy. Best, Ekaitz
Re: freecad 19.1 / flatpak
Hi there,
On Wednesday, March 31, 2021 11:15 AM, Christophe Pisteur
wrote:
> Hello,
>
> After having failed to install freecad from an Appimage on guix system
> (1), a new version of freecad (freecad 19.1) has just been realised and
> proposed on flathub (2).
>
> Unfortunately, the addon manager does not work with this version of
> freecad on flathub on guix system. But it is in the addons that we find
> one of the main functions until a mature version can be included in the
> main project: the assembly modules (A2+, Assembly2, Assembly4).
>
> Here are the error messages:
>
> 1. When launching freecad from terminal:
> $ flatpak run org.freecadweb.FreeCAD
> FreeCAD 0.19, Libs: 0.19R14555 (Git shallow)
> © Juergen Riegel, Werner Mayer, Yorik van Havre and others 2001-2021
> FreeCAD is free and open-source software licensed under the terms of
> LGPL2+ license.
> FreeCAD wouldn't be possible without FreeCAD community.
>
>
> # ###
>
> # # # #
>
>
>
> ## # # # #
>
> =
>
> # # # # # # # # #
>
> # # # # #
>
>
>
> # # # # # # # # ## ##
>
> ==
>
> # ### # # ## ##
>
> ===
>
> Traceback (most recent call last):
> File "/app/freecad/Mod/Tux/NavigationIndicatorGui.py", line 55, in
> changeEvent
> onTooltip()
> NameError: name 'onTooltip' is not defined
> connect failed: Aucun fichier ou dossier de ce type
>
> 2) When you click on the "install" button from the addon manager:
> Traceback (most recent call last):
> File "/app/freecad/Mod/AddonManager/addonmanager_workers.py", line
> 769, in run
> repo = git.Repo.clone_from(self.repos[idx][1], clonedir)
> File "/app/lib/python3.8/site-packages/git/repo/base.py", line 1032,
> in clone_from
> return cls._clone(git, url, to_path, GitCmdObjectDB, progress,
> multi_options, **kwargs)
> File "/app/lib/python3.8/site-packages/git/repo/base.py", line 973,
> in _clone
> finalize_process(proc, stderr=stderr)
> File "/app/lib/python3.8/site-packages/git/util.py", line 329, in
> finalize_process
> proc.wait(**kwargs)
> File "/app/lib/python3.8/site-packages/git/cmd.py", line 408, in wait
> raise GitCommandError(self.args, status, errstr)
> git.exc.GitCommandError: Cmd('git') failed due to: exit code(128)
> cmdline: git clone -v https://github.com/Zolko-123/FreeCAD_Assembly4
> /home/christophe/.FreeCAD//Mod/Assembly4
> stderr: 'Cloning into '/home/christophe/.FreeCAD//Mod/Assembly4'...
> fatal: unable to find remote helper for 'https'
>
> I tried with several addons: same results.
>
> Any Idea?
>
> Christophe
>
> (1) https://lists.gnu.org/archive/html/help-guix/2021-02/msg00035.html
> (2) https://flathub.org/apps/details/org.freecadweb.FreeCAD
According to a random blogpost in the wild you can see this:
https://blog.hostonnet.com/git-unable-to-find-remote-helper-for-https
Looks like your git is not installed with SSL support. Probably
the package comes with its own git and it's not capable enough.
Try to install it separately or something. I'm not sure if
flatpak is going to use it or not, because I have no idea about
how flatpak works, but the problem you have is because of the
git version your freecad is using.
Hope that helps,
Ekaitz
Re: guix pull errors
Before the `reconfigure` I had a 10 days old system. Now it seems to work fine. I had a hard time make it upgrade tbh, it was like stuck all the time with the same error. So it was just a network issue and it is fixed now, right? Great! Good to know, thanks!
Re: Blender freezes i3wm
Weirdly enough, yesterday I sculpted for 2h in a row and everything went well. After that, I opened the model I worked on and it freezed my screen again, but this time some buttons of other screens started to appear on top of the frozen window. Very weird. Anyway, this looks like it's impossible to debug so I'm going to act as it didn't happen.
guix pull errors
Hi, I'm getting tons of guix pull errors these days. They look like they are related with network issues because they disappear if I insist (many times) and they appear in different packages. They look like this: ``` guix/ui.scm:2164:12: In procedure run-guix-command: Bad Read-Header-Line header: # substitution of /gnu/store/c39c4jhgg1pq3sw083pcmp42gba05kr2-evince-3.36.5 failed ``` I'm finding them all the time. I think they are very cryptic... Maybe we can do something about that. Best, Ekaitz
Re: Blender freezes i3wm
‐‐‐ Original Message ‐‐‐ On Saturday, March 27, 2021 8:33 PM, Luis Felipe wrote: > On Saturday, March 27, 2021 2:45 PM, Luis Felipe > luis.felipe...@protonmail.com wrote: > > > I can try it here to see if I get the same result, but I haven't used > > Blender in a long time and never did sculpting. > > I just tried the following and couldn't reproduce the issue: > > 1. Start Blender > 2. Delete the default cube > 3. Add a UV Sphere (Shift+A → Mesh → UV Sphere) > 4. Click the Sculpting tab > 5. Click and drag on the sphere to sculpt I tried to replicate my own issue and I can't. So it looks like it is solved. I don't know. Good news then! Thanks to all for the help and the support! Much appreciated. Ekaitz
Re: Blender freezes i3wm
Btw, I can actually exit the session and stuff like that but when I'm back it's still frozen. Does this say something to you?
Re: Blender freezes i3wm
Hi all, I just tried with Gnome Desktop and I was able to replicate the same issue, it could be related with the hardware or with X11? I needed to push it a little bit further this time but I remember to sculpt the models I used for the testing in this actual device or the previous laptop, which was just older and lower quality... I am pretty sure this didn't happen in older versions like the 2.7... :( The way I manage to replicate is entering the Sculpting mode and rotating the model. It just freezes the screen (but not the mouse pointer!). Just to test I managed to use my muscle memory to run a terminal and reboot the computer using hotkeys and imagining how should the screen look like. It actually rebooted the computer as expected. Also I see no increment on the processing or the temperature of the CPU. So it's like the window manager or something has issues to manage this but I don't get what kind of debugging information can I obtain from the fact that the mouse pointer can be moved around. Does anyone have any suggestion on that? > I haven't used it for a long time in a while, so, not sure. I'll do > some more modeling soon, I'll try to push it then. > I'm using Sway currently, which seems rather stable. Maybe try that? > Don't even need to reconfigure, just guix environment --ad-hoc sway > (and maybe some more packages) and dbus-run-session sway on a different > vtty. I'll give it a try to this too, just in case it's only related with X. Thanks for the advice. I'll keep you posted!
