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

Comparing AnyEvent/Changes (file contents):
Revision 1.15 by root, Wed Nov 1 01:22:19 2006 UTC vs.
Revision 1.116 by root, Mon May 26 17:45:05 2008 UTC

1Revision history for Perl extension AnyEvent. 1Revision history for Perl extension AnyEvent.
2 2
3TODO: testsuite, now that we can be sure that a model is available at all times 34.05 Mon May 26 19:44:06 CEST 2008
4 - some platforms ignore the rfc and prepend an extra sa_len member to
5 the sockaddr structure, cater for those. also use sockaddr_family,
6 didn't know it was there, but comes in handy for the workaround.
7 - undo the import WIN32 hack.
8
94.04 Mon May 26 08:03:31 CEST 2008
10 - try to work around yet another windows bug: failed connects are reported
11 as if it were out-of-band data. windows users: you suck. Only
12 EV, Glib, Event and the pure perl backend can handle this condition.
13 - optimize write algorithm in AnyEvent::Handle.
14 - properly parse PERL_ANYEVENT_PROTOCOLS and use the info correctly.
15 - tcp_connect now properly iterates through all targets.
16 - check wether IPv6 sockets can actually be created, otherwise, disable ipv6
17 support entirely. also hardcode some AF_INET6 constants for the
18 sake of perl 5.8.
19 - vastly improved nameserver/suffix detection algorithm for windows
20 (still a hack...)
21 - try to use a fallback dns server if no dns servers could be found.
22 - splatter around some local $SIG{__DIE__}'s to work around bad
23 die handlers.
24
254.03 Sun May 25 05:05:57 CEST 2008
26 - work around perl bugs on windows, where perl returns undocumented
27 error codes for sysread, syswrite etc.
28 - fix AnyEvent::Handle::unshift_read to really unshift and not push.
29 - AnyEvent::Handle could sometimes stumble over 0-byte writes.
30 - fix unpack_sockaddr, now works for AF_INET6 sockaddr's.
31 - updated documentation.
32 - lots of workarounds for perl 5.8 quirks.
33 - implement a more flexible type system for both reads and writes
34 in AnyEvent::Handle.
35 - support IPv6 in tcp_server.
36 - pass local hostname and port to the prepare callback in tcp_server.
37 - make types extendable in AnyEvent::Handle.
38 - croak, not die, in AnyEvent::Handle;
39
404.0 Sat May 24 19:58:08 CEST 2008
41 - added (experimental) AnyEvent::DNS module.
42 - added (experimental) AnyEvent::Socket module.
43 - added (experiemntal) support for TLS/SSL in AnyEvent::Handle.
44 - make condvar->end call ->send as documented.
45 - make condvar constructor support the documented cb => argument.
46 - overload condvars so you can use them instead of code references.
47 - support more of the POSIX forms instead of a pure dotted quad for
48 verbatim ip addresses.
49 - AnyEvent::Socket::inet_aton will use AnyEvent::DNS now and not fork anymore.
50 - try very hard not to freeze on broken windows perls.
51 - make on_eof optional in AnyEvent::Handle.
52 - support service names in tcp_connect and tcp_server.
53 - work around netbsd/openbsd bug causing unix sockets not to
54 become writable untill empty.
55 - only make a dummy read for error detection when getpeername
56 returns ENOTCONN.
57 - defend AnyEvent::Util::inet_aton against Coro::LWP brutal
58 override.
59 - new AnyEvent::Guard::cancel.
60
613.5 Sat May 17 23:17:14 CEST 2008
62 - gracefully deal with systems not defining CLOCK_MONOTONIC
63 (as opposed to not supporting the clock itself, which worked in 3.41).
64 - added AnyEvent::Util::tcp_connect and AnyEvent::Util::tcp_server
65 helpers, to replace the ill-designed AnyEvent::Socket class.
66 - added some examples (eg/connect, eg/handle).
67 - added AnyEvent::Util::guard.
68 - AnyEvent::Util now exports most of its functions.
69 - fixed a bug in *_read_chunk, where the callback was called with $self as
70 first and second argument.
71 - simplified and fixed regex handling in AnyEvent::Handle::*_read_line.
