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

Re: RFS: Upload of updated NLP software stack (CLST, Radboud University Nijmegen)



Hi Mattia et al,

Do you or someone else happen to have some time to help with this still? Specifically
the migration from cdbs to debhelper..? Then I can fix the rest of the
various packages I prepared and hopefully get it all uploaded soonish.

I got set up on salsa in the mean time. I suppose it's best we change
all Vcs-Browser and Vcs-git links to point to there explicitly instead
of anonscm.debian.org?

Regards,

--

Maarten van Gompel
    Language Machines research group
    Centre for Language and Speech Technology
    Radboud Universiteit Nijmegen

proycon@anaproy.nl
https://proycon.anaproy.nl
https://github.com/proycon

GnuPG key:  0x39FE11201A31555C
XMPP: proycon@anaproy.nl  Matrix: @proycon:matrix.org
Telegram:   proycon       IRC: proycon (freenode)
Discord:    proycon#8272
Twitter:    https://twitter.com/proycon
ORCIRD:     https://orcid.org/0000-0002-1046-0006
Keybase:    https://keybase.io/proycon
Bitcoin:    1BRptZsKQtqRGSZ5qKbX2azbfiygHxJPsd


Quoting Maarten van Gompel (2018-01-02 13:19:56)
> Quoting Mattia Rizzolo (2017-12-13 16:56:42)
> > > * https://anonscm.debian.org/git/debian-science/packages/ticcutils.git
> >
> > this url is 404 (but the one mentioned in the Vcs-Git is fine).
>
> Right, should be libticcutils, historical discrepancy in the package
> name vs repo name ..  :)
>
> > quick review:
> > * please bump std-ver (of course, do all the changes required to catch
> >   up with the last policy version) (related: are you subscribed to
> >   debian-devel-annount@lists.debian.org?  You probably should.)
>
> Ok, updated std-ver and subscribed.
>
> > * you are using debhelper 10, but still mentioning dh-autoreconf
> > * please stop using versioned -dev package, they are a PITA requiring
> >   sourceful uploads at every transition for no gain.  Take the occasion
> >   to do one last -dev rename and just remove that version.  I have no
> >   idea why that was done in e8e0fa87dae625cfe33831999755344d3dbf5d98,
> >   but "strong compatibility" doesn't really mean anything.
>
> Ok, I wouldn't know either, perhaps Joost remembers? I just repeated the
> same old trick :) But I now followed your recommendation and switched
> back to an unversioned -dev package name.
>
> > * this is library transition, so it has to go to experimental
> >   https://wiki.debian.org/Teams/ReleaseTeam/Transitions
>
> Ok, adapted
>
> > * Can I recommend to let go of cdbs and use dh instead? (there are good
> >   chances that doing that would also multiarchify the library)
>
> Yes, but I have no idea how whatsoever, I inherited the package like this, so
> I'd have to be taken through all the necessary steps.. It would just be a small
> rewrite of debian/rules I suppose? Is it easy enough to do for this
> iteration? It might be necessary now anyways, becasue after bumping the compat level, I run into:
>
>  dh_systemd_enable: dh_systemd_enable is no longer used in compat >= 11, please use dh_installsystemd instead
>  /usr/share/cdbs/1/rules/debhelper.mk:233: recipe for target 'binary-install/libticcutils-dev' failed
>
> So I suppose letting go off cdbs would solve this?
>
> > * pristine-tar is not updated, neither is the upstream branch
>
> Oops, forgot to push --all , done now.
>
> > * I see there is no symbols file, would it be a good idea to add one?
> >   Some simple starting is at https://wiki.debian.org/UsingSymbolsFiles
>
> I guess we might just as well add one now yeah. If we can tackle the
> cdbs problem first then I'll have a look at this afterwards.
>
> > This also means I'll just review them one at time as well…
>
> Perfect, after getting this one right again I should hopefully be able to fix
> similar issues in the others.
>
> > > I did not add the debian/$latestversion tag yet in the git repo, probably best
> > > to add when its verified to be all right for upload? Other tags should be
> > > present.
> >
> > Personally, I tend to prefer to do the debian/* tag myself while
> > uploading.
>
> Agreed, that's what I was counting on indeed.


Reply to: