Raymond Wiker <Raymond.Wiker@...> writes:
> I missed out that there were two calls to (gnu)make in
> install.sh. Also, in my checked-out version of sbcl, there is an empty
> subtree sbc-aclrep (presumably the result of a keyboarding
> misadventure :-) Adding an explicit check for a Makefile and not just
> a directory solves this.
Thanks. With luck, install.sh and make-target-contrib.sh now both
behave sensibly as of sbcl-0.7.12.36.
What doesn't behave sensibly is binary-distribution.sh... what is a
binary distribution? It should probably have the contrib/ fasls, at
the very least, in it. (maybe, for asdf-using contribs, it should
have a concatenated fasl if possible...).
I've taken the liberty of committing, at the same time, my
implementation of <http://www.cliki.net/ROTATE-BYTE>, in the hope that
it'll be useful either as is or as an example of how to use the
compiler interfaces to optimize otherwise unoptimizeable code. If
anyone out there has, for instance, an md5 library, why not see if
using this contrib helps? :-)
http://www-jcsu.jesus.cam.ac.uk/~csr21/ +44 1223 510 299/+44 7729 383 757
(set-pprint-dispatch 'number (lambda (s o) (declare (special b)) (format s b)))
(defvar b "~&Just another Lisp hacker~%") (pprint #36rJesusCollegeCambridge)