72 - fix manpage condvars examples to use new syntax.
73 - considerably improve inet_aton with EV::ADNS backend to support
74 cname chains (common dns configuration bug, e.g. with www.google.de)
75 and aaaa records if no a record could be found.
76
773.41 Sun May 11 19:53:13 CEST 2008
78 - work around a bug in perl 5.8.8's local. Reported by Yi Ma Mao.
79
803.4 Sun May 11 00:29:25 CEST 2008
81 - complete re-implementation of condvars with a hopefully much mroe useful
82 API (backward compatible functions exist but are undocumented).
83 - AnyEvent::Handle fully rewritten with a hopefully more powerful API
84 (still subject to change).
85 - take advantage of CLOCK_MONOTONIC, if available, in AnyEvent::Impl::Perl.
86 - provide AnyEvent::post_detect and @AnyEvent::post_detect, which
87 allows module authors to avoid forcing event loop detection.
88 (used by Coro::AIO, Coro::BDB, Coro::AnyEvent for example).
89 - remove coro backends: Coro now provides generic support for AnyEvent
90 via Coro::AnyEvent.
91
923.3 Mon Apr 28 09:51:06 CEST 2008
93 - added AnyEvent::Handle, AnyEvent::Socket and AnyEvent::Util modules.
94 - fix a bug in the pure perl backend that kept watchers alive
95 when multiple watchers were registered for the same fh/poll combo.
96 - add a benchmark section showing AnyEvent overhead and comparing
97 the different event loops with each other.
98 - prefer pure perl over tk when autoprobing, as it's about as fast,
99 but doesn't crash with many watchers.
100 - declare Qt support non-experimental.
101 - clarify the confusing section about the file descriptor being
102 kept alive.
103 - document the race between loading of an event module and
104 child processes exiting.
105 - support POE as "backend" (with some caveats, POE is not generic
106 enough, and darn slow).
107 - support Wx and Prima through POE.
108 - optimise perl backend to use 20% less memory and take advantage
109 of typical timeout behaviour. It can now compete with
110 select/poll-based C event loops in most cases (it is usually
111 faster than Event and Glib :).
112 - roughly cut EV memory use in half and increase its speed by 30%,
113 by removing undocumented functionality. Did something similar
114 to other event loops where possible.
115
1163.2 Thu Apr 24 10:10:40 CEST 2008
117 - do not die when anyevent watchers are destroyed while running callbacks
118 in the pureperl backend (could only happen when two watchers are registered
119 for the same fh/poll combo).
120 - support autoloading for child watchers, was broken in all versions.
121 - implement PERL_ANYEVENT_MODEL env variable.
122 - (experimental) implement interface to the (very crashy/buggy) Event::Lib module.
123 - (experimental) implement interface to the Qt module (cannot be autoprobed).
124 - this release is IO:AnyEvent-proof.
125
1263.12 Tue Apr 22 07:11:46 CEST 2008
127 - reinstate AUTHOR section that got lost somehow.
128 - do not hang in the testsuite with badly broken perls (activestate,
129 strawberry...), but instead diagnose the problem and continue.
130 - use INT instead of CHLD in an attempt to support broken windows perls
131 better (this decreases the test quality, unfortunately).
132 - do not send the signal to the process group
133 (no problem for CHLD, bad for INT :).
134
1353.11 Sat Apr 19 06:57:31 CEST 2008
136 - major documentation rework.
137 - document the fact that child watchers only watch for zombies.
138 - fix the child watcher example.
139
1403.1 Wed Apr 16 17:09:01 CEST 2008
141 - work around recurring bugs in Tk by dup'ing filehandles, the
142 only method with good success chances on Tk (the bugs apparently
143 don't get fixed anytime soon).
144 - lift the restriction of only one watcher per fh direction
145 (as the Tk bug workaround also lifts it and only Tk imposed
146 such strong limits).
147 - changed probe order to prefer coro adaptors.
148 - explain why recursion into the event loop is not supported
149 unless the backend supports it (only Coro::EV does without
150 any restrictions...).
151 - add simple manpages for all backend modules.
152
1533.0 Mon Apr 7 21:30:23 CEST 2008
154 - Coro::Signal changed semantics, roll our own, also cleaning
155 up the Coro implementation in general.
