Re: https://github.com/Clozure/ccl/releases
Grégory Vanuxem <[email protected]> Mon, 6 May 2024 05:55:51 +0200
| Newsgroups | gmane.lisp.openmcl.devel |
|---|---|
| Message-ID | <CAHnU2dZzSThpNqumqNbVU0_rPHm9Qpb72DVyep50f8bCd77Gew@mail.gmail.com> |
Absolutely no knowledge* Le lun. 6 mai 2024 =C3=A0 05:49, Gr=C3=A9gory Vanuxem <[email protected]>= a =C3=A9crit : > > Hello folks, > > The first time I wanted to try out Clozure CL I postponed its use > because I was lacking concrete information to build and/or install > CCL. > I just rebuilt a new image with code cloned from GitHub, and looking > at the documentation of how to build CCL I remembered what I was > asking myself at that time. Even if I did what I will say I was > unsure. I had absolutely knowledge about Clozure CL, just that it was > a Common Lisp implementation that conforms to ANSI standards. > > So if you don't know like me at that time, the name of the Clozure CL > executable, I think it could be handy to replace in the file mentioned > in object: > > -- After unpacking the binaries, you can rebuild CCL by evaluating > (rebuild-ccl :full t) as usual. > ++ After unpacking the binaries, you can rebuild CCL by evaluating > (rebuild-ccl :full t) with the usual lx${arch} executable. > > Or something like that. It is really not a common name for an > executable, I guess some history is involved here, but even a shell > script, imagine CCL is installed in /usr/bin for example, you have to > use 'uname' or something else to evaluate CL code by CCL. This is not > very UNIX-likish. > > In fact I did not try CCL, and only tried it one or two years later, > when I discovered 'roswell' [1] that helped me to use it > conventionally, that is with help of someone who really knows Clozure > CL and really knows how to build it. And even with 'roswell', it is > not possible to build it, contrary to SBCL for example. > > Just my two cents, > > - Greg > > [1] https://roswell.github.io/