[tex-live] TL and win32

[gnwiii at gmail.com]

On 12/22/06, Siep Kroonenberg <siepo at cybercomm.nl> wrote:

> I agree with Staszek that you have to be careful not to interfere
> with a possible existing Ghostscript installation.
>
> Have a look at the way MikTeX incorporates its own private copy of
> Ghostscript: it is renamed mgs.exe, ignores environment variables
> and registry settings and needs to get all its configuration
> information from the command-line. And it uses fonts from the texmf
> tree.

MikTeX's mgs has been a major source of confusion.  Part of the
problem was that the URW fonts were available in a single "legacy"
package as well as in separate packages.  Depending on which packaging
you choose, the Fontmap.miktex file needs to be adjusted, but if you
happen to have ghostscript installed in c:\gs, then mgs would use the
fontmap
and fonts from there because mgs didn't remove the old default gs
library path.  While 90% of the problems can be solved by using the
Fontmap.miktex that matches the installed URW fonts, there are many
people with eps figures that render with gs (sometimes on a different
platform) but not mgs due to font errors.

You can't only think about Windows in isolation.   It is very common
to have multi-author documents where some authors are using Win32 and
others are using linux.  On linux they probably have Valek Flipov's
fonts with cyrillic glyphs as system fonts and being used for figures
and pdf images created outside tex.   We can't expect documents with
figures that can be formatted on one platform to work on another.
What we need is:

1.  better tools to diagnose font problems resulting from
substitutions (Adobe MM fonts used for Times/Helvetica on one system,
URW fonts used on another, different versions of URW fonts used in
different apps or on different platforms, ...)

2.  fixes and workarounds for common situations (e.g, Win32 user
collaborating with linux user)

-- 
George N. White III <aa056 at chebucto.ns.ca>
Head of St. Margarets Bay, Nova Scotia