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

Comparing AnyEvent/Changes (file contents):
Revision 1.167 by root, Tue Jul 8 23:44:51 2008 UTC vs.
Revision 1.535 by root, Fri Mar 30 01:49:28 2012 UTC

1Revision history for Perl extension AnyEvent. 1Revision history for Perl extension AnyEvent.
2 2
34.2 3TODO: document TCP_*** constants
4TODO: add debug/trace logging to some modules
5TODO: AnyEvent::IO
6TODO: PERL_ANYEVENT_IO_MODEL docs
7tODO: maxim souffle handle storable
8TODO: document import tags aeio, also ae_ naming
9TODO: t/io
10lib/AnyEvent/DNS.pm: open my $fh, "<", $resolv_conf
11lib/AnyEvent/DNS.pm: if (open my $fh, "ipconfig /all |") {
12lib/AnyEvent/Log.pm: open my $fh, ">>", $path
13lib/AnyEvent/Log.pm: open my $fh, ">>", $path
14lib/AnyEvent/Socket.pm: open my $fh, "<",
15lib/AnyEvent/TLS.pm: or croak "$dh_file: failed to open DH parameter file: $!";
16
176.2
18 - documented Rocco fabricating statements about AnyEvent and me.
19 - new module AnyEvent::IO, that is a frontend to either a pure-perl
20 synchronous I/O implementation (AnyEvent::IO::Perl), or to the
21 asynchronous I/O extension IO::AIO (AnyEvent::IO::IOAIO).
22 - load /etc/hosts only when dns fails has no answer.
23 - stat /etc/hosts on every access and reload it if it changed.
24 - load /etc/hosts via AnyEvent::IO - potentially asynchronous.
25 - the default logging level was not properly documented in a variety of
26 places, this has hopefully been rectified.
27 - updated uts46data.pl for unicode 6.1.0.
28 - made log messages genearted by AnyEvent submodules not
29 include the package name anymore, as it will be logged
30 by default already.
31 - upgrade to the trick used by common-sense 3.5 to work
32 around extra warning torture/breakage under perl 5.15.x.
33 - log messages by anyevent are now ucfirst, are usually full
34 sentences and do no longer include the package name.
35
366.14 Tue Jan 31 20:00:24 CET 2012
37 - AnyEvent::Impl::Tk was broken due to a mysterious "die" inside,
38 probably an editing mistake (reported by Darin McBride).
39
406.13 Thu Jan 12 07:27:01 CET 2012
41 - AnyEvent::Util::fork_call checked for POSIX availability in the wrong way,
42 causing it to fail if POSIX wasn't loaded already (analysed by Rock Power).
43 - AnyEvent::Handle::push_read (line => did pass $1 directly, so
44 regex-matching inside the callback would change the parameter, despite
45 $1 being dynamically scoped per-block. this perl bug is now being
46 worked around (testcase by Cindy Wang).
47
486.12 Mon Dec 12 13:21:10 CET 2011
49 - $! was clobbered by subsequent calls in tcp_connect, due to
50 the postpone not saving/restoring it (analyzed by Richard Garnish).
51
526.11 Tue Nov 22 10:36:05 CET 2011
53 - Tk cannot create windows when tainted unless you set a title, so
54 set a dummy title for AnyEvent's dummy mainwindow. How dummy.
55 - escape any nonprintable/nonascii characters when stringifying
56 backtraces.
57 - log the reason loading an interface module fails to load at level debug.
58
596.1 Tue Oct 4 19:44:30 CEST 2011
60 - INCOMPATIBLE CHANGE: the default log level is now "4" (error
61 and above), and some messages inside AnyEvent have been elevated
62 to higher log levels to print by default.
63 - AnyEvent::log, unlike AnyEvent::Log::log, did not always exit on fatal.
64 - work around yet another signal race bug, newly introduced by POE. My
65 how that module sucks.
66 - add google public ipv6 dns as fallback servers, so AE::DNS can finally
67 support ipv6-only operation even with defaults (also use ::1 instead
68 of 127.0.0.1).
69 - give better guidelines on how to chose a logging level.
70 - AnyEvent::Log can now cap (limit) log message priority by context.
71 - AE::log will now save and restore $! and $@, so callers don't have to.
72 - AE::Strict now checks that registered fds are still valid at regular
73 intervals.
74 - skip frequently called 'require' calls if the module appears loaded.
75 - increase signal latency test timer considerably for poor little osx,
76 which can't be bothered with basic posix functionality such as signal
77 delivery.
78 - mention some debugging aids in the tutorial.
79 - new env variable PERL_ANYEVENT_MAX_SIGNAL_LATENCY.
80 - private (%namespace) log contexts in PERL_ANYEVENT_LOG now have the name
81 of the context as title.
82 - AnyEvent::Strict would not restore read-write access to $_ when a callback
83 died - this is a user bug, but since AE::Strict exists to help dbeugging...
84 - detect the extremely evil IO::Async::Loop::AnyEvent module and refuse
85 to work with it - the author is well aware of the issues but refuses
86 to fix them. This extreme step has been taken because the module actively
87 attacks the goals of AnyEvent and makes it harder for module authors to use
88 AnyEvent.
89
906.02 Fri Aug 26 20:08:31 CEST 2011
91 - log_to_syslog now accepts facility strings.
92 - log_to_syslog would not allow facilities - "$facility|$level"
93 *literally* means that string. sheesh.
94 - the CPAN parser got confused and thought we provide Sys::Syslog, try
95 to work around.
