Comment 9 for bug 1887142

> I don't know how I could/should evaluate how bad this is compared to having a non-working package such as pdfsam, so I would appreciate any feedback related to the SRU process on such decisions (maybe just having the regressions described is good enough?).

Just whatever best effort you can manage. I don't think there's any specific objective measure that can be used. Thank you for noting this in the bug description. I think describing these are important because if a user does get regressed, it does help that they can find out that we considered their case when making a decision.

In this case I think that given no such user is known and no other packages in the archive use libsejda-data the likelyhood of regression in this manner is unlikely, so bumping this is the least worst option to fix pdfsam (which surely has many more direct users).