Hi Gerald, On Wed, Oct 04, 2023 at 12:17:48AM +0200, Gerald Pfeifer wrote: > On Tue, 19 Sep 2023, Mark Wielaard wrote: > >> Although there were some positive responses (on list and on irc) it is > >> sometimes hard to know if there really is consensus for these kind of > >> infrastructure tweaks. But I believe there is at least no sustained > >> opposition to changing the gcc-patches mailman setting as proposed > >> above. > > This change is now done for gcc-patches. > > Yeah, yeah, yeah. Thank you!
Thanks to the fsf-tech team who explained the setup they are using for lists.gnu.org and everybody helping to test in the bug: https://sourceware.org/bugzilla/show_bug.cgi?id=29713 It seems to work well, but it does mean disabling most of the things mailman can do, like filtering out HTML attachements, adjusting headers, adding footers or subject prefixes, etc. And we did break at least one workflow for people who were using DKIM signing and the Errors-To header. > >> And if there are no complaints at Cauldron we could do the same for > >> the other patch lists the week after. > > Sadly I missed Cauldron - have there been any complaints there? No, the opposite. People seemed happy and there were some examples shown where (on other lists, like binutils) From rewriting caused issues for some tools relying on patchwork.sourceware.org. > Can you adjust the g...@gcc.gnu.org list and others @gcc.gnu.org as well? > I for one would love to see that. Being somewhat conservative doing that in steps. Last week we switched the other gcc lists that receive patches (gcc-rust, libstdc++, fortran and jit). This week looking at non-gcc lists on sourceware that use patchwork (newlib, binutils, elfutils-devel, gdb-patches and libabigail). Then if nothing breaks horribly more general lists if people want. Cheers, Mark
