[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: ggi debs for testing



> hm, I am using 2.2.7 (vger actually, it just identifies itself as 2.2.7) after
> many suggestions by Hartmut :-) It runs fine in all respects, but I keep having the same

:-)  And the vger source is broken for me, it is powermac centric for the mac-kbd, i use
ps/2. 

> Anyways. Hartmut, libggi+stuff maintainer told me that I should upload without changing
> the changelog (I did an NMU binary-only upload), and the changelog entry I added was
> quite minimalistic :-). He says unless I have made some changes to the source code, I should
> let the autobuilder to handle it. Come to think of it, there are new ggi source packages.
> Should I take them or let the autobuilder handle them?

The autobuilder take only new uploaded versions, but if you do an binary-upload for powerpc
he means it is all up-to-date. 

If we follow strictly policy and we change a little bit to the source we must! do an MNU with
source upload. But we will this not do without asking the maintainer of this package. But if
the maintainer is MIA we must wait ~3 months. This is ok for one or two packages, but not for 20,
50, 100 or 200 packages. 

So we break policy and compile the package and upload it regulary (if possible, that means if
the version of this package is not yet available for powerpc; if we must upload a newer version
we increment the debian version: from foo-4.3 to foo-4.4) and post only the source-diff as a
bug-report. 

The autobuilder handle only source-packages.

If you have the permission from the maintainer to do an NMU, you should change then the source
and the debian/changelog (dpkg-buildpackage without -B).

> Last but not least. You probably noticed that I uploaded netstd and netbase. Anything more
> for me? I think I'm getting lazy... :-)

Ok :-)  Let me search for important packages.  

BTW: we are down from 20.4% to 18.1% of failed packages; yeah!

Greetings,

      Hartmut 


Reply to: