I agree that ASDF 1.634 has reached the state of an ASDF 2 candidate, at least codewise if not documentationwise.
If no one objects, I'm OK with enabling #+asdf2.
Want me to do it, or can you do it and release?
[ François-René ÐVB Rideau | Reflection&Cybernethics | http://fare.tunes.org ] Majority, n.: That quality that distinguishes a crime from a law.
On 13 March 2010 16:55, Robert Goldman rpgoldman@sift.info wrote:
On 3/13/10 Mar 13 -11:35 AM, Robert Goldman wrote:
The member of *features* that indicates new asdf is
:asdf2
This is probably the source of the confusion.
Probably need a section of the manual on this.
I see that :asdf2 has been removed from *features*, according to the manual, until ASDF reaches 2.0 official release.
I would like to appeal this decision; I think it's a mistake.
The current 1.6xx releases of ASDF are effectively release candidates for ASDF 2.0.
While we are in testing, we need to be able to validate that ASDF2 and ASDF "classic" are compatible.
In particular, we need to start writing ASDF files with code like
#+asdf2 .... #-asdf2
<classic asdf compatibility code>
For example, I would like to test with
#+asdf2
<asdf-output-translations code here> #-asdf2 <asdf-binary-locations code here>
Removing :asdf2 makes this kind of testing impossible, and indeed makes it impossible pre-release to check expedients for backward compatibility of any kind.
I would like to revise the manual to say basically what I have said above and restore the
(pushnew :asdf2 *features*)
to asdf.lisp.
will this upset anyone?
thanks, r