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

Comparing AnyEvent/Changes (file contents):
Revision 1.215 by root, Sat Feb 7 22:18:02 2009 UTC vs.
Revision 1.336 by root, Wed Aug 12 15:50:44 2009 UTC

1Revision history for Perl extension AnyEvent. 1Revision history for Perl extension AnyEvent.
2 2
35.11 Wed Aug 12 17:49:37 CEST 2009
4 - fixed a bug in run_cmd with input from filename.
5 - create a json coder object when none is given in write direction
6 to reduce dependency on JSON.
7
85.1 Tue Aug 11 03:17:32 CEST 2009
9 - tcp_connect now reports the correct error in $!, not an unrelated
10 one (was broken since 4.91).
11 - AE::Handle did not properly clear rtimeout/wtimeout watchers
12 when retrying a tcp_connect.
13 - new functions: AnyEvent::Util::run_cmd and close_all_fds_except.
14 - fixed wrong documentation in AE::Socket.
15
165.01 Mon Aug 10 03:16:32 CEST 2009
17 - last release broke EV child watchers :/.
18
195.0 Sun Aug 9 17:32:09 CEST 2009
20 - officially introduce and document the AE API.
21 - lowering the timeout at runtime did not have immediate effect
22 in AnyEvent::Handle.
23 - AE::Handle now has separate and independent read and write timeout
24 settings.
25 - probe for Guard module when AnyEvent::Util is loaded, not at
26 runtime (Event doesn't like eval "use" from callbacks, and it avoids
27 exporting the probe function).
28 - Impl::IOAsync's io watchers did not use the filehandle emulation
29 code to work around it's watcher limits.
30 - work around IO::Async's problems with overloaded objects.
31 - try to disable wearnings inside IO::Async, it's just too buggy.
32 - do no longer use Event by default, only EV or Perl, Event is too buggy.
33 - hackishly provide fast AE:: interface when EV or Perl is the backend.
34 - inherit VERSION from AnyEvent::VERSION in most modules.
35
364.91 Thu Aug 6 15:42:45 CEST 2009
37 - AE::Handle::starttls could get out of sync when the read buffer
38 already contains some TLS handshake.
39 - AE::Handle did not properly free the TLS session, causing bigger
40 memory leaks in Net::SSLeay than expected (reported by toaster).
41 - AE::Socket::tcp_connect will now artificially delay invoking
42 the callback to avoid returning after invoking the callback.
43 - convert many internal watcher uses to AE API.
44
454.9 Sat Aug 1 11:07:01 CEST 2009
46 - Glib silently fails when registering a timeout with
47 a negative value, so avoid that.
48 - call condvar callback immediately when it is set
49 after the condition is already signalled.
50 - check rbuf_max condition only after trying to consume
51 data.
52 - normalise signal numbers to names when using the ae's
53 signal handling, but do not document this (yet).
54 - pure perl signal emulation did not properly set nonblocking
55 mode on the signal pipe on win32 (this is unlikely to have caused
56 any issues).
57 - new module: AnyEvent::Debug.
58 - AnyEvent::Strict now checks that a signal was specified
59 by name (not name or number).
60 - reduce memory footprint in typical cases by ~50kb by
61 only compiling signal code when necessary.
62 - add AnyEvent::Handle->rbuf_max.
63 - grab the AE namespace for future new API, implement stubs for
64 said future/alternative API.
65 - new function: AnyEvent::Socket::format_hostport.
66
674.881 Tue Jul 28 12:51:53 CEST 2009
68 - work around a bug in local in pre-5.10 perls, causing
69 AnyEvent::Handle to recurse when it should not
70 (analyzed by elmex).
71
724.88 Tue Jul 28 04:04:37 CEST 2009
73 - re-bless the handle into a dummy package after calling
74 AnyEvent::Handle::destroy, so the user does not need to check
75 for errors after every push_write etc.
76 - do not attempt to run t/02_signals.t on obviously broken
77 platforms.
78
794.87 Sun Jul 26 02:06:16 CEST 2009
80 - do not attempt to linger when there is no longer a valid fh.
81 - enforce tls mode to be either accept or connect, do not simply
82 segfault in Net::SSLeay.
83 - AnyEvent::Handle can now call tcp_connect itself (new parameters
84 connect, on_prepare, on_connect and on_connect_error). Updated
85 tutorial accordingly.
86 - add AnyEvent::Impl::Irssi backend.
87
884.86 Mon Jul 20 23:52:29 CEST 2009
89 - since the verbose warning is not enough, explicitly document
90 that versions before 1.33 of Net::SSLeay are not secure.
91 - work around signal handling races in Event and (...) Event::Lib.
