[asdf-devel] Testing with RMCL.... [OT]

[Robert Goldman]

On 4/22/12 Apr 22 -10:21 AM, Stelian Ionescu wrote:
> On Sun, 2012-04-22 at 10:18 -0500, Robert Goldman wrote:
>> On 4/22/12 Apr 22 -9:55 AM, Faré wrote:
>>> On Sun, Apr 22, 2012 at 08:54, Pascal Costanza <pc at p-cos.net> wrote:
>>>> Why is it that ASDF requires restarting lisp in a pristine state?
>>>>
>>> Mostly historical reasons. Its test suite predates most test-suite software.
>>> No one has rewritten it to use anything else.
>>> I would gladly accept a patch that replaces that with, say, fiveam or stefil.
>>> Proper use of clear-system and delete-package can help.
>>
>> I suppose it could be rewritten to do that.  I have never thought it was
>> worth the trouble, since running the tests would damage the state of any
>> lisp process in which it was run (because of changing the internal state
>> of ASDF), but such a rewrite would make ASDF testable on LW personal and
>> RMCL, which would be a worthwhile objective.
>>
>> Sadly, FiveAM seems to have been abandoned and is in need of a new
>> maintainer.  None of the bese projects darcs repos has been modified in
>> almost 2 years....
> 
> Somebody complained about that on bordeaux-threads-devel, so I picked it
> up: http://fenlix.dreamwidth.org/2255.html
> 

That's great.  Any chance of you taking over the fiveam page on
common-lisp.net and fixing it so it's no longer misleading?  Right now
it's a decoy pointer to bad stuff.

If Marco has given you ownership, taking over the common-lisp.net page
for bese.it would be a kindness, if only to point out the transfer.

[let's take this off-list now, if there's anything more.]

Best,
r