[tlbuild] patches for TeX Live

[Mojca Miklavec]

(redirecting from tex-live mailing list from "synctex patches in TeX Live")

On Thu, Aug 25, 2011 at 00:14, Karl Berry wrote:
>    My question is: assuming that a linux distribution is releasing TeX
>    Live 2011 now - does it make sense to include the following patch?
>    http://tug.org/svn/texlive?view=revision&revision=23644
>
> As far as I know, yes.  Perhaps Peter will read this in due course and
> confirm.
>
> I haven't tried to build the sources with just that
> patch and only that patch applied, but looking at it, I find it hard to
> imagine that it interacts with anything else.
>
> (Of course, it would make sense for any kind of distribution, of course,
> not just [GNU/]Linux.  The patch is not platform-specific.  For the
> record ...)

In the meantime we tested. Compilation runs fine and the resulting PDF
works much better.

>    doesn't provide any bugfix release with the most critical bugs being
>    fixed.
>
> If it would be useful for the distro people, I could apply that patch to
> the branch.

That would definitely help. But see below.

> That would be a topic for the tldistro list.

So I'm forwarding to that one instead ...

> I rarely get
> any feedback from any distro people other than Norbert, so I don't know
> what it is useful and is a waste of my time, so of course I do nothing
> by default.

I was volunteering for a "context beta tester" for two distributions
this year. I have a few more longer comments and suggestions (for
another thread and for another time - it would be a long list), but in
the context of this particular issue ...

People who package TeX Live have no time following each and every bug
report in TeX Live. They might patch what they find by accident or if
something doesn't build at all. I asked the Mandriva developer to
apply a tiny patch for ConTeXt and the patch mentined above. TeX
Live's SVN is being constantly improved and patched. Trunk contains
both new package + vital bugfixes, but those bugfixes never
proliferate into older/stable branches, so there is no easy way for
packagers to apply just the vital patches. If I was packaging TeX
Live, I would definitely welcome bugfix releases (due to the big size,
most probably patch files would make more sense than the whole
repository zipped again).

(Many distributions pick a random shapshot of TeX Live's SVN instead
of taking .tar.gz files. If bugs appear in 2011, there would be at
least a chance to get rid of some of them without the need to use a
random snapshot of TeX Live repository from February 2012.)

But that means that people behind development (those who submit
patches) would need to treat bugfixes and improvements differently and
only apply necessary patches to stable branch without adding new
functionality. For example: I applied hotfixes for ConTeXt to trunk at
the end of July and those patches could go to 2011 branch as well.
Many other patches should only stay in trunk.

The main question is: is there enough "workforce" for such workflow?
Should I submit patches to 2011 branch in future? Would anyone create
patch files and put them on FTP every now and then?

(In ConTeXt we definitely cannot afford to provide bugfix releases for
"stable ConTeXt" even if we are in a much more desperate need for
that. This may be done once per year for TeX Live, but not on regular
basis.)



Apart from that: Akira keeps updating windows binaries from time to
time. It would be nice to be able to do the same for other platforms
when bugs are revealed. TeX Live now has XeTeX and LuaTeX binaries for
all platforms with a known bug and a known cure. We have the
infrastructure ready to update binaries. Wouldn't it make sense if
those who built the binaries would have an option to compile and send
a patched version of both binaries? Compiling from trunk would not be
the smartest choice since trunk contains other patches as well. If
some important patches would be applied to source in 2011 branch,
binary builders could use that one for recompiling. Yes, I'm aware.
Compiling takes time and one would never get all the platforms
updated. But at least that would give a chance to those who are
willing to compile to fix the bugs on their own platforms.

Mojca

(and sorry for my horrible English :)