Hi Piet,
On Mar 23, 2006, at 4:54 PM, Piet van Oostrum wrote:
>>>>>> Gregory Wright <gwright at comcast.net> (GW) wrote:
>
>> GW> Hi Piet,
>
>> GW> Yes, DP installs a shell script wrapper for sbcl. It's not a
>> bug,
>> GW> since there is at least one other piece of software in DP which
>> GW> requires the wrapper. (It's there so that 'sbcl' will run
>> sbcl the
>> GW> same way 'clisp' on the command line runs clisp.)
>
>> GW> I can check if the other software still requires this. I don't
>> off the
>> GW> top of my head remember what it was; I'll have to check through
>> GW> the old bug reports.
>
> As far as I could see, sbcl should do the proper thing without the
> wrapper,
> as the install procedure chooses the proper default sbcl.core file.
> If not,
> the shell script should check for any --core arguments and not
> include the
> default --core parameter if one is found. Otherwise the invocation
> sbcl
> --core somefile will not work as described in the man page. But as
> I said,
> I think it should work without the wrapper.
>
SBCL will work without the wrapper, it's something else that depends on
scbl that requires the wrapper. This situation is more common than one
might guess in darwinports. I either need to wrap sbcl or to patch
another
program. The wrapper is usually the least invasive. It's usually
not a problem
for more sophisticated users to figure out what is going on.
The wrapper, IIRC, was put in to make sure sbcl worked under one of
emacs'
lisp modes. If that was ilisp, I'm happy to get rid of the wrapper,
since
ilisp is dead. If it was one of the other lisp development
environments,
I'm not going to do it since that would make sbcl unusable in that
environment.
I'm not inclined to try to patch a lot of elisp code just to avoid
the wrapper.
So if the wrapper is truly unnecessary, I will remove it, otherwise
it will stay.
If it stays, I'll patch the manpage to reflect the difference.
This is one of the choices that are necessarily made in a packaging
system
where different developers make incompatible choices.
>> GW> Maxima 5.9.3 builds under darwinports using clisp by default
>> GW> and sbcl as a variant. The issue you note is automatically taken
>> GW> care of by the portfile. The portfile is not yet updated because
>> GW> maxima 5.9.3 does not yet work with TeXmacs. (I think the
>> TeXmacs
>> GW> fix is trivial, just updating the recognized version numbers, but
>> GW> it has to be tested.)
>
> I installed 5.9.3 manually. By the way, I don't see sbcl mentioned
> in the
> maxima Portfile.
The 5.9.3 portfile is not out yet. 5.9.3 causes TeXmacs to barf when
run as
a maxima front end. As I mentioned, the fix may be simple.
Best Wishes,
Greg
> --
> Piet van Oostrum <piet at cs.uu.nl>
> URL: http://www.cs.uu.nl/~piet [PGP 8DAE142BE17999C4]
> Private email: piet at vanoostrum.org
>
> _______________________________________________
> Maxima mailing list
> Maxima at math.utexas.edu
> http://www.math.utexas.edu/mailman/listinfo/maxima