[tex-live] Perl for Windows

[gnwiii at gmail.com]

On 6/19/06, Reinhard Kotucha <reinhard.kotucha at web.de> wrote:
> >>>>> "Hans" == Hans Hagen <pragma at wxs.nl> writes:
>
>   >> It is pretty simple to do such things under Linux.
>   >>
>   > sure, and equally simple to do it under windows, however ... for
>   > who do we make tex live? windows (and i assume also mac) users
>   > want to run tthing swithout the need to revert to shell
>   > programming; fabrice made xemtex (fptex ++) for schools (teachers
>   > and such) and they (normally) want something that installs as easy
>   > as any program under windows: click a setup program, want some
>   > minutes, and get working; if you'd tell them -in case of
>   > conflixts- to open a shell (cmd) and key in some magic code, tex
>   > no longer is a valid option

TL can't meet the needs of all users.  What it can do is

1) strive to overcome barriers that make it harder for people like
Fabrice to create good distros for specific requirements (in
particular, most derviative distros end up being platform-specific),
and

2) strive to play well with existing user applications, including utilities like
gs, perl, etc.

Analyzing the problems that arise in Win32 is important, not just
because many TeX users want to use Win32, but because the things we
learn are likely to help with problems we will encounter when new
platforms appear.  The ability to use a platform tool after checking
for the required functionality is helpful because it provides
information that can be used to fix broken external tools and because
the tests will be useful in other situations (e.g., porting to a
totally new OS ten years from now).

This discussion is useful if it leads to some general agreement over
the approach to 3rd party tools:

0) if 3rd party tools aren't readily available through packages, they
can be considered for inclusion (e.g., sam2p)

1) don't break a user's existing tools unless the user agrees

2) try to use existing tools by checking against defined requirements
(and avoid tools where it is difficult to analyze the configuration).

3) strive for portability in core components

This thread deals mainly with points 1+2, but the others are important
because we need a replacement for teTeX, which violates (3) due to the
reliance on shell scripts.

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