Re: Blender freezes i3wm
Hi! > I may have ran into something similar, but didn't record the exact > error at the time. > What guix commit are you on? And are you using drivers from The > Forbidden Channel? I've worked on Blender 2.83.9 and Blender 2.91.0 and both had the same issue. I have to admit that I didn't update the window manager in a while. And yes, I use the drivers from the forbidden channel. Were you able to fix the issue?
Blender freezes i3wm
Hi, I'm recently finding issues running blender. Trying to sculpt a simple shape and rotate the view freezes i3wm but the mouse cursor is still responsive (it even detects when it's on top of a text area or a clickable area) but the screen is frozen and I can open terminals using ctrl-alt-F[2-6] that are actually responsive. Did you experience anything similar? I'm on an intel i7-10510U cpu with the integrated graphics card (that is more than capable to handle such models). Any idea of what can I do to fix this? Thanks! Ekaitz
Re: setup c/c++ development environment
Hi, ‐‐‐ Original Message ‐‐‐ On Thursday, March 11, 2021 9:03 PM, Fredrik Salomonsson wrote: > > > Hi Ekaitz, > > CC:ing the rest of the help-guix mailing list, I hope you don't mind. Yeah sorry I messed up with the email. > Ekaitz Zarraga eka...@elenq.tech writes: > > > I have a question with all this you shared. > > Is this compiling only the needed parts or it recompiles the whole > > directory from scratch every time you run `guix build`? > > To my knowledge it will recompile everything every time you run > `guix build`. > > So this method might be better suited for CI (continuous integration). > > But benefit with the `guix.scm` file is that you can also use the same > file to setup the environment you need to build your package. > > `guix environment -l guix.scm` > or > `guix environment -l build-aux/guix.scm` > > Depending how you have layed out the guix.scm file. > > That will use guix to setup the environment needed to build the your > package. And in that shell you can use: > > meson build > meson compile -C build > meson test -C build > > Or equivalent for other build systems. Which will support partial > builds. > > Note that guix might have issues with build generated files for some > build systems when mixing guix environment and guix build. E.g. run > `guix environment`, generate the build files and then run `guix build`. > It seems fine with meson, even when not filtering out the build files. > But I have had issues with GNU autotools and permissions, > `make distclean` before calling `guix build` is a quick workaround. A > better option is probably to filter out those files when collecting the > source. > > I hope that answers your question. So the workflow with guix build is not very good for development but for CI and stuff like that because you are going to rebuild everything. Thanks for the explanations. Ekaitz
Re: Android development in Guix
> This is terrible, but it is the only way I found to run it somewhat properly. > Running it in an environment also isolates it from the system and prevents it > from doing any damage to my home directory or my personal data. I hope this > is useful to you! LOL! > HTH :) It totally does. Thanks a lot :)
Android development in Guix
Hi, I'm trying to setup some Android development tooling in my Guix machine and I wonder if someone has done it before. I tried, even if I don't like it, to run android studio using OpenJDK and Icedtea and I'm getting this error: ``` JDK Required: Cannot load a JDK class: com.sun.jdi.Field Please ensure you run the IDE on JDK rather than JRE. ``` This is the perfect excuse to use something else that fits better with our philosophy but I'm not sure about what to do. Does anyone have a solution to the error or an alternative that is more acceptable? I'm open to any suggestion you have to make Android development: programming languages, tooling... I don't really have any precondition. I'm just looking for the most free-software friendly way to make simple and reasonable Android development that fits Guix. Thanks!
Re: Defining custom services in /etc/config.scm
‐‐‐ Original Message ‐‐‐
On Sunday, January 24, 2021 8:09 PM, Sergiu Ivanov wrote:
> Hello,
>
> I'm trying to start redshift [0] as a shepherd service, and this is how
> I defined the redshift service in my /etc/config.scm:
>
> (define evry-geolocation "48.63681:2.42777")
> (define redshift-service
> (shepherd-service
> (provision '(redshift))
> (requirement '(xorg-server))
> (documentation "Start Redshift as a service")
> (start #~(make-forkexec-constructor
> `("redshift" "-l" ,evry-geolocation)))
> (stop #~(make-kill-destructor
>
> However, I can't find a way to add this to the services field of the
> operating-system declaration.
>
> When I try:
>
> (services (append
> (list redshift-service
> (service xfce-desktop-service-type)
> (set-xorg-configuration
> (xorg-configuration (keyboard-layout keyboard-layout
> %desktop-services))
>
> I get the message "Wrong type argument #< ...".
>
> I understand that I should somehow construct a (?) from my
> , but I don't see a way to do it.
>
> I've also seen a couple mailing list discussions [1,2] and a blog post
> [3] about doing what I want in "pure shepherd", i.e. independently of
> Guix, but these solutions seem to be doing more than what I need.
>
> How can I add a custom shepherd service to the operating-system
> definition in /etc/config.scm?
>
> -
>
> Sergiu
>
> [0] http://jonls.dk/redshift/
> [1] https://lists.gnu.org/archive/html/guix-devel/2018-02/msg00047.html
> [2] https://www.mail-archive.com/guix-devel@gnu.org/msg46278.html
> [3] https://guix.gnu.org/blog/2020/gnu-shepherd-user-services/
Hi Sergiu,
I recently made my redshift configuration using shepherd user
services, which is what people in IRC recommended me.
You can find it here:
https://github.com/ekaitz-zarraga/dotfiles/tree/guix/shepherd/shepherd
Also, it forces me to run the shepherd as a user using my i3 config.
I also wanted to configure it in my config.scm file but people
convinced me not to do so.
I'm not sure why your case shouldn't work though.
Best,
Ekaitz
Re: How to define my dns servers in resolv
здравствуйте, I'll try to do my best to help but I'm not the correct person for it. I'll just try to give you some guidance you can follow to find more info. ‐‐‐ Original Message ‐‐‐ On Saturday, January 16, 2021 2:40 PM, znavko--- via wrote: > Hello! I cannot agree limitations my provider setup > with the orders of Russian government to forbid access > to this or that website, for example, > protonmail.com, tutanota.com, vpnbook.com are prohibited. > Does any civilized country do such a harm? I won't even try to answer to this. > I wish to use my own DNS servers from free and secure services. > I want to start just to type them into /etc/resolv.conf You probably can. > And may be further actions will be dnsmasq with wpa_supplicant > with your help. > > Can I deal with resolv.conf without dnsmasq? Yes, I think so, but I'm not sure about how all this network stack works: `network-manager-configuration` lets you choose a dns field. In that you can say if you want the network manager to edit resolv.conf or not. That might be helpful. See: https://guix.gnu.org/manual/en/html_node/Networking-Services.html#Networking-Services > Can you advise me manual page or example where resolv.conf is > configured?? I did not find it except notes in dnsmasq configurations. You may be able to set it with `etc-service-type` ? Good luck :)
Re: Problem on packaging qt-light extension for Chicken
Hi, > --8<---cut here---start->8--- > phase `unpack' succeeded after 0.0 seconds starting phase`patch-usr-bin-file' > phase `patch-usr-bin-file' succeeded after 0.0 seconds starting > phase`patch-source-shebangs' > patch-shebang: ./qt-light/chicken-compile-qt-extension.scm: changing > `/bin/sh' > to`/gnu/store/pwcp239kjf7lnj5i4lkdzcfcxwcfyk72-bash-minimal-5.0.16/bin/sh' > phase `patch-source-shebangs' succeeded after 0.0 seconds starting > phase`setup-environment' > phase `setup-environment' succeeded after 0.0 seconds starting phase`build' > > Error: extension not cached > command "chicken-install" "-cached" "-no-install" "qt-light" failed with > status 70 > --8<---cut here---end--->8--- > When I saw the message I was puzzled and tried to work around your issues with no success. I talked with who started the chicken buildsystem, and I was told the error is related with some dependencies missing. The error message should be clearer on this but it makes sense, the non-cached extension is not qt-light but its deps. The egg says: --8<---cut here---start->8--- (dependencies bind protobj matchable miscmacros shell) --8<---cut here---end--->8--- So you probably need to package that first. Hope this helps. Good luck! Ekaitz
Re: Custom package, specifying cmake source directory?
Hi again, Let me re-answer because I think I screwed in a couple of points. So, you are not trying to use other build directory but a different source directory. That's a little bit different. Instead of making a `chdir` you need to touch the `srcdir` variable. In that case, you can change the current directory in a previous phase, adding a new phase before the `configure` step that just jumps to source code just like `hypre` does. Run `guix edit hypre` to see it: ``` scheme (add-before 'configure 'chdir-src (lambda _ (chdir "src"))) ``` Probably this last option is the best, I'm not sure if it works. It should. Sorry for the misunderstanding! Good luck, Ekaitz
Re: Custom package, specifying cmake source directory?
Hi Matthew, ‐‐‐ Original Message ‐‐‐ On Monday, December 14, 2020 3:55 AM, Matthew Brooks wrote: > I'm working on a custom package definition, but when it gets built the cmake > files are in "source/src" instead of "source", and so the build system errors > out when trying to run cmake. How do I specify the source directory for the > cmake build system? > From grepping around in the definition for the cmake build system, it appears > that it expects the "source" variable to point to the appropriate location, > but I'm not at all familiar with scheme or guile, and I have no idea if > that's correct or how to change it. I'll try to answer you. I'm never played with this, but I think we can find something together. Cmake buildsystem has an interesting keyword argument called #:out-of-source? that chooses to compile the project inside sources or in a separate build directory. You need something similar, but not that. I tracked down that variable and I finally got into Cmake's configure step: I'm reading the contents of the file `guix/build/cmake-build-system.scm`. There you can find the definition of the `configure` step. You don't really need to do a lot in you package. You can replace the configure step with your own version of it, using the `sources/src` directory instead of what the default version uses. In order to replace the configure phase you need to do add a `#:phases` `argument` to your package and use a `modify-phases` call to `replace 'configure` with your own. (all the parts marked with backticks are literal code, grep them in the gnu/packages/ directory and you'll find examples). Your version could have the exact same contents that the default but with different directory. You just need to make a `chdir` to the place you like, just like this block (line 46 in guix/build/cmake-build-system.scm) does: ``` scheme (when out-of-source? (mkdir "../build") (chdir "../build")) (format #t "build directory: ~s~%" (getcwd)) ``` Remove all of it and change it with `(chdir "src")` or something like that and it should work. That's most of it. I just gave you a high-level explanation but you can ask me more if you need. Just ask anything you need. Best, Ekaitz
Re: Submitting new packages
‐‐‐ Original Message ‐‐‐ On Thursday, November 5, 2020 12:03 PM, Barnabás Béres wrote: > Hey Guixers, > > I’m quite new to Guix but I would like to contribute. I created two font > packages (fort awesome and nerd fonts). Can you please help my how can send > in and later maintain my packages? > > Thanks for your help! > > Best regards, > Barnabas Beres Hi, Pretty exciting moment! Your first contribution! There's some documentation about contributing here: https://guix.gnu.org/manual/en/html_node/Contributing.html But contributing is "as simple as" creating a patch and sending it to guix-patches mailing list. You can use `git format-patch` or `git send-mail` for that. That's pretty much it, if you have specific questions please don't hesitate to ask. The documentation is quite clear but it lacks some weird details you might miss in the first pass but we can try to help you here or in guix-devel mailing list too. Hope this helps. Best, Ekaitz
Re: Help with .deb, .rpm and guix packaging.
