Post-MegaRelease projects

[Ben Cooksley]

On Sat, Feb 24, 2024 at 10:27 PM Konstantin Kharlamov <Hi-Angel at yandex.ru>
wrote:

> On Sat, 2024-02-24 at 22:16 +1300, Ben Cooksley wrote:
>
> On Sat, Feb 24, 2024 at 8:37 PM Konstantin Kharlamov <Hi-Angel at yandex.ru>
> wrote:
>
> On Sat, 2024-02-24 at 16:31 +1300, Ben Cooksley wrote:
>
> On Fri, Feb 23, 2024 at 11:12 PM Sune Vuorela <nospam at vuorela.dk> wrote:
>
> On 2024-02-22, Nate Graham <nate at kde.org> wrote:
> > I've started pondering post-megarelease projects. We've spent so long on
> > porting and bugfixing that I think it might be useful to shift gears to
> > feature work, and I'd like to brainstorm potential large-scale projects
> > and gauge the level of interest in putting resources into them soon.
>
> A bit more from the devops end that I'd love to see people tackle:
>
>  - Ensure frameworks and app unit tests interacting with windows can run
>    on Windows.
>    More details: The following fails on our windows CI
>    https://invent.kde.org/sune/windows-test-thingie/-/blob/master/main.cpp
>    I find it weird that we are spending resources on putting things in
>    the windows store and making apps available on windows, but we can't
>    actually have passing tests in our CI.
>
>
> This unfortunately will not be easy to solve.
>
> One of the key things that we've learned out of doing CI, as has been
> showcased by FreeBSD in particular, is that the builders need to be
> ephemeral - that is only around for the build in question that is being run.
> We're currently accomplishing this by using containers - via Podman (for
> Linux/Android/FreeBSD) and Docker (for Windows).
>
> Containers also offer us the advantage of allowing people to easily
> reproduce the CI environment on their local system without too much trouble.
>
> For Windows however, Microsoft has limited the container stack to not
> allow anything GUI related to work. The underlying libraries may be there,
> but the equivalent display server components are not operational.
>
> To complicate things further, on Windows certain permissions are
> restricted to the interactive console and are not possible to do as either
> a scheduled task or as a system service.
> Usage of existing Windows automation frameworks such as Powershell
> Remoting or SSH will therefore not work if we want things to perfectly
> replicate a end user environment - because those will run the command(s) as
> part of a non-interactive session (even if the user we connect as is the
> same one logged in on the desktop console).
>
>
> Idk if it's a silly question, but… If Windows native containers have so
> many restrictions, why not just use Linux containers with WINE inside?
>
>
> Because Wine is not Windows either, and there could be subtle differences
> in how things run / interact with the system.
> Plus some of our software would like to test certain system level
> infrastructure (like say KDE Connect).
>
>
> Out of curiosity, what does this infrastructure include? I thought KDE
> connect only uses network sockets and system tray.
>

No idea, I saw their commentary on debugging issues they were having in
their unit tests in #kde-devel.
Those issues were due to a lack of permissions, specifically around the
interactive console - that is how I know some of our tests need those
additional permissions and why running as a scheduled task / system service
will not be sufficient for "fully working" CI tests on Windows.


>
> Plus, we have to have native Windows to compile things anyway as we need
> to use MSVC (otherwise you have no Qt Web Engine support, as that cannot be
> built with MingW)
>
>
> But I presume it can be built with Clang? In particular, Google Chrome on
> Windows is being built with Clang — and Web Engine is basically a fork of
> Chromium.
>

Qt 6 as a whole does not list Clang as a supported compiler - see
https://doc.qt.io/qt-6/supported-platforms.html

Given Windows is a bit strange in the first place, i'd be quite reluctant
to step outside of what they list as supported.


>
> (and you cannot really mix MingW / MSVC binaries due to incompatible
> compiler ABIs for C++ code)
>
>
> Well, if for testing purposes Qt was pre-built with Clang, I guess there
> won't be any ABI issues.
>

Would require that everything else we have is rebuilt, and that all
downstream users of our Craft cache also switch to Clang.
Note that like most open source projects, we don't know the full extent to
which Craft is used outside of our own project.

Historically we have seen through Microsoft provided data that dependencies
which we built and signed have shown up in surprising places (such as
Snoretoast - which was used by something Node.js based at one point or
another I believe)

Cheers,
Ben
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mail.kde.org/pipermail/kde-devel/attachments/20240224/37902864/attachment.htm>