156 - rename Coro backend to CoroEvent.
157 - add some decision helping paragraph to the manpage that should
158 help people to decide whether AnyEvent is the right thing for them.
159
1602.9 Mon Jan 28 13:31:54 CET 2008
161 - update for EV 3.0 API changes.
162
1632.8 Sun Nov 25 15:06:03 CET 2007
164 - waitpid can validly return 0. accept this fact of life
165 instead of reporting it to any watchers.
166
1672.7 Fri Nov 23 11:41:14 CET 2007
168 - force use of AnyEvent::Impl::Perl in testsuite, there is
169 too much breakage outside AnyEvent.
170 - deliver signals synchronously in AnyEvent::Impl::Perl
171 or any other event loop that relies on AnyEvents child
172 watcher emulation. *Could* help with hanging testsuite
173 (except when Event or EV are installed).
174
1752.6 Fri Nov 9 20:36:35 CET 2007
176 - fix bug in testsuite.
177 - move EV adaptor modules to AnyEvent.
178 - add Coro+EV adaptor module.
179
1802.55 Tue Nov 6 17:41:32 CET 2007
181 - add EV to the list of supported event models.
182 - do not auto-reset pid watchers, pass pid and status to them.
183 - allow a pid of zero to watch for all children in child watchers.
184
1852.54 Wed Jul 18 17:36:23 CEST 2007
186 - work around a perl bug that results in
187 BEGIN not safe after errors--compilation aborted without
188 any discernible reason or error message by once
189 again not use'ing strict.
190
1912.53 Sun Jul 8 10:51:53 CEST 2007
192 - make Glib interface work again (spotted by elmex).
193
1942.52 Wed Mar 7 18:36:16 CET 2007
195 - child watchers stopped working permanently when
196 all current child watchers were destroyed.
197
1982.51 Mon Dec 11 21:33:24 CET 2006
199 - work around bugs in perl where eval "require Module" returns
200 true even if the module couldn't be loaded :/.
201
2022.5 Mon Dec 11 02:15:28 CET 2006
203 - avoid $AUTOLOAD because many perls corrupt it.
204 - AnyEvent::detect forces autodetection.
205 - implement signal watchers (experimental)
206 (Tk does not support async signals (<= 804.027 at least),
207 so they might get delayed indefinitely on Tk).
208 - implement child watchers (experimental).
209 - moved default condvar implementation into base module, simplifying
210 most implementation modules.
211
2122.1 Fri Nov 24 15:50:48 CET 2006
213 - better docs.
214 - simple testsuite.
215 - added AnyEvent->one_event method for special purposes.
4 216
52.0 Wed Nov 1 02:21:30 CET 2006 2172.0 Wed Nov 1 02:21:30 CET 2006
6 - INCOMPATIBLE CHANGE: poll can now either be "r" or "w" but not both. 218 - INCOMPATIBLE CHANGE: poll can now either be "r" or "w" but not both.
7 - INCOMPATIBLE CHANGE: io watcher callbacks have no arguments anymore. 219 - INCOMPATIBLE CHANGE: io watcher callbacks have no arguments anymore.
8 - EXPERIMENTAL: 220 - EXPERIMENTAL:
… …
37 - fixed documentation. 249 - fixed documentation.
38 - clarified some corner cases. 250 - clarified some corner cases.
39 - vastly improved documentation with a long example. 251 - vastly improved documentation with a long example.
40 252
410.2 Thu Dec 1 22:18:49 CET 2005 2530.2 Thu Dec 1 22:18:49 CET 2005
42 - check wether a valid method was called, to avoid endless loops. 254 - check whether a valid method was called, to avoid endless loops.
43 - move condvars into their own class (cleaner, safer). 255 - move condvars into their own class (cleaner, safer).
44 - don't die when some modules are not available. 256 - don't die when some modules are not available.
45 - Tk is too broken to allow for words. Will work once Tk works. 257 - Tk is too broken to allow for words. Will work once Tk works.
46 258
470.1 Thu Dec 1 19:44:57 CET 2005 2590.1 Thu Dec 1 19:44:57 CET 2005

Diff Legend

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