‐‐‐ Original Message ‐‐‐ On Thursday, October 29, 2020 5:34 PM, Aniket Patil wrote: > Hi, > There are GNU/Linux distros which are Debian-based or fedora(Redhat) based > etc. What I know to my knowledge is GNU packages are .tar packages with > .deb and .rpm file format, am I right? What I am wondering about how guix > makes it happen? I mean are we bypassing the process of making the package > in .deb and .rpm format and releasing it, or what exactly are we doing? To > my knowledge, we are building from source code as we can see in the recipe. > So that means we are bypassing the process of making packages explicitly > for the distro, aren't we? > > Where can I learn more about this? I am going through guix manual but I > don't know which chapter to look for. Also, where can I learn more about > packaging software on GNU/Linux distribution? Like .deb or .rpm > > Regards, > Aniket. Hi Aniket, In the distros you mention, software is pre-compiled and the .tar packages have a directory structure that defines where do the packaged files need to be installed. Those packages don't define how the software has to be compiled and they don't include the sources. Guix is not bypassing those systems: Guix is one of those systems. Guix is a package manager. The main difference is that Guix describes how is the source compiled and installed in a recipe. When the user installs a package using: guix install packagename The package sources are downloaded, compiled and installed in the system, following the recipe. Guix also has ways to keep packages separated to avoid dependency collisions, ad-hoc environments, time traveling, package installation history with a reasonable rollback system... and some other interesting features. Guix is also a software distribution built on top of the Guix package manager that extends those interesting tools further and defines a uniform way to describe your operating system. Hope this helps. If you have further questions, don't hesitate to ask them. Regards, Ekaitz
Re: Guix Bluetooth Headset
ElenQ Technology Ethical Innovation ‐‐‐ Original Message ‐‐‐ On Saturday, September 26, 2020 4:04 PM, Guillaume Le Vaillant wrote: > > > Hi, > > I have a headset that can be used either with a cord or with Bluetooth, > and I got it to almost work with Bluetooth. The only problem I have is > that I want it to use the A2DP profile but most of the time it decides > to use the HSP profile and refuses to switch to A2DP, therefore the > sound is not as good. > > In my 'config.scm' file I had to modify the pulseaudio service to use > a custom configuration file (in attachment): > > --8<---cut here---start->8--- > (operating-system > ... > (services > (cons* ... > (bluetooth-service) > (modify-services %desktop-services > (pulseaudio-service-type config => > > (pulseaudio-configuration >(inherit config) >(script-file (local-file "/etc/guix/default.pa" > > > --8<---cut here---end--->8--- > > I'm also using the vanilla Linux kernel instead of Linux-libre, but I'm > not sure if it makes a difference for Bluetooth headsets. > > HTH! Thanks for example! The main problem I see to this approach is that it doesn't look very accessible... I'll try to make it work in my speaker. Best, Ekaitz
Re: Guix Bluetooth Headset
Well... I don't think both goals are incompatible. I'm totally fine with Guix being Free Software only distribution and I encourage the project to keep that path because I strongly believe it's the right thing to do. That doesn't mean the user needs to have a deep understanding of the software layers involved in a system. I'd say it's unrelated. The issues with the device compatibility is a different story that should get better at some point, and there's NonGuix out there too that can solve some issues for those users that can accept binary blobs. The problem with the bluetooth for instance it's not a matter of hardware support neither. I have the drivers for the device installed, but I have no clue about how to configure the software to make all the different pieces work together. That's something we can improve without compromising the Ethical values of Guix, which are, in my opinion unquestionable at this point. This case of the sound system is a clear example of a software stack that have been too complex for ages in every single distro but distros managed well enough to hide that complexity from the users (because they were no actual reasons to expose it) and Guix didn't manage to do it that well. I'm 100% sure there's a working configuration we can already use, but it's *really* hard to find it. That discoverability issue is what carries most of the friction with the user in Guix. Of course, in my opinion. Hope we can use this thread to improve this point with some documentation, code or packages. I'm open to help in any of those areas. Best, Ekaitz ‐‐‐ Original Message ‐‐‐ On Saturday, September 26, 2020 3:17 PM, Yasuaki Kudo wrote: > I totally agree - while diverging a little from my original bluetooth > question, it seems that we need to have an "unofficial" "Guix-Compromised" > downstream distribution. > > The frustrating thing is that with the zero-tolerance attitude of Guix > project (while well-meaning and I applaud such high standards), it is > difficult to even discuss such things as tactical abandonment of > libre-software purity in order dramatically improve the actual end-user > experience. > > We only effectively discuss delivering the best possible user-experience > (while ignoring the ethical questions) when we can employ all possible means > including using bad software form companies we dislike. > > Having said that, I think the Guix' stance is actually very defendable and we > would be barking up the wrong tree to complain to the Guix team. > > This is why I think we need a proper downstream distribution of Guix with the > goal of delivering the best Linux experience albeit with perhaps degraded > software purity. > > -Yasu > > > On Sep 26, 2020, at 21:03, Ekaitz Zarraga eka...@elenq.tech wrote: > > Hi all, > > I'm a full-time Guix user but I don't really think I have average > > requirements for my computing. > > I tried to connect to a bluetooth speaker once, there's a thread in the > > mailing lists about it... I didn't manage to do it. > > I tried through Gnome first, and it literally doesn't show any Bluetooth > > device. I was able to use some bluetooth controllers like Bluez or so and I > > managed to pair the device (and Gnome still didn't recognize the bluetooth > > but did recognize the pairing, amazing stuff), but I didn't manage to make > > the music sound through it. > > I'd like to help more but this is what I got. > > This is the kind of things that are discouraging. Guix is a good thing but > > the user experience is sometimes plainly horrible, for stuff that should be > > simple. > > > > > I am curious how many people actually use Guix for the user-facing > > > computers - do people use Ubuntu something and use Guix as a mere package > > > manager or install Guix on a headless (virtual) computer and remotely use > > > it? > > > > I am curious too, because I only use Guix for the only computer I own, but > > I'm not sure if it could be reasonable for anyone else, mostly because > > sometimes Guix users are supposed to know too much about internal software > > layers and there's no help about those times where you don't know enough. > > This bluetooth case is very accurate case to explain this kind of > > situation. Sound has a very weird software stack, if you combine it with > > bluetooth everything gets complicated. If you don't know about it you are > > in trouble. > > Sorry for the depressing thoughts, but this is how I feel. > > I hope you find your way to make it work and share it. > > Best, > > Ekaitz > > ‐‐‐ Original Message ‐‐‐ > > > > > On Friday, September 25, 20
Re: Guix Bluetooth Headset
Hi all, I'm a full-time Guix user but I don't really think I have average requirements for my computing. I tried to connect to a bluetooth speaker once, there's a thread in the mailing lists about it... I didn't manage to do it. I tried through Gnome first, and it literally doesn't show any Bluetooth device. I was able to use some bluetooth controllers like Bluez or so and I managed to pair the device (and Gnome still didn't recognize the bluetooth but did recognize the pairing, amazing stuff), but I didn't manage to make the music sound through it. I'd like to help more but this is what I got. This is the kind of things that are discouraging. Guix is a good thing but the user experience is sometimes plainly horrible, for stuff that should be simple. > I am curious how many people actually use Guix for the user-facing computers > - do people use Ubuntu something and use Guix as a mere package manager or > install Guix on a headless (virtual) computer and remotely use it? I am curious too, because I only use Guix for the only computer I own, but I'm not sure if it could be reasonable for anyone else, mostly because sometimes Guix users are supposed to know too much about internal software layers and there's no help about those times where you don't know enough. This bluetooth case is very accurate case to explain this kind of situation. Sound has a very weird software stack, if you combine it with bluetooth everything gets complicated. If you don't know about it you are in trouble. Sorry for the depressing thoughts, but this is how I feel. I hope you find your way to make it work and share it. Best, Ekaitz ‐‐‐ Original Message ‐‐‐ On Friday, September 25, 2020 4:16 PM, Yasuaki Kudo wrote: > Hi Joshua, > > My bluetooth USB device itself seems to be recognized by the kernel but there > is this whole layers of software I don't understand - (when I used to use > NixOS on the same hardware, it worked - I think it was a combination of > something called PulseAudi and Bluetooth) > > So I just want to know whether someone actually has a working configuration I > can just copy. > > I am curious how many people actually use Guix for the user-facing computers > - do people use Ubuntu something and use Guix as a mere package manager or > install Guix on a headless (virtual) computer and remotely use it? > > Cheers, > Yasu > > > On Sep 25, 2020, at 21:53, Joshua Branson jbra...@dismail.de wrote: > > Have you tried looking at h-node.org? > > That will give you some hints at what hardware works well with > > GNU/Linux. > > I personally have a RIG headset with integrated microphone. It was a > > bit of an impulse buy. It doesn't use bluetooth. It has one chord. I > > can use it to record my voice, or listen to music. But I can't seem to > > do both at the same time with it. :( > > Thanks, > > Joshua > > -- > > Joshua Branson > > Sent from Emacs and Gnus > > https://gnucode.me
Re: nnn with icons
‐‐‐ Original Message ‐‐‐ On Thursday, August 20, 2020 1:51 PM, Fulbert wrote: > On Thu, Aug 20, 2020 at 11:21:30AM +0000, Ekaitz Zarraga wrote: > > > Hi, > > Hello Ekaitz > > > Is anyone here using nnn terminal file browser? > > I don't, but… > > > I installed but I don't see icons. Do you know what should I do to make the > > icons work well with it? > > It says¹ "To see the icons correctly, install icons-in-terminal²." > ¹ https://asciinema.org/a/353811 > ² https://github.com/sebastiencs/icons-in-terminal Yes but as nnn is packaged for guix I supposed someone uses it and has the icons set. The icons-in-terminal package is not packaged for guix (or I can't find it)... > > Thanks! > > You are welcome ! Thanks again for your time!