96 - renamed AnyEvent::Impl::FLTK2 to FLTK.
97
986.01 Fri Aug 26 07:04:11 CEST 2011
99 - INCOMPATIBLE CHANGE: AnyEvent::DNS resource records now include
100 the TTL as fourth element - this affects ->resolve, ->request
101 and ->dns_unpack indirectly. Convenience functions such as
102 AnyEvent::DNS::{a,txt,mx,srv} are not affected, neither is
103 AnyEvent::Socket.
104 - INCOMPATIBLE CHANGE: backend authors now should not implement
105 one_event or loop, but instead the AnyEvent::CondVar::_wait and _poll
106 methods.
107 - INCOMPATIBLE CHANGE: verbosity levels have been redefined, see
108 AnyEvent (PERL_ANYEVENT_VERBOSE) and AnyEvent::Log for details.
109 - DEPRECIATION: please change
110 "use AnyEvent::Impl::Perl" to "use AnyEvent::Loop" and
111 "AnyEvent::Impl::Perl::loop" to "AnyEvent::Loop::run"
112 in your programs, for future compatibility.
113 - *blush* idle watcher emulation was borked.
114 - AnyEvent::Strict now checks the AE::xxx functions as well.
115 - querying too long domain names would barf deep inside AnyEvent::DNS - since
116 this is hard to avoid for applications, AnyEvent:.DNS::request now checks for
117 for too long domain names explicitly and passes an undef to the callback
118 instead (reported by James Bromberger).
119 - make AnyEvent::Util more robust against EINTR (analyzed by Andrew Suffield).
120 - new functions: AnyEvent::postpone, AnyEvent::log.
121 - new module: AnyEvent::Log, for simple logging needs.
122 - new env variable, PERL_ANYEVENT_LOG, for fine-grained logging config.
123 - new env variable, PERL_ANYEVENT_DEBUG_SHELL, to automatically start
124 a debug shell.
125 - new env variable, PERL_ANYEVENT_DEBUG_WRAP, to automatically enable
126 watcher instrumentation/wrapping (see AnyEvent::Debug::wrap).
127 - new env variable, PERL_ANYEVENT_HOSTS, to override the place where
128 /etc/hosts can be found.
129 - an empty PERL_ANYEVENT_RESOLV_CONF now selects the default
130 configuration.
131 - env variables can now be specified using AE_xyz instead of
132 PERL_ANYEVENT_xyz, subject to some rules.
133 - lots of new AnyEvent::Debug functionality.
134 - AnyEvent::Strict now checks for modifications of $_ by the callback
135 (by setting it to readonly while executing the callback, if possible).
136 - IO::Async sometimes cancels the wrong timer when you call cancel_timer.
137 Invest extra overhead to work around this peculiar behaviour.
138 - split the pure perl backend into a loop module and an impl
139 module (AnyEvent::Loop and AnyEvent::Impl::Perl).
140 - some read types (regex, netstring, json, storable) could
141 errornously keep an AE::Handle object alive on their own
142 (reported by Mohammad Toossi).
143 - AnyEvent::Socket::resolve_sockaddr and all functions using it now
144 supports /etc/hosts overrides.
145 - add more workarounds around child watcher bugs in IO::Async
146 and (newly introduced) in POE.
147 - use glib child watchers and try to work around its limitations,
148 if possible (based on a patch by Kevin Ryde).
149 - add support for FLTK (fltk2), via AnyEvent::Impl::FLTK2.
150 - many read types in AnyEvent::Handle would malfunction during
151 transitions to TLS because of a "delete $self->{rbuf}".
152 (reported by Gerald Galster).
153 - improve condvar blocking wait performance (probably only noticable
154 with faster backends), also make it easier for event loops that
155 do not support blocking waits.
156 - PERL_ANYEVENT_MODEL now also allows full module names, although
157 the usefulness of this feature is not yet clear.
158 - add tests for all supported event loops, conditional on PERL_ANYEVENT_LOOP_TESTS.
159 - add t/80_ssltext.t.
160 - no longer use AUTOLOAD for the initial loading (for no very good reason).
161 - tcp_server now tries to unlink unix domain sockets when it gets destroyed.
162 - do not run event loop detection multiple times when programs call their cached
163 copies of AnyEvent::detect.
164 - parse_hostport now accepts absolute paths als unix domain sockets.
165 - suppress the idiotic warning inside POE by patching the POE::Kernel
166 run flags instead of asking the user to jump through contortions.
167 - add Task::Weaken as dependency (as recommends only at this time).
168 - replace PF_UNSPEC by 0, for good old minix.
169
1706.0 (not officially released, but was on CPAN for about 15 minutes :/)
171
1725.34 Thu May 12 09:59:41 CEST 2011
173 - AE::TLS didn't support both cert_file and key at the same time,
174 due to some copy&paste bug (reported and analyzed by vti).
175 - improve AE::Handle timeout documentation slightly.
176
1775.33 Mon Feb 7 21:51:41 CET 2011
178 - never appeared on CPAN, for unknown reasons.
179 - AnyEvent::Handle on_connect_error - do not try to destroy a handle
180 object when it's already gone (analyzed by Ryan Bullock).
181 - update idna algorithm(s) to UTS#46 version 6.0.0.
182 - preserve trailing dot in idn_to_ascii - testsuite says to preserve,
183 uts46 says it's an error, and testsuite makes more sense.
184 - remove leading dots in idn_to_ascii - testsuite says to remove,
185 uts46 says it's an error, and testsuite makes more sense.
186 - fixed a bug in idn ignored character processing.
187 - passes the uts46 6.0.0 testsuite except for three cases which
188 apparently are bugs in the testsuite itself.
189 - passes the uts46 6.0.1 testsuite except for eleven cases which
190 apparently are bugs in the testsuite itself.
191
1925.31 Sun Jan 23 11:49:19 CET 2011
193 - work around a POE bug where POE would first poll for new events and
194 invoke handlers BEFORE handling the message queue.
195 - fix parsing of "domain" statement in resolv.conf (Johannes Plunien).
196 - stop_read now does just that in TLS mode. see its
197 documentation for caveats.
198 - update to common::sense 3.4 warning set (some warnings in perl impose
199 an arbitrary 32kb limit on regex matches, which affects
200 AnyEvent::Handle).
201 - AE::Handle had a broken rbuf_max method and missing wbuf_max method.
202 - new experimental AE::Handle::resettls method.
203
2045.3 Fri Dec 31 05:48:13 CET 2010
205 - major IOAsync patch by Paul Evans - 0.33 is required, as the major
206 issues with IO::Async have been solved in that version, including
207 the availabilty of a default loop!
208 - implement new wbuf_max parameter for AnyEvent::Handle.
209 - added Cocoa::EventLoop backend by Daisuke Murase.
210 - the default testsuite now honors PERL_ANYEVENT_MODEL (Paul Evans).
211 - clarify handle on_prepare documentation.
212 - set CLOEXEC flag on parent-side fd's in AnyEvent::Util::run_cmd,
213 to avoid leaking them to any child processes.
214 - try to force files into the arch-dependend lib directory, for the
215 half-"turly obsessive compulsive person"(s) that depend on it.
216
2175.29 Sun Dec 5 10:49:21 CET 2010
218 - convert EV backend to EV 4.00 API (so better upgrade EV too).
219 - AE::Handle: implement dynamic read_size adjustment, add
220 max_read_size parameter, reduce default read_size to 2048.
221 - add some FAQ entries for servers and/or on_eof vs. on_error.
222 - work around OS bugs (cygwin again) when lingering in AE::Handle.
223 - reduce memory usage slightly when connect was used in AE::Handle.
224 - wrok around more segfaults in Tk.
225 - document tls_ctx => undef as valid.
226 - detect WNOHANG value at compiletime.
227 - include some probably linux-only support for building AnyEvent
228 as part of the perl core.
229 - improve compatibility of t/08_idna.t to perl 5.8.
230 - make AnyEvent::Strict truly optional (in case somebody builds a
231 minimal perl...).
232
2335.28 Wed Oct 13 04:14:23 CEST 2010
234 - due to a glitch, AnyEvent's internal getprotobyname wasn't
235 actually used in tcp_server.
236 - implement AnyEvent::CondVar->new method, which was documented but
237 missing.
238 - support autoloading of read types in unshift_read.
239 - AnyEvent::Strict: do the test for invalid fh arguments differently,
240 resulting in more informative error messages.
241 - new AnyEvent::FAQ manpage.
242 - clarify that you can actually call push_read etc. in on_read.
243 - put keys and certs in /tmp, not into the current directory,
244 and document this in the SECURITY CONSIDERATIONS sections
245 of AnyEvent::TLS.
246 - support ";" as resolv.conf comment character in AnyEvent::DNS, as
247 requested by Ron Isaacson.
248 - document $AnyEvent::DNS::RESOLVER variable.
249 - incorporated major typo patches by Abhijit Menon-Sen.
250 - AnyEvent::Handle now croaks when negative timeouts are passed.
251 - add a shitload of TCP_xxx sockopt name constants to AnyEvent::Util,
252 undocumented, but who knows when they might come in handy.
253
2545.271 Tue Jun 8 12:05:46 CEST 2010
255 - backport to perl 5.8.x.
256
2575.27 Sun Jun 6 12:12:05 CEST 2010
258 - postpone differently in AnyEvent::Socket now, as
259 when not, canceling the connection attempt might fail
260 (found by Felix Antonius Wilhelm Ostmann).
261 - explicitly check for non-stream sockets in AE::Handle, too many
262 clueless people fell into the trap of this somehow working.
263 - simplified and reworked the "OTHER MODULES" section.
264 - better/more condvar examples.
265
2665.261 Wed Apr 28 16:13:36 CEST 2010
267 - AF_INET6 was not properly used from Socket6 during configuration
268 time (found by Andy Grundman).
269
2705.26 Mon Apr 12 04:49:35 CEST 2010
271 - don't generate spurious readyness notifications when select
272 returns EINTR in the pure perl backend (analysed and testcase
273 by Bas Denissen).
274 - give same examples in SYNOPSIS sections of AnyEvent and AE.
275 - provider faster implementations for AE::time/now/now_update for
276 EV and Perl backends.
277 - provide faster AE::cv and AE::time implementations for all
278 backends.
279 - no longer support lower/mixed-case signal names (this was
280 never documented nor universally supported).
281 - some more memory µ-optimisations, and somewhat less messy
282 code to cope with both AnyEvent and AE APIs.
283
2845.251 Sat Mar 13 00:58:21 CET 2010
285 - make test failed if it was run before make install on perl <=
286 5.8.8, and older versions of perl make our live very difficult,
287 so write constants.pl during Makefile.PL time.