92 - try to align signal-race timer to full-second boundaries.
93 - work around Tk not liking negative timeouts.
94 - don't complain of different grades of environmental unfriendlyness
95 in IO::Async.
96
974.85 Sat Jul 18 06:16:14 CEST 2009
98 - nail the signal race problem in perl once and for all
99 (see $AnyEvent::MAX_SIGNAL_LATENCY).
100 - take advantage of Async::Interrupt if it is available.
101 - load Time::HiRes and Guard modules on demand only.
102 - add optional/recommended modules section to AnyEvent
103 documentation.
104 - reduce memory usage considerably (and reduce startup penalty)
105 by not using "strict", "warnings" and "overload" modules.
106 - work around buggy windows/openbsd perls and provide EBADMSG
107 and EPROTO ourselves when missing.
108 - improve perl 5.6 compatibility of the core event loop.
109 - made Net::SSLeay version 1.33 a soft requirement.
110
1114.83 Fri Jul 17 16:56:26 CEST 2009
112 - implement AnyEvent::Socket::getprotobyname.
113 - AnyEvent::CondVar's will now detect recursive blocking
114 waits and will croak, as too many people fall into
115 this trap.
116 - AnyEvent::Handle will now call ->destroy on itself after
117 executing the on_error callback, instead of doing some
118 half-baked internal shutdown, for fatal errors.
119 - clarify on_eof behaviour w.r.t. the read queue and
120 on_read callbacks.
121 - ignore some possible spurious wake-ups in tcp_connect.
122
1234.82 Sat Jul 11 00:34:55 CEST 2009
124 - POE and Event backends didn't accept some callable objects as
125 callbacks.
126 - use Config module instead of POSIX module to detect signal names
127 in AnyEvent::Strict and AnyEvent::Impl::EventLib, as the POSIX
128 module doesn't even have all POSIX signals :/.
129 - use more workarounds around the many refcnt/corruption bugs in
130 Event::Lib.
131 - work around a race condition in perl's select, causing t/03_child.t
132 to rarely fail.
133
1344.81 Thu Jul 9 10:30:30 CEST 2009
135 - AnyEvent::Handle didn't properly diagnose write errors
136 (it expected -1 from syswrite, how lame... :).
137 - support file descriptors in addition to file handles
138 in AnyEvent->io.
139 - new env variables: PERL_ANYEVENT_RESOLV_CONF,
140 PERL_ANYEVENT_MAX_OUTSTANDING_DNS, PERL_ANYEVENT_CA_FILE
141 and PERL_ANYEVENT_CA_PATH.
142 - provide a sensible synopsis section for AnyEvent::TLS.
143 - add a "supported backends" section to the manpage.
144 - added simple io watcher test to testsuite, using a
145 portable_socketpair.
146 - tried to improve the stability of the Event::Lib backend,
147 YMMV.
148
1494.8 Mon Jul 6 23:45:16 CEST 2009
150 - AnyEvent::DNS did not properly follow CNAME records with
151 uppercase targets.
152 - AnyEvent::DNS would errornously return AAAA records
153 with v4 mapped addresses (a faulty record) as ipv4 addresses,
154 causing AnyEvent::Socket to throw an exception.
155 - added new module AnyEvent::TLS for easier SSL/TLS context
156 creation, with many options including hostname verification,
157 secure default configuration, lots of documentation and,
158 predefined diffie-hellman keys for perfect forward security
159 and much more. get it while it's still fresh!
160 - use AnyEvent::TLS in AnyEvent::Handle for context management.
161 - load AnyEvent::Handle only on demand in AnyEvent::DNS,
162 so AnyEvent::Socket users have smaller memory footprint
163 in the common case.
164 - add AnyEvent::Handle->push_shutdown method.
165 - add an additional error message parameter to AnyEvent::Handle's
166 on_error callback (for TLS, $! is still available).
167 - add AnyEvent::Handle on_starttls/on_stoptls callbacks.
168 - make AnyEvent::Handle more robust against early conenction
169 failures (during new), and return C<undef> in that case
170 from the constructor.
171 - AnyEvent::Handle will now try to load only JSON::XS first,
172 then fall back to JSON.
173 - format_ipv4/format_ipv6 are now exported by default, for symmetry,
174 and because it was documented that way.
175
1764.451 Fri Jul 3 00:28:58 CEST 2009
177 - do not clear rbuf when shutting down an AnyEvent::Handle
178 object - doing so breaks AnyEvent::HTTP.
179
1804.45 Mon Jun 29 22:59:26 CEST 2009
181 - a write error could cause AnyEvent::Handle to create
182 an I/O watcher with an undefined $fh.
