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

Comparing AnyEvent/Changes (file contents):
Revision 1.297 by root, Fri Jul 24 08:40:35 2009 UTC vs.
Revision 1.425 by root, Fri Dec 24 02:46:21 2010 UTC

1Revision history for Perl extension AnyEvent. 1Revision history for Perl extension AnyEvent.
2 2
3TODO: maybe support "peerip" verificstion, i.e. getpeername diretcly as 3TODO: document TCP_*** constants
4additional peername? 4TODO: catch AE::xxx for anyevent strict and anyevent debug
5TODO: anyevent::trace? or debug tracing all watchers?
6TODO: anyevent strict checking for $_ changes?
5 7
8TODO: wbuf_max for anyevent
9TODO: fix with the rest: POE + IO::Async? No go. Tk + Event? No go. Again: if
10 - set CLOEXEC flag on parent-side fd's in AnyEvent::Util::run_cmd,
11 to avoid leaking them to any child processes.
12
135.29 Sun Dec 5 10:49:21 CET 2010
14 - convert EV backend to EV 4.00 API (so better upgrade EV too).
15 - AE::Handle: implement dynamic read_size adjustment, add
16 max_read_size parameter, reduce default read_size to 2048.
17 - add some FAQ entries for servers and/or on_eof vs. on_error.
18 - work around OS bugs (cygwin again) when lingering in AE::Handle.
19 - reduce memory usage slightly when connect was used in AE::Handle.
20 - wrok around more segfaults in Tk.
21 - document tls_ctx => undef as valid.
22 - detect WNOHANG value at compiletime.
23 - include some probably linux-only support for building AnyEvent
24 as part of the perl core.
25 - improve compatibility of t/08_idna.t to perl 5.8.
26 - make AnyEvent::Strict truly optional (in case somebody builds a
27 minimal perl...).
28
295.28 Wed Oct 13 04:14:23 CEST 2010
30 - due to a glitch, AnyEvent's internal getprotobyname wasn't
31 actually used in tcp_server.
32 - implement AnyEvent::CondVar->new method, which was documented but
33 missing.
34 - support autoloading of read types in unshift_read.
35 - AnyEvent::Strict: do the test for invalid fh arguments differently,
36 resulting in more informative error messages.
37 - new AnyEvent::FAQ manpage.
38 - clarify that you can actually call push_read etc. in on_read.
39 - put keys and certs in /tmp, not into the current directory,
40 and document this in the SECURITY CONSIDERATIONS sections
41 of AnyEvent::TLS.
42 - support ";" as resolv.conf comment character in AnyEvent::DNS, as
43 requested by Ron Isaacson.
44 - document $AnyEvent::DNS::RESOLVER variable.
45 - incorporated major typo patches by Abhijit Menon-Sen.
46 - AnyEvent::Handle now croaks when negative timeouts are passed.
47 - add a shitload of TCP_xxx sockopt name constants to AnyEvent::Util,
48 undocumented, but who knows when they might come in handy.
49
505.271 Tue Jun 8 12:05:46 CEST 2010
51 - backport to perl 5.8.x.
52
535.27 Sun Jun 6 12:12:05 CEST 2010
54 - postpone differently in AnyEvent::Socket now, as
55 when not, canceling the connection attempt might fail
56 (found by Felix Antonius Wilhelm Ostmann).
57 - explicitly check for non-stream sockets in AE::Handle, too many
58 clueless people fell into the trap of this somehow working.
59 - simplified and reworked the "OTHER MODULES" section.
60 - better/more condvar examples.
61
625.261 Wed Apr 28 16:13:36 CEST 2010
63 - AF_INET6 was not properly used from Socket6 during configuration
64 time (found by Andy Grundman).
65
665.26 Mon Apr 12 04:49:35 CEST 2010
67 - don't generate spurious readyness notifications when select
68 returns EINTR in the pure perl backend (analysed and testcase
69 by Bas Denissen).
70 - give same examples in SYNOPSIS sections of AnyEvent and AE.
71 - provider faster implementations for AE::time/now/now_update for
72 EV and Perl backends.
73 - provide faster AE::cv and AE::time implementations for all
74 backends.
75 - no longer support lower/mixed-case signal names (this was
76 never documented nor universally supported).
77 - some more memory ยต-optimisations, and somewhat less messy
78 code to cope with both AnyEvent and AE APIs.
