ViewVC Help
View File | Revision Log | Show Annotations | Download File
/cvs/AnyEvent/Changes
Revision: 1.112
Committed: Mon May 26 06:03:20 2008 UTC (15 years, 11 months ago) by root
Branch: MAIN
Changes since 1.111: +1 -0 lines
Log Message:
*** empty log message ***

File Contents

# Content
1 Revision history for Perl extension AnyEvent.
2
3 - try to work around yet another windows bug: failed connects are reported
4 as if it were out-of-band data. windows users: you suck. Only
5 EV, Glib, Event and the pure perl backend can handle this condition.
6 - optimize write algorithm in AnyEvent::Handle.
7 - properly parse PERL_ANYEVENT_PROTOCOLS and use the info correctly.
8 - tcp_connect now properly iterates through all targets.
9 - check wether IPv6 sockets can actually be created, otherwise, disable ipv6
10 support entirely. also hardcode some AF_INEt6 constants for the
11 sake of perl 5.8.
12 - vastly improved nameserver/suffix detection algorithm for windows
13 (still a hack...)
14 - try to use a fallback dns server if no dns servers could be found.
15 - splatter around some local $SIG{__DIE__}'s to work around bad
16 die handlers.
17
18 4.03 Sun May 25 05:05:57 CEST 2008
19 - work around perl bugs on windows, where perl returns undocumented
20 error codes for sysread, syswrite etc.
21 - fix AnyEvent::Handle::unshift_read to really unshift and not push.
22 - AnyEvent::Handle could sometimes stumble over 0-byte writes.
23 - fix unpack_sockaddr, now works for AF_INET6 sockaddr's.
24 - updated documentation.
25 - lots of workarounds for perl 5.8 quirks.
26 - implement a more flexible type system for both reads and writes
27 in AnyEvent::Handle.
28 - support IPv6 in tcp_server.
29 - pass local hostname and port to the prepare callback in tcp_server.
30 - make types extendable in AnyEvent::Handle.
31 - croak, not die, in AnyEvent::Handle;
32
33 4.0 Sat May 24 19:58:08 CEST 2008
34 - added (experimental) AnyEvent::DNS module.
35 - added (experimental) AnyEvent::Socket module.
36 - added (experiemntal) support for TLS/SSL in AnyEvent::Handle.
37 - make condvar->end call ->send as documented.
38 - make condvar constructor support the documented cb => argument.
39 - overload condvars so you can use them instead of code references.
40 - support more of the POSIX forms instead of a pure dotted quad for
41 verbatim ip addresses.
42 - AnyEvent::Socket::inet_aton will use AnyEvent::DNS now and not fork anymore.
43 - try very hard not to freeze on broken windows perls.
44 - make on_eof optional in AnyEvent::Handle.
45 - support service names in tcp_connect and tcp_server.
46 - work around netbsd/openbsd bug causing unix sockets not to
47 become writable untill empty.
48 - only make a dummy read for error detection when getpeername
49 returns ENOTCONN.
50 - defend AnyEvent::Util::inet_aton against Coro::LWP brutal
51 override.
52 - new AnyEvent::Guard::cancel.
53
54 3.5 Sat May 17 23:17:14 CEST 2008
55 - gracefully deal with systems not defining CLOCK_MONOTONIC
56 (as opposed to not supporting the clock itself, which worked in 3.41).
57 - added AnyEvent::Util::tcp_connect and AnyEvent::Util::tcp_server
58 helpers, to replace the ill-designed AnyEvent::Socket class.
59 - added some examples (eg/connect, eg/handle).
60 - added AnyEvent::Util::guard.
61 - AnyEvent::Util now exports most of its functions.
62 - fixed a bug in *_read_chunk, where the callback was called with $self as
63 first and second argument.
64 - simplified and fixed regex handling in AnyEvent::Handle::*_read_line.
65 - fix manpage condvars examples to use new syntax.
66 - considerably improve inet_aton with EV::ADNS backend to support
67 cname chains (common dns configuration bug, e.g. with www.google.de)
68 and aaaa records if no a record could be found.
69
70 3.41 Sun May 11 19:53:13 CEST 2008
71 - work around a bug in perl 5.8.8's local. Reported by Yi Ma Mao.
