Lennart Poettering
Lennart Poettering
I would really prefer if we'd fold this into ConditionFirmware= somehow. What's the usecase for this? Cna you elaborate? The thing is, we already allow to nicely match DMI from...
Do you actually have a real-life usecase for the version comapre stuff? otherwise, please drop, and just do fnmatch
hmmm, ok, if you do control the hw then version compares are indeed ok. But in that case, I think we should probably be systematic it also do the regular...
and maybe == for simple string compares
> Perl uses `=~` and `!~` for positive and negative regular expression matches, thus maybe the same operator symbols could be used for fnmatch matches. They look a bit better...
> Why do we need to use `strverscmp_improved()` for equality / inequality comparison of versions at all? Where is the difference to use `fnmatch()` in this case? ``` strverscmp_improved("4711*", "4711")...
There are SMBIOS fields that are textual strings, and SMBIOS fields that can conceptually be considered version numbers. For the former fnmatch() makes sense, for the latter strverscmp_improved() matches. Since...
hmm, not sure about that. the thing is that we probably should teach os-release condition checks and similar the same semantics, but there you can't just add a prefix, because...
please, let's not go overboard with this. nesting concepts is clearly too much. So you have a point that `strverscmp_improved() == 0` is different from `streq()`. But I wonder if...