- page 1 of 370
- next ->
|gh||Thu Oct 23 14:00:32 UTC 2014|
This bug is a pretty good example of why the "rebase internal patch" is a rather nasty hack - see 'Note [Rebase representation]' in src/Darcs/Patch/Rebase.hs
|ganesh||Tue Oct 21 18:24:11 UTC 2014|
Abstract the applyPatches function from the apply command, and make pull use it.
This means the following changes to apply:
It means the following changes to pull:
|ganesh||Tue Oct 21 16:08:25 UTC 2014|
|ganesh||Tue Oct 21 14:06:58 UTC 2014|
|gh||Tue Oct 21 14:13:36 UTC 2014|
|gh||Fri Oct 17 17:40:32 UTC 2014|
|ganesh||Thu Oct 16 17:43:17 UTC 2014|
|gh||Wed Oct 15 20:14:30 UTC 2014|
|gh||Wed Oct 15 21:05:34 UTC 2014|
|gh||Wed Oct 15 21:04:57 UTC 2014|
|gh||Tue Oct 14 12:18:32 UTC 2014|
|gh||Wed Oct 8 18:58:29 UTC 2014|
This test was introduced in 2006 to assert that darcs shouldn't exit with an error if stdout disappears while it's running (e.g. darcs changes | head).
It looks like at some point this test stopped being useful, because it wasn't actually producing enough output to actually fail, e.g. because of buffering on the other side of the pipe.
It also seems that at some point darcs did start exiting with an error in this scenario. This may have been due to GHC library changes in the handling of EPIPE - see GHC bugs #1619 and #2699 for some context.
Then the output format of 'darcs changes' was changed by "show patch hash in UI", which tipped the amount of output back over the edge and caused the test to fail.
In general the typical behaviour for UNIX commands is to exit with an error when their output pipe disappears. The exit code is typically 141 because the command receives SIGPIPE. In our either darcs or the GHC runtime or both are ignoring SIGPIPE, so the eventual exit code is 2 (see Darcs.Util.SignalHandler).
I think this behaviour is fairly reasonable so I am just deleting the test.
|ganesh||Sat Sep 27 14:30:18 UTC 2014|
|Sun Sep 28 10:28:35 UTC 2014|
|ganesh||Sat Sep 27 23:14:29 UTC 2014|
|ganesh||Sat Sep 27 22:54:50 UTC 2014|
|ganesh||Sat Sep 27 22:36:06 UTC 2014|
It causes problems on e.g. MacOSX which doesn't have many pkgconfig files.
|ganesh||Sat Sep 27 21:53:57 UTC 2014|
Interactively entered authors now end up in the global prefs file. To avoid clobbering the user's prefs if they run the script by hand, the test is now skipped if a global authors file exists.
|ganesh||Sat Sep 27 16:14:19 UTC 2014|
|Sat Sep 27 09:49:39 UTC 2014|
|Sat Sep 27 15:59:17 UTC 2014|
|ganesh||Sat Sep 27 09:32:49 UTC 2014|
This reduces the time spent preprocessing, cutting the time for a "null" cabal build from 17s to 12s for me.
|ganesh||Fri Sep 26 20:26:05 UTC 2014|
For some reason 'host' resolves to 127.0.53.53 for me right now, might be that something changed in the root zones or some behaviour of Google's DNS server.
|ganesh||Fri Sep 26 19:22:48 UTC 2014|
This avoids clashes with the Prelude's join and is a better name anyway Also rename various tests and utility functions to reflect this name change
|ganesh||Sat Sep 27 08:17:10 UTC 2014|
|ganesh||Sat Sep 27 07:52:55 UTC 2014|
|aristid||Fri Sep 26 17:39:05 UTC 2014|
|aristid||Fri Sep 26 16:43:34 UTC 2014|
|aristid||Sat Sep 27 07:37:53 UTC 2014|
|aristid||Fri Sep 26 16:45:01 UTC 2014|
- page 1 of 370
- next ->