Re: [q-lang-users] Proposal for two minor rearrangements

[Albert G. <Dr....@t-...>]

John Cowan wrote:
> I think it better for the code to break than to suddenly run 20 or 200
> times slower because you excluded clib for some reason.

I don't agree with that. If you're running without clib, it's a
conscious decision, and most users who want to do that for some reason
will probably prefer the rest of the standard library to just keep
working without having to jump through a lot of hoops.

> After all, code will break if you exclude other standard modules, too.

That's true, but the rest of the stdlib has been specifically designed
to not depend on clib. (Well, that's not 100% true any more, but AFAICS
it's just a single function from clib which is extended in rational.q,
and you can easily comment that one out.)

The rationale behind this is, if someone wants to port Q to some exotic
Gizmo XYZ device which doesn't have a full POSIX interface and thus clib
might not even compile there for some reason, then you still have the
option to make a minimal Q implementation work there, without much ado.
I think that this is important enough to justify the cost of having a
little cruft in the library (it's just a few definitions anyway).

However, if that's a real showstopper for anyone, then I could add a
--with(out)-clib option to the configury. Opinions?

Albert

-- 
Dr. Albert Gr"af
Dept. of Music-Informatics, University of Mainz, Germany
Email:  Dr....@t-..., ag...@mu...
WWW:    http://www.musikinformatik.uni-mainz.de/ag