[asdf-devel] Please test 2.20.16

Faré fahree at gmail.com
Sat Apr 21 20:13:01 UTC 2012


> Yeah, I was on the wrong branch.  With the master branch, make lisp
> fails because when cmucl compiles asdf, there are a few notes which I
> guess are warnings.  If you modify compile-asdf.lisp to ignore the
> warnings (like it already does for ecl and scl), then the tests can
> proceed.  At the end I get:
>
Interesting.
In 2.20.15 (last Monday), I had to ignore similar warnings on SCL.
There used to not be warnings.
No "WARNING" message shows, only compiler notes.
Are they the warnings? Isn't it a bug that
they should count as warnings rather than say mere style-warnings?
In any case, I've just disabled bork-on-warning on CMUCL as well as SCL.

> -#---------------------------------------
> Using cmulisp -noinit -batch
> Ran 39 tests:
>  39 passing and 0 failing
> all tests apparently successful
> -#---------------------------------------
>
> This is with CMU Common Lisp snapshot-2012-04 (20C Unicode) running on
> OSX 10.6.8.  This is, of course, the "official" build from common-lisp.net.
>
I don't see that snapshot on the download page;
but browsing the site by editing existing linked URLs, I find
http://common-lisp.net/project/cmucl/downloads/snapshots/2012/04/

And yes, it works for me, too, with the 2012-04 snapshot. Yay!
But as reported a few days ago, it was borking for me on vanilla 20c.

Can you link the working snapshot from the download page?
That beats offering a failing release as first choice.

> Where do you get you cmucl binaries from?  What OS are you running on?
>
I for one always use the unicode tarballs on Linux;
AFAIK, rpgoldman always uses the unicode tarballs on MacOS X.

An unidentified already fixed bug sure beats
one that's identified but unfixed.

—♯ƒ • François-René ÐVB Rideau •Reflection&Cybernethics• http://fare.tunes.org
What's funny with equality is how everyone has a different idea of it.  —#f




More information about the asdf-devel mailing list