183 - special-case mapped ipv4 addresses in both
184 AnyEvent::Socket::format_address and parse_address,
185 to treat them just like ipv4 addresses.
186 - updated and overhauled the AnyEvent::Intro doc.
187 - implement AnyEvent::Socket::format_ipv4/ipv6.
188 - slightly speed up portable_pipe/socketpair.
189 - expand condvar begin/end documentation.
190
1914.42 Fri Jun 26 08:32:18 CEST 2009
192 - preliminary and neccesarily incomplete support for IO::Async.
193 - reset SIGCHLD to DEFAULT when AnyEvent is loaded, in
194 case it was set to IGNORE, to ensure we can catch child statuses
195 even when the calling env acts stupidly.
196 - updated benchmarks with IO::Async, which performs very well.
197
1984.412 Wed Jun 24 01:35:57 CEST 2009
199 - support an "untaint" attribute for AnyEvent::DNS
200 and set it on the default resolver.
201 - implement and document AnyEvent::Impl::Perl::loop.
202 - remove all anyevent-env variables from %ENV
203 when running in tainted mode.
204 - mention and extend the IO::Lambda benchmark.
205
2064.411 Sun Jun 7 18:48:02 CEST 2009
207 - do not try to use F_SETFD on windows.
208
2094.41 Thu May 14 06:40:11 CEST 2009
210 - work around issues in older perls (5.8.0?) when
211 a signal handler is deleted from the %SIG hash.
212 - use POSIX::_exit in child test, to avoid running
213 destructors.
214 - speed up CHLD handling by relying on SIGCHLD being
215 synchronously delivered, even when we roll our own
216 implementation.
217 - AnyEvent::DNS: add the "dname" resource record name
218 for cosmetic reasons.
219
2204.4 Sun Apr 26 20:12:33 CEST 2009
221 - implemented idle watchers, where applicable.
222 - AnyEvent->time died when Event backend was in use.
223 - fix a memleak in the Tk backend.
224 - sped up Tk timer handling.
225 - clip negative "after" values to 0 in AnyEvent::Impl::Event
226 to avoid spamming out warning messages.
227 - fix Qt timers without interval.
228 - avoid Qt zero-timeout special case, in old libqt's.
229
2304.352 Mon Apr 20 16:31:11 CEST 2009
231 - fix AnyEvent::Strict error messages for child watchers.
232 - fix/update Makefile.PL "recommends" versions.
233 - add AnyEvent->now_update.
234
2354.351 Sat Apr 11 07:56:14 CEST 2009
236 - actually make the signal pipe work under win32.
237 - localise $! in signal handler to avoid changing $!, although
238 perl itself does not seem to save/restore errno either.
239 - set the cloexec flag on the signal pipe (normally set by perl too).
240
2414.35 Fri Mar 27 11:48:20 CET 2009
242 - event models relying on AnyEvent's signal watcher emulation did
243 invoke the callback asynchronously, contradicting documentation
244 and causing signals to get lost (this includes AnyEvent's own
245 event loop). AnyEvent now uses the standard pipe trick to make
246 callback execution synchronous to the event loop.
247 - AnyEvent::Handle didn't free TLS context data on DESTROY
248 (patch by Pavel Shaydo).
249 - work around the ever-incompatibly-changing API of MakeMaker.
250 - document that changing global variables without restoring them
251 is a bad idea in Perl (noted by Adam Rosenstein).
252 - AnyEvent::Strict now barfs if ->io is passed a file.
253
2544.34 Thu Feb 12 18:32:45 CET 2009
255 - separately buffer TLS read data, as otherwise the read queue could
256 deadlock as receiving data is not expected while draining the
257 read queue (which cna only happen with TLS).
3 - raise EBADMSG error on JSON decoding errors. 258 - raise EBADMSG error on JSON decoding errors.
4 - fix some minor manpage bugs (reported by Maximilian Gaß). 259 - fix some minor manpage bugs (reported by Maximilian Gaß).
5 - speed up select bitmask parsing quite a bit in the pure perl 260 - speed up select bitmask parsing quite a bit in the pure perl
6 backend. 261 backend.
262 - use CORE::select instead of just select in AnyEvent::Impl::Perl.
7 263
84.331 Tue Jan 6 21:07:25 CET 2009 2644.331 Tue Jan 6 21:07:25 CET 2009
9 - socketpair fails on many vista machines because vista has 265 - socketpair fails on many vista machines because vista has
10 completely broken accept/getpeername and getsockname functions, 266 completely broken accept/getpeername and getsockname functions,
11 so we provide our own socketpair emulation that kind of works 267 so we provide our own socketpair emulation that kind of works

Diff Legend

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