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

Re: More fine-grained control in texlive?



Rogério Brito <rbrito@ime.usp.br> wrote:

>> So let's have a look.  Which (binary) packages do you see that are
>> 120MB or even larger?
>
> I'm not speaking here for Alexandre, but the *installed* size of, at
> least, tl-fonts-extra is around 120MB. Doing a dpkg -c on the package, I
> see that we have mostly documentation on these packages.

I see, I didn't look at the installed size.  So this might be a
candidate for splitting.  

But why do you say it's mostly documentation?

$ dlocate -L texlive-fonts-extra | grep doc | wc -l
1194
(605)[20:34:56]frank@riesling:~$ dlocate -L texlive-fonts-extra | grep -v doc | wc -l
7301

and the 1194 needs to be divided by 2, because we have a symlink for
every file (maybe this should better be changed to directory symlinks,
but well).

> Since I am in the New Maintainer queue now (Yay!) I think that I can try
> to help you more effectively. In other words, read: I would be willing
> to help with the tex-maint group.

You already have helped us, and of course you are welcome to contribute
more in the future!

>> This is already now a problem with tl-humanities and tl-science which
>> have been split of tl-latex-extra.
>
> Is there a problem with tl-science? I am mostly concerned with that,
> since it includes the algorithms package.

I'm not sure whether there's a real problem with tl-science.  But there
was one with tl-humanities for sure, one componend depended on something
in an other package (tl-latex-extra?).  Upstream considered how to fix
this, and found out that moving the dependency to tl-humanities didn't
make sense (would have pulled in even more, and/or was not related to
humanities).  So the result was that things were left unchanged, and the
message is that it is hard to properly split texlive-latex-* both from a
point of view of dependency logic connected with application areas, and
in terms of manpower (people who check dependencies).

Is that clearer?

Regards, Frank
-- 
Frank Küster
Single Molecule Spectroscopy, Protein Folding @ Inst. f. Biochemie, Univ. Zürich
Debian Developer (teTeX/TeXLive)



Reply to: