- Dec 04, 2016
-
-
Francois-Rene Rideau authored
-
- Dec 02, 2016
-
-
Francois-Rene Rideau authored
-
Francois-Rene Rideau authored
Make a proper forward reference to run-program in the degenerate version of launch-program::%posix-send-signal used on some implementations.
-
- Dec 01, 2016
-
-
Robert Goldman authored
Split launch-program.lisp off of run-program.lisp run-program.lisp was way too big. Even split in two, each part is a mindful. See merge request !56
-
Robert Goldman authored
uiop/run-program: fix abcl version identifier for development releases Without this patch, Elias' recent work on UIOP/RUN-PROGRAM won't be utilized on any ABCL which is built from development source. ABCL uses the convention that development versions start appending strings separated via #\- characters to the primary value returned by CL:LISP-IMPLEMENTATION-VERSION (e.g. '1.5.0-dev'). Such values cause the UIOP/UTILITY:PARSE-VERSION function to return nil, meaning that this is not a suitable conditional for whether LAUNCH-PROGRAM is invoked. This patch uses the value of `UIOP/OS:IMPLEMENTATION-IDENTIFIER` to identify version. Unknown whether this would work on MKCL/ECL, which might simplify the code path here. See merge request !57
-
Francois-Rene Rideau authored
- Nov 30, 2016
-
-
Mark authored
-
Robert Goldman authored
Bad system name Signal a WARNING when a system is defined with a bad name that will make asdf fail to find it, e.g. foo-test in foo.asd instead of e.g. foo/test. See merge request !51
-
Robert Goldman authored
Use offset to ensure that touch changes the timestamp in a predictable way.
-
- Nov 29, 2016
-
-
Mark authored
-
- Nov 27, 2016
-
-
Francois-Rene Rideau authored
-
Francois-Rene Rideau authored
It was part of a bug report, superseded by test/test-xach-update-bug.script.
-
Francois-Rene Rideau authored
-
Francois-Rene Rideau authored
No operation initargs This MR build upon !33 and wholly eliminate the non-sensical and unsupported initargs to operation. A casualty is `make-build`, as inherited from asdf-ecl (2005 era) and apparently still used by ECL. @dkochmanski will be pissed, but he'll have to learn about using `program-op` and `program-system` instead — which are supported (and using secondary systems, if for some reason he doesn't want to use `program-system` with primary systems). While we're pissing off @dkochmanski, I'm also removing the trivial and useless function `build-system` in the same directory. @epipping I don't think you're interested in reviewing defsystem changes, so I'm assigning directly to @rgoldman for review. @rgoldman considering the controversy, you may want to either merge fast before next release and be done with it, or make an announcement at this release and do it for the next release, which lets the wound fester longer, but gives more time for users to adapt (assuming there's more than a single one, which is not that obvious). See merge request !34
-
- Nov 26, 2016
-
-
Francois-Rene Rideau authored
Push monolithic-op to the back of inheritance, to preserve class-allocated slot bundle-type slot in a class in front of the inheritance list. Now that make-build was remove, adjust some ECL tests to use program-op instead.
-
Francois-Rene Rideau authored
Call finalize-inheritance in coerce-class on Allegro. Indeed the previous commit uses class meta-objects rather than their naming symbols as canonical representation for operations; now mark-component-preloaded is called at compile-toplevel and uses those representations, but on Allegro the classes haven't been finalized yet, and the finalization isn't implicitly triggered by subtypep. Moreover, comparing a class object to a symbol with subtypep fails, so normalize the super variable to a class object.
-
Francois-Rene Rideau authored
ASDF never supported operation initargs, and its component-operation-times always assumed that two operations of the same class were equivalent. Make it explicitly so and enforce it. Do NOT pass keys from operate to make-instance operation. Catch any attempt to create an operation with initargs and raise an error. Make sure no operation class has any slot that isn't :allocation :class (with some MOP magic we could do it in at class definition time, but that's not portable enough and we don't want to reimplement closer-mop or depend on it.) This is a preliminary to allowing CONS instead of NODE-FOR to identify actions, whether or not we want to allow some form of initargs in the future. Remove unused COMPILE-OP-FLAGS and COMPILE-OP-PROCLAMATIONS. Remove MAKE-BUILD that depended on the unsupported operation initargs. Use PROGRAM-OP on a PROGRAM-SYSTEM instead. Remove the confusingly misnamed and not-so-useful BUILD-SYSTEM function; it can be done better in userland.
-
- Nov 25, 2016
-
-
Robert Goldman authored
Fix typo This commit corrects a typo in the document string of `uiop/filesystem:probe-file*`. See merge request !55
-
Robert Goldman authored
Run program windows Various fixes for Windows. See merge request !52
-
Robert Goldman authored
-
Robert Goldman authored
RUN-SHELL-COMMAND requires string commands to RUN-PROGRAM to function, and these are not supported on the Windows version of ECL. Signal NOT-SUPPORTED-ERROR.
-
Robert Goldman authored
Previous error string implied that UIOP *never* supported string commands on Windows, but actually we do support them on some lisp implementations.
-
- Nov 24, 2016
-
-
dkim authored
-
- Nov 23, 2016
-
-
Francois-Rene Rideau authored
Always go through CMD in that case, including on Allegro and CLISP, since the previous behavior is simultaneously non-portable, error-prone and limited; better to provide a portable well-defined abstraction. Users who want more can use the (bad) implementation-specific interfaces, or IOlib. Simplify the call to RUN-PROGRAM. Add a PARAMETER-ERROR to %NORMALIZE-COMMAND on unsupported platforms.
-
Elias Pipping authored
We cannot use `exit` on Windows with :force-shell nil (it's a builtin); use `cmd /c exit` instead. Using `exit` on Unix with :force-shell nil may or may not work (it's a builtin but CL implementations typically invoke a shell anyway). `true` and `false` are guaranteed to exist by the POSIX standard.
-
- Nov 22, 2016
-
-
Francois-Rene Rideau authored
-
Elias Pipping authored
-
Robert Goldman authored
Documentation tweaks See merge request !54
-
Francois-Rene Rideau authored
-
Robert Goldman authored
Use of PATHNAME-MATCH caused spurious test failures on clisp.
-
Robert Goldman authored
-
Elias Pipping authored
With ECL it causes problems. On other platforms it should not be necessary. This gets run-program with ECL on windows far closer to working again
-
Elias Pipping authored
-
Francois-Rene Rideau authored
This requires a patch to sb-impl::prepare-args relative to sbcl 1.3.10.
-
Francois-Rene Rideau authored
-