79
805.251 Sat Mar 13 00:58:21 CET 2010
81 - make test failed if it was run before make install on perl <=
82 5.8.8, and older versions of perl make our live very difficult,
83 so write constants.pl during Makefile.PL time.
84
855.25 Sat Mar 13 00:23:14 CET 2010
86 - fix a race condition in AnyEvent::Handle that would cause
87 a "bio_read: unsupported method" error (in _another_
88 ssl connection) after user code threw an exception.
89 - added AnyEvent::Handle->destroyed method.
90 - speed up AnyEvent::Socket::format_address and ::format_ipv6.
91 - the AnyEvent::Util::fh_nonblocking function worked only
92 by lucky accident on win32.
93 - smaller and faster AnyEvent::Util::fh_nonblocking.
94 - when the (required!) Time::HiRes module is missing, AnyEvent
95 did not fall back to built-in time properly.
96 - do not load Fcntl at runtime, saving memory and loading time.
97 - precompile a number of constants and use them instead of runtime
98 detection and eval.
99 - free detection code after detection and similar memory optimisations.
100 - Perl backend timer interval best effort drift has been improved
101 (same algorithm as EV).
102 - update unicode idna mapping table.
103
1045.24 Tue Jan 5 11:39:43 CET 2010
105 - cygwin never reports errors from failed connects. 1.5 just gives
106 you continous readyness and EAGAIN, 1.7 is even more broken and
107 just hangs. work around both issues in a cygwin-specific hack.
108 - improve idle watcher documentation slightly.
109
1105.23 Sun Dec 20 23:48:00 CET 2009
111 - support IDNs in resolve_sockaddr, and therefore in tcp_connect.
112 - implement punycode_encode/decode, idn_nameprep,
113 idn_to_ascii and idn_to_unicode operations in AnyEvent::Util.
114 - provide $AE::VERSION.
115 - removed traces of "no strict 'refs'".
116
1175.22 Sat Dec 5 03:51:13 CET 2009
118 - downgrade-or-fail in AnyEvent::Handle::push_write, to
119 diagnose encoding failures earlier and more succinctly.
120 (this works around bugs in perl, throwing away encoding info
121 when passing scalar data to extensions).
122 - add more examples to AnyEvent::Socket manpage.
123 - upgrade internal warning set to the same as common::sense 2.03.
124 - use pack "n/a*" for pre-5.8.9 perl compatibility in AnyEvent::DNS
125 (John Beppu).
126 - AnyEvent::Socket::inet_aton now properly supports ipv6, as documented.
127 - add google public dns servers to fallback server set and make sure
128 we load-balance properly between the three sets. also add all
129 fallback dns servers, not just a random one, to each dns config.
130
1315.21 Thu Nov 19 02:48:47 CET 2009
132 - fix a problem where socket constants were called with parameters
133 (spotted by David Friedland).
134 - fork_call never use'd POSIX (reported by Daisuke Maki).
135 - improve perl 5.6 compatibility further (but it still won't work
136 unless you rip out everything but the core).
137 - prefer Net::DNS::Resolver over ipconfig, if installed, on win32.
138 uses 10MB of RAM, but doesn't open a console window. *sigh*.
139
1405.202 Wed Oct 14 22:35:44 CEST 2009
141 - AnyEvent::DNS would unexpectedly clobber $_ under windows
142 (analysed by Matthias Waldorf).
143 - AnyEvent::Handle::run_cmd can now pass the PID of the
144 newly-created process, which is much less useful than it might
145 sound (based on patch by Yann Kerherve).
146
1475.201 Tue Sep 29 12:09:25 CEST 2009
148 - AnyEvent:Handle::on_starttls/on_stoptls methods were broken
149 (reported by Torsten Foertsch).
150 - common::sense 2.0 could cause tcp_server to throw an exception
151 (analysed by elmex).
152
1535.2 Mon Sep 14 07:04:49 CEST 2009
154 - INCOMPATIBLE CHANGE: do no longer support register_read_type
155 and register_write_type in AnyEvent::Handle, instead support
156 package names (the facility was mostly abused).
157 - implement "packagename-as-read/write type" support in
158 AnyEvent::Handle.
159 - AnyEvent::Handle: new options "keepalive" and "oobinline".
160 - oobinline set by default to avoid security issues.
161 - the pure-perl event loop backend wrongly detected times() fallback
162 support (spotted by Pavel Boldin).
