Hi, I happen to be attempting the rebase at the same time d-;
On 2026-04-14 at 15:02+02:00, Andreas Enge wrote: > Am Tue, Apr 14, 2026 at 01:22:47PM +0100 schrieb Sharlatan Hellseher: > > I tried to rebase and faced with the same issues of git conflicts: > > - - rust-crates has a long list of conflicts > > For these usually there are only additions; so I have taken the habit > to solve them by just removing all the lines "<<<...", "===..." and > ">>>...". At worst, we reintroduce definitions that were already > removed. > > I think that at some point in time, we should (by a script?) remove all > unused crate entries. Preferably in one go and when no team branches are > active (if this ever happens). cc @efraim as the master of crates. This is not sufficient as make reports lots of shadow definitions of crates. On 2026-04-14 at 15:02+02:00, Andreas Enge wrote: > Am Tue, Apr 14, 2026 at 01:22:47PM +0100 schrieb Sharlatan Hellseher: > > - - commit removing python build system messes up with some othere sorting > > used-module. > > These are the ones where I dropped the ball, because moreover they look > very confusing. Maybe the simplest solution would be to drop > commit edf6898cf416d6c16ed84c2b44931958544cb417 > Author: Nicolas Graves <[email protected]> > Date: Tue Mar 24 13:13:39 2026 +0100 > guix: build-system: Set pypi-uri in pyproject, drop python module. > > and to add it again on top; the commit itself looks very clear, just > the git algorithm creates strange conflicts. > All the > - #:use-module (guix build-system python) > could simply be created by sed or a similar script. You don't need to drop the commits, but git restore -s HEAD -- gnu/packages/... # the conflict files and then manually remove occurances of guix build-system python. I think the last rebase missed a couple of them. Best wishes, Phong
