Book logo xindy

A Flexible Indexing System

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: xindy for DOS ?

	Hi Joergen,

> For years I have been using my own index processor for processing my
> LaTeX indexes.  It is certainly not very advanced or flexible, but it
> has one great advantage: it knows the Danish sorting rules.
> But now it seems that I should switch to xindy!

Indeed, xindy has been developed to overcome many of the
language-specific problems with makeindex, and we are currently
exploring further enhancements for international support.

> But as were are using DOS computers here at my department, the
> crucial question is: is someone porting xindy to DOS ?
> (or, since the homepage text hints that the answer is no, maybe I
> should ask: how difficult is it to make a DOS version?)

We have no working port to DOS, yet. The problem is that xindy is
mostly written in Lisp with some parts implemented in C for
performance reasons. The implementation was done on UNIX boxes first,
but with the other platforms in mind. The Lisp we use runs on DOS as
well, this is why we have chosen it. The OS/2 port is finished and
will be shipped out next week, hopefully. As far as I know, OS/2
supports cross-compiling to DOS and Windows to some extend and I think
that based on our experience with OS/2 there seem to be few problems
with a port which will require a 32bit extender in all cases, thus
from i386 upwards.

You are the second one asking for a DOS-port and I'm still looking for
people doing this port. Anybody who has already ported UNIX-software
to DOS using the GNU C compiler will be able to do the port in one or
two days with help from me and probably the OS/2 coordinator Joerg
Diederich. The OS/2 port took a little bit longer, but Joerg has done
a good piece of work. So if there are any volunteers in your
department, tell me :) I personally will not be able to do so, since
I'm not experienced in DOS-programming and have not enough resources
to do so (we started a first attempt at the German TeX conference and
it looked promising, though we had to stop it due to lack of time).
Only a machine with enough memory (let's say 16MB) is needed to do the

I think it's worth the effort and the system itself has already
reached a very stable state and only few problems have been reported
so far, thus it's ready for use.

Best regards


Roger Kehr
Computer Science Department          Technical University of Darmstadt