<div dir="ltr"><div dir="ltr">On Thu, Feb 20, 2025 at 4:01 AM Joshua Goins <<a href="mailto:josh@redstrate.com">josh@redstrate.com</a>> wrote:</div><div class="gmail_quote gmail_quote_container"><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex">On Tuesday, February 18, 2025 7:15:34 PM Eastern Standard Time Justin Zobel <br>
wrote:<br>
> On 19/2/25 10:26, Joshua Goins wrote:<br>
> > Someone noticed a build failure on some systems (NixOS and Slackware it<br>
> > seems) due to C++ funkiness. They fixed it, and I cherry-picked it to<br>
> > 6.3. Can we get a respin or a 6.3.1.1? Thanks!<br>
> <br>
> If it only affects two distros, I think having them carrying the patch<br>
> downstream is more suitable than re-spinning (something in my opinion we<br>
> should never do) or creating a new tarball, e.g. 6.3.1.1.<br>
<br>
I thought about this, but the fact that it's not compilable is kinda bad for <br>
something like plasma-desktop (it also reflects badly on KDE if someone has to <br>
patch our newly released software to make it work.) As of writing, the extra <br>
release happened anyway.<br>
<br>
> What is different about these two distros that only they were affected?<br>
> Perhaps it is something we can add tests or CI jobs for.<br>
<br>
This would be good IMO, I'm worried that something like this slipped past us. <br>
I'm still not certain what compilers trigger this, the commit message says <br>
GCC14. So maybe we need a more recent GCC and/or clang?<br></blockquote><div><br></div><div>Nope, our current set of images are using GCC 14:</div><div><br></div><div><span style="font-family:monospace"><span style="color:rgb(0,0,0)">user@45085f386ad8:/> g++ --version
</span><br>g++ (SUSE Linux) 14.2.1 20241007 [revision 4af44f2cf7d281f3e4f3957efce10e8b2ccb2ad3]<br></span></div><div><br></div><div>It must be something else those distros are doing in their default compiler flags.</div><div><br></div><div>There is only so much the CI system can try to pick up i'm afraid, and it will never be able to cover everything as CI has to run a rolling release distribution which means it will likely be running newer software than many distros.</div><div><br></div><blockquote class="gmail_quote" style="margin:0px 0px 0px 0.8ex;border-left:1px solid rgb(204,204,204);padding-left:1ex">
<br>
tangent: really wish this kind of stuff would stop happening in C++. why is it <br>
that some code can compile and link fine in one setup, and then completely <br>
break in another? 🙂 I *know* the reason, it's just frustrating.<br>
<br>
Josh<br>
<br></blockquote><div><br></div><div>Thanks,</div><div>Ben </div></div></div>