ViewVC Help
View File | Revision Log | Show Annotations | Download File
/cvs/Coro/Changes
(Generate patch)

Comparing Coro/Changes (file contents):
Revision 1.476 by root, Sun Dec 7 15:33:21 2008 UTC vs.
Revision 1.507 by root, Wed Jul 1 11:33:25 2009 UTC

… …
2 2
3TODO: should explore PerlIO::coroaio (perl leaks like hell). 3TODO: should explore PerlIO::coroaio (perl leaks like hell).
4TODO: maybe implement a default message channel, very much like Erlang's 4TODO: maybe implement a default message channel, very much like Erlang's
5 actor model (which is cool in a lot of important aspects (failures!), 5 actor model (which is cool in a lot of important aspects (failures!),
6 but very lacking in others (higher level ipc)). 6 but very lacking in others (higher level ipc)).
7TODO: unready_all
8TODO: myhttpd header parsing
9TODO: channel->maxsize(newsize)?
7 10
8new -> ready 11 - backport to windows process emulation code again (patch by
9new -> canceled 12 Yasuhiro MATSUMOTO).
10ready -> running 13 - slightly update Coro::MakeMaker.
11ready -> canceled
12running -> ready
13running -> blocked
14running -> canceled
15blocked -> ready
16blocked -> canceled
17 14
155.15 Tue Jun 30 10:28:06 CEST 2009
16 - deprecate Coro::Socket, document how to get ipv6 support via
17 AnyEvent::Socket instead.
18 - implement signal->wait ($cb) interface, similar to semaphores.
19 - work around SvOK not supporting getmagic, so we have to getmagic
20 to test for undef :/ (reported by Matthias Waldorf).
21 - load Coro::AnyEvent in all modules using AnyEvent.
22 - work around perl corrupting our internal data structures,
23 reported by Tokuhiro Matsuno.
24 - enable per-coroutine real and cpu time gathering
25 (Coro::State::enable_times).
26
275.14 Wed Jun 24 01:37:48 CEST 2009
28 - provide explicit functions to "cede" to the event loop in Coro::AnyEvent,
29 as this seems to have been a difficult concept (poll, sleep, idle,
30 idle_upto).
31 - add Coro::AnyEvent::readable/writable functions.
32 - clarify Coro::EV/Event/AnyEvent manpages.
33 - free per-thread global scalars in the thread calling ->cancel, to
34 avoid crashes when $_, $@ etc., are magical but some of those
35 had already been freed.
36 - "unexperimentalise" the callback interface for Coro::Semaphore.
37 - speed up ready queue management/context switching by using a linked
38 list instead of an array (~5-10%).
39 - implement "watch" command in Coro::Debug shells.
40 - for fun, implement time-slicing as an example in the manpage.
41 - if AnyEvent detects EV or Event, but we don't have Coro::EV or
42 Coro::Event, use the normal AnyEvent handling instead of dieing
43 (the same is true for Coro::Handle).
44 - properly document Coro::EV::timed_io_once.
45 - avoid unneccessary ->cancel calls in Coro::Handle.
46 - maybe make it work on mingw32 with win32 backend
47 (based on patch by Yasuhiro Matsumoto).
48
495.132 Fri May 29 09:00:39 CEST 2009
50 - do not keep a reference to the argument itself in
51 Coro::Semaphore::guard, as it could change later.
52 - support SO_RCVBUF/SO_SNDBUF nonstandard Coro::Socket options,
53 should support a prepare callback.
54
555.131 Mon Mar 16 23:20:37 CET 2009
56 - implement and document Coro->suspend, Coro->resume.
57 - fix Coro::Select implementation to not (often) close
58 the passed file descriptors (testcase provided by pippijn).
59
605.13 Mon Dec 15 21:51:42 CET 2008
61 - EXPERIMENTAL: implement dynamic winds (on_enter/on_leave).
18 - don't set diehook to C<undef> but instead to NULL, to avoid 62 - don't set diehook to C<undef> but instead to NULL, to avoid
19 spurious warnings. 63 spurious warnings.
64 - fix a lot of bugs in Coro::SemaphoreSet.
65 - Coro::SemaphoreSet will less often create a semaphore needlessly.
66 - add Coro::SemaphoreSet::count and wait methods.
67 - take advantage of the new Guard module.
68 - deprecate Coro::guard.
69 - try to fix the dreaded 01_unblock tests once more. I hate it when
70 testsuites need more fixing than the code they are supposed to test.
71 - croak in more cases when a required callback isn't resolvable.
72 - fix some minor issues in Coro::State->call/eval.
73 - use current coroutine context instead of a temporary one
74 when temporarily switching to another coroutine.
75 - do not call C-level on_destroy handlers during global destruction,
76 to avoid needless segfaults.
20 77
215.12 Sun Dec 7 13:30:38 CET 2008 785.12 Sun Dec 7 13:30:38 CET 2008
22 - add default config for MirOS, which seems to be a bug-to-bug 79 - add default config for MirOS, which seems to be a bug-to-bug
23 compatible fork of openbsd ("world domination by releasing 80 compatible fork of openbsd ("world domination by releasing
24 openbsd cvs before the openbsd folks do it" or so :). 81 openbsd cvs before the openbsd folks do it" or so :).
25 - free_padlist did destroy the names pad, not good, but didn't 82 - free_padlist did destroy the names pad, not good, but didn't
26 seem to bother perl - this could fix issues such as eval "" 83 seem to bother perl - this could fix issues such as eval ""
27 inside a function called form multiple coroutines. 84 inside a function called from multiple coroutines.
28 - use a different method to detect destruction time. 85 - use a different method to detect destruction time.
29 - be more careful when freeing padlists just before global 86 - be more careful when freeing padlists just before global
30 destruction. 87 destruction.
31 - fixed and expanded the call/cc example. 88 - fixed and expanded the call/cc example.
32 - renamed _terminate to _coro_run. 89 - renamed _terminate to _coro_run.

Diff Legend

– Removed lines
+ Added lines
< Changed lines
> Changed lines