[bug#66870] Review

[bug#66870] Review

[Gabriel Wicki]

Hi

So i've de-frizzled the patches by renaming #66870 to "[PATCH 0/6] Add
yabridge."

[PATCH 1/6] looks good to me!  Have you checked the packages depending
on asio still build and work?

You can check the dependents with this command:

$ guix refresh -l asio
Building the following 8 packages would ensure 16 dependent packages are rebuilt: emacs-ob-sclang@0.1-1.cd3f3c8 cl-collider@2018.7.15-0.a469088 ecl-cl-collider@2018.7.15-0.a469088 gnome-arcade@0.240 widelands@1.1 musikcube@3.0.1 jami@20230323.0 restbed@4.8

I checked and they all still build fine :)

[PATCH 2/6] LGTM.

[PATCH 3/6] Looks goot, although I am a bit confused WRT the comment for
the (recursive #t) clause in the (source ..) field (indicating we need
to recursively copy the repo for the tests to complete) and argument
(#:tests? #f) indicating that the tests wouldn't be run.

[PATCH 4/6] LGTM.

[PATCH 5/6] LGTM.

[PATCH 6/6] Wow, now this is a big one!  Apart from the following
comments this looks good to me!

 - I'm not sure whether your (let ((arch))) clause isn't an abuse;  you
test whether (%current-system) is either x86_64-linux or aarch64-linux
and set it to x86_64-unix and i386-unix otherwise.  Does this actually
work for aarch64-linux systems?  You could make use of the
target-64-bit? function in (guix utils) and only allow the package to
build on x86 systems.  Also, you inherit wine64's supported-systems
field - which only lists x86_64-linux.  Building for i368 seems to be
unsupported by our wine packages (or wine in general?)

 - I'm a bit confused about your patch series adding clap 1.1.9 -
although yabridge seems to need 1.1.7.  Why not add clap@1.1.7 with this
series and with the next upgrade of yabridge also update clap?
 
 - Similar with input vst3sdk: Why not add it as a separate package?
If I interpret the situation correctly, using default definitions of
these inputs could also let you simplify your native-inputs list to the
new style.



All packages built fine and running $(guix style) didn't change
anything.
$(guix lint) on the patched packages gave the following output:

~/src/guix/gnu/packages/networking.scm:3431:15: asio@1.28.0: permanent
redirect from https://think-async.com/Asio to
https://think-async.com/Asio/
fetching CVE database for 2023...
~/src/guix/gnu/packages/networking.scm:3417:5: asio@1.28.0:
source not archived on Software Heritage and missing from the Disarchive
database
~/src/guix/gnu/packages/cpp.scm:2311:13: function2@4.2.3:
can be upgraded to 4.2.4
~/src/guix/gnu/packages/audio.scm:2724:13: clap@1.1.9: can
be upgraded to 1.1.10
~/src/guix/gnu/packages/audio.scm:6002:0: yabridge@5.0.5:
line 6002 is way too long (95 characters)
~/src/guix/gnu/packages/audio.scm:6013:0: yabridge@5.0.5:
line 6013 is way too long (100 characters)



Please send in a updated patch-set.

Thanks for your time and effort!
g

[bug#66870] Review

[sughosha via Guix-patches via]

Hi,

Thanks for the review.

> So i've de-frizzled the patches by renaming #66870 to "[PATCH 0/6] Add
> yabridge."

Thank you.

> [PATCH 1/6] looks good to me! Have you checked the packages depending
> on asio still build and work?

I checked only musikcube, jami and widelands and they worked fine. 
Others
must work as well.

> [PATCH 3/6] Looks goot, although I am a bit confused WRT the comment 
> for
> the (recursive #t) clause in the (source ..) field (indicating we need
> to recursively copy the repo for the tests to complete) and argument
> (#:tests? #f) indicating that the tests wouldn't be run.

Removed "(recursive #t)".

> - I'm not sure whether your (let ((arch))) clause isn't an abuse; you
> test whether (%current-system) is either x86_64-linux or aarch64-linux
> and set it to x86_64-unix and i386-unix otherwise. Does this actually
> work for aarch64-linux systems? You could make use of the
> target-64-bit? function in (guix utils) and only allow the package to
> build on x86 systems. Also, you inherit wine64's supported-systems
> field - which only lists x86_64-linux. Building for i368 seems to be
> unsupported by our wine packages (or wine in general?)

Removed the substitute as it is no more needed.

> - I'm a bit confused about your patch series adding clap 1.1.9 -
> although yabridge seems to need 1.1.7. Why not add clap@1.1.7 with this
> series and with the next upgrade of yabridge also update clap?

Moved clap@1.1.7 to a seperate definition.

> - Similar with input vst3sdk: Why not add it as a separate package?
> If I interpret the situation correctly, using default definitions of
> these inputs could also let you simplify your native-inputs list to the
> new style.

This vst3sdk is not the original one. It is found in a different
repository than the original one and is suppose to clone as subproject. 
So
it does not make sense to package it in a seperate definition.

> $(guix lint) on the patched packages gave the following output:
> 
> ~/src/guix/gnu/packages/networking.scm:3431:15: asio@1.28.0: permanent
> redirect from https://think-async.com/Asio to
> https://think-async.com/Asio/
> fetching CVE database for 2023...
> ~/src/guix/gnu/packages/networking.scm:3417:5: asio@1.28.0:
> source not archived on Swith "(let ((arch)))"oftware Heritage and 
> missing from the Disarchive
> database
> ~/src/guix/gnu/packages/cpp.scm:2311:13: function2@4.2.3:
> can be upgraded to 4.2.4
> ~/src/guix/gnu/packages/audio.scm:2724:13: clap@1.1.9: can
> be upgraded to 1.1.10
> ~/src/guix/gnu/packages/audio.scm:6002:0: yabridge@5.0.5:
> line 6002 is way too long (95 characters)
> ~/src/guix/gnu/packages/audio.scm:6013:0: yabridge@5.0.5:
> line 6013 is way too long (100 characters)

Fixed and updated these packages.

I will submit the v2 patches with the above changes.

Regards,
Sughosha

[bug#66870] Review

[Sughosha via Guix-patches via]

[-- Attachment #1: Type: text/plain, Size: 204 bytes --]

A small thing to remove in [PATCH v2 7/7]:
in yabridge definition, "(modules '((guix build utils)))" inside "(source)" is not needed. If you can remove it before merging, it would be great.

Thank you.

[-- Attachment #2: Type: text/html, Size: 273 bytes --]