<beerman>
but squid still wouldn't build with gcc12.1
<jue>
wrt pear, the files you mentioned are part of then footprint already?
<jue>
thanks
<beerman>
mh, I dunno, they stayed in my container at least, and were not removed. Maybe I need to double check on that but I got an error installing the port from a clean state becuase it would overwrite those files
<jaeger>
I don't have any issue with going to gcc 12 for 3.7 if we want to do that, just need to do more build tests over here. I'm aware of the binutils conflict, it's only in 3.7 and I have local updates already, just not pushed. Thanks for the heads up on slim, will check it out.
<jaeger>
ugh, wine has some new regression causing it to fail with -j<x>, very annoying
<beerman>
I have a few of those too, unrar yesterday
<jaeger>
Looks like it's only in 3.7 so I guess a gcc 11+ interaction... still annoying :P
<jaeger>
well, shit, now I can't even build wine-staging 7.8 with -j1, 7.7 did work
<beerman>
Oo
<jaeger>
Hrmm, this might be a binutils 2.38 issue rather than wine, doing some more testing
<jaeger>
yeah, it's a binutils issue, so for now I can keep mingw-w64-binutils at 2.37 instead of 2.38
<jaeger>
(in 3.7)
<beerman>
mh. i haven't dug into that in detail
<stenur>
btw my own cacerts triggered an update last Saturday, EC-ACC is gone (Agencia Catalana de Certificacio). Granted this surely does not make such a difference, but still.
<stenur>
in my humble opinion system should probably come to a point where everything works again first. How often have i seen new compilers do a miscompilation here and there, NetBSD and anything cross-compilable is full of quirks for this.