This fixes build in qt514 case.
The usual way here is to provide patches for each qt5 version
separately. No other module adds them in this generic way.
The problem is when you combine the approaches; qtModule will only
take the list from the module and ignore the version-specific list.
At some point, I'd like to make another attempt at
71f1f4884b ("openssl: stop static binaries referencing libs"), which
was reverted in 195c7da07d. One problem with my previous attempt is
that I moved OpenSSL's libraries to a lib output, but many dependent
packages were hardcoding the out output as the location of the
libraries. This patch fixes every such case I could find in the tree.
It won't have any effect immediately, but will mean these packages
will automatically use an OpenSSL lib output if it is reintroduced in
future.
This patch should cause very few rebuilds, because it shouldn't make
any change at all to most packages I'm touching. The few rebuilds
that are introduced come from when I've changed a package builder not
to use variable names like openssl.out in scripts / substitution
patterns, which would be confusing since they don't hardcode the
output any more.
I started by making the following global replacements:
${pkgs.openssl.out}/lib -> ${lib.getLib pkgs.openssl}/lib
${openssl.out}/lib -> ${lib.getLib openssl}/lib
Then I removed the ".out" suffix when part of the argument to
lib.makeLibraryPath, since that function uses lib.getLib internally.
Then I fixed up cases where openssl was part of the -L flag to the
compiler/linker, since that unambigously is referring to libraries.
Then I manually investigated and fixed the following packages:
- pycurl
- citrix-workspace
- ppp
- wraith
- unbound
- gambit
- acl2
I'm reasonably confindent in my fixes for all of them.
For acl2, since the openssl library paths are manually provided above
anyway, I don't think openssl is required separately as a build input
at all. Removing it doesn't make a difference to the output size, the
file list, or the closure.
I've tested evaluation with the OfBorg meta checks, to protect against
introducing evaluation failures.
python38 appears to work just as well, so it seems better to use that rather
than python2. This also resolves some build flakiness seen due to parallel
invocations of python2 that are fixed in python3.
Note that the scripts aren't compatible with python39 or later, some patching
would be required to resolve that.
Since NixOS 21.11 it seems as if QT uses Wayland if possible[1].
However, my `pinentry-qt` flavor stopped floating because it's now
running in Wayland-mode rather than in XWayland mode where this seems to
be fine.
I wanted to add a rule to my `sway(1)`-config for that, but realized
that `pinentry` is missing an `app_id` to match:
$ swaymsg -t get_tree | jq '.nodes[2].nodes[1].nodes[1].nodes[1].app_id'
""
This is because `QWaylandWindow::initWindow()` uses the application's
`baseName` to determine the app window. Unfortunately the `baseName`
drops all chars of the filename after the first dot[2]. This means that
every wrapped Nix package (i.e. `pkgs.foo` with `$out/bin/.foo-wrapped`)
will have an empty-string as baseName because the first character of the
filename is a dot. Since we're using the `wrapQtAppsHook` quite
excessively, a lot of programs are affected by this.
In order to work around this, I implemented a small patch for
`qtwayland` that strips away the `nixpkgs`-specific `.(name)-wrapped` of
a filename if needed and then sets the `app_id` to the expected
`baseName`. This is useful to make e.g. `sway`-configs with
`for_window`[3]-expressions from other distros compatible.
With this change, the `app_id` is set as I'd expect it:
$ swaymsg -t get_tree | jq '.nodes[2].nodes[1].nodes[1].nodes[1].app_id'
"pinentry-qt"
Even though we'll need the patch to get e.g. `foo` from `.foo-wrapped`,
I decided to file a bug-report against upstream[4].
[1] https://nixos.org/manual/nixos/stable/release-notes.html#sec-release-21.11
[2] https://doc.qt.io/qt-5/qfileinfo.html#baseName
[3] https://man.archlinux.org/man/sway.5.en
[4] https://bugreports.qt.io/browse/QTBUG-99137
I hope this will improve the situation on aarch64-linux.
I don't think it could make anything worse.
Recent staging-next timeouts:
https://hydra.nixos.org/build/141551270https://hydra.nixos.org/build/141547098
For now let's not touch meta.timeout, as there's less consensus and the
setting seems less reliable in practice, too.
The default 10h timeout caused the cancellation of builds on
aarch64-linux builders.
Perhaps counterintuitively, this wastes resources because it
requires a restart, but never completes.
https://hydra.nixos.org/build/136917190
