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

Comparing Coro/Changes (file contents):
Revision 1.469 by root, Fri Nov 28 23:30:55 2008 UTC vs.
Revision 1.516 by root, Wed Jul 22 03:02:07 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
8TODO: CF_RUNNING makes sense only for Coros, not States 115.161 Wed Jul 22 04:47:38 CEST 2009
12 - Coro::AnyEvent::poll could have a different prototype when EV was
13 used as backend (analyzed by Tatsuhiko Miyagawa).
14 - Coro::AnyEvent errornously initialised the event loop when loaded,
15 not on demand.
16 - try to workaround rare */t/01_unblock.t failures.
17
185.16 Tue Jul 21 01:44:37 CEST 2009
19 - Coro::AnyEvent failed to hook into the event loop
20 when no threads had been readied between detecting
21 the event loop and actually running it.
22 - considerably speed up Coro::Select by taking avdantage
23 of AnyEvent > 4.8 and some other optimisations.
24 - implement paragraph readline mode in Coro::Handle
25 (based on patches by Zsbán Ambrus).
26 - replace WSAEINPROGRESS by WSAEWOULDBLOCK (reported
27 and analyzed by Yasuhiro MATSUMOTO).
28 - clarified libcoro license and copyright.
29 - someone stole my EXTRA_META!!!
30 - implement Coro::Select::patch_pp_sselect and it's brother,
31 for hardcode select overriding.
32
335.151 Mon Jul 6 05:41:57 CEST 2009
34 - backport to windows process emulation code again (patch by
35 Yasuhiro MATSUMOTO).
36 - slightly update Coro::MakeMaker.
37
385.15 Tue Jun 30 10:28:06 CEST 2009
39 - deprecate Coro::Socket, document how to get ipv6 support via
40 AnyEvent::Socket instead.
41 - implement signal->wait ($cb) interface, similar to semaphores.
42 - work around SvOK not supporting getmagic, so we have to getmagic
43 to test for undef :/ (reported by Matthias Waldorf).
44 - load Coro::AnyEvent in all modules using AnyEvent.
45 - work around perl corrupting our internal data structures,
46 reported by Tokuhiro Matsuno.
47 - enable per-coroutine real and cpu time gathering
48 (Coro::State::enable_times).
49
505.14 Wed Jun 24 01:37:48 CEST 2009
51 - provide explicit functions to "cede" to the event loop in Coro::AnyEvent,
52 as this seems to have been a difficult concept (poll, sleep, idle,
53 idle_upto).
54 - add Coro::AnyEvent::readable/writable functions.
55 - clarify Coro::EV/Event/AnyEvent manpages.
56 - free per-thread global scalars in the thread calling ->cancel, to
57 avoid crashes when $_, $@ etc., are magical but some of those
58 had already been freed.
59 - "unexperimentalise" the callback interface for Coro::Semaphore.
60 - speed up ready queue management/context switching by using a linked
61 list instead of an array (~5-10%).
62 - implement "watch" command in Coro::Debug shells.
63 - for fun, implement time-slicing as an example in the manpage.
64 - if AnyEvent detects EV or Event, but we don't have Coro::EV or
65 Coro::Event, use the normal AnyEvent handling instead of dieing
66 (the same is true for Coro::Handle).
67 - properly document Coro::EV::timed_io_once.
68 - avoid unneccessary ->cancel calls in Coro::Handle.
69 - maybe make it work on mingw32 with win32 backend
70 (based on patch by Yasuhiro Matsumoto).
71
725.132 Fri May 29 09:00:39 CEST 2009
73 - do not keep a reference to the argument itself in
74 Coro::Semaphore::guard, as it could change later.
75 - support SO_RCVBUF/SO_SNDBUF nonstandard Coro::Socket options,
76 should support a prepare callback.
77
785.131 Mon Mar 16 23:20:37 CET 2009
79 - implement and document Coro->suspend, Coro->resume.
80 - fix Coro::Select implementation to not (often) close
81 the passed file descriptors (testcase provided by pippijn).
82
835.13 Mon Dec 15 21:51:42 CET 2008
84 - EXPERIMENTAL: implement dynamic winds (on_enter/on_leave).
85 - don't set diehook to C<undef> but instead to NULL, to avoid
86 spurious warnings.
87 - fix a lot of bugs in Coro::SemaphoreSet.
88 - Coro::SemaphoreSet will less often create a semaphore needlessly.
89 - add Coro::SemaphoreSet::count and wait methods.
90 - take advantage of the new Guard module.
91 - deprecate Coro::guard.
92 - try to fix the dreaded 01_unblock tests once more. I hate it when
93 testsuites need more fixing than the code they are supposed to test.
94 - croak in more cases when a required callback isn't resolvable.
95 - fix some minor issues in Coro::State->call/eval.
96 - use current coroutine context instead of a temporary one
97 when temporarily switching to another coroutine.
98 - do not call C-level on_destroy handlers during global destruction,
99 to avoid needless segfaults.
100
1015.12 Sun Dec 7 13:30:38 CET 2008
9 - add default config for MirOS, which seems to be a bug-to-bug 102 - add default config for MirOS, which seems to be a bug-to-bug
10 compatible fork of openbsd ("world domination by releasing 103 compatible fork of openbsd ("world domination by releasing
11 openbsd cvs before the openbsd folks do it" or so :). 104 openbsd cvs before the openbsd folks do it" or so :).
12 - free_padlist did destroy the names pad, not good, but didn't 105 - free_padlist did destroy the names pad, not good, but didn't
13 seem to bother perl. 106 seem to bother perl - this could fix issues such as eval ""
107 inside a function called from multiple coroutines.
14 - use a different method to detect destruction time. 108 - use a different method to detect destruction time.
15 - be more careful when freeing padlists just before global 109 - be more careful when freeing padlists just before global
16 destruction. 110 destruction.
17 - fixed and expanded the call/cc example. 111 - fixed and expanded the call/cc example.
112 - renamed _terminate to _coro_run.
113 - new method Coro::Channel->shutdown.
114 - try pthreads on openbsd <4.4 (broken sigaltstack, will
115 pthreads fare better?).
116 - be less picky about destroying "a" running coroutine.
18 117
195.11 Tue Nov 25 21:49:05 CET 2008 1185.11 Tue Nov 25 21:49:05 CET 2008
20 - DEBUGGING in 5.10.0 is a mess: it constantly flags perfectly 119 - DEBUGGING in 5.10.0 is a mess: it constantly flags perfectly
21 working code with failed assertions, introducing more bugs than 120 working code with failed assertions, introducing more bugs than
22 it fixes, requiring elaborate workarounds :( 121 it fixes, requiring elaborate workarounds :(

Diff Legend

Removed lines
+ Added lines
< Changed lines
> Changed lines