Runtime determination of share directories?

Robert Dodier wrote:
> On 1/22/09, Raymond Toy <toy.raymond at gmail.com> wrote:
>
>   
>>  Wouldn't it be nicer if we could do this at startup and just find all of
>>  the directories in share directory?  Then we wouldn't have to maintain
>>  this list at all, and users can easily add share packages themselves and
>>  have them automatically available.  (Well, perhaps we could keep a few
>>  of the core share packages in the list, but most would be determined at
>>  runtime).
>>     
>
> Yes, that's a good idea. Have at it!
>   
I have a version(s) that works for cmucl, clisp, and ecl.  (The code
can't be the same because the directory function is not the same.)  Some
extra directories get included the lapack/fortran and lapack/lisp, but
that's probably ok.  If not, they're easy enough to prune.

>   
>>  I played with this a while back, but never finished it.  I think,
>>  however, that all of the lisps we support can do this using the Lisp
>>  DIRECTORY function with a bit of extra glue to return the right stuff.
>>  GCL might be a bit more difficult, but I think it can be done there too.
>>     
>
> I believe it is within the realm of possibility that we might stop
> working around GCL's idiosyncrasies.
>   
Kind of sad, but that's the way it goes.
> I think it would be worth the time to compare Maxima + ECL against
> Maxma + GCL on various platforms, considering ECL as a possible
> replacement for GCL. If someone wants to do that & report to the
> mailing list, that would be terrific.
>   
I have been regularly building with gcl and ecl, and occassionally run
the testsuite for all.  IIRC, there is only one unusual test result with
ecl.

I'll post the results later.

Ray