nnn with icons
Hi, Is anyone here using nnn terminal file browser? I installed but I don't see icons. Do you know what should I do to make the icons work well with it? Thanks!
Re: what about custom licenses?
Hi, You can check guix/licenses.scm file. It contains how licenses are defined in Guix so you can build your own. There are projects with custom licenses there like wxwidgets. I hope this helps. Best, Ekaitz ‐‐‐ Original Message ‐‐‐ On Saturday, August 1, 2020 11:44 AM, Adam Kandur via wrote: > > > hi everyone! i packed some cl libs. but they have custom (foss) licenses. how > i can describe it in guix?
Re: Icecat - WebGL support
‐‐‐ Original Message ‐‐‐
On Sunday, June 7, 2020 12:59 PM, Ekaitz Zarraga wrote:
>
>
> ‐‐‐ Original Message ‐‐‐
> On Saturday, June 6, 2020 9:27 PM, Julien Lepiller jul...@lepiller.eu wrote:
>
> > Le 6 juin 2020 15:14:07 GMT-04:00, Ekaitz Zarraga eka...@elenq.tech a écrit
> > :
> >
> > > Wait a minute! It's working!
> > > But the only way to work is with the LD_LIBRARY_PATH set. I installed
> > > mesa and set the LD_LIBRARY_PATH from my command line, an icecat run
> > > from the command line works, the one run from the desktop environment
> > > doesn't.
> > > So... guix is not setting the LD_LIBRARY_PATH correctly for Mesa and
> > > that's why IceCat is unable to find it?
> >
> > I see icecat is wrapped for pulseaudio to work, so we might as well wrap
> > mesa.
>
> It already has Mesa as an input, too.
> Looks like it's just a matter of the LD_LIBRARY_PATH not being set.
>
> Thoughts?
Hi all,
I made this change in the package and it seems to work.
If it looks fine I'll send the patch.
Thanks!
diff --git a/gnu/packages/gnuzilla.scm b/gnu/packages/gnuzilla.scm
index 7bf9850e97..570398ecd9 100644
--- a/gnu/packages/gnuzilla.scm
+++ b/gnu/packages/gnuzilla.scm
@@ -1112,11 +1112,13 @@ from forcing GEXP-PROMISE."
(lib (string-append out "/lib"))
(gtk (assoc-ref inputs "gtk+"))
(gtk-share (string-append gtk "/share"))
+(mesa (assoc-ref inputs "mesa"))
+(mesa-lib (string-append mesa "/lib"))
(pulseaudio (assoc-ref inputs "pulseaudio"))
(pulseaudio-lib (string-append pulseaudio "/lib")))
(wrap-program (car (find-files lib "^icecat$"))
`("XDG_DATA_DIRS" prefix (,gtk-share))
- `("LD_LIBRARY_PATH" prefix (,pulseaudio-lib)))
+ `("LD_LIBRARY_PATH" prefix (,pulseaudio-lib ,mesa-lib)))
#t))
(home-page "https://www.gnu.org/software/gnuzilla/;)
(synopsis "Entirely free browser derived from Mozilla Firefox")
Re: Icecat - WebGL support
‐‐‐ Original Message ‐‐‐
On Thursday, June 11, 2020 10:42 PM, Julien Lepiller wrote:
> Le 11 juin 2020 16:26:19 GMT-04:00, Ekaitz Zarraga eka...@elenq.tech a écrit :
>
> > ‐‐‐ Original Message ‐‐‐
> > On Sunday, June 7, 2020 12:59 PM, Ekaitz Zarraga eka...@elenq.tech
> > wrote:
> >
> > > ‐‐‐ Original Message ‐‐‐
> > > On Saturday, June 6, 2020 9:27 PM, Julien Lepiller jul...@lepiller.eu
> > > wrote:
> > >
> > > > Le 6 juin 2020 15:14:07 GMT-04:00, Ekaitz Zarraga eka...@elenq.tech
> > > > a écrit :
> > >
> > > > > Wait a minute! It's working!
> > > > > But the only way to work is with the LD_LIBRARY_PATH set. I
> > > > > installed
> > >
> > > > > mesa and set the LD_LIBRARY_PATH from my command line, an icecat
> > > > > run
> > >
> > > > > from the command line works, the one run from the desktop
> > > > > environment
> > >
> > > > > doesn't.
> > > > > So... guix is not setting the LD_LIBRARY_PATH correctly for Mesa
> > > > > and
> > >
> > > > > that's why IceCat is unable to find it?
> > > >
> > > > I see icecat is wrapped for pulseaudio to work, so we might as well
> > > > wrap mesa.
> > >
> > > It already has Mesa as an input, too.
> > > Looks like it's just a matter of the LD_LIBRARY_PATH not being set.
> > > Thoughts?
> >
> > Hi all,
> > I made this change in the package and it seems to work.
> > If it looks fine I'll send the patch.
> > Thanks!
