| 1 | Revision history for Perl extension AnyEvent. |
1 | Revision history for Perl extension AnyEvent. |
| 2 | |
2 | |
| 3 | 4.341 |
3 | 5.01 Mon Aug 10 03:16:32 CEST 2009 |
|
|
4 | - last release broke EV child watchers :/. |
|
|
5 | |
|
|
6 | 5.0 Sun Aug 9 17:32:09 CEST 2009 |
|
|
7 | - officially introduce and document the AE API. |
|
|
8 | - lowering the timeout at runtime did not have immediate effect |
|
|
9 | in AnyEvent::Handle. |
|
|
10 | - AE::Handle now has separate and independent read and write timeout |
|
|
11 | settings. |
|
|
12 | - probe for Guard module when AnyEvent::Util is loaded, not at |
|
|
13 | runtime (Event doesn't like eval "use" from callbacks, and it avoids |
|
|
14 | exporting the probe function). |
|
|
15 | - Impl::IOAsync's io watchers did not use the filehandle emulation |
|
|
16 | code to work around it's watcher limits. |
|
|
17 | - work around IO::Async's problems with overloaded objects. |
|
|
18 | - try to disable wearnings inside IO::Async, it's just too buggy. |
|
|
19 | - do no longer use Event by default, only EV or Perl, Event is too buggy. |
|
|
20 | - hackishly provide fast AE:: interface when EV or Perl is the backend. |
|
|
21 | - inherit VERSION from AnyEvent::VERSION in most modules. |
|
|
22 | |
|
|
23 | 4.91 Thu Aug 6 15:42:45 CEST 2009 |
|
|
24 | - AE::Handle::starttls could get out of sync when the read buffer |
|
|
25 | already contains some TLS handshake. |
|
|
26 | - AE::Handle did not properly free the TLS session, causing bigger |
|
|
27 | memory leaks in Net::SSLeay than expected (reported by toaster). |
|
|
28 | - AE::Socket::tcp_connect will now artificially delay invoking |
|
|
29 | the callback to avoid returning after invoking the callback. |
|
|
30 | - convert many internal watcher uses to AE API. |
|
|
31 | |
|
|
32 | 4.9 Sat Aug 1 11:07:01 CEST 2009 |
|
|
33 | - Glib silently fails when registering a timeout with |
|
|
34 | a negative value, so avoid that. |
|
|
35 | - call condvar callback immediately when it is set |
|
|
36 | after the condition is already signalled. |
|
|
37 | - check rbuf_max condition only after trying to consume |
|
|
38 | data. |
|
|
39 | - normalise signal numbers to names when using the ae's |
|
|
40 | signal handling, but do not document this (yet). |
|
|
41 | - pure perl signal emulation did not properly set nonblocking |
|
|
42 | mode on the signal pipe on win32 (this is unlikely to have caused |
|
|
43 | any issues). |
|
|
44 | - new module: AnyEvent::Debug. |
|
|
45 | - AnyEvent::Strict now checks that a signal was specified |
|
|
46 | by name (not name or number). |
|
|
47 | - reduce memory footprint in typical cases by ~50kb by |
|
|
48 | only compiling signal code when necessary. |
|
|
49 | - add AnyEvent::Handle->rbuf_max. |
|
|
50 | - grab the AE namespace for future new API, implement stubs for |
|
|
51 | said future/alternative API. |
|
|
52 | - new function: AnyEvent::Socket::format_hostport. |
|
|
53 | |
|
|
54 | 4.881 Tue Jul 28 12:51:53 CEST 2009 |
|
|
55 | - work around a bug in local in pre-5.10 perls, causing |
|
|
56 | AnyEvent::Handle to recurse when it should not |
|
|
57 | (analyzed by elmex). |
|
|
58 | |
|
|
59 | 4.88 Tue Jul 28 04:04:37 CEST 2009 |
|
|
60 | - re-bless the handle into a dummy package after calling |
|
|
61 | AnyEvent::Handle::destroy, so the user does not need to check |
|
|
62 | for errors after every push_write etc. |
|
|
63 | - do not attempt to run t/02_signals.t on obviously broken |
|
|
64 | platforms. |
