[PATCH v2 00/21] Introduce FIXME_ONCE macro

[David Kahurani]

On Mon, Oct 18, 2021 at 4:18 PM Rémi Bernon <rbernon at codeweavers.com> wrote:

> On 10/18/21 2:57 PM, David Kahurani wrote:
> > The following patches introduce the FIXME_ONCE macro
> > which suppresses repeated FIXMEs into WARNings. The current
> > FIXME macro tends to be insufficient in cases where
> > a developer wishes to suppresses fixmes other than the first.
> > It is based on the vkd3d version.
> >
> > v2: Fix programming mistakes, fixed a test and some formating
> >
> > David Kahurani (21):
>
> Hi David,
>
> I understand this can look better at first but there's additional things
> to take into consideration, and especially what a change like this implies.
>
> First of all, this is changing a header used in absolutely all Wine
> source code. This means that such change will require other developers
> to fully rebuild Wine, and things like ccache won't help at all.
>
> This better has to be really well justified, and maybe subject to prior
> discussion and general agreement unless there's no doubt it's needed
> (like adding Windows declarations in public header files).
>
> I don't think this is the case here: the change introduces a new macro
> that will only save one line of code everywhere it's used.
>
> Second, every patch sent on this mailing list is going to be built and
> eventually tested by an automated test bot. Like for the first point, as
> this series is modifying a global header used everywhere else, the bot
> is going to have to rebuild everything from scratch.
>
> This will take a certain amount of time and resources... but moreover,
> it's going to have to do it for every single patch of the series, and
> that is going to take ages. As you may have noticed, your previous
> attempt took around 18h to build and test all the patches, and during
> that period of time the test bot is going to be fairly unavailable for
> any other developer or patch (there's some priorities involved but still).
>
> In conclusion, I think I can say this is very unlikely to be accepted,
> because of the limited usefulness I described in the first part, but
> even if it was you should probably start sending much much smaller
> batches if you change global headers.
>

I have considered all that you described. This change was suggested here on
the list a while back and I took it up. The problem with the patchset is
that they go into different subsystems which forces me to split them up.
They can obviously be bundled into one single patch or maybe just one patch
demonstrating the use.


> Cheers,
> --
> Rémi Bernon <rbernon at codeweavers.com>
>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.winehq.org/pipermail/wine-devel/attachments/20211018/058c8ab3/attachment.htm>