> > diff --git a/gnu/packages/gnuzilla.scm b/gnu/packages/gnuzilla.scm
> > index 7bf9850e97..570398ecd9 100644
> > --- a/gnu/packages/gnuzilla.scm
> > +++ b/gnu/packages/gnuzilla.scm
> > @@ -1112,11 +1112,13 @@ from forcing GEXP-PROMISE."
> > (lib (string-append out "/lib"))
> > (gtk (assoc-ref inputs "gtk+"))
> > (gtk-share (string-append gtk "/share"))
> >
> > - (mesa (assoc-ref inputs "mesa"))
> >
> >
> > - (mesa-lib (string-append mesa "/lib"))
> > (pulseaudio (assoc-ref inputs "pulseaudio"))
> > (pulseaudio-lib (string-append pulseaudio "/lib")))
> > (wrap-program (car (find-files lib "^icecat$"))
> >`("XDG_DATA_DIRS" prefix (,gtk-share))
> >
> >
> >
> > - `("LD_LIBRARY_PATH" prefix (,pulseaudio-lib)))
> >
> >
> >
> > - `("LD_LIBRARY_PATH" prefix (,pulseaudio-lib
> >
> >
> >
> > ,mesa-lib)))
> > #t))
> > (home-page "https://www.gnu.org/software/gnuzilla/;)
> > (synopsis "Entirely free browser derived from Mozilla Firefox")
>
> Looks great! Please send the patch :)
Done! #41815
Re: Icecat - WebGL support
‐‐‐ Original Message ‐‐‐ On Saturday, June 6, 2020 9:27 PM, Julien Lepiller wrote: > Le 6 juin 2020 15:14:07 GMT-04:00, Ekaitz Zarraga eka...@elenq.tech a écrit : > > > Wait a minute! It's working! > > But the only way to work is with the LD_LIBRARY_PATH set. I installed > > mesa and set the LD_LIBRARY_PATH from my command line, an icecat run > > from the command line works, the one run from the desktop environment > > doesn't. > > So... guix is not setting the LD_LIBRARY_PATH correctly for Mesa and > > that's why IceCat is unable to find it? > > I see icecat is wrapped for pulseaudio to work, so we might as well wrap mesa. It already has Mesa as an input, too. Looks like it's just a matter of the LD_LIBRARY_PATH not being set. Thoughts?
Re: Icecat - WebGL support
Wait a minute! It's working! But the only way to work is with the LD_LIBRARY_PATH set. I installed mesa and set the LD_LIBRARY_PATH from my command line, an icecat run from the command line works, the one run from the desktop environment doesn't. So... guix is not setting the LD_LIBRARY_PATH correctly for Mesa and that's why IceCat is unable to find it?
Re: Icecat - WebGL support
‐‐‐ Original Message ‐‐‐ On Saturday, June 6, 2020 6:07 PM, Tobias Geerinckx-Rice wrote: > Ekaitz, > > Ekaitz Zarraga 写道： > > > I'm unable to use WebGL in IceCat. I've activated all the > about:config > > options related to it and also the preferences > checkbox. > > Here's[0] a bug report about this, and a solution that let me Enter the Cube > — from a fresh profile. Unfortunately, it's implied there's some forgotten > magic setting on my system that makes it work. > > > On the other hand, ungoogled-chromium handles WebGL by default > in my > > machine. No settings changed. > > Horrible… > > Kind regards, > > T G-R > > [0]: http://issues.guix.gnu.org/34135#10 Hi, Thanks for taking the time to answer. I tried everything in the issue there and it doesn't work. I don't like chromium but looks like I need to use it for certain things. I'd like to help correcting IceCat's issue but I don't even know how to start. This makes me sad. Best, Ekaitz
Icecat - WebGL support
Hi, I'm unable to use WebGL in IceCat. I've activated all the about:config options related to it and also the preferences checkbox. But when I visit https://get.webgl.org/ it always says: "Hmm. While your browser seems to support WebGL, it is disabled or unavailable. If possible, please ensure that you are running the latest drivers for your video card." On the other hand, ungoogled-chromium handles WebGL by default in my machine. No settings changed. What am I doing wrong? ElenQ Technology Ethical Innovation
Re: ffmpeg and alsa/pulseaudio
‐‐‐ Original Message ‐‐‐ On Friday, May 29, 2020 3:34 AM, Leo Famulari wrote: > On Tue, Apr 07, 2020 at 11:54:35AM +0000, Ekaitz Zarraga wrote: > > > Hi, > > I'm trying to make a screencast with ffmpeg and I'm having issues. > > I had a screen recording script on the past, based on alsa, but it's not > > working in guix because it looks like ffmpeg wasn't compiled with alsa. Is > > that true? Is there something I missing to activate alsa on ffmpeg? > > This is the command I was running: > > ffmpeg \ > > -y -f x11grab -draw_mouse 1 -framerate 25 -video_size 1920x1080 -i $DISPLAY > > \ > > -f alsa -ac 2 -i default \ > > -pix_fmt yuv420p -c:v libx264 -preset veryfast \ > > -c:a libmp3lame -q:v 1 -s 1920x1080 -f matroska \ > > output.mkv > > I tried this but it failed with ":0: Invalid argument". That's the > $DISPLAY variable. Do you know what it should say? Hi Leo, Thanks for the answer. I have :1 in DISPLAY and it's not failing because of that. More on the issue: I had to use a different microphone for this just because of this issue. Looks like when I use a USB microphone and use pulseaudio engine it breaks somehow and records audio with wrong timestamps. An analog microphone works well. With alsa I found a working configuration but it doesn't work :( Thanks!
Thunderbolt support in GuixSD?
Hi, Does anyone use a computer with a thunderbolt port in GuixSD? Is there support for that? Looks like the kernel already has support for them and freedesktop has developed a daemon for it: https://gitlab.freedesktop.org/bolt/bolt But doesn't look like it's packaged for Guix. Is it? Thanks!
