Hi Again, On 8/29/21 1:35 AM, Félix Sipma wrote: > So, let's ask Roger what he thinks we should do. > > I guess we should revert this, as 2.0.6-1 introduces regressions. I don't think doing so will be very optimal, I'd have agreed with you if we were in freeze times, but not now. It'd be better is we simply fix the reverse dependencies now. The Release 1.1.2 of md5-simd fixes the problem with latest cpuid. I tested the new version with ruby-team/meta[1] and it does not seem to break minio-go-v7 so I believe it should be safe to simply upload this. Output: autopkgtest golang-github-minio-minio-go-v7 ... PASS rebuild golang-github-minio-minio-go-v7 ... PASS Or, you could cherry pick relevant patches from upstream - whatever you prefer more. Personally, updating feels much cheaper. > Given this, it won't migrate to testing anyway... It would, if we fix the situation. Similar fix has been done for also new upstream version of golang-github-klauspost-reedsolomon which is the other failing dependency [1]: https://salsa.debian.org/ruby-team/meta Cheers, Nilesh -- Debian Developer, Uploading ⢀⣴⠾⠻⢶⣦⠀ ⣾⠁⢠⠒⠀⣿⡁ Nilesh Patra ⢿⡄⠘⠷⠚⠋ nilesh@debian.org / nileshpatra.info ⠈⠳⣄
Attachment:
OpenPGP_signature
Description: OpenPGP digital signature