Re: git: c89b6f279989 - main - lang/php8[12]: Fix build on aarch64 with ZTS

On Wed, Apr 12, 2023 at 02:21:40PM +0000, Alexey Dokuchaev wrote:
> On Wed, Apr 12, 2023 at 03:55:00PM +0200, Mathieu Arnold wrote:
> > On Wed, Apr 12, 2023 at 01:43:43PM +0000, Alexey Dokuchaev wrote:
> > > On Wed, Apr 12, 2023 at 05:39:24AM +0000, Muhammad Moinur Rahman wrote:
> > > > commit c89b6f2799891129e89d1b4fedf8e25865241060
> > > > 
> > > >   lang/php8[12]: Fix build on aarch64 with ZTS
> > > > 
> > > >   - Refresh patches
> > > 
> > > What good did it bring other than cloaking the actual fix with tons
> > > of needless, unrelated changes?  You've poured a glass of honey into
> > > a bucket of tar. :(
> > 
> > Because that is something you have to do everytime a port is updated,
> 
> No, you don't, and you shouldn't.  Patches come and go all the time,
> blindly following this "rule" just brings useless noise to the commit
> diff, nothing else.
> 
> > refresh the patches so that the line offsets are correct wrt current
> > sources, to make sure patch(1) can just apply the patches and not
> > resort to guesswork.
> 
> If patch(1) cannot apply some patch, it will complain and the port
> won't build, so existing patches are applicable by definition.
> 
> But we have already argued about this in the past.  I prefer cleaner
> commit diffs, you don't, oh well.

I don't care about the commit diff of patches, really, patches of
patches are mostly unreadable.
On the other hand, we, as a group, care about patches that are correct,
not patches that happen to apply, maybe at the correct place, by the
grace of some random heuristic in patch(1).

Please stop badgering committers about rules you decided on your own and
that only apply for you. I would also add something along the lines of
"please follow our policies" but we both know you will not.

-- 
Mathieu Arnold