Idris Samawi Hamid ادريس سماوي حامد wrote:

On Fri, 26 Dec 2008 09:47:15 -0700, وفا خلیقی، Vafa Khalighi wrote:

...

We should have a sample complex bidirectional text which contain everything in it (boxes, LTR and RTL texts, poems, colors, diagrams, footnotes and ...) and then compare the produced outputs.

Of course

...

Please also note that I am using LaTeX not context.

As Hartmut points out, our tests at the end of the day need to be done with primitives alone.

it depends what you test; primitive behaviour or typesetting; you can only show real typeset text when you use some kind of high level wrapper (load fonts, deal with language aspects); tex itself is just machinery; it's like a car: we can compare wheels and engines but it's the car that has to bring you from a to b ----------------------------------------------------------------- Hans Hagen | PRAGMA ADE Ridderstraat 27 | 8061 GH Hasselt | The Netherlands tel: 038 477 53 69 | fax: 038 477 53 74 | www.pragma-ade.com | www.pragma-pod.nl -----------------------------------------------------------------

...

Furthermore you all talk about LTR and RTL, but what about vertical typesetting, in the Chinese or in >the Mongolian way? Obviously TeX-e-Parsi doesn't care about that.

It's true. and I think Omega did not handle that well either.

Actually, Omega did a reasonable job, the bottleneck was that no conforming font was available. The examples I sent in my other mail use exclusively Omega features, apart for the part where I play with the font callback (the 2nd step in my file). I did experiments with Manju in Omega half a dozen years ago, using Oliver Corff's MonTeX / ManjuTeX system as a starting point. I wrote OTP, I modified the Type 1 fonts and the metrics and it worked fine, but the files have long been lost. Arthur