[tex-live] Stable vs. Unstable/Testing Update Repositories?

Reinhard Kotucha reinhard.kotucha at web.de
Wed Feb 24 04:19:22 CET 2010

On 23 February 2010 C.M. Connelly wrote:

 > Near the end of 2009, tlmgr came out for TL 2008, and I installed
 > it and used it on my machine.  There weren't many updates, and I
 > (possibly wrongly) assumed that the updates were ``safe'' and
 > ``recommended''.
 > But TeX Live 2009 includes tlmgr as part of the base distribution,
 > and talks about it in the docs as if it's something ordinary
 > people should use.  MacTeX 2009 goes further by installing TeX
 > Live Utility.app right next to TeXShop.app along with a nice
 > document telling them how to use it.  Curious users are going to
 > try that app out to see what it does.
 >     KB> 2) There is no feasible way for anyone involved (authors,
 >     KB> CTAN, TL) to know what a given update breaks and what
 >     KB> doesn't.  So there's no feasible way to have branches (in
 >     KB> addition, it would be a huge amount of work and a huge
 >     KB> additional complication).
 > The same argument could be made about any of the Linux
 > distributions, which have thousands of potentially interacting
 > packages from hundreds of upstream developers.  Yet they manage by
 > implementing some system to handle the testing stage -- Debian has
 > sid (unstable) and testing in addition to the stable release;
 > Fedora has its releases as well as rawhide for development (and is
 > moving towards having a separate prerelease testing repository,
 > too).  Even better -- these distros typically have stable releases
 > that only get security updates or occasional feature updates.
 > The same options could be used for TeX Live.  There could be a
 > time-based progression, where new packages come in from CTAN and
 > are added to a bleeding-edge TL repo for people to play with, and
 > after some amount of time, those packages would qualify to be
 > moved into the main TL repo that's considered safe to update from.
 > New versions of a package reset the clock.  Adding some active
 > testing would make that system even safer, of course, but would
 > make things more complex.
 > An even easier (and arguably better) option is to simply have a
 > stable release repository, that only gets bugfixes or well-tested
 > feature changes, and a second development repository where the
 > newest stuff comes directly in from CTAN, and publicize the stable
 > repositories.  An added advantage here is that you can continue to
 > maintain that release for as long as you want, and even keep it
 > around for people who aren't ready to update to the next year's
 > release.
 > [...] 
 > I'd be even happier if there was a stable repository for each TL
 > release that only had bugfixes added to it and could be used by
 > ordinary users without fear of breaking their systems.

Oh yeah, everything was absolutetely stable and well tested in the
good old days when there was no tlmgr which now breaks everything.

Pure nonsense.  TeX Live ...., 2003, 2004, 2005, 2007 were snapshots
of what was on CTAN at the time TeX Live was released.  Nothing was
tested at all.  And if something didn't work you had to wait for next
year's release.

Now we have tlmgr.  All it does is to provide a snapshot of what is
currently on CTAN.  It's possible that a particular update breaks your
system, but if a problem occurs, all you have to do is to report the
problem and run tlmgr again a few days later.  In the "good old days"
you had to wait for a new TeX Live release.  Everything is better than
in the past.  Definitely.

I agree with you that something well tested like "Debian stable" is
useful somehow.  I really don't want to prevent you from maintaining
something like this.  But please keep in mind that TeX Live now
consists of more than 91,000 files and this number is rapidly growing.
If you think that it's sensible to provide a repository where all these
files are well tested, please go ahead.

I personally think that tlmgr is a great thing.  If you are missing
the "good old days", there is a simple workaround: Don't use tlmgr
and wait for TeX Live 2010 if you encounter a problem.  It might be
fixed then or not.  If you don't use tlmgr, everything behaves exactly
as in the "good old days".

 > I'm happy with the policy being ``updating is for elite TeX
 > hackers only; ordinary users should install the released version
 > and never update until next year's release'', but it would be nice
 > if the tools to subvert that policy weren't prominently included
 > with the distributions and used to provide other attractive
 > functionality.

I'm not sure I understand you correctly.  Do you think it's sufficient
if the TeX Live installer provides a [back-to-the-stoneage] button or
do you prefer a solution à la Khmer Rouge?


Reinhard Kotucha			              Phone: +49-511-3373112
Marschnerstr. 25
D-30167 Hannover	                      mailto:reinhard.kotucha at web.de
Microsoft isn't the answer. Microsoft is the question, and the answer is NO.

More information about the tex-live mailing list