GUIX_LOCPATH message over and over again
Hi all, I made a guix pull the other day and since then I'm getting the classic: guile: warning: failed to install locale hint: Consider installing the `glibc-utf8-locales' or `glibc-locales' package and defining `GUIX_LOCPATH', along these lines: guix package -i glibc-utf8-locales export GUIX_LOCPATH="$HOME/.guix-profile/lib/locale" See the "Application Setup" section in the manual, for mo... But what it says doesn't solve anything. The error persists. What am I missing? IIRC it went away in previous time just doing what the error message says. Also, isn't there any way to make it just work in GuixSD? Thanks! ElenQ Technology Ethical Innovation
Re: Guix on the MNT Reform
‐‐‐ Original Message ‐‐‐ On Friday, May 8, 2020 5:06 PM, Christopher Lemmer Webber wrote: > I'm very excited to see the MNT Reform launch: > > https://www.crowdsupply.com/mnt/reform > https://www.crowdsupply.com/mnt/reform/updates/the-campaign-is-live > > Completely hackable laptop; all the designs (that are possible) are > libre, and you can even 3d print to replace many of the parts. > > However my impression is that running Guix on ARM is not necessarily > straightforward. I haven't tried it yet. > > Most important specs: > > CPU: NXP/Freescale i.MX8MQ with 4x ARM Cortex-A53 cores (1.5 GHz), > 1x Cortex-M4F core. CPU and RAM are on exchangeable SO-DIMM sized > module. > RAM: 4GB LPDDR4 memory > GPU: Vivante GC7000Lite GPU with mainline Linux drivers and OpenGL > 2.1, ES 2.0 > > (My other main worry is: that's pretty light on RAM these days...! > But hey, maybe incentive for me to cut the fat in the programs I use a > bit more...) > > I am thinking of getting one and running Guix on it. It would be nice > to know that I wouldn't be alone. :) Anyone else planning to get one > and joing me on the journey on maybe the most hacker-oriented laptop > ever? > > - Chris Hey Chris, It's a great project, I've been checking the schematics and stuff myself and I liked it a lot. It's a great reference design to build on top of. I can't afford one atm but I can do my best to help you play with it. I'm really interested on if it's able to run Guix and I'd like to know about your experience with it. Please keep me posted on your journey! Best, Ekaitz
Re: Guixsd and meshlab ?
> > I see you submitted it in the meantime: > > https://issues.guix.gnu.org/issue/40492 > > Almost there! :-) > > Thanks, > Ludo’. Yeah, still have to fight a little with it. Thanks!
ffmpeg and alsa/pulseaudio
Hi, I'm trying to make a screencast with ffmpeg and I'm having issues. I had a screen recording script on the past, based on alsa, but it's not working in guix because it looks like ffmpeg wasn't compiled with alsa. Is that true? Is there something I missing to activate alsa on ffmpeg? This is the command I was running: ffmpeg \ -y -f x11grab -draw_mouse 1 -framerate 25 -video_size 1920x1080 -i $DISPLAY \ -f alsa -ac 2 -i default \ -pix_fmt yuv420p -c:v libx264 -preset veryfast \ -c:a libmp3lame -q:v 1 -s 1920x1080 -f matroska \ output.mkv I tried to make it with pulseaudio replacing alsa by pulse in the command but the audio is distorted. It sounds very fast. Any idea? Do you use ffmpeg for screen recording too? Thanks
Re: Guixsd and meshlab ?
‐‐‐ Original Message ‐‐‐ On Tuesday, April 7, 2020 1:25 AM, Olivier Farges wrote: > Hi Guix, > > has anyone ever had the opportunity to install meshlab and could possibly > provide a corresponding .scm file? > > Regards, > > - > > Olivier Farges Hi Olivier, I didn't but I'm interested on it and I would take the time to make one. If you are not in a hurry and can wait a couple of days I can make the package for you. Regards
Re: Bluetooth + GNOME issues
> Reading the bluetooth-service documentation I see a part that says, "Users > need to be in the ‘lp’ group to access the D-Bus service." The user defined > in your config.scm doesn't seem to be in that group. I don't have a clue, but > could that be the problem? I changed my group and the problem persists. I can see the bluetooth in the system tray but the Settings don't show any device and "Not interested in Bluez" keeps happening. How isn't it interested? I don't understand.
Re: Bluetooth + GNOME issues
‐‐‐ Original Message ‐‐‐ On Thursday, March 5, 2020 4:52 PM, sirgazil wrote: > On Thu, 05 Mar 2020 05:10:44 -0500 Ekaitz zarragaeka...@elenq.tech wrote > > > > Hi there > > > > > > I'm having issues to make bluetooth work with GNOME that I don't know how > > to fix. > > > > > > I managed to run bluez and even connect to a speaker via bluetoothctl. I > > was unable to make it sound though, because it doesn't appear as a possible > > audio output. > > > > > > This is my current configuration: > > > > > > https://gitlab.com/ekaitz-zarraga/guix-configuration/-/blob/master/config.scm > > > > > > Am I missing something? > > This seems like a known issue. Maybe > checkhttps://issues.guix.gnu.org/issue/32166 to confirm? It's the same thing I'd say but as I'm using a bluetooth speaker I'm not able to make the it sound. I don't really care about having a settings menu if there's a way for me to make it sound but I'm unable to find it.
Re: Bluetooth + GNOME issues
‐‐‐ Original Message ‐‐‐ On Thursday, March 5, 2020 2:47 PM, Pierre Neidhardt wrote: > Can you connect to your dongle manually? > > See > https://wiki.archlinux.org/index.php/Bluetooth > for detailed instructions. > > > > Pierre Neidhardt > https://ambrevar.xyz/ Yes I actually did via `bluetoothctl` and it worked. Once I was connected GNOME recognized the connection but the settings didn't show it. It only appeared in the system tray. Also I was unable to make it work, because the device (it's a speaker) didn't appear as an extra audio output.
Re: Bluetooth + GNOME issues
‐‐‐ Original Message ‐‐‐ On Thursday, March 5, 2020 10:10 AM, Ekaitz Zarraga wrote: > Hi there > > I'm having issues to make bluetooth work with GNOME that I don't know how to > fix. > > I managed to run bluez and even connect to a speaker via bluetoothctl. I was > unable to make it sound though, because it doesn't appear as a possible audio > output. > > This is my current configuration: > > https://gitlab.com/ekaitz-zarraga/guix-configuration/-/blob/master/config.scm > > Am I missing something? > > Thanks! Just to add some extra information. Running gnome-control-center makes these messages. It's working, its finding all the stuff but I see nothing in the screen but "No bluetooth available, please insert a dongle" Ekaitz@tuxedo ~$ gnome-control-center -v 15:12:57.0053 Bluetooth:DEBUG: Not adding device 'BlueZ 5.52' 15:12:57.0053 Bluetooth:DEBUG: Not interested in device 'BlueZ 5.52' 15:12:57.0053 Bluetooth:DEBUG: Default adapter changed to: /org/bluez/hci0 15:12:57.0363bluetooth-cc-panel:DEBUG: Updating airplane mode: BluetoothHasAirplaneMode 0, BluetoothHardwareAirplaneMode 0, BluetoothAirplaneMode 0, AirplaneMode 0 15:12:57.0363bluetooth-cc-panel:DEBUG: No Bluetooth available 15:12:57.0363 Bluetooth:DEBUG: Default adapter changed to: /org/bluez/hci0 15:12:57.0364bluetooth-cc-panel:DEBUG: Updating airplane mode: BluetoothHasAirplaneMode 0, BluetoothHardwareAirplaneMode 0, BluetoothAirplaneMode 0, AirplaneMode 0 15:12:57.0364bluetooth-cc-panel:DEBUG: No Bluetooth available 15:12:57.0365 Bluetooth:DEBUG: Unhandled UUID 694d6174-7444-4553-4341-886b0f59922c (0x694d6174) 15:12:57.0365 Bluetooth:DEBUG: Adding device (null) (/org/bluez/hci0/dev_88_6B_0F_59_92_2C) 15:12:57.0365 Bluetooth:DEBUG: Saving device type Unknown for 88:6B:0F:59:92:2C 15:12:57.0365 Bluetooth:DEBUG: Adding device AY-SPKF4 (/org/bluez/hci0/dev_F4_4E_FD_63_1F_44) 15:12:57.0365 Bluetooth:DEBUG: Saving device type Headset for F4:4E:FD:63:1F:44 15:12:57.0366 Bluetooth:DEBUG: Not interested in device 'BlueZ 5.52' 15:12:57.0366 Bluetooth:DEBUG: Not interested in device 'BlueZ 5.52' 15:12:57.0366 Bluetooth:DEBUG: obexd appeared, registering agent 15:12:57.0369 Bluetooth:DEBUG: Not interested in device 'BlueZ 5.52' 15:12:57.0369 Bluetooth:DEBUG: Not interested in device 'BlueZ 5.52' 15:12:58.0862 Bluetooth:DEBUG: Adding device (null) (/org/bluez/hci0/dev_58_AB_68_2A_49_91) 15:12:58.0862 Bluetooth:DEBUG: Saving device type Unknown for 58:AB:68:2A:49:91 15:12:59.0232 Bluetooth:DEBUG: Adding device (null) (/org/bluez/hci0/dev_D0_03_DF_C9_03_13) 15:12:59.0232 Bluetooth:DEBUG: Saving device type Unknown for D0:03:DF:C9:03:13 15:12:59.0401 Bluetooth:DEBUG: Unhandled property: rssi (gnome-control-center:1552): dconf-DEBUG: 15:12:59.869: unwatch_fast: "/org/gnome/control-center/" (active: 1, establishing: 0) (gnome-control-center:1552): dconf-DEBUG: 15:12:59.870: sync 15:12:59.0870 cc-object-storage:DEBUG: Destroying cached objects Ekaitz@tuxedo ~$ gnome-control-center -v 15:13:25.0859 Bluetooth:DEBUG: Not adding device 'BlueZ 5.52' 15:13:25.0859 Bluetooth:DEBUG: Not interested in device 'BlueZ 5.52' 15:13:25.0859 Bluetooth:DEBUG: Default adapter changed to: /org/bluez/hci0 15:13:26.0363bluetooth-cc-panel:DEBUG: Updating airplane mode: BluetoothHasAirplaneMode 0, BluetoothHardwareAirplaneMode 0, BluetoothAirplaneMode 0, AirplaneMode 0 15:13:26.0363bluetooth-cc-panel:DEBUG: No Bluetooth available 15:13:26.0363 Bluetooth:DEBUG: Default adapter changed to: /org/bluez/hci0 15:13:26.0364bluetooth-cc-panel:DEBUG: Updating airplane mode: BluetoothHasAirplaneMode 0, BluetoothHardwareAirplaneMode 0, BluetoothAirplaneMode 0, AirplaneMode 0 15:13:26.0364bluetooth-cc-panel:DEBUG: No Bluetooth available 15:13:26.0364 Bluetooth:DEBUG: Unhandled UUID 694d6174-7444-4553-4341-886b0f59922c (0x694d6174) 15:13:26.0364 Bluetooth:DEBUG: Adding device (null) (/org/bluez/hci0/dev_88_6B_0F_59_92_2C) 15:13:26.0364 Bluetooth:DEBUG: Saving device type Unknown for 88:6B:0F:59:92:2C 15:13:26.0365 Bluetooth:DEBUG: Adding device AY-SPKF4 (/org/bluez/hci0/dev_F4_4E_FD_63_1F_44) 15:13:26.0365 Bluetooth:DEBUG: Saving device type Headset for F4:4E:FD:63:1F:44 15:13:26.0365 Bluetooth:DEBUG:
Bluetooth + GNOME issues
Hi there I'm having issues to make bluetooth work with GNOME that I don't know how to fix. I managed to run bluez and even connect to a speaker via bluetoothctl. I was unable to make it sound though, because it doesn't appear as a possible audio output. This is my current configuration: https://gitlab.com/ekaitz-zarraga/guix-configuration/-/blob/master/config.scm Am I missing something? Thanks!
Re: AVR toolchain and AVR libc
> I found the bug report I was looking for. Hope that helps! > > https://issues.guix.info/issue/24416 Hi, Thanks for the link, but I'm not sure if that solves anything I didn't solve by the proposition in the bug report I opened: https://issues.guix.info/issue/39794 The idea was that taking the link you sent me in the past, it says "we don't use CROSS_* variables anymore". I understand that in the sense of "You have to use CPATH instead" so that's the change I proposed but I'm not sure if that's the best way to solve the issue or if I should follow the path proposed in the link. Anyway, a second problem appears when it tries to find `stubs32.h`, which is supposed to be part of multilib... so it should be part of avr-libc in this case but it isn't even installed (I searched for it in my Guix profile with no success, but I find `stubs.h`). Now I need to find why isn't it installed and make it work. Any idea about how to make `stubs32.h` to be installed? I think that's the only thing missing.
Re: AVR toolchain and AVR libc
BTW, is there any resource point to why the package of the avr-toolchain 5.5 is broken? I may spend some time trying to make it work. ElenQ Technology Ethical Innovation ‐‐‐ Original Message ‐‐‐ On Sunday, February 23, 2020 6:54 PM, Ekaitz Zarraga wrote: > > This is the compiler treating the strict prototype warning as an > > > error. You could get around this by using the ALLOW_WARNINGS=yes > > makeflag supported by the Makefile of the QMK firmware project. > > make ALLOW_WARNINGS=yes > > It worked! > Thanks a lot.
Re: AVR toolchain and AVR libc
> This is the compiler treating the strict prototype warning as an > error. You could get around this by using the ALLOW_WARNINGS=yes > makeflag supported by the Makefile of the QMK firmware project. > > make ALLOW_WARNINGS=yes > It worked! Thanks a lot.