72
73 3.4 Sun May 11 00:29:25 CEST 2008
74 - complete re-implementation of condvars with a hopefully much mroe useful
75 API (backward compatible functions exist but are undocumented).
76 - AnyEvent::Handle fully rewritten with a hopefully more powerful API
77 (still subject to change).
78 - take advantage of CLOCK_MONOTONIC, if available, in AnyEvent::Impl::Perl.
79 - provide AnyEvent::post_detect and @AnyEvent::post_detect, which
80 allows module authors to avoid forcing event loop detection.
81 (used by Coro::AIO, Coro::BDB, Coro::AnyEvent for example).
82 - remove coro backends: Coro now provides generic support for AnyEvent
83 via Coro::AnyEvent.
84
85 3.3 Mon Apr 28 09:51:06 CEST 2008
86 - added AnyEvent::Handle, AnyEvent::Socket and AnyEvent::Util modules.
87 - fix a bug in the pure perl backend that kept watchers alive
88 when multiple watchers were registered for the same fh/poll combo.
89 - add a benchmark section showing AnyEvent overhead and comparing
90 the different event loops with each other.
91 - prefer pure perl over tk when autoprobing, as it's about as fast,
92 but doesn't crash with many watchers.
93 - declare Qt support non-experimental.
94 - clarify the confusing section about the file descriptor being
95 kept alive.
96 - document the race between loading of an event module and
97 child processes exiting.
98 - support POE as "backend" (with some caveats, POE is not generic
99 enough, and darn slow).
100 - support Wx and Prima through POE.
101 - optimise perl backend to use 20% less memory and take advantage
102 of typical timeout behaviour. It can now compete with
103 select/poll-based C event loops in most cases (it is usually
104 faster than Event and Glib :).
105 - roughly cut EV memory use in half and increase its speed by 30%,
106 by removing undocumented functionality. Did something similar
107 to other event loops where possible.
108
109 3.2 Thu Apr 24 10:10:40 CEST 2008
110 - do not die when anyevent watchers are destroyed while running callbacks
111 in the pureperl backend (could only happen when two watchers are registered
112 for the same fh/poll combo).
113 - support autoloading for child watchers, was broken in all versions.
114 - implement PERL_ANYEVENT_MODEL env variable.
115 - (experimental) implement interface to the (very crashy/buggy) Event::Lib module.
116 - (experimental) implement interface to the Qt module (cannot be autoprobed).
117 - this release is IO:AnyEvent-proof.
118
119 3.12 Tue Apr 22 07:11:46 CEST 2008
120 - reinstate AUTHOR section that got lost somehow.
121 - do not hang in the testsuite with badly broken perls (activestate,
122 strawberry...), but instead diagnose the problem and continue.
123 - use INT instead of CHLD in an attempt to support broken windows perls
124 better (this decreases the test quality, unfortunately).
125 - do not send the signal to the process group
126 (no problem for CHLD, bad for INT :).
127
128 3.11 Sat Apr 19 06:57:31 CEST 2008
129 - major documentation rework.
130 - document the fact that child watchers only watch for zombies.
131 - fix the child watcher example.
132
133 3.1 Wed Apr 16 17:09:01 CEST 2008
134 - work around recurring bugs in Tk by dup'ing filehandles, the
135 only method with good success chances on Tk (the bugs apparently
136 don't get fixed anytime soon).
137 - lift the restriction of only one watcher per fh direction
138 (as the Tk bug workaround also lifts it and only Tk imposed
139 such strong limits).
140 - changed probe order to prefer coro adaptors.
141 - explain why recursion into the event loop is not supported
142 unless the backend supports it (only Coro::EV does without
143 any restrictions...).
144 - add simple manpages for all backend modules.
145
146 3.0 Mon Apr 7 21:30:23 CEST 2008
147 - Coro::Signal changed semantics, roll our own, also cleaning
148 up the Coro implementation in general.
149 - rename Coro backend to CoroEvent.
150 - add some decision helping paragraph to the manpage that should
151 help people to decide whether AnyEvent is the right thing for them.
152
153 2.9 Mon Jan 28 13:31:54 CET 2008
154 - update for EV 3.0 API changes.
155
156 2.8 Sun Nov 25 15:06:03 CET 2007
157 - waitpid can validly return 0. accept this fact of life
158 instead of reporting it to any watchers.
159
160 2.7 Fri Nov 23 11:41:14 CET 2007
161 - force use of AnyEvent::Impl::Perl in testsuite, there is
162 too much breakage outside AnyEvent.
163 - deliver signals synchronously in AnyEvent::Impl::Perl
164 or any other event loop that relies on AnyEvents child
165 watcher emulation. *Could* help with hanging testsuite
166 (except when Event or EV are installed).
167
168 2.6 Fri Nov 9 20:36:35 CET 2007
169 - fix bug in testsuite.
170 - move EV adaptor modules to AnyEvent.
171 - add Coro+EV adaptor module.
172
173 2.55 Tue Nov 6 17:41:32 CET 2007
174 - add EV to the list of supported event models.
175 - do not auto-reset pid watchers, pass pid and status to them.
176 - allow a pid of zero to watch for all children in child watchers.
177
178 2.54 Wed Jul 18 17:36:23 CEST 2007
179 - work around a perl bug that results in
180 BEGIN not safe after errors--compilation aborted without
181 any discernible reason or error message by once
182 again not use'ing strict.
183
184 2.53 Sun Jul 8 10:51:53 CEST 2007
185 - make Glib interface work again (spotted by elmex).
186
187 2.52 Wed Mar 7 18:36:16 CET 2007
188 - child watchers stopped working permanently when
189 all current child watchers were destroyed.
190
191 2.51 Mon Dec 11 21:33:24 CET 2006
192 - work around bugs in perl where eval "require Module" returns
193 true even if the module couldn't be loaded :/.
194
195 2.5 Mon Dec 11 02:15:28 CET 2006
196 - avoid $AUTOLOAD because many perls corrupt it.
197 - AnyEvent::detect forces autodetection.
198 - implement signal watchers (experimental)
199 (Tk does not support async signals (<= 804.027 at least),
200 so they might get delayed indefinitely on Tk).
201 - implement child watchers (experimental).
202 - moved default condvar implementation into base module, simplifying
203 most implementation modules.
204
205 2.1 Fri Nov 24 15:50:48 CET 2006
206 - better docs.
207 - simple testsuite.
208 - added AnyEvent->one_event method for special purposes.
209
210 2.0 Wed Nov 1 02:21:30 CET 2006
211 - INCOMPATIBLE CHANGE: poll can now either be "r" or "w" but not both.
212 - INCOMPATIBLE CHANGE: io watcher callbacks have no arguments anymore.
213 - EXPERIMENTAL:
214 new pure-perl model added, AnyEvent now always finds a suitable
215 event model.
216 - improved documentation.
217 - reduced cpu overhead.
218
219 1.02 Fri Jan 13 14:15:40 CET 2006
220 - add COPYING to clarify license.
221
222 1.01 Sun Jan 8 05:48:09 CET 2006
223 - correctly cancel Event timers, avoid memory leaks.
224
225 1.0 Sun Jan 8 05:40:58 CET 2006
226 - allow third-party interfaces to be added at runtime.
227 - fix version number.
228
229 0.9 Fri Dec 30 02:25:41 CET 2005
230 - fixed errors in the documentation.
231 - going back to using $VERSION as "load check",
232 as suggested by Jörn Reder. Let's see why I stopped using
233 this method.
234 - add 'hup' to the list of events to receive for glib
235 and generate 'rw' events for it, as some versions of
236 glib seem to require this, while this breaks other
237 (older) versions of glib.
238 - implemented PERL_ANYEVENT_VERBOSE env variable.
239
240 0.3 Sun Dec 4 10:43:33 CET 2005
241 - no changes to the code.
242 - fixed documentation.
243 - clarified some corner cases.
244 - vastly improved documentation with a long example.
245
246 0.2 Thu Dec 1 22:18:49 CET 2005
247 - check whether a valid method was called, to avoid endless loops.
248 - move condvars into their own class (cleaner, safer).
249 - don't die when some modules are not available.
250 - Tk is too broken to allow for words. Will work once Tk works.
251
252 0.1 Thu Dec 1 19:44:57 CET 2005
253 - spiced up and cleaned up for release.
254
255 0.01 Tue Sep 21 18:12:09 CEST 2004
256 - original version; copied from Convert::Scalar.
257