Re: [Texmacs-dev] Current work, nogencc and tree in generic programming

[Joris van der Hoeven]

On Sun, 28 Apr 2002, Stéphane Payrard wrote:

> We badly need developpers and extenders.
> This means focusing on the core quality of TeXmacs -- the typesetting -- 
> and dropping the NIH stuff. THe NIH stuff may be superior to 
> what already exist (for example, using gennc will give faster compilation
>  and gives cleaner code than standard C++ templating)
> but it is an added cost when learning TeXmacs.
> 
> So, I think that David removal of gencc is a good move.

Yes, but like most good moves, it requires a lot of time,
and David and I are not done yet.

> Also it is the occasion to add 'const's to make the code faster.

Good idea, but *PLEASE*, everything in it's time.

> Personnally, like you may have seen from my nodes in the wiki, I am
> now focusing on using gtk as the GUI. It is the occasion to add
> perl as a scripting language. And to do it in way that facilate the inclusion
> of other scripting languages.

> Lisp is nice and popular in the mathematical
> communauty but the programmer communauty is  more interested in
> Ruby, Python and Perl.

Emacs uses Lisp and does seem to be quite popular too.
I do want to support Ruby, Python, Perl and whatever you like.
I will include all useful contributions in this respect.

But I also issue a warning: since I absolutely have no time for
thinking deeply about this now, I reserve myself the possibility
to make any changes in potential contributions on this subject
when my ideas will become clearer or to simply remove
solutions which I do not like.

I have to keep to my own developement priorities, which are
the result of interactions with many users and developers.
All contributions are welcome, but remember that I have to
check the consistency of everything: using several extension
languages in a consistent way is a complex affair.
It is not the priority for the next four months and
I may not have time to help you on this issue at all.

So: please do whatever you think is useful. If it works and
clearly does not raise major integration problems,
then I will include the contribution. But my time will
go for 95% to prioritary projects: if a non prioritary project
is too complex, I may not have time to help you or to support
the contribution immediately.

> BTW: that'w now more than three months I am poking the TeXmacs source
> and I am yet unable to find where to patch it to generate formulae on
> off-screen pixmaps. This shows either that I am slow witted or that
> there is a lot of material to cover before being able to contribute.

Yes, certain aspects (like the one you mention) are complex.
With the help of David, we are working to simplify and improve
the C++ code, but this is a long and delicate process.

This is why I try to direct developers to prioritary and feasable
projects, which can be fully understood in a reasonable amount of time.
Current examples are: CAS interfaces, converters, (soon) style files,
reorganization of menus and keyboard, the scheme interface, etc.
If you want to do something else, then you are basically on your own,
or with other developers which have more time than me.

I am sorry if my remarks deceive you, but at this point you must
understand that there are hundreds of issues to be covered inside TeXmacs
(simply take a look at the bug and wish lists); I cannot address them all 
at once and I need an agenda to which I should remain faithful as much as
possible.

What about the French documentation and the Html converters?