Re: small remark on packaging consistency
Hi,
Le mardi 10 décembre 2019 à 21:08 +0100, Patrice Duroux a écrit :
> Two tiny remarks related to the recent python3-testpath upgrade to
> 0.4.4+dfsg-1.
Oops! I did it again?
> 1. Two open bugs were almost the same (or at least to me):
>
> #945214 assigned to python3-pip and reassigned to python3-testpath on
> Sun, 08
> Dec 2019 12:45:05 GMT
>
> and
>
> #945887 assigned to jupyter-notebook and reassigned to python3-
> testpath on Sat,
> 30 Nov 2019 16:42:02 GMT
>
> But why it was only the first that was closed with the last upgrade?
Well, I sent the merging mail, but didn't get the answer that they
couldn't be merged (differing severity) before I actually uploaded, so
I marked it fixed in this version later.
> 2. Also regarding this bug why still using some d/rules treatment
> using an extra
> debian template file to generate an .egg-info file while the upstream
> version is
> now providing a PKG-INFO file that is actually not part of the Debian
> salsa
> repository for this package?
I just unbroke what was there, and didn't look further.
This PKG-INFO file doesn't appear in the salsa repo for the Debian
packaging because it doesn't appear in upstream's tarball (git import-
orig --uscan), and I just had a look and didn't see it in upstream's
github repo either. Or did I miss it?
> Other DMPT packages do not seem to do the same to generate and
> provide this .egg-info file.
I have now looked further : "git blame" says Ximin Luo added it (commit
ab9d454b) and the documented reason is: "Add an .egg-info file so pip
recognises it. (Closes: #880241)" ; so it looks like there was a valid
reason and it's a nice contribution...
If there's something to do to improve the packaging, I'm open to
suggestions (and contributions!).
Cheers,
JP
Reply to: