Hi,
I wonder if this might due a change I did during the process: I imported
a tar.xz file in the pristine-tar (which is now the default format
provided by gitlab), but the previous version was a tar.gz
Could this explain the issue ?
Manik.
Le 11/01/2026 à 19:19, Étienne Mollier a écrit :
Hi again,
Étienne Mollier, on 2026-01-11:
Something off also occurred to the upstream branch, as it's head
is outdated from before camitk 5.2 in trixie:
$ git log upstream | head
commit 0c9ad36ae84636b3e74d690f147de30ebae8d34a
Author: Andreas Tille <[email protected]>
Date: Mon Jan 17 09:33:52 2022 +0100
New upstream version 5.0.2
commit ddac4758554c4f89e61d280594a49ec5594a5b52
Author: Emmanuel Promayon <[email protected]>
Date: Mon Jul 16 20:21:30 2018 +0000
The upstream tag made it to the master branch, which may explain
how Salsa CI eventually managed to have most items pass anyways.
I'm afraid the git repository of camitk in salsa does not look
in good shape right now. :(
I have been investigating in my backups whether I would have the
repository in a recoverable state, but it seems that the problem
affecting the upstream branch is not new; possibly even not as
problematic as I first thought. The pristine-tar issue remains,
I let you proceed with gbp-pristine-tar(1). If this proves to
be difficult, please don't hesitate to ping back and I'll see
whether I manage to harmonize the git branches again.
In hope this helps,
