[OS X TeX] altpdflatex - back compatibility please

[Joachim Kock]

Dear Gerben,

thanks a lot for looking into this, and thanks in general for your good
work with the tetex distribution.

> The altpdftex->simpdftex change has been made because altpdftex needs a
> separate command file for every type of format (altpdftex, altpdflatex,
> etc.) and this clutters your binary directory and is also problematic to
> maintain: if you enable or disable formats (via fmtutil, texconfig or the
> i-Package configure phase) the required extra altpdf* links are not
> automatically created.

I understand that although these scripts are extremely useful, the original
design was not thought-through enough, and perhaps you did not imagine how
popular and wide-spread the scripts would be.

However, the problem that new-format versions of the script are not
generated automatically is not a problem that ever broke any functionality.
If altpdfsmartlatex never existed, then neither did anyone ever expect it
to exist.  Cluttering the bin/ directory shouldn't really be a problem for
anyone, at least not a problem compared to the advantage and noble concern
of back compatibility.


> When we were dicussing adding this, the result was that altpdf* would be
> replaced by a single command with an argument.  That needs to to be
> maintained and it is cleaner.

I agree that the simpdftex design is much better.  But there is nothing to
maintain: just freeze the old alt* scripts as they are and keep them, in
respect of people using them.  That's extremely simple and should not be a
problem for anyone.  (You might freeze them together with a warning, in the
style of the latex 2.09 compatibility.)  If some day the scripts break,
too bad, but before that happens we will probably have had another couple
of years of transition, and the problem would be much smaller.


> So, last year, simpdftex was introduced and altpdf*tex were retained
> because of backward compatibility. The frontend designers were informed
> that the change had taken place and they were requested to change their
> frontends to use simpdftex instead of altpdf*tex. It was then said that
> altpdf*tex would be retired later.

But why retire them at all as long as they work?

> The problem we are seeing now is that the front end designers (busy bees
> as they are) forgot about it and I did not remind them periodically).

This is not a question of frontend developers being busy bees: the question
is that the frontend is supposed to work with existing tetex distributions
(and not just the newest ones).  TeXShop and other frontends should not be
forced to require the user to have a less-than-one-year-old tetex
installation!

Perhaps you forget how stable a thing tex (and gwtex) is.  For 90 percent
of the tex users, there is no reason to upgrade, and there may be many old
installations around.  Certainly on this list, people are very advanced tex
users and we are all very thankful for the constant development and
improvements, and the new bells and whistles, but the average user (not
subscribed to this list) may not need more than standard OT1 encoded CM
fonts, and a few AMS packages.  This has been working perfectly well ever
since your first distribution back in the previous millenium.

By removing those scripts, the frontend writer wishing to use the scripts
is forced to either formally require a new version of tetex (gwtex) or
write complicated instructions in the manual or 'readme', explaining that
in certain distributions you have to do one thing, and in other
distributions you have to do some other thing, and it is akward to describe
precisely what versions because there are so many components involved and
it is not clear which version number you should refer to or where the user
can find this information.


> Re-introducing a back compatibility script would defeat the object of the
> transition completely so I am not going to do it.

Sorry for being stupid, but how does the error message work?  Isn't it a
script?  Why does this error script not defeat the objective of the
transition?  Why is it better to have an error script (breaking existing
functionality) than having the current symlinks?

I don't see any problem with the population of the bin/ directory.  Nobody
ever opens this directory anyway -- we are not talking a polished OSX
folder where the presence of more than six items would be intimidating!



Compare with some other famous transitions: you can still write
\documentstyle{article} in your source file and it will work.  You can
still say latex in the command line and it will invoke pdflatex with the
appropriate parameters...  Do these tricks defeat the objective of the
transitions?

> Backwards compatibility is not truly lost.  You can still compile all
> your old documents, the only thing that has changed is that the
> convenience script for running tex, dvips and distiller has been renamed
> and changed its command line interface.  In my book, backwards
> compatibility is not lost of you still can get the same results even if
> you have to do different things to get those results.  

Which book?  And which edition?  :-)

> But it is of course possible to disagree on the interpretation of
> backward compatibility.

The transitions I mentioned above represent different views on backward 
compatibility than yours.  But of course they were also more important
transitions...


Cheers,
Joachim.




-------------------------------------------------------------------------
Start.no tilbyr nå raskere bredbånd til lavere pris.
Sjekk http://www.start.no/bredband/ for mer informasjon
------------------------- Info --------------------------
Mac-TeX Website: http://www.esm.psu.edu/mac-tex/
          & FAQ: http://latex.yauh.de/faq/
TeX FAQ: http://www.tex.ac.uk/faq
List Archive: http://tug.org/pipermail/macostex-archives/