08:10 <davidlt> rwmjones_, this is what I wanted to avoid:

08:10 <davidlt> https://koji.fedoraproject.org/koji/buildinfo?buildID=2410274

08:10 <davidlt> https://src.fedoraproject.org/rpms/gdal/commits/rawhide

08:10 <davidlt> riscv64 change landed in gdal. It also landed in riscv64, but it's not incl. in changelog

08:11 <davidlt> It's not enough to look into Koji or/and RPM changelog, but also checking dist-git is important

08:12 <davidlt> I want riscv64 change land with a separate NVR

08:12 <davidlt> So it's visible, etc.

08:13 <davidlt> It costs more time for me to check a package. Also things should land in F40 faster. Taking a new NVR from F41/Rawhide doesn't help me a lot.

08:14 <davidlt> My script checks packages between two koji using tags. If I build NVR from F41 it will attempt to build the one from F40.

08:14 <davidlt> As soon as ".riscv64" is lost from NVR it will attempt to sync packages between tags.

08:15 <davidlt> Now if the change didn't land in F40, but it builds the package gets broken.

08:15 <davidlt> I only get to notice that my analyzing failed build logs.

08:16 <davidlt> Anyways, sooner or later I will loose track of all this.

09:55 <rwmjones_> ack

10:45 <davidlt> rwmjones_, https://src.fedoraproject.org/rpms/ppl/pull-request/1#

10:45 <davidlt> this is stuck for 6 months now

10:45 <davidlt> Yet it's trivial enough

10:54 <rwmjones_> looking

11:33 <rwmjones_> (still doing ppl)

11:51 <davidlt> rwmjones_, just image the day 16-core SiFive P670 exist

12:47 <rwmjones_> davidlt: I've done ppl

12:47 <rwmjones_> I'm waiting for that day

13:05 <davidlt> Jakub is bumping GCC

13:06 <davidlt> https://src.fedoraproject.org/rpms/gcc/c/8ed2af4dd025fe882557d740b4bc0ca63b7cff19?branch=rawhide

13:06 <davidlt> 104 minutes ago: https://gcc.gnu.org/git/?p=gcc.git;a=log;h=refs/vendors/redhat/heads/gcc-14-branch