Hi all, Ole Streicher, on 2025-07-14: > Am 14.07.25 um 07:48 schrieb Lucas Nussbaum: > > In #1109178 (emmax: fails to dist-upgrade from bookworm (related to > > libatlas3-base removal)), Etienne added a Breaks to libatlas3-base to > > work-around the issue. > > > > Maybe the break could be set in libpacke instead. > > liblapacke already has a > > Breaks: [...], libatlas3-base (<< 3.10.3-14) > > which should work for a direct update from bookworm (Bookworm has > 3.10.3-13), but not from a later testing (latest version in testing was > 3.10.3-14). > > However, this break was already detected by apt: > > > > > Broken liblapacke:amd64 Breaks on libatlas3-base:amd64 < 3.10.3-13 @ii > > > > mK > (< 3.10.3-14) > > > > Considering libatlas3-base:amd64 1 as a solution to liblapacke:amd64 0 > > > > Holding Back liblapacke:amd64 rather than change libatlas3-base:amd64 > > where I don't understand the solution taken by apt. However, even apt > considers liblapacke as "broken"; and I don't see why adding a break to > psfex & Co. would solve this.
It doesn't. I wrapped up the Breaks in emmax yesterday but mis-tested the upgrade and thought that resolved the issue, but the upgrade from bookworm to sid of emmax still doesn't happen today. I reopen #1109178. > I'd really tend to re-assing these bugs to lapack. I tend to agree. Have a nice day, :) -- .''`. Étienne Mollier <emoll...@debian.org> : :' : pgp: 8f91 b227 c7d6 f2b1 948c 8236 793c f67e 8f0d 11da `. `' sent from my alarm clock `-
signature.asc
Description: PGP signature
