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

Comparing AnyEvent/Changes (file contents):
Revision 1.32 by root, Fri Nov 9 19:35:26 2007 UTC vs.
Revision 1.114 by root, Mon May 26 06:18:53 2008 UTC

1Revision history for Perl extension AnyEvent. 1Revision history for Perl extension AnyEvent.
2 2
32.6 3 - undo the import WIN32 hack.
4
54.04 Mon May 26 08:03:31 CEST 2008
6 - try to work around yet another windows bug: failed connects are reported
7 as if it were out-of-band data. windows users: you suck. Only
8 EV, Glib, Event and the pure perl backend can handle this condition.
9 - optimize write algorithm in AnyEvent::Handle.
10 - properly parse PERL_ANYEVENT_PROTOCOLS and use the info correctly.
11 - tcp_connect now properly iterates through all targets.
12 - check wether IPv6 sockets can actually be created, otherwise, disable ipv6
13 support entirely. also hardcode some AF_INET6 constants for the
14 sake of perl 5.8.
15 - vastly improved nameserver/suffix detection algorithm for windows
16 (still a hack...)
17 - try to use a fallback dns server if no dns servers could be found.
18 - splatter around some local $SIG{__DIE__}'s to work around bad
19 die handlers.
20
214.03 Sun May 25 05:05:57 CEST 2008
22 - work around perl bugs on windows, where perl returns undocumented
23 error codes for sysread, syswrite etc.
24 - fix AnyEvent::Handle::unshift_read to really unshift and not push.
25 - AnyEvent::Handle could sometimes stumble over 0-byte writes.
26 - fix unpack_sockaddr, now works for AF_INET6 sockaddr's.
27 - updated documentation.
28 - lots of workarounds for perl 5.8 quirks.
29 - implement a more flexible type system for both reads and writes
30 in AnyEvent::Handle.
31 - support IPv6 in tcp_server.
32 - pass local hostname and port to the prepare callback in tcp_server.
33 - make types extendable in AnyEvent::Handle.
34 - croak, not die, in AnyEvent::Handle;
35
364.0 Sat May 24 19:58:08 CEST 2008
37 - added (experimental) AnyEvent::DNS module.
38 - added (experimental) AnyEvent::Socket module.
39 - added (experiemntal) support for TLS/SSL in AnyEvent::Handle.
40 - make condvar->end call ->send as documented.
41 - make condvar constructor support the documented cb => argument.
42 - overload condvars so you can use them instead of code references.
43 - support more of the POSIX forms instead of a pure dotted quad for
44 verbatim ip addresses.
45 - AnyEvent::Socket::inet_aton will use AnyEvent::DNS now and not fork anymore.
46 - try very hard not to freeze on broken windows perls.
47 - make on_eof optional in AnyEvent::Handle.
48 - support service names in tcp_connect and tcp_server.
49 - work around netbsd/openbsd bug causing unix sockets not to
50 become writable untill empty.
51 - only make a dummy read for error detection when getpeername
52 returns ENOTCONN.
53 - defend AnyEvent::Util::inet_aton against Coro::LWP brutal
54 override.
55 - new AnyEvent::Guard::cancel.
56
573.5 Sat May 17 23:17:14 CEST 2008
58 - gracefully deal with systems not defining CLOCK_MONOTONIC
59 (as opposed to not supporting the clock itself, which worked in 3.41).
60 - added AnyEvent::Util::tcp_connect and AnyEvent::Util::tcp_server
61 helpers, to replace the ill-designed AnyEvent::Socket class.
62 - added some examples (eg/connect, eg/handle).
63 - added AnyEvent::Util::guard.
64 - AnyEvent::Util now exports most of its functions.
65 - fixed a bug in *_read_chunk, where the callback was called with $self as
66 first and second argument.
67 - simplified and fixed regex handling in AnyEvent::Handle::*_read_line.
68 - fix manpage condvars examples to use new syntax.
69 - considerably improve inet_aton with EV::ADNS backend to support
70 cname chains (common dns configuration bug, e.g. with www.google.de)
71 and aaaa records if no a record could be found.
72
733.41 Sun May 11 19:53:13 CEST 2008
74 - work around a bug in perl 5.8.8's local. Reported by Yi Ma Mao.
75
763.4 Sun May 11 00:29:25 CEST 2008
77 - complete re-implementation of condvars with a hopefully much mroe useful
78 API (backward compatible functions exist but are undocumented).
79 - AnyEvent::Handle fully rewritten with a hopefully more powerful API
80 (still subject to change).
81 - take advantage of CLOCK_MONOTONIC, if available, in AnyEvent::Impl::Perl.
82 - provide AnyEvent::post_detect and @AnyEvent::post_detect, which
83 allows module authors to avoid forcing event loop detection.
84 (used by Coro::AIO, Coro::BDB, Coro::AnyEvent for example).
85 - remove coro backends: Coro now provides generic support for AnyEvent
86 via Coro::AnyEvent.
87
883.3 Mon Apr 28 09:51:06 CEST 2008
89 - added AnyEvent::Handle, AnyEvent::Socket and AnyEvent::Util modules.
90 - fix a bug in the pure perl backend that kept watchers alive
91 when multiple watchers were registered for the same fh/poll combo.
92 - add a benchmark section showing AnyEvent overhead and comparing
