This is the mail archive of the
guile@sources.redhat.com
mailing list for the Guile project.
Re: Language translation proposal 1.4
"Joshua Rosen" <rozzin@jello.geekspace.com> writes:
> in that model, the client *must* know the language that the module is
> written in
You can't plug your headphones into the power supply system. But you
*can* decide to plug it into your own cd player your neighbour's
stereo or video recorder, a tape recorder etc.
If you don't know the language a file contains, `file' will tell
you what it is.
> If a module decides, internally, what translator to use, then modules all
> `just look like modules', and using the module is simplified.
These are two different things. Of course it should be
possible to create new modules by linking other modules together.
But this has nothing to do with the implicit module connections
we're talking about.
> because the software using a module only sees the *interface* to the
> module, and knows nothing about the implementation of the module
> (...and what language
Don't forget the type that a module has [yes, okay, not in the current
module system]. (F <I1> <P>) is insensitive to the call (f <I42>
<P>). All you get is "no applicable method", meaning: You can't plug
the parameter module f into F.
All I am proposing is that the :language and :file options are in the
wrong place. Instead of (define-module m :language .. :file) they
belong to the (use-modules ...) clause. Simple fix, big effect.
Jost