163
1645.12 Tue Sep 1 20:26:50 CEST 2009
165 - be more lenient when parsing resolv.conf files, as some people
166 use hashmarks as comment indicator inside directives (reported
167 by Michael S. Fischer).
168 - use same set of warnings as common::sense 2.0.
169 - fix a potential 32 bit overflow issue due to perl having problems
170 with large hex constants in the Perl backend.
171 - do not use the slower 5.6 version fo the enc_name code in AnyEvent::DNS
172 by default.
173 - fix some prototypes.
174 - (optionally) require Async::Interrupt 1.02.
175 - replace opendns fallbacks by verizon/level3 public dns servers.
176
1775.112 Fri Aug 21 13:59:04 CEST 2009
178 - AnyEvent::Handle could cause hard crashes in Net::SSLeay when
179 using starttls with an illegal/uncreatable tls context.
180 - port to broken win32 perls.
181
1825.111 Sun Aug 16 18:44:36 CEST 2009
183 - perl errornously requires sockaddr_un structures returned by the
184 kernel to have a specific length. work around this bug by padding
185 sockaddr_un structures (found to fail on OS X, as perl only uses
186 the correct code for GNU/Linux, but likely to fail on other
187 OSes as well).
188
1895.11 Wed Aug 12 17:49:37 CEST 2009
190 - fixed a bug in run_cmd with input from filename.
191 - create a json coder object when none is given in write direction
192 to reduce dependency on JSON.
193
1945.1 Tue Aug 11 03:17:32 CEST 2009
195 - tcp_connect now reports the correct error in $!, not an unrelated
196 one (was broken since 4.91).
197 - AE::Handle did not properly clear rtimeout/wtimeout watchers
198 when retrying a tcp_connect.
199 - new functions: AnyEvent::Util::run_cmd and close_all_fds_except.
200 - fixed wrong documentation in AE::Socket.
201
2025.01 Mon Aug 10 03:16:32 CEST 2009
203 - last release broke EV child watchers :/.
204
2055.0 Sun Aug 9 17:32:09 CEST 2009
206 - officially introduce and document the AE API.
207 - lowering the timeout at runtime did not have immediate effect
208 in AnyEvent::Handle.
209 - AE::Handle now has separate and independent read and write timeout
210 settings.
211 - probe for Guard module when AnyEvent::Util is loaded, not at
212 runtime (Event doesn't like eval "use" from callbacks, and it avoids
213 exporting the probe function).
214 - Impl::IOAsync's io watchers did not use the filehandle emulation
215 code to work around it's watcher limits.
216 - work around IO::Async's problems with overloaded objects.
217 - try to disable wearnings inside IO::Async, it's just too buggy.
218 - do no longer use Event by default, only EV or Perl, Event is too buggy.
219 - hackishly provide fast AE:: interface when EV or Perl is the backend.
220 - inherit VERSION from AnyEvent::VERSION in most modules.
221
2224.91 Thu Aug 6 15:42:45 CEST 2009
223 - AE::Handle::starttls could get out of sync when the read buffer
224 already contains some TLS handshake.
225 - AE::Handle did not properly free the TLS session, causing bigger
226 memory leaks in Net::SSLeay than expected (reported by toaster).
227 - AE::Socket::tcp_connect will now artificially delay invoking
228 the callback to avoid returning after invoking the callback.
229 - convert many internal watcher uses to AE API.
230
2314.9 Sat Aug 1 11:07:01 CEST 2009
232 - Glib silently fails when registering a timeout with
233 a negative value, so avoid that.
234 - call condvar callback immediately when it is set
235 after the condition is already signalled.
236 - check rbuf_max condition only after trying to consume
237 data.
238 - normalise signal numbers to names when using the ae's
239 signal handling, but do not document this (yet).
240 - pure perl signal emulation did not properly set nonblocking
241 mode on the signal pipe on win32 (this is unlikely to have caused
242 any issues).
243 - new module: AnyEvent::Debug.
244 - AnyEvent::Strict now checks that a signal was specified
245 by name (not name or number).
246 - reduce memory footprint in typical cases by ~50kb by
247 only compiling signal code when necessary.
248 - add AnyEvent::Handle->rbuf_max.
249 - grab the AE namespace for future new API, implement stubs for
250 said future/alternative API.
251 - new function: AnyEvent::Socket::format_hostport.
252
2534.881 Tue Jul 28 12:51:53 CEST 2009
254 - work around a bug in local in pre-5.10 perls, causing
255 AnyEvent::Handle to recurse when it should not
256 (analysed by elmex).