93 the different event loops with each other.
94 - prefer pure perl over tk when autoprobing, as it's about as fast,
95 but doesn't crash with many watchers.
96 - declare Qt support non-experimental.
97 - clarify the confusing section about the file descriptor being
98 kept alive.
99 - document the race between loading of an event module and
100 child processes exiting.
101 - support POE as "backend" (with some caveats, POE is not generic
102 enough, and darn slow).
103 - support Wx and Prima through POE.
104 - optimise perl backend to use 20% less memory and take advantage
105 of typical timeout behaviour. It can now compete with
106 select/poll-based C event loops in most cases (it is usually
107 faster than Event and Glib :).
108 - roughly cut EV memory use in half and increase its speed by 30%,
109 by removing undocumented functionality. Did something similar
110 to other event loops where possible.
111
1123.2 Thu Apr 24 10:10:40 CEST 2008
113 - do not die when anyevent watchers are destroyed while running callbacks
114 in the pureperl backend (could only happen when two watchers are registered
115 for the same fh/poll combo).
116 - support autoloading for child watchers, was broken in all versions.
117 - implement PERL_ANYEVENT_MODEL env variable.
118 - (experimental) implement interface to the (very crashy/buggy) Event::Lib module.
119 - (experimental) implement interface to the Qt module (cannot be autoprobed).
120 - this release is IO:AnyEvent-proof.
121
1223.12 Tue Apr 22 07:11:46 CEST 2008
123 - reinstate AUTHOR section that got lost somehow.
124 - do not hang in the testsuite with badly broken perls (activestate,
125 strawberry...), but instead diagnose the problem and continue.
126 - use INT instead of CHLD in an attempt to support broken windows perls
127 better (this decreases the test quality, unfortunately).
128 - do not send the signal to the process group
129 (no problem for CHLD, bad for INT :).
130
1313.11 Sat Apr 19 06:57:31 CEST 2008
132 - major documentation rework.
133 - document the fact that child watchers only watch for zombies.
134 - fix the child watcher example.
135
1363.1 Wed Apr 16 17:09:01 CEST 2008
137 - work around recurring bugs in Tk by dup'ing filehandles, the
138 only method with good success chances on Tk (the bugs apparently
139 don't get fixed anytime soon).
140 - lift the restriction of only one watcher per fh direction
141 (as the Tk bug workaround also lifts it and only Tk imposed
142 such strong limits).
143 - changed probe order to prefer coro adaptors.
144 - explain why recursion into the event loop is not supported
145 unless the backend supports it (only Coro::EV does without
146 any restrictions...).
147 - add simple manpages for all backend modules.
148
1493.0 Mon Apr 7 21:30:23 CEST 2008
150 - Coro::Signal changed semantics, roll our own, also cleaning
151 up the Coro implementation in general.
152 - rename Coro backend to CoroEvent.
153 - add some decision helping paragraph to the manpage that should
154 help people to decide whether AnyEvent is the right thing for them.
155
1562.9 Mon Jan 28 13:31:54 CET 2008
157 - update for EV 3.0 API changes.
158
1592.8 Sun Nov 25 15:06:03 CET 2007
160 - waitpid can validly return 0. accept this fact of life
161 instead of reporting it to any watchers.
162
1632.7 Fri Nov 23 11:41:14 CET 2007
164 - force use of AnyEvent::Impl::Perl in testsuite, there is
165 too much breakage outside AnyEvent.
166 - deliver signals synchronously in AnyEvent::Impl::Perl
167 or any other event loop that relies on AnyEvents child
168 watcher emulation. *Could* help with hanging testsuite
169 (except when Event or EV are installed).
170
1712.6 Fri Nov 9 20:36:35 CET 2007
4 - fix bug in testsuite. 172 - fix bug in testsuite.
5 - move EV adaptor modules to AnyEvent. 173 - move EV adaptor modules to AnyEvent.
6 - add Coro+EV adaptor module. 174 - add Coro+EV adaptor module.
7 175
82.55 Tue Nov 6 17:41:32 CET 2007 1762.55 Tue Nov 6 17:41:32 CET 2007
… …
77 - fixed documentation. 245 - fixed documentation.
78 - clarified some corner cases. 246 - clarified some corner cases.
79 - vastly improved documentation with a long example. 247 - vastly improved documentation with a long example.
80 248
810.2 Thu Dec 1 22:18:49 CET 2005 2490.2 Thu Dec 1 22:18:49 CET 2005
82 - check wether a valid method was called, to avoid endless loops. 250 - check whether a valid method was called, to avoid endless loops.
83 - move condvars into their own class (cleaner, safer). 251 - move condvars into their own class (cleaner, safer).
84 - don't die when some modules are not available. 252 - don't die when some modules are not available.
85 - Tk is too broken to allow for words. Will work once Tk works. 253 - Tk is too broken to allow for words. Will work once Tk works.
86 254
870.1 Thu Dec 1 19:44:57 CET 2005 2550.1 Thu Dec 1 19:44:57 CET 2005

Diff Legend

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