Hi, I want to become provenpackager because I want to make fedora better and now I'm working on adding NeuroImaging and NeuroScience tools into Fedora (https://fedoraproject.org/wiki/Changes/NeuroFedora) and sometimes (often) I need to rebuild or update non-updated packages (like joblib, scikit-learn, etc.) or add python3 subpackage or .... Now I'm requesting ACLs/sending patches/opening bugs for this and waiting unknown time while those will be accepted. It is main reason for now, but I did/doing/will do the same for non-related to Neuro... packages.
Today's example of patch: https://bugzilla.redhat.com/show_bug.cgi?id=1286349
Right now I'm maintainer of 90 packages: https://admin.fedoraproject.org/pkgdb/packager/ignatenkobrain/
I've mailed sponsors to provide feedback for this request. I am +1 myself.
+1
+1.
Signed, his sponsor.
About
sometimes (often) I need to rebuild or update non-updated packages (like joblib, scikit-learn, etc.) or add python3 subpackage or
and:
.... Now I'm requesting ACLs/sending patches/opening bugs for this and waiting unknown time while those will be accepted.
We really really should get some leadership/committee people to work on the Provenpackager policy:
https://fedoraproject.org/wiki/Provenpackager_policy
As well as:
https://fedoraproject.org/wiki/Who_is_allowed_to_modify_which_packages
Particularly note how those documents tell you should establish contact with the current "owners" of a package before taking further action. The 2nd doc even tells how long you should wait before taking action.
Today's example of patch: https://bugzilla.redhat.com/show_bug.cgi?id=1286349
More relevant and interesting would be examples of cases where you've had to wait an "unknown time", e.g. encountering a non-responsive maintainer.
Watching how some people go in and update arbitrary packages in dist git without paying attention to activity in bugzilla, is reason to be concerned.
Replying to [comment:6 mschwendt]:
About sometimes (often) I need to rebuild or update non-updated packages (like joblib, scikit-learn, etc.) or add python3 subpackage or Sometimes it relates to rebuild stuff (e.g. libgit2 soname changes. There was only one person who really carried about this - zbyzsek). and: .... Now I'm requesting ACLs/sending patches/opening bugs for this and waiting unknown time while those will be accepted. We really really should get some leadership/committee people to work on the Provenpackager policy: https://fedoraproject.org/wiki/Provenpackager_policy As well as: https://fedoraproject.org/wiki/Who_is_allowed_to_modify_which_packages Particularly note how those documents tell you should establish contact with the current "owners" of a package before taking further action. The 2nd doc even tells how long you should wait before taking action. Yes, I read those pages. And that bug which metioned in start post was about broken scipy which used by 50+ packages (repoquery is still running, but I guess there are much more than 150). Today's example of patch: https://bugzilla.redhat.com/show_bug.cgi?id=1286349 More relevant and interesting would be examples of cases where you've had to wait an "unknown time", e.g. encountering a non-responsive maintainer. All of maintainers was responding after 1-2.5 weeks so I can't find any totally unresponsive maintainers. https://bugzilla.redhat.com/show_bug.cgi?id=1219815 I asked to update in May (with patch attached) and forgot about it and in august maintainer was responded in mail and then in BZ. Watching how some people go in and update arbitrary packages in dist git without paying attention to activity in bugzilla, is reason to be concerned. I am going to bugzilla first in 90% of cases. In other 10% I'm sending patches directly to mail to owners of package and they giving me ACLs ;)
sometimes (often) I need to rebuild or update non-updated packages (like joblib, scikit-learn, etc.) or add python3 subpackage or Sometimes it relates to rebuild stuff (e.g. libgit2 soname changes. There was only one person who really carried about this - zbyzsek).
Particularly note how those documents tell you should establish contact with the current "owners" of a package before taking further action. The 2nd doc even tells how long you should wait before taking action. Yes, I read those pages. And that bug which metioned in start post was about broken scipy which used by 50+ packages (repoquery is still running, but I guess there are much more than 150).
More relevant and interesting would be examples of cases where you've had to wait an "unknown time", e.g. encountering a non-responsive maintainer. All of maintainers was responding after 1-2.5 weeks so I can't find any totally unresponsive maintainers. https://bugzilla.redhat.com/show_bug.cgi?id=1219815 I asked to update in May (with patch attached) and forgot about it and in august maintainer was responded in mail and then in BZ.
Watching how some people go in and update arbitrary packages in dist git without paying attention to activity in bugzilla, is reason to be concerned. I am going to bugzilla first in 90% of cases. In other 10% I'm sending patches directly to mail to owners of package and they giving me ACLs ;)
{{{ $ sudo dnf repoquery --whatrequires scipy --alldeps --tree | wc -l 1170 }}}
To be precised.
We need more provenpackagers. Igor has a significant potential, so that's why I'm adding +1 here.
+1 from me.
+1 from me
+1 from me, too. Worked on some packages with Igor and it was always a pleasure.
Its been a week and I only see +1s.
Use your powers for good.
Thanks!
Login to comment on this ticket.