- Apr 22, 2014
-
-
Francois-Rene Rideau authored
-
Francois-Rene Rideau authored
-
- Apr 20, 2014
-
-
Francois-Rene Rideau authored
-
Francois-Rene Rideau authored
-
Francois-Rene Rideau authored
- Apr 16, 2014
-
-
Francois-Rene Rideau authored
-
Francois-Rene Rideau authored
- Apr 15, 2014
-
-
Robert P. Goldman authored
-
Robert Goldman authored
-
Robert Goldman authored
-
Robert P. Goldman authored
-
Robert P. Goldman authored
New semantics is that a component with no specified version fails all version-satisfies tests.
-
Robert P. Goldman authored
-
Robert P. Goldman authored
Previously, if a component's version slot was unbound, it would match *all* version specifications. This is obviously wrong, because it means you can't detect when a system goes from unversioned to versioned. Changed this to an interpretation that an unbound version slot matches NO version specifications.
-
Robert Goldman authored
- Apr 14, 2014
-
-
Francois-Rene Rideau authored
-
Francois-Rene Rideau authored
-
Robert P. Goldman authored
Bundle operations don't work on Mac OS X on either lispworks (can't restart the image properly) or on ECL (build fails, I believe because of an ECL build bug).
-
Robert P. Goldman authored
Previously, we would detect Unix first (correctly), then short-circuit, and never detect Mac OSX. I would like to rewrite DETECT-OS. The code seems opaque. But this is correct, so committing it now, since it used to be broken.
-
Francois-Rene Rideau authored
-
- Apr 13, 2014
-
-
Francois-Rene Rideau authored
-
Francois-Rene Rideau authored
Instead, refactor how we call make-hello-world.
-
Francois-Rene Rideau authored
for bugs in the make-hello-world subprocess.
-
Francois-Rene Rideau authored
-
- Apr 07, 2014
-
-
Francois-Rene Rideau authored
-
Francois-Rene Rideau authored
Better tests. Thanks to Robert Brown.
-
Francois-Rene Rideau authored
- Apr 06, 2014
-
-
Francois-Rene Rideau authored
-
Francois-Rene Rideau authored
- Mar 28, 2014
-