[omega] Draft of paper for EuroTeX 2005

Yannis Haralambous yannis.haralambous at enst-bretagne.fr
Tue Mar 1 13:58:06 CET 2005


> I do wonder about the implementation.  This is so far from TeX, I
> surmise that it's basically writing a new program from scratch.  Sure,
> assorted algorithms from TeX can be re-used, but actual code?  Not that
> this is necessarily a bad thing, in and of itself.  (Indeed, one might
> argue it's long overdue -- all those global variables, etc.)  However,
> we know that it's devilishly difficult to retain compatibility with TeX
> without actually starting with tex.web.

Well it's actually not *that* far from TeX. Internally we are using 
data structures for signs. Char_nodes, lig_nodes and disc_nodes are 
replaced by sign nodes, but for the moment we haven't touched any other 
type of node. We would like this to be a series of short modifications 
of TeX's code. [I must admit I'm afraid of projects where everything 
has to be rewritten from scratch]. As for compatibility with TeX, we 
are introducing a new strata of information which is orthogonal with 
the other ones, the only precaution to take is to be sure that in the 
absence of sign information we still get the same result, and this 
shouldn't be very hard. The hard part IMHO will be to retrieve 
information from OpenType fonts and insert it into sign nodes so that 
the paragraph builder doesn't need to search for ligature and kerning 
programs.

> If you have plans along these lines you can share, I'd be most
> interested in hearing.

At some time we will release a highly experimental version with very 
limited features. But since we have to take a lot of decisions we would 
like to discuss the ideas presented in the paper. Also it would be 
interesting to hear from LaTeX people what the possible LaTeX 
applications of signs could be.

>
> Best,
> karl
>
>
--
+--------------------------------------------------------------------+
| Yannis Haralambous, Ph.D.      yannis.haralambous at enst-bretagne.fr |
| Directeur d'Études                   http://omega.enstb.org/yannis |
|                                          Tel. +33 (0)2.29.00.14.27 |
|                                          Fax  +33 (0)2.29.00.12.82 |
| Département Informatique                                           |
| École Nationale Supérieure des Télécommunications de Bretagne      |
| Technopôle de Brest Iroise, CS 83818, 29238 Brest CEDEX 3, France  |
+--------------------------------------------------------------------+
                          ...pour distinguer l'extérieur d'un aquarium,
                                         mieux vaut n'être pas poisson

                         ...the ball I threw while playing in the park
                                        has not yet reached the ground



More information about the omega mailing list