Context Navigation

← Previous Period
Next Period →

Timeline

Feb 23, 2013:

7:29 PM Changeset [f3ac1ae] by David Benjamin <davidben@mit.edu>: Merge 104a4eb952a68c48da10585deab68e021a648ca6 into bbe7d4a8ffa8c6c6b49ff192f54df4c385dfc1cb
7:28 PM Changeset [104a4eb]release-1.10 by David Benjamin <davidben@mit.edu>: Accept validsettings in new_variable_string If people reeeaaally want to, I suppose they could pass in "<path>" or something silly. Whatever.
5:44 PM Changeset [6a8b519]release-1.10 by David Benjamin <davidben@mit.edu>: Remove pointless return code from owl_variable_add_defaults
5:44 PM Changeset [4953c44]release-1.10 by David Benjamin <davidben@mit.edu>: Change set_fromstring to die on failure Returning zero/non-zero is awkward. Leave a note for future improvement to pass the error message along, although the default one is pretty reasonable already.
5:44 PM Changeset [4419d3d]release-1.10 by David Benjamin <davidben@mit.edu>: Add unit tests for perl-side variables
5:44 PM Changeset [f0a7c09]release-1.10 by David Benjamin <davidben@mit.edu>: Remove read-only variable check Those are impossible now and were impossible before this branch (except for the old "OWL_VARIABLE_OTHER" thing, which was unused). Yay deleting code.
5:44 PM Changeset [366badc]release-1.10 by David Benjamin <davidben@mit.edu>: Add some unit tests for enum variables So we don't break them again.
5:44 PM Changeset [bbc31e4]release-1.10 by David Benjamin <davidben@mit.edu>: Whitespace that I don't feel like rebasing Sigh. Tabs and spaces.
5:44 PM Changeset [e26abb6]release-1.10 by David Benjamin <davidben@mit.edu>: Add new_variable_enum This seems to work. Yay.
5:44 PM Changeset [523146b]release-1.10 by David Benjamin <davidben@mit.edu>: Add new_variable_full Route everything through that.

Feb 21, 2013:

4:42 PM Changeset [63d3e4e]release-1.10 by David Benjamin <davidben@mit.edu>: Make GClosure invokes more mechanical We don't really need some of those g_value_unsets, but we may as well avoid making assumptions. Someone else might copy-and-paste the code later and get it wrong.
4:42 PM Changeset [f41956d]release-1.10 by David Benjamin <davidben@mit.edu>: Merge owl_variable_free and owl_variable_cleanup We don't do the init/cleanup thing anymore.
4:42 PM Changeset [1a4508b]release-1.10 by David Benjamin <davidben@mit.edu>: Fix the unit tests The arguments got reordered slightly to match the nasty macros.
4:42 PM Changeset [4584d1f]release-1.10 by David Benjamin <davidben@mit.edu>: Pipe all C-created variables through function calls Gets us some long-overdue type checking in the various type-specific callbacks.

Feb 20, 2013:

1:06 AM Milestone BarnOwl 1.7 completed
1:06 AM Milestone BarnOwl 1.6 completed

Feb 19, 2013:

8:29 PM Changeset [3b9ca71]release-1.10 by David Benjamin <davidben@mit.edu>: Allow perl variables to participate in argumentless set/unset
8:29 PM Changeset [69f74c2]release-1.10 by David Benjamin <davidben@mit.edu>: And now... the moment you've all been waiting for... Expose OWL_VARIABLE_OTHER variables to perl. Rewrite all perl variable entry points to use that one. From now on, C only thinks of these variables as strings. (This has one regression to be fixed in a later commit: you can't set/unset with no arguments since we don't know how to set from boolean.)
8:29 PM Changeset [e5210c9]release-1.10 by David Benjamin <davidben@mit.edu>: dlsym the functions we need after use Glib and before use BarnOwl After use Glib so the library is loaded. Before use BarnOwl so we try to use them before they're available.
8:29 PM Changeset [897fc1c]release-1.10 by David Benjamin <davidben@mit.edu>: Fix error messages a bit
8:29 PM Changeset [5001a3d]release-1.10 by David Benjamin <davidben@mit.edu>: g_strdup validsettings This'll allow perl to set it too.
8:29 PM Changeset [353719a]release-1.10 by David Benjamin <davidben@mit.edu>: Start of perlvariables iteration N+1 New plan: the C <-> perl interface deals entirely with strings. Storage is entirely perl-side. C does not know and doesn't care about the underlying perl-side types. Note that this means an "int" variable in C land and an "int" variable in perl land have no relation.
8:28 PM Changeset [06e04a9]release-1.10 by David Benjamin <davidben@mit.edu>: perlvariables - iteration N Later commits reworked a lot of this, but the use of GValue over void* was kept.

Feb 18, 2013:

2:58 AM Changeset [b7a74a8] by Anders Kaseorg <andersk@mit.edu>: Merge bbe7d4a8ffa8c6c6b49ff192f54df4c385dfc1cb into 6a209963395d62d98c3d6b52daa6fb2e3054ae1b
2:53 AM Changeset [bbe7d4a]release-1.10 by Anders Kaseorg <andersk@mit.edu>: owl_zephyr_loadsubs: Don’t leak memory on error opening ~/.zephyr.subs Found by cppcheck. Signed-off-by: Anders Kaseorg <andersk@mit.edu>