|
|
65 | |
|
|
66 | 4.87 Sun Jul 26 02:06:16 CEST 2009 |
|
|
67 | - do not attempt to linger when there is no longer a valid fh. |
|
|
68 | - enforce tls mode to be either accept or connect, do not simply |
|
|
69 | segfault in Net::SSLeay. |
|
|
70 | - AnyEvent::Handle can now call tcp_connect itself (new parameters |
|
|
71 | connect, on_prepare, on_connect and on_connect_error). Updated |
|
|
72 | tutorial accordingly. |
|
|
73 | - add AnyEvent::Impl::Irssi backend. |
|
|
74 | |
|
|
75 | 4.86 Mon Jul 20 23:52:29 CEST 2009 |
|
|
76 | - since the verbose warning is not enough, explicitly document |
|
|
77 | that versions before 1.33 of Net::SSLeay are not secure. |
|
|
78 | - work around signal handling races in Event and (...) Event::Lib. |
|
|
79 | - try to align signal-race timer to full-second boundaries. |
|
|
80 | - work around Tk not liking negative timeouts. |
|
|
81 | - don't complain of different grades of environmental unfriendlyness |
|
|
82 | in IO::Async. |
|
|
83 | |
|
|
84 | 4.85 Sat Jul 18 06:16:14 CEST 2009 |
|
|
85 | - nail the signal race problem in perl once and for all |
|
|
86 | (see $AnyEvent::MAX_SIGNAL_LATENCY). |
|
|
87 | - take advantage of Async::Interrupt if it is available. |
|
|
88 | - load Time::HiRes and Guard modules on demand only. |
|
|
89 | - add optional/recommended modules section to AnyEvent |
|
|
90 | documentation. |
|
|
91 | - reduce memory usage considerably (and reduce startup penalty) |
|
|
92 | by not using "strict", "warnings" and "overload" modules. |
|
|
93 | - work around buggy windows/openbsd perls and provide EBADMSG |
|
|
94 | and EPROTO ourselves when missing. |
|
|
95 | - improve perl 5.6 compatibility of the core event loop. |
|
|
96 | - made Net::SSLeay version 1.33 a soft requirement. |
|
|
97 | |
|
|
98 | 4.83 Fri Jul 17 16:56:26 CEST 2009 |
|
|
99 | - implement AnyEvent::Socket::getprotobyname. |
|
|
100 | - AnyEvent::CondVar's will now detect recursive blocking |
|
|
101 | waits and will croak, as too many people fall into |
|
|
102 | this trap. |
|
|
103 | - AnyEvent::Handle will now call ->destroy on itself after |
|
|
104 | executing the on_error callback, instead of doing some |
|
|
105 | half-baked internal shutdown, for fatal errors. |
|
|
106 | - clarify on_eof behaviour w.r.t. the read queue and |
|
|
107 | on_read callbacks. |
|
|
108 | - ignore some possible spurious wake-ups in tcp_connect. |
|
|
109 | |
|
|
110 | 4.82 Sat Jul 11 00:34:55 CEST 2009 |
|
|
111 | - POE and Event backends didn't accept some callable objects as |
|
|
112 | callbacks. |
|
|
113 | - use Config module instead of POSIX module to detect signal names |
|
|
114 | in AnyEvent::Strict and AnyEvent::Impl::EventLib, as the POSIX |
|
|
115 | module doesn't even have all POSIX signals :/. |
|
|
116 | - use more workarounds around the many refcnt/corruption bugs in |
|
|
117 | Event::Lib. |
|
|
118 | - work around a race condition in perl's select, causing t/03_child.t |
|
|
119 | to rarely fail. |
|
|
120 | |
|
|
121 | 4.81 Thu Jul 9 10:30:30 CEST 2009 |
|
|
122 | - AnyEvent::Handle didn't properly diagnose write errors |
|
|
123 | (it expected -1 from syswrite, how lame... :). |
|
|
124 | - support file descriptors in addition to file handles |
|
|
125 | in AnyEvent->io. |
|
|
126 | - new env variables: PERL_ANYEVENT_RESOLV_CONF, |
|
|
127 | PERL_ANYEVENT_MAX_OUTSTANDING_DNS, PERL_ANYEVENT_CA_FILE |
|
|
128 | and PERL_ANYEVENT_CA_PATH. |
|
|
129 | - provide a sensible synopsis section for AnyEvent::TLS. |
|
|
130 | - add a "supported backends" section to the manpage. |
|
|
131 | - added simple io watcher test to testsuite, using a |
|
|
132 | portable_socketpair. |
|
|
133 | - tried to improve the stability of the Event::Lib backend, |
|
|
134 | YMMV. |
|
|
135 | |
|
|
136 | 4.8 Mon Jul 6 23:45:16 CEST 2009 |
|
|
137 | - AnyEvent::DNS did not properly follow CNAME records with |
|
|
138 | uppercase targets. |
|
|
139 | - AnyEvent::DNS would errornously return AAAA records |
|
|
140 | with v4 mapped addresses (a faulty record) as ipv4 addresses, |
|
|
141 | causing AnyEvent::Socket to throw an exception. |
|
|
142 | - added new module AnyEvent::TLS for easier SSL/TLS context |
|
|
143 | creation, with many options including hostname verification, |
|
|
144 | secure default configuration, lots of documentation and, |
|
|
145 | predefined diffie-hellman keys for perfect forward security |
|
|
146 | and much more. get it while it's still fresh! |
|
|
147 | - use AnyEvent::TLS in AnyEvent::Handle for context management. |
|
|
148 | - load AnyEvent::Handle only on demand in AnyEvent::DNS, |
|
|
149 | so AnyEvent::Socket users have smaller memory footprint |
|
|
150 | in the common case. |
|
|
151 | - add AnyEvent::Handle->push_shutdown method. |
|
|
152 | - add an additional error message parameter to AnyEvent::Handle's |
|
|
153 | on_error callback (for TLS, $! is still available). |
|
|
154 | - add AnyEvent::Handle on_starttls/on_stoptls callbacks. |
|
|
155 | - make AnyEvent::Handle more robust against early conenction |
|
|
156 | failures (during new), and return C<undef> in that case |
|
|
157 | from the constructor. |
|
|
158 | - AnyEvent::Handle will now try to load only JSON::XS first, |
|
|
159 | then fall back to JSON. |
|
|
160 | - format_ipv4/format_ipv6 are now exported by default, for symmetry, |
|
|
161 | and because it was documented that way. |
|
|
162 | |
|
|
163 | 4.451 Fri Jul 3 00:28:58 CEST 2009 |
|
|
164 | - do not clear rbuf when shutting down an AnyEvent::Handle |
|
|
165 | object - doing so breaks AnyEvent::HTTP. |
|
|
166 | |
|
|
167 | 4.45 Mon Jun 29 22:59:26 CEST 2009 |
|
|
168 | - a write error could cause AnyEvent::Handle to create |
|
|
169 | an I/O watcher with an undefined $fh. |
|
|
170 | - special-case mapped ipv4 addresses in both |
|
|
171 | AnyEvent::Socket::format_address and parse_address, |
|
|
172 | to treat them just like ipv4 addresses. |
|
|
173 | - updated and overhauled the AnyEvent::Intro doc. |
|
|
174 | - implement AnyEvent::Socket::format_ipv4/ipv6. |
|
|
175 | - slightly speed up portable_pipe/socketpair. |
|
|
176 | - expand condvar begin/end documentation. |
|
|
177 | |
|
|
178 | 4.42 Fri Jun 26 08:32:18 CEST 2009 |
|
|
179 | - preliminary and neccesarily incomplete support for IO::Async. |
|
|
180 | - reset SIGCHLD to DEFAULT when AnyEvent is loaded, in |
|
|
181 | case it was set to IGNORE, to ensure we can catch child statuses |
|
|
182 | even when the calling env acts stupidly. |
|
|
183 | - updated benchmarks with IO::Async, which performs very well. |
|
|
184 | |
|
|
185 | 4.412 Wed Jun 24 01:35:57 CEST 2009 |
|
|
186 | - support an "untaint" attribute for AnyEvent::DNS |
|
|
187 | and set it on the default resolver. |
|
|
188 | - implement and document AnyEvent::Impl::Perl::loop. |
|
|
189 | - remove all anyevent-env variables from %ENV |
|
|
190 | when running in tainted mode. |
|
|
191 | - mention and extend the IO::Lambda benchmark. |
|
|
192 | |
|
|
193 | 4.411 Sun Jun 7 18:48:02 CEST 2009 |
