Re: [Axiom-developer] Some latex packages that may help with multiple individual files in books

[C Y]

--- Ralf Hemmecke <address@hidden> wrote:

> 1) If you need to edit the pamphlet files to make them work in the 
> context of a big book. That sounds as if you have too much time to
> spare.

Not really.  I take it you regard the idea as frivolous?
 
> 2) Where are our priorities? Improving LaTeX or improving
> computational mathematics?

As a literate project I would think typesetting technology and
conventations are very important to our effort.  Also different members
of the project with different skill sets may have varying priorities -
those less skilled as mathematical researchers can attack other
problems.

> I could imagine to impose restrictions on the LaTeX packages that are
> approved be used inside an Axiom pamphlet. But do you know all the
> LaTeX packages and the preferences of authors that you feel able to
> decide which package to allow and which package to forbid? I don't.

Surely it could be discussed as it came up.  I don't imagine there
would be more than a dozen or so key packages (some of which are
already on our radar screen) and if over time a few more came up we
could deal with it on a case by case basis.

> Hmm, yes, you could call it a bug. In fact, I know LaTeX not well
> enough to understand why it is so difficult to set up a wrapper file
> around a collection of latex files so that each file would appear
> (nearly) identical to the way it would appear as dvi when compiled
> separately.

Because latex packages sometimes change the way output is produced from
commands.  If two different packages redefine the same command, there
is a collision and one of the papers won't get what it expects.  More
difficult, sometimes there are mutually exclusive options which must be
resolved one way or another between packages.

> The collection should, of course produce running page numbers.

There are also desirable features such as a common table of contents
and index.


> What I don't believe is that you can convince package authors to take
> every effort to keep their package compatible with any other package
> out there. Some people just write a package because they need it for
> their work and then suddenly it becomes useful for others. The was
> probably only a question that the package need to be compatible with
> packages that the original author used.

Probably.  But then, for the purposes of Axiom, why not take that
package to the next level and make it a more general tool?  I fully
expect that Axiom will eventually push the bounds of LaTeX's
typesetting abilities, and I think this is a good thing.  Human
readable mathematical documents are a very interesting problem in
themselves - Knuth solved many of those problems well but some things
like automatic line breaking are still imperfectly solved, at best.

Anyway.  It's possible I am underestimating the difficulties involved,
but the only way to know for sure is experimentation.

Cheers,
CY

__________________________________________________
Do You Yahoo!?
Tired of spam?  Yahoo! Mail has the best spam protection around 
http://mail.yahoo.com