From: Marco Antoniotti <marcoxa@cs...> - 2005-08-16 21:29:07
On Aug 16, 2005, at 5:09 PM, Sam Steingold wrote:
> The following message is a courtesy copy of an article
> that has been posted to gmane.lisp.clocc.general as well.
>> * Marco Antoniotti <znepbkn@...> [2005-08-16 16:20:34 -0400]:
>>> I would much rather have one de-facto standard SDF than two similar
>>> incompatible ones.
>> Of course. The right way to go ahead and do this is to write up a
>> spec first and then to go ahead and re-implement.
> I don't want to re-implement anything.
> I want to use one of the two existing facilities.
Well. Neither does TRT IMO. I just feel that MK:DEFSYSTEM despit its
clunkiness is still ahead.
>> All in all I have been debating myself whether to switch to ASDF.
> that does it.
>> Somehow, I just haven't. Plus, I have some semi-working code for
>> MK:DEFSYSTEM 5.0 that I have been using locally, but I am not
>> satisfied with that either and time is a tyrant.
> it appears that defsystem 3 is frozen and defsystem 4 is vaporware.
> now you are saying that you have v5?!
3.x is very stable. Apart from your EVAL message and the GCL crowd who
cannot get their head around the fact that ANSI got out in 1994 :)
there have not been complaints recently. And I know it is still been
used, so it is not a lack of users that can account for all the silence
v4 is effectively dead.
I have a v5 in my computer, but I have not released it at all. I
decided that I will never release anything until I have decent
documentation for it.
The main features are a truly open dependency checking and traversal
framework, plus the ability to compile C/C++/ Java/Fortran et al "out
of the box".
Unfortunately I always find that there are some more fundamental things
to do with CL than working on MK5.
Marco Antoniotti http://bioinformatics.nyu.edu
NYU Courant Bioinformatics Group tel. +1 - 212 - 998 3488
715 Broadway 10th FL fax. +1 - 212 - 998 3484
New York, NY, 10003, U.S.A.