- Dec 05, 2015
-
-
Eric Timmons authored
Package inferred systems now inherit around-compile from their top-level system. Added test in package-inferred-system-test suite.
-
Francois-Rene Rideau authored
- Dec 01, 2015
-
-
Francois-Rene Rideau authored
Undelete function load-sysdef, which must continue to exist as long as people use the current version of SLIME (to be fixed imminently as of 2015-12-01), though it is OK for it to just issue an error. Indeed, swank-asdf supposes that any backward-compatibility function it defines will exist forever, by unconditionally calling fdefinition on them. Now, it depends on load-asd, the backward-compatible definition of which (for asdf older than 2.26.149) depends on load-sysdef, that it also defines for versions older than 2.16.
- Nov 29, 2015
-
-
Francois-Rene Rideau authored
It's used by swank-asdf.
- Nov 28, 2015
-
-
Francois-Rene Rideau authored
This helps with e.g. not needlessly triggering find-system hooks, or invalidating the loadedness status because there was a system update. https://github.com/quicklisp/quicklisp-client/pull/122#issuecomment-159744105
-
- Nov 21, 2015
-
-
Francois-Rene Rideau authored
ccl, gcl, lispworks think that (probe-file #p".bashrc") has its pathname-name and pathname-type be NIL and "bashrc" whereas other implementations (at least abcl allegro clisp cmucl ecl mkcl sbcl xcl) think that it has them be ".bashrc" and NIL. cl -l $l -i "(let ((p (probe-file \".bashrc\"))) \ (format t \"FOO $l: ~S ~S~%\" (pathname-name p) (pathname-type p)))" \ |& grep FOO
-
Francois-Rene Rideau authored
And fix tests for exceptions.
- Nov 19, 2015
-
-
Francois-Rene Rideau authored
Fix DIRECTORY* on Genera (contribution by Gary Palter, lp#1518044): UIOP uses FS:DIRECTORY-LIST on Genera to enumerate the contents of the source repositories. However, if the directory doesn't exist, FS:DIRECTORY-LIST signals an error, whereas UIOP expects to get NIL.
-
Robert Goldman authored
* 'master' of common-lisp.net:asdf/asdf: Remove some ASDF 2 driver functions. Get rid of remnant ASDF 2 backward-compatibility Don't individually invalidate functions Bump version to 3.1.6.2 Fix make wc. Stop supporting old asdf/foo aliases for uiop. Use #+cmucl instead of #+cmu Move all code out of header.lisp Include 3.1.5 and 3.1.6 in default upgrade tags. bundle: drop support for abcl < 1.3.0 (2014-03-14) Fix typo in Makefile for upgrade methods
-
Robert Goldman authored
Fix for ticket 1517799.
- Nov 18, 2015
-
-
Francois-Rene Rideau authored
Not used in Quicklisp.
-
Francois-Rene Rideau authored
These backward-compatibility functions aren't required now that we wholly punt on upgrade from ASDF 2. Also not used by any system in Quicklisp.
-
Francois-Rene Rideau authored
It's not required anymore, since we now wholly punt when upgrading from an ancient ASDF 1 or 2, regardless of the implementation.
-
Francois-Rene Rideau authored
-
Francois-Rene Rideau authored
A few are kept, for backward-compatibility with systems in quicklisp.
-
Francois-Rene Rideau authored
It's been the official way of calling CMUCL since 2011.
-
Francois-Rene Rideau authored
-
Francois-Rene Rideau authored
-
Francois-Rene Rideau authored
-
Francois-Rene Rideau authored
-
Robert Goldman authored
-
Debian packaging will be continued shortly at git://git.debian.org/git/pkg-common-lisp/cl-asdf.git In the meantime, the branch 'debian' in this repo should contain the relevant files.
-
Francois-Rene Rideau authored
-
- Nov 17, 2015
-
-
Francois-Rene Rideau authored
Attila offered these tests, that currently fail at HEAD, for which I have a patch (to be included in next commit). They check that an error when loading defsystem-depends-on dependencies doesn't cause an incomplete system to be registered but never properly parsed, that thereafter pollute the build. A real solution would imply to fix lp#1500578 though: ASDF should handle change in defsystem-depends-on dependencies https://bugs.launchpad.net/asdf/+bug/1500578
-
Francois-Rene Rideau authored
-
Francois-Rene Rideau authored
Note that the method-qualifier is optional.
-
Francois-Rene Rideau authored
I tried to make them obsolete back in the 2.27 days, but too many people use them, and I now agree they are sometimes better than to have a separate defmethod form, notably for e.g. perform methods on test-op. Remove source-code comment that declared them obsolete, and update the explanation for the slot in defclass component.
-
Francois-Rene Rideau authored
I initially introduced all that make-pathname* madness on the road to ASDF 2.27, back when I was trying to normalize pathname components so I may have pathname-indexed entries in the asdf-cache (to minimize costly filesystem access from file-write-date and such, that could also introduce inconsistency in the build plan). Problem is, there was no good way (even non-portable) to normalize pathnames in a way that works when you use all of parse-namestring, make-pathname, directory, probe-file, translate-logical-pathname, merge-pathnames, etc. In the end (as late as b4dc78de, before 3.0.2.13!), I conceded defeat, for it was an uphill battle, and instead asdf-cache is indexing things using namestrings instead, which is more stable (see asdf/cache:normalize-namestring). The hacks in make-pathname* are therefore not needed anymore, and the function can be done away with. This commit guts the contents of make-pathname* and marks it deprecated. Just use make-pathname instead. It also updates uiop and asdf to not use make-pathname* but make-pathname. Some future commits may move it to backward-driver and eventually start issuing warnings of some sort when it's used.
- Nov 02, 2015
-
-
Francois-Rene Rideau authored
-
Francois-Rene Rideau authored
-
Francois-Rene Rideau authored