[Python-modules-team] Bug#915856: sphinx: Failed cross building with Build-Depends on python3-sphinx

Dmitry Shachnev mitya57 at debian.org
Fri Jan 25 21:55:28 GMT 2019


On Fri, Jan 25, 2019 at 05:41:56PM +0100, Helmut Grohne wrote:
> Hi Dmitry,
>
> On Fri, Jan 25, 2019 at 11:14:23AM +0300, Dmitry Shachnev wrote:
> > Does this mean that packages that are not using autodoc (like ncmpc) can
> > already build-depend on python3-sphinx:native to become cross-buildable?
>
> Yes, that is my understanding. My plan was postponing such patches to
> simplify stretch-backports. I'm not opposed to them in general, I just
> didn't want to cause unnecessary work.

@Kaliko: please test if it works for you.

> > Does cmake sound like a good start? It already has docs in a separate
> > package, just needs the B-D adjusted.
>
> Yes, cmake is on the list above precisely due to sphinx. It also is a
> key package, so it fits my description very well.
>
> I certainly appreciate you putting effort in resolving these issues. I
> also appreciate X-Debbugs-Cc (to record the bug numbers).
>
> Debian unstable currently has almost 30000 source packages. Almost 14000
> of them build architecture-dependent binary packages. (The others are
> irrelevant to cross building). Around 7500 have cross-satisfiable
> dependencies. I didn't work that much on satisfiability problems lately,
> so this is only around 55%. Having you work here would be a boon. Of the
> 3900 packages that I tried building, 2900 do cross build. This is a much
> higher ratio of almost 75%. Getting those 55% up would be awesome. If
> you fix half of the sphinx users, we're 1% closer to perfection. :)

Sorry, I have quite little time for Debian at all, and from it very little
time I can devote to Sphinx (as I also have Qt and many other packages).
And most of that time comes into test-rebuilding reverse dependencies when
upstream releases a new version that breaks everything...

So I cannot promise that I will take care of half of those packages.
I will take care only of some of them, those that I am familiar with or
those that are easy to fix. Fortunately cmake matches these patterns.

P.S. Maybe I should merge this bug with #818115? They both discuss cross-
building issues.

--
Dmitry Shachnev
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 833 bytes
Desc: not available
URL: <http://alioth-lists.debian.net/pipermail/python-modules-team/attachments/20190126/f13c5125/attachment.sig>


More information about the Python-modules-team mailing list