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

Comparing AnyEvent/Changes (file contents):
Revision 1.68 by root, Sat May 10 01:02:25 2008 UTC vs.
Revision 1.570 by root, Tue Jul 30 22:13:08 2013 UTC

1Revision history for Perl extension AnyEvent. 1Revision history for Perl extension AnyEvent.
2 2
3TODO: implement condvars... 3TODO: document TCP_*** constants
4TODO: add debug/trace logging to some modules
5TODO: use the eval-with-cb-call trick also to speed up JSON?
6TODO: FAQ: common pitfalls? invoke-before-return
7TODO: FAQ: scope-ids? hosts file order?
8TODO: AnyEvent::Handle with AnyEvent::IO?
9
10TODO: invalid. and localhost. specialcasing inside AEDNS and not AESocket (rfc6761)
11TODO: maybe always try to downgrade in push_write.
12TODO: maybe handle \r in lines?
13TODO: expose default formatting implementation (ft &c) in AE::log, also allow overridign fatal exit
14TODO: José Micó (3.0K) 1 inbox Error connecting to hosts resolved from /etc/hosts
15TODO: Daniel Carosone (5.2K) 1 inbox AnyEvent::IO vs AnyEvent::AIO (doc patch)
16TODO: maybe implement env variable to give hosts precedence
17 - uts46data.pl couldn't be found due to wrong naming of the file
18 (reported by Fulko Hew).
19 - AnyEvent::Debug shell can now run coro shell commands, if
20 available.
21 - t/63* tests were wrongly in MANIFEST.
22 - mention IO::AIO as optional dependency.
23 - kernel.org's finger server went MIA, switch to freebsd.org and icculus.org.
24
257.04 Thu Nov 15 02:23:58 CET 2012
26 - debugging code left in.
27
287.03 Thu Nov 15 02:16:11 CET 2012
29 - AnyEvent::Socket::inet_aton did not work when dns resolution was
30 used to find the addresses (analyzed and patch by Maxime Soulé).
31 - fix a memory leak in the /etc/hosts lookup code when hosts don't
32 resolve and are not in hosts (analyzed by sten).
33
347.02 Tue Aug 14 04:27:58 CEST 2012
35 - AnyEvent::Util::run_cmd could block indefinitely (analyzed and test
36 program by Yorhel).
37 - verified that AnyEvent::Socket follows RFC5952.
38 - try to parse "ADDR#PORT" in addition to "ADDR PORT".
39
407.01 Sun May 13 01:03:17 CEST 2012
41 - fail with EPROTO in AnyEvent::Handle wqhen TLS is requested but
42 not available, instead of throwing an exception.
43 - use File::Spec to get the tmpdir in t/*, to avoid needless
44 failures on (most, not mine :) windows boxes.
45 - new handle read types: tls_detect and tls_autostart.
46
477.0 Fri Apr 13 06:33:30 CEST 2012
48 - child watchers are broken in POE 1.352 (also many earlier
49 versions) and there seems to be no way to work around it, as POE
50 itself is inherently racy. Document this regression and add a delay
51 in t/68_poe_03_child.t for the time being.
52 - new module AnyEvent::IO, that is a frontend to either a pure-perl
53 synchronous I/O implementation (AnyEvent::IO::Perl), or to the
54 asynchronous I/O extension IO::AIO (AnyEvent::IO::IOAIO).
55 - load /etc/hosts only when DNS has no answer.
56 - stat /etc/hosts on every access and reload it if it changed.
57 - load /etc/hosts via AnyEvent::IO - potentially asynchronous.
58 - fix a buggy croak in the dh parameter reading in AnyEvent::TLS.
59 - AnyEvent::Log log_to_file and log_to_path now use AnyEvent::IO. As
60 a side effect, they now use true appending as opposed to libc
61 appending, and the file might not have been opened when the function
62 returns.
63 - the default logging level was not properly documented in a variety of
64 places, this has hopefully been rectified.
65 - updated uts46data.pl for unicode 6.1.0.
66 - made log messages generated by AnyEvent submodules not
67 include the package name anymore, as it will be logged
68 by default already.
69 - upgrade to the trick used by common-sense 3.5 to work
70 around extra warning torture/breakage under perl 5.15.x.
71 - log messages by anyevent are now ucfirst, are usually full
72 sentences and do no longer include the package name.
73 - the storable read type would sometimes throw an exception instead
74 of causing EBADMSG (analyzed by Maxime Soulé).
75 - work around a bug in openssl 1.0.1 which enforces a minimum rsa
76 keysize (reported by Paul Howarth).
77 - documented Rocco fabricating statements about AnyEvent and me.
78
796.14 Tue Jan 31 20:00:24 CET 2012
80 - AnyEvent::Impl::Tk was broken due to a mysterious "die" inside,
81 probably an editing mistake (reported by Darin McBride).