|
|
194 | - do not try to use F_SETFD on windows. |
|
|
195 | |
|
|
196 | 4.41 Thu May 14 06:40:11 CEST 2009 |
|
|
197 | - work around issues in older perls (5.8.0?) when |
|
|
198 | a signal handler is deleted from the %SIG hash. |
|
|
199 | - use POSIX::_exit in child test, to avoid running |
|
|
200 | destructors. |
|
|
201 | - speed up CHLD handling by relying on SIGCHLD being |
|
|
202 | synchronously delivered, even when we roll our own |
|
|
203 | implementation. |
|
|
204 | - AnyEvent::DNS: add the "dname" resource record name |
|
|
205 | for cosmetic reasons. |
|
|
206 | |
|
|
207 | 4.4 Sun Apr 26 20:12:33 CEST 2009 |
|
|
208 | - implemented idle watchers, where applicable. |
|
|
209 | - AnyEvent->time died when Event backend was in use. |
|
|
210 | - fix a memleak in the Tk backend. |
|
|
211 | - sped up Tk timer handling. |
|
|
212 | - clip negative "after" values to 0 in AnyEvent::Impl::Event |
|
|
213 | to avoid spamming out warning messages. |
|
|
214 | - fix Qt timers without interval. |
|
|
215 | - avoid Qt zero-timeout special case, in old libqt's. |
|
|
216 | |
|
|
217 | 4.352 Mon Apr 20 16:31:11 CEST 2009 |
|
|
218 | - fix AnyEvent::Strict error messages for child watchers. |
|
|
219 | - fix/update Makefile.PL "recommends" versions. |
|
|
220 | - add AnyEvent->now_update. |
|
|
221 | |
|
|
222 | 4.351 Sat Apr 11 07:56:14 CEST 2009 |
|
|
223 | - actually make the signal pipe work under win32. |
|
|
224 | - localise $! in signal handler to avoid changing $!, although |
|
|
225 | perl itself does not seem to save/restore errno either. |
|
|
226 | - set the cloexec flag on the signal pipe (normally set by perl too). |
|
|
227 | |
|
|
228 | 4.35 Fri Mar 27 11:48:20 CET 2009 |
| 4 | - event models relying on AnyEvent's signal watcher emulation did |
229 | - event models relying on AnyEvent's signal watcher emulation did |
| 5 | invoke the callback asynchronously, contradicting documentation |
230 | invoke the callback asynchronously, contradicting documentation |
| 6 | and causing signals to get lost (this includes AnyEvent's own |
231 | and causing signals to get lost (this includes AnyEvent's own |
| 7 | event loop). AnyEvent now uses the standard pipe trick to make |
232 | event loop). AnyEvent now uses the standard pipe trick to make |
| 8 | callback execution synchronous to the event loop. |
233 | callback execution synchronous to the event loop. |
|
|
234 | - AnyEvent::Handle didn't free TLS context data on DESTROY |
|
|
235 | (patch by Pavel Shaydo). |
| 9 | - work around the ever-incompatibly-changing API of MakeMaker. |
236 | - work around the ever-incompatibly-changing API of MakeMaker. |
| 10 | - document that changing global variables without restoring them |
237 | - document that changing global variables without restoring them |
| 11 | is a bad idea in Perl (noted by Adam Rosenstein). |
238 | is a bad idea in Perl (noted by Adam Rosenstein). |
|
|
239 | - AnyEvent::Strict now barfs if ->io is passed a file. |
| 12 | |
240 | |
| 13 | 4.34 Thu Feb 12 18:32:45 CET 2009 |
241 | 4.34 Thu Feb 12 18:32:45 CET 2009 |
| 14 | - separately buffer TLS read data, as otherwise the read queue could |
242 | - separately buffer TLS read data, as otherwise the read queue could |
| 15 | deadlock as receiving data is not expected while draining the |
243 | deadlock as receiving data is not expected while draining the |
| 16 | read queue (which cna only happen with TLS). |
244 | read queue (which cna only happen with TLS). |