257
2584.88 Tue Jul 28 04:04:37 CEST 2009
259 - re-bless the handle into a dummy package after calling
260 AnyEvent::Handle::destroy, so the user does not need to check
261 for errors after every push_write etc.
262 - do not attempt to run t/02_signals.t on obviously broken
263 platforms.
264
2654.87 Sun Jul 26 02:06:16 CEST 2009
6 - do not attempt to linger when there is no longer a valid fh. 266 - do not attempt to linger when there is no longer a valid fh.
7 - enforce tls mode to be either accept or connect, do not simply 267 - enforce tls mode to be either accept or connect, do not simply
8 segfault in Net::SSLeay. 268 segfault in Net::SSLeay.
269 - AnyEvent::Handle can now call tcp_connect itself (new parameters
270 connect, on_prepare, on_connect and on_connect_error). Updated
271 tutorial accordingly.
9 - add AnyEvent::Impl::Irssi backend. 272 - add AnyEvent::Impl::Irssi backend.
10 273
114.86 Mon Jul 20 23:52:29 CEST 2009 2744.86 Mon Jul 20 23:52:29 CEST 2009
12 - since the verbose warning is not enough, explicitly document 275 - since the verbose warning is not enough, explicitly document
13 that versions before 1.33 of Net::SSLeay are not secure. 276 that versions before 1.33 of Net::SSLeay are not secure.
14 - work around signal handling races in Event and (...) Event::Lib. 277 - work around signal handling races in Event and (...) Event::Lib.
15 - try to align signal-race timer to full-second boundaries. 278 - try to align signal-race timer to full-second boundaries.
16 - work around Tk not liking negative timeouts. 279 - work around Tk not liking negative timeouts.
17 - don't complain of different grades of environmental unfriendlyness 280 - don't complain of different grades of environmental unfriendlyness
18 in IO::ASync. 281 in IO::Async.
19 282
204.85 Sat Jul 18 06:16:14 CEST 2009 2834.85 Sat Jul 18 06:16:14 CEST 2009
21 - nail the signal race problem in perl once and for all 284 - nail the signal race problem in perl once and for all
22 (see $AnyEvent::MAX_SIGNAL_LATENCY). 285 (see $AnyEvent::MAX_SIGNAL_LATENCY).
23 - take advantage of Async::Interrupt if it is available. 286 - take advantage of Async::Interrupt if it is available.
… …
2204.3 Fri Oct 3 09:18:43 CEST 2008 4834.3 Fri Oct 3 09:18:43 CEST 2008
221 - AnyEvent will now install a no-op signal handler for SIGPIPE, 484 - AnyEvent will now install a no-op signal handler for SIGPIPE,
222 unless one has been installed already. 485 unless one has been installed already.
223 - warn about not loaidng AnyEvent::Impl::POE early enough 486 - warn about not loaidng AnyEvent::Impl::POE early enough
224 (patch by Adam Rosenstein). 487 (patch by Adam Rosenstein).
225 - fixed a great number of bugs and corner cases in AnyEvent::Handle: 488 - fixed a great number of bugs and corner cases in AnyEvent::Handle.
226 - fix a bug in where in SSL connect mode, the 489 - fix a bug in where in SSL connect mode, the
227 client would first wait for some data by the server and 490 client would first wait for some data by the server and
228 otherwise hang. (reported and analyzed in an absolutely 491 otherwise hang. (reported and analysed in an absolutely
229 exemplary manner by Adam Rosenstein). 492 exemplary manner by Adam Rosenstein).
230 - fix a bug in where SSL EOF would not be treated 493 - fix a bug in where SSL EOF would not be treated
231 as stream EOF, putting the connection into a hung state. 494 as stream EOF, putting the connection into a hung state.
232 (reported and analyzed in an absolutely exemplary manner by Adam 495 (reported and analysed in an absolutely exemplary manner by Adam
233 Rosenstein). 496 Rosenstein).
234 - fix a potential segfault when the TLS context 497 - fix a potential segfault when the TLS context
235 would go missing in a read callback (e.g. due to stoptls) - 498 would go missing in a read callback (e.g. due to stoptls) -
236 Net::SSLeay of course makes no type checking whatsoever. 499 Net::SSLeay of course makes no type checking whatsoever.
237 - AnyEvent::Handle will not stop the read watcher in TLS mode. 500 - AnyEvent::Handle will not stop the read watcher in TLS mode.

Diff Legend

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