288
2895.25 Sat Mar 13 00:23:14 CET 2010
290 - fix a race condition in AnyEvent::Handle that would cause
291 a "bio_read: unsupported method" error (in _another_
292 ssl connection) after user code threw an exception.
293 - added AnyEvent::Handle->destroyed method.
294 - speed up AnyEvent::Socket::format_address and ::format_ipv6.
295 - the AnyEvent::Util::fh_nonblocking function worked only
296 by lucky accident on win32.
297 - smaller and faster AnyEvent::Util::fh_nonblocking.
298 - when the (required!) Time::HiRes module is missing, AnyEvent
299 did not fall back to built-in time properly.
300 - do not load Fcntl at runtime, saving memory and loading time.
301 - precompile a number of constants and use them instead of runtime
302 detection and eval.
303 - free detection code after detection and similar memory optimisations.
304 - Perl backend timer interval best effort drift has been improved
305 (same algorithm as EV).
306 - update unicode idna mapping table.
307
3085.24 Tue Jan 5 11:39:43 CET 2010
309 - cygwin never reports errors from failed connects. 1.5 just gives
310 you continous readyness and EAGAIN, 1.7 is even more broken and
311 just hangs. work around both issues in a cygwin-specific hack.
312 - improve idle watcher documentation slightly.
313
3145.23 Sun Dec 20 23:48:00 CET 2009
315 - support IDNs in resolve_sockaddr, and therefore in tcp_connect.
316 - implement punycode_encode/decode, idn_nameprep,
317 idn_to_ascii and idn_to_unicode operations in AnyEvent::Util.
318 - provide $AE::VERSION.
319 - removed traces of "no strict 'refs'".
320
3215.22 Sat Dec 5 03:51:13 CET 2009
322 - downgrade-or-fail in AnyEvent::Handle::push_write, to
323 diagnose encoding failures earlier and more succinctly.
324 (this works around bugs in perl, throwing away encoding info
325 when passing scalar data to extensions).
326 - add more examples to AnyEvent::Socket manpage.
327 - upgrade internal warning set to the same as common::sense 2.03.
328 - use pack "n/a*" for pre-5.8.9 perl compatibility in AnyEvent::DNS
329 (John Beppu).
330 - AnyEvent::Socket::inet_aton now properly supports ipv6, as documented.
331 - add google public dns servers to fallback server set and make sure
332 we load-balance properly between the three sets. also add all
333 fallback dns servers, not just a random one, to each dns config.
334
3355.21 Thu Nov 19 02:48:47 CET 2009
336 - fix a problem where socket constants were called with parameters
337 (spotted by David Friedland).
338 - fork_call never use'd POSIX (reported by Daisuke Maki).
339 - improve perl 5.6 compatibility further (but it still won't work
340 unless you rip out everything but the core).
341 - prefer Net::DNS::Resolver over ipconfig, if installed, on win32.
342 uses 10MB of RAM, but doesn't open a console window. *sigh*.
343
3445.202 Wed Oct 14 22:35:44 CEST 2009
345 - AnyEvent::DNS would unexpectedly clobber $_ under windows
346 (analysed by Matthias Waldorf).
347 - AnyEvent::Handle::run_cmd can now pass the PID of the
348 newly-created process, which is much less useful than it might
349 sound (based on patch by Yann Kerherve).
350
3515.201 Tue Sep 29 12:09:25 CEST 2009
352 - AnyEvent:Handle::on_starttls/on_stoptls methods were broken
353 (reported by Torsten Foertsch).
354 - common::sense 2.0 could cause tcp_server to throw an exception
355 (analysed by elmex).
356
3575.2 Mon Sep 14 07:04:49 CEST 2009
358 - INCOMPATIBLE CHANGE: do no longer support register_read_type
359 and register_write_type in AnyEvent::Handle, instead support
360 package names (the facility was mostly abused).
361 - implement "packagename-as-read/write type" support in
362 AnyEvent::Handle.
363 - AnyEvent::Handle: new options "keepalive" and "oobinline".
364 - oobinline set by default to avoid security issues.
365 - the pure-perl event loop backend wrongly detected times() fallback
366 support (spotted by Pavel Boldin).
367
3685.12 Tue Sep 1 20:26:50 CEST 2009
369 - be more lenient when parsing resolv.conf files, as some people
370 use hashmarks as comment indicator inside directives (reported
371 by Michael S. Fischer).
372 - use same set of warnings as common::sense 2.0.
373 - fix a potential 32 bit overflow issue due to perl having problems
374 with large hex constants in the Perl backend.
375 - do not use the slower 5.6 version fo the enc_name code in AnyEvent::DNS
376 by default.
377 - fix some prototypes.
378 - (optionally) require Async::Interrupt 1.02.
379 - replace opendns fallbacks by verizon/level3 public dns servers.
380
3815.112 Fri Aug 21 13:59:04 CEST 2009
382 - AnyEvent::Handle could cause hard crashes in Net::SSLeay when
383 using starttls with an illegal/uncreatable tls context.
384 - port to broken win32 perls.
385
3865.111 Sun Aug 16 18:44:36 CEST 2009
387 - perl errornously requires sockaddr_un structures returned by the
388 kernel to have a specific length. work around this bug by padding
389 sockaddr_un structures (found to fail on OS X, as perl only uses
390 the correct code for GNU/Linux, but likely to fail on other
391 OSes as well).
392
3935.11 Wed Aug 12 17:49:37 CEST 2009
394 - fixed a bug in run_cmd with input from filename.
395 - create a json coder object when none is given in write direction
396 to reduce dependency on JSON.
397
3985.1 Tue Aug 11 03:17:32 CEST 2009
399 - tcp_connect now reports the correct error in $!, not an unrelated
400 one (was broken since 4.91).
401 - AE::Handle did not properly clear rtimeout/wtimeout watchers
402 when retrying a tcp_connect.
403 - new functions: AnyEvent::Util::run_cmd and close_all_fds_except.
404 - fixed wrong documentation in AE::Socket.
405
4065.01 Mon Aug 10 03:16:32 CEST 2009
407 - last release broke EV child watchers :/.
408
4095.0 Sun Aug 9 17:32:09 CEST 2009
410 - officially introduce and document the AE API.
411 - lowering the timeout at runtime did not have immediate effect
412 in AnyEvent::Handle.
413 - AE::Handle now has separate and independent read and write timeout
414 settings.
415 - probe for Guard module when AnyEvent::Util is loaded, not at
416 runtime (Event doesn't like eval "use" from callbacks, and it avoids
417 exporting the probe function).
418 - Impl::IOAsync's io watchers did not use the filehandle emulation
419 code to work around it's watcher limits.
420 - work around IO::Async's problems with overloaded objects.
421 - try to disable wearnings inside IO::Async, it's just too buggy.
422 - do no longer use Event by default, only EV or Perl, Event is too buggy.
423 - hackishly provide fast AE:: interface when EV or Perl is the backend.
424 - inherit VERSION from AnyEvent::VERSION in most modules.
425
4264.91 Thu Aug 6 15:42:45 CEST 2009
427 - AE::Handle::starttls could get out of sync when the read buffer
428 already contains some TLS handshake.
429 - AE::Handle did not properly free the TLS session, causing bigger
430 memory leaks in Net::SSLeay than expected (reported by toaster).
431 - AE::Socket::tcp_connect will now artificially delay invoking
432 the callback to avoid returning after invoking the callback.
433 - convert many internal watcher uses to AE API.
434
4354.9 Sat Aug 1 11:07:01 CEST 2009
436 - Glib silently fails when registering a timeout with
437 a negative value, so avoid that.
438 - call condvar callback immediately when it is set
439 after the condition is already signalled.
440 - check rbuf_max condition only after trying to consume
441 data.
442 - normalise signal numbers to names when using the ae's
443 signal handling, but do not document this (yet).
444 - pure perl signal emulation did not properly set nonblocking
445 mode on the signal pipe on win32 (this is unlikely to have caused
446 any issues).
447 - new module: AnyEvent::Debug.
448 - AnyEvent::Strict now checks that a signal was specified
449 by name (not name or number).
450 - reduce memory footprint in typical cases by ~50kb by
451 only compiling signal code when necessary.
452 - add AnyEvent::Handle->rbuf_max.
453 - grab the AE namespace for future new API, implement stubs for
454 said future/alternative API.
455 - new function: AnyEvent::Socket::format_hostport.
456
4574.881 Tue Jul 28 12:51:53 CEST 2009
458 - work around a bug in local in pre-5.10 perls, causing
459 AnyEvent::Handle to recurse when it should not
460 (analysed by elmex).
461
4624.88 Tue Jul 28 04:04:37 CEST 2009
463 - re-bless the handle into a dummy package after calling
464 AnyEvent::Handle::destroy, so the user does not need to check
465 for errors after every push_write etc.
466 - do not attempt to run t/02_signals.t on obviously broken
467 platforms.
468
4694.87 Sun Jul 26 02:06:16 CEST 2009
470 - do not attempt to linger when there is no longer a valid fh.
471 - enforce tls mode to be either accept or connect, do not simply
472 segfault in Net::SSLeay.
473 - AnyEvent::Handle can now call tcp_connect itself (new parameters
474 connect, on_prepare, on_connect and on_connect_error). Updated
475 tutorial accordingly.
476 - add AnyEvent::Impl::Irssi backend.
477
4784.86 Mon Jul 20 23:52:29 CEST 2009
479 - since the verbose warning is not enough, explicitly document
480 that versions before 1.33 of Net::SSLeay are not secure.
481 - work around signal handling races in Event and (...) Event::Lib.
482 - try to align signal-race timer to full-second boundaries.
483 - work around Tk not liking negative timeouts.
484 - don't complain of different grades of environmental unfriendlyness
485 in IO::Async.
486
4874.85 Sat Jul 18 06:16:14 CEST 2009
488 - nail the signal race problem in perl once and for all
489 (see $AnyEvent::MAX_SIGNAL_LATENCY).
490 - take advantage of Async::Interrupt if it is available.
491 - load Time::HiRes and Guard modules on demand only.
492 - add optional/recommended modules section to AnyEvent
493 documentation.
494 - reduce memory usage considerably (and reduce startup penalty)
495 by not using "strict", "warnings" and "overload" modules.
496 - work around buggy windows/openbsd perls and provide EBADMSG
497 and EPROTO ourselves when missing.
498 - improve perl 5.6 compatibility of the core event loop.
499 - made Net::SSLeay version 1.33 a soft requirement.
500
5014.83 Fri Jul 17 16:56:26 CEST 2009
502 - implement AnyEvent::Socket::getprotobyname.
503 - AnyEvent::CondVar's will now detect recursive blocking
504 waits and will croak, as too many people fall into
505 this trap.
506 - AnyEvent::Handle will now call ->destroy on itself after
507 executing the on_error callback, instead of doing some
508 half-baked internal shutdown, for fatal errors.
509 - clarify on_eof behaviour w.r.t. the read queue and
510 on_read callbacks.
511 - ignore some possible spurious wake-ups in tcp_connect.
512
5134.82 Sat Jul 11 00:34:55 CEST 2009
514 - POE and Event backends didn't accept some callable objects as
515 callbacks.
516 - use Config module instead of POSIX module to detect signal names
517 in AnyEvent::Strict and AnyEvent::Impl::EventLib, as the POSIX
518 module doesn't even have all POSIX signals :/.
519 - use more workarounds around the many refcnt/corruption bugs in
520 Event::Lib.
521 - work around a race condition in perl's select, causing t/03_child.t
522 to rarely fail.
523
5244.81 Thu Jul 9 10:30:30 CEST 2009
525 - AnyEvent::Handle didn't properly diagnose write errors
526 (it expected -1 from syswrite, how lame... :).
527 - support file descriptors in addition to file handles
528 in AnyEvent->io.
529 - new env variables: PERL_ANYEVENT_RESOLV_CONF,
530 PERL_ANYEVENT_MAX_OUTSTANDING_DNS, PERL_ANYEVENT_CA_FILE
531 and PERL_ANYEVENT_CA_PATH.
532 - provide a sensible synopsis section for AnyEvent::TLS.
533 - add a "supported backends" section to the manpage.
534 - added simple io watcher test to testsuite, using a
535 portable_socketpair.
536 - tried to improve the stability of the Event::Lib backend,
537 YMMV.
538
5394.8 Mon Jul 6 23:45:16 CEST 2009
540 - AnyEvent::DNS did not properly follow CNAME records with
541 uppercase targets.
542 - AnyEvent::DNS would errornously return AAAA records
543 with v4 mapped addresses (a faulty record) as ipv4 addresses,
544 causing AnyEvent::Socket to throw an exception.
545 - added new module AnyEvent::TLS for easier SSL/TLS context
546 creation, with many options including hostname verification,
547 secure default configuration, lots of documentation and,
548 predefined diffie-hellman keys for perfect forward security
549 and much more. get it while it's still fresh!
550 - use AnyEvent::TLS in AnyEvent::Handle for context management.
551 - load AnyEvent::Handle only on demand in AnyEvent::DNS,
552 so AnyEvent::Socket users have smaller memory footprint
553 in the common case.
554 - add AnyEvent::Handle->push_shutdown method.
555 - add an additional error message parameter to AnyEvent::Handle's
556 on_error callback (for TLS, $! is still available).
557 - add AnyEvent::Handle on_starttls/on_stoptls callbacks.
558 - make AnyEvent::Handle more robust against early conenction
559 failures (during new), and return C<undef> in that case
560 from the constructor.
561 - AnyEvent::Handle will now try to load only JSON::XS first,
562 then fall back to JSON.
563 - format_ipv4/format_ipv6 are now exported by default, for symmetry,
564 and because it was documented that way.
565
5664.451 Fri Jul 3 00:28:58 CEST 2009
567 - do not clear rbuf when shutting down an AnyEvent::Handle
568 object - doing so breaks AnyEvent::HTTP.
569
5704.45 Mon Jun 29 22:59:26 CEST 2009
571 - a write error could cause AnyEvent::Handle to create
572 an I/O watcher with an undefined $fh.
573 - special-case mapped ipv4 addresses in both
574 AnyEvent::Socket::format_address and parse_address,
575 to treat them just like ipv4 addresses.
576 - updated and overhauled the AnyEvent::Intro doc.
577 - implement AnyEvent::Socket::format_ipv4/ipv6.
578 - slightly speed up portable_pipe/socketpair.
579 - expand condvar begin/end documentation.
580
5814.42 Fri Jun 26 08:32:18 CEST 2009
582 - preliminary and neccesarily incomplete support for IO::Async.
583 - reset SIGCHLD to DEFAULT when AnyEvent is loaded, in
584 case it was set to IGNORE, to ensure we can catch child statuses
585 even when the calling env acts stupidly.
586 - updated benchmarks with IO::Async, which performs very well.
587
5884.412 Wed Jun 24 01:35:57 CEST 2009
589 - support an "untaint" attribute for AnyEvent::DNS
590 and set it on the default resolver.
591 - implement and document AnyEvent::Impl::Perl::loop.
592 - remove all anyevent-env variables from %ENV
593 when running in tainted mode.
594 - mention and extend the IO::Lambda benchmark.
595
5964.411 Sun Jun 7 18:48:02 CEST 2009
597 - do not try to use F_SETFD on windows.
598
5994.41 Thu May 14 06:40:11 CEST 2009
600 - work around issues in older perls (5.8.0?) when
601 a signal handler is deleted from the %SIG hash.
602 - use POSIX::_exit in child test, to avoid running
603 destructors.
604 - speed up CHLD handling by relying on SIGCHLD being
605 synchronously delivered, even when we roll our own
606 implementation.
607 - AnyEvent::DNS: add the "dname" resource record name
608 for cosmetic reasons.
609
6104.4 Sun Apr 26 20:12:33 CEST 2009
611 - implemented idle watchers, where applicable.
612 - AnyEvent->time died when Event backend was in use.
613 - fix a memleak in the Tk backend.
614 - sped up Tk timer handling.
615 - clip negative "after" values to 0 in AnyEvent::Impl::Event
616 to avoid spamming out warning messages.
617 - fix Qt timers without interval.
618 - avoid Qt zero-timeout special case, in old libqt's.
619
6204.352 Mon Apr 20 16:31:11 CEST 2009
621 - fix AnyEvent::Strict error messages for child watchers.
622 - fix/update Makefile.PL "recommends" versions.
623 - add AnyEvent->now_update.
624
6254.351 Sat Apr 11 07:56:14 CEST 2009
626 - actually make the signal pipe work under win32.
627 - localise $! in signal handler to avoid changing $!, although
628 perl itself does not seem to save/restore errno either.
629 - set the cloexec flag on the signal pipe (normally set by perl too).
630
6314.35 Fri Mar 27 11:48:20 CET 2009
632 - event models relying on AnyEvent's signal watcher emulation did
633 invoke the callback asynchronously, contradicting documentation
634 and causing signals to get lost (this includes AnyEvent's own
635 event loop). AnyEvent now uses the standard pipe trick to make
636 callback execution synchronous to the event loop.
637 - AnyEvent::Handle didn't free TLS context data on DESTROY
638 (patch by Pavel Shaydo).
639 - work around the ever-incompatibly-changing API of MakeMaker.
640 - document that changing global variables without restoring them
641 is a bad idea in Perl (noted by Adam Rosenstein).
642 - AnyEvent::Strict now barfs if ->io is passed a file.
643
6444.34 Thu Feb 12 18:32:45 CET 2009
645 - separately buffer TLS read data, as otherwise the read queue could
646 deadlock as receiving data is not expected while draining the
647 read queue (which cna only happen with TLS).
648 - raise EBADMSG error on JSON decoding errors.
649 - fix some minor manpage bugs (reported by Maximilian Gaß).
650 - speed up select bitmask parsing quite a bit in the pure perl
651 backend.
652 - use CORE::select instead of just select in AnyEvent::Impl::Perl.
653
6544.331 Tue Jan 6 21:07:25 CET 2009
655 - socketpair fails on many vista machines because vista has
656 completely broken accept/getpeername and getsockname functions,
657 so we provide our own socketpair emulation that kind of works
658 (AnyEvent::Util::portable_pipe).
659 - new function: AnyEvent::Util::portable_socketpair.
660 - take advantage of the Guard module if it exists.
661
6624.33 Fri Nov 21 02:35:40 CET 2008
663 - AnyEvent::Strict did errornously flag a fileno of 0 as
664 illegal.
665 - reduce memory usage and slightly speed up the pure perl backend
666 by only storing the file descriptor, not the file handle.
667 - add missing autocork method to AnyEvent::Handle (reported
668 by Adam Rosenstein).
669 - AnyEvent::DNS->resolve errornously documented an $rcode result
670 argument, but there isn't (spotted by Henrik Krohns).
671 - the naked truth about Tk - it's basically unmaintained.
672
6734.32 Mon Nov 3 22:46:32 CET 2008
674 - fix AnyEvent::Socket::resolve_sockaddr to properly support
675 unix sockets again and choose a proper default for
676 the socket type.
677 - fix call to resolve_sockaddr in tcp_connect.
678
6794.31 Thu Oct 30 04:41:48 CET 2008
680 - implemented AnyEvent::Handle->destroy method.
681 - hint about unexpected effects in TLS mode.
682 - speed up AnyEvent::Impl::Perl by using more arrays
683 instead of hashes without the slightest loss of readability :)
684 - work around a perl argument refcounting bug.
685 - with some perl patches applied, AnyEvent now seems leak-free.
686
6874.3 Fri Oct 3 09:18:43 CEST 2008
688 - AnyEvent will now install a no-op signal handler for SIGPIPE,
689 unless one has been installed already.
690 - warn about not loaidng AnyEvent::Impl::POE early enough
691 (patch by Adam Rosenstein).
692 - fixed a great number of bugs and corner cases in AnyEvent::Handle.
693 - fix a bug in where in SSL connect mode, the
694 client would first wait for some data by the server and
695 otherwise hang. (reported and analysed in an absolutely
696 exemplary manner by Adam Rosenstein).
697 - fix a bug in where SSL EOF would not be treated
698 as stream EOF, putting the connection into a hung state.
699 (reported and analysed in an absolutely exemplary manner by Adam
700 Rosenstein).
701 - fix a potential segfault when the TLS context
702 would go missing in a read callback (e.g. due to stoptls) -
703 Net::SSLeay of course makes no type checking whatsoever.
704 - AnyEvent::Handle will not stop the read watcher in TLS mode.
705 - AnyEvent::Handle->stoptls will send a TLS close notify
706 instead of simply destroying the stream now.
707 - fix a bug where large blocks of data written in TLS mode
708 would not be sent unless triggered by receives.
709 - on_drain will now take the tls write buffer into account.
710 - SSL operations have been streamlined a bit and should be faster
711 now, more to come, though.
712 - added a FAQ section, will document nontrivial issues in there.
713
7144.234 Mon Sep 29 04:08:13 CEST 2008
715 - fix child watcher documentation: only child exits will be reported,
716 no trace events.
717 - mention SIGPIPE in AnyEvent::Handle.
718 - perl backend will now use POSIX::times as monotonic clock
719 when available and the monotonic clock option is not.
720 - run even without Time::HiRes in some cases now, but do not
721 provide sub-second accuracy (all *supported* perl versions have
722 Time::HiRes).
723
7244.233 Fri Aug 22 01:48:05 CEST 2008
725 - fix a bug in the testsuite, causing a hang.
726 - clarified lots of AnyEvent::Handle constructor arguments.
727
7284.232 Thu Aug 21 20:44:25 CEST 2008
729 - fix a bug in AnyEvent::Handle that could cause two fatal error
730 callbacks on EOF, which caused AnyEvent::HTTP to signal success twice
731 (testcase by Leon Brocard).
732 - clarified on_eof/on_error documentation a bit.
733
7344.231 Tue Jul 29 13:12:15 CEST 2008
735 - remove some debugging code left in AnyEvent::Util::fork_call
736 (and no, it's impossible to implement with the broken windows
737 perls without resource leaks or worse).
738
7394.23 Tue Jul 29 12:19:59 CEST 2008
740 - document the first parameter passed to condvar callbacks to be
741 the callback.
742 - add AnyEvent::Socket::{ntoa,aton} aliases.
743 - optimize the AE::Handle->push_read (line) for the default
744 eol marker.
745 - optimize push_read (packstring|storable) for small packets.
746 - invoke on_error callback when no on_eof callback is set.
747 - fix a bug in push_read (storable) of unknown impact.
748
7494.22 Sun Jul 20 16:34:13 CEST 2008
750 - new function AnyEvent::Socket::parse_hostport.
751 - as the bulkheads at microsoft can't even get getprotobyname reliably
752 working on their shitty fucking broken os we need to hardcode
753 some common protocol numbers in AnyEvent::Socket. How can
754 people even bother with such a pile of shit as windows.
755
7564.21 Thu Jul 17 14:40:05 CEST 2008
757 - INCOMPATIBLE CHANGE: AnyEvent::Util::fork_call now has a prototype
758 which is incompatible to previous invocation syntax.
759 - work around more windows perl bugs in fork_call. I cannot imagine
760 why anybody would bother implementing fork in such an obviously
761 unusable way.
762 - avoid calling $do_retry if the request has already finished in
763 AnyEvent::DNS, thus avoiding an exception (reported by Anatoly
764 K. Sharifulin).
765
7664.2 Sat Jul 12 22:42:11 CEST 2008
4 - support an "interval" argument for repeating timers. 767 - support an "interval" argument for repeating timers.
5 - not importing anything from Event causes it to malfunction
6 and not use Time::HiRes, work around this bug.
7 - fix a bug in the pure perl event loop that caused timers to 768 - fix a bug in the pure perl event loop that caused timers to
8 be delayed under some circumstances. 769 be delayed under some circumstances.
9 - implement optional argument checking via PERL_ANYEVENT_STRICT. 770 - implement optional argument checking via PERL_ANYEVENT_STRICT.
771 - not importing anything from Event causes it to malfunction
772 and not use Time::HiRes, work around this bug.
773 - more bug workarounds against the endlessly broken Tk module,
774 seems to work now on 32 bit machines, but less so on 64 bit.
10 - minor optimisations applied to most backends. 775 - minor optimisations applied to most backends.
11 776
124.161 Fri Jul 4 14:24:48 CEST 2008 7774.161 Fri Jul 4 14:24:48 CEST 2008
13 - fixed the t/handle/02_write.t, it read after write, with the 778 - fixed the t/handle/02_write.t, it read after write, with the
14 assumption that all data could be written before read 779 assumption that all data could be written before read
… …
168 - support service names in tcp_connect and tcp_server. 933 - support service names in tcp_connect and tcp_server.
169 - work around netbsd/openbsd bug causing unix sockets not to 934 - work around netbsd/openbsd bug causing unix sockets not to
170 become writable until empty. 935 become writable until empty.
171 - only make a dummy read for error detection when getpeername 936 - only make a dummy read for error detection when getpeername
172 returns ENOTCONN. 937 returns ENOTCONN.
173 - defend AnyEvent::Util::inet_aton against Coro::LWP brutal 938 - defend AnyEvent::Util::inet_aton against Coro::LWP's brutal
174 override. 939 override.
175 - new AnyEvent::Guard::cancel. 940 - new AnyEvent::Guard::cancel.
176 941
1773.5 Sat May 17 23:17:14 CEST 2008 9423.5 Sat May 17 23:17:14 CEST 2008
178 - gracefully deal with systems not defining CLOCK_MONOTONIC 943 - gracefully deal with systems not defining CLOCK_MONOTONIC
… …
192 957
1933.41 Sun May 11 19:53:13 CEST 2008 9583.41 Sun May 11 19:53:13 CEST 2008
194 - work around a bug in perl 5.8.8's local. Reported by Yi Ma Mao. 959 - work around a bug in perl 5.8.8's local. Reported by Yi Ma Mao.
195 960
1963.4 Sun May 11 00:29:25 CEST 2008 9613.4 Sun May 11 00:29:25 CEST 2008
197 - complete re-implementation of condvars with a hopefully much mroe useful 962 - complete re-implementation of condvars with a hopefully much more useful
198 API (backward compatible functions exist but are undocumented). 963 API (backward compatible functions exist but are undocumented).
199 - AnyEvent::Handle fully rewritten with a hopefully more powerful API 964 - AnyEvent::Handle fully rewritten with a hopefully more powerful API
200 (still subject to change). 965 (still subject to change).
201 - take advantage of CLOCK_MONOTONIC, if available, in AnyEvent::Impl::Perl. 966 - take advantage of CLOCK_MONOTONIC, if available, in AnyEvent::Impl::Perl.
202 - provide AnyEvent::post_detect and @AnyEvent::post_detect, which 967 - provide AnyEvent::post_detect and @AnyEvent::post_detect, which

Diff Legend

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