Re[2]: Modules

[Giuseppe Bilotta <bourbaki@bigfoot.com>]

Saturday, January 19, 2002 Marco Kuhlmann wrote:

MK> * Berend de Boer <berend@pobox.com> (2002-01-19 19:07:18 +0100):

>> 1. What is a well-defined module interface? You have \usemodule. A
>>    module is just any ConTeXt file for now.

MK> Exactly. But what about

MK>   - module parameters

ConTeXt is extremely parameter driven: just have a look at how
parameters are parsed by ConTeXt core code and use the same on
your module: define an appropriate \setupwhatever relative to your
function.

MK>   - standard \setup keywords

Have a look at mult-con.tex

MK>   - namespaces

I don't know. I suppose this means automatic namespaces; prefixing
internal variables with the module name would ensure uniqueness.

MK>   - usage of standard library macros

I'm not sure I follow this. What exactly do you mean by this?

>> 2. You can use any ConTeXt code, why should a module be limited?

MK> For portability reasons, for example. If there were some kind
MK> of standard library, then one would rather want to use that
MK> instead of self-hacked things, because it will still work and
MK> the sizes of the modules will be smaller. Of course, in
MK> general, you can do whatever you want. This is merely a
MK> question of maintenance and system integrity -- which LaTeX
MK> lacks, for example.

But there are lots of "tools" available from ConTeXt internals:
just have a look at the syst-* modules.

>> 3. Distribution mechanism isn't there, but I'm sure Hans is quite
>>    willing to put up links on his ConTeXt page to 3rd party
>>    stuff. Perhaps a good idea, what do you think Hans?

MK> It certainly would be. However, it should also be on CTAN. And
MK> Giuseppe's comments on things like TeXUtil hooks etc. are worth
MK> considering, in my opinion.

IIRC Hans already he wanted to set up an "other contributions" in
the ConTeXt site (I'd have to search the old posts ...)

And of course I do indeed think Giuseppe's comments are worth
considering ;-)

--
Giuseppe "Oblomov" Bilotta