82
836.13 Thu Jan 12 07:27:01 CET 2012
84 - AnyEvent::Util::fork_call checked for POSIX availability in the wrong way,
85 causing it to fail if POSIX wasn't loaded already (analysed by Rock Power).
86 - AnyEvent::Handle::push_read (line => did pass $1 directly, so
87 regex-matching inside the callback would change the parameter, despite
88 $1 being dynamically scoped per-block. this perl bug is now being
89 worked around (testcase by Cindy Wang).
90
916.12 Mon Dec 12 13:21:10 CET 2011
92 - $! was clobbered by subsequent calls in tcp_connect, due to
93 the postpone not saving/restoring it (analyzed by Richard Garnish).
94
956.11 Tue Nov 22 10:36:05 CET 2011
96 - Tk cannot create windows when tainted unless you set a title, so
97 set a dummy title for AnyEvent's dummy mainwindow. How dummy.
98 - escape any nonprintable/nonascii characters when stringifying
99 backtraces.
100 - log the reason loading an interface module fails to load at level debug.
101
1026.1 Tue Oct 4 19:44:30 CEST 2011
103 - INCOMPATIBLE CHANGE: the default log level is now "4" (error
104 and above), and some messages inside AnyEvent have been elevated
105 to higher log levels to print by default.
106 - AnyEvent::log, unlike AnyEvent::Log::log, did not always exit on fatal.
107 - work around yet another signal race bug, newly introduced by POE. My
108 how that module sucks.
109 - add google public ipv6 dns as fallback servers, so AE::DNS can finally
110 support ipv6-only operation even with defaults (also use ::1 instead
111 of 127.0.0.1).
112 - give better guidelines on how to chose a logging level.
113 - AnyEvent::Log can now cap (limit) log message priority by context.
114 - AE::log will now save and restore $! and $@, so callers don't have to.
115 - AE::Strict now checks that registered fds are still valid at regular
116 intervals.
117 - skip frequently called 'require' calls if the module appears loaded.
118 - increase signal latency test timer considerably for poor little osx,
119 which can't be bothered with basic posix functionality such as signal
120 delivery.
121 - mention some debugging aids in the tutorial.
122 - new env variable PERL_ANYEVENT_MAX_SIGNAL_LATENCY.
123 - private (%namespace) log contexts in PERL_ANYEVENT_LOG now have the name
124 of the context as title.
125 - AnyEvent::Strict would not restore read-write access to $_ when a callback
126 died - this is a user bug, but since AE::Strict exists to help dbeugging...
127 - detect the extremely evil IO::Async::Loop::AnyEvent module and refuse
128 to work with it - the author is well aware of the issues but refuses
129 to fix them. This extreme step has been taken because the module actively
130 attacks the goals of AnyEvent and makes it harder for module authors to use
131 AnyEvent.
132
1336.02 Fri Aug 26 20:08:31 CEST 2011
134 - log_to_syslog now accepts facility strings.
135 - log_to_syslog would not allow facilities - "$facility|$level"
136 *literally* means that string. sheesh.
137 - the CPAN parser got confused and thought we provide Sys::Syslog, try
138 to work around.
139 - renamed AnyEvent::Impl::FLTK2 to FLTK.
140
1416.01 Fri Aug 26 07:04:11 CEST 2011
142 - INCOMPATIBLE CHANGE: AnyEvent::DNS resource records now include
143 the TTL as fourth element - this affects ->resolve, ->request
144 and ->dns_unpack indirectly. Convenience functions such as
145 AnyEvent::DNS::{a,txt,mx,srv} are not affected, neither is
146 AnyEvent::Socket.
147 - INCOMPATIBLE CHANGE: backend authors now should not implement
148 one_event or loop, but instead the AnyEvent::CondVar::_wait and _poll
149 methods.
150 - INCOMPATIBLE CHANGE: verbosity levels have been redefined, see
151 AnyEvent (PERL_ANYEVENT_VERBOSE) and AnyEvent::Log for details.
152 - DEPRECIATION: please change
153 "use AnyEvent::Impl::Perl" to "use AnyEvent::Loop" and
154 "AnyEvent::Impl::Perl::loop" to "AnyEvent::Loop::run"
155 in your programs, for future compatibility.
156 - *blush* idle watcher emulation was borked.
157 - AnyEvent::Strict now checks the AE::xxx functions as well.
158 - querying too long domain names would barf deep inside AnyEvent::DNS - since
159 this is hard to avoid for applications, AnyEvent:.DNS::request now checks for
160 for too long domain names explicitly and passes an undef to the callback
161 instead (reported by James Bromberger).
162 - make AnyEvent::Util more robust against EINTR (analyzed by Andrew Suffield).
163 - new functions: AnyEvent::postpone, AnyEvent::log.
164 - new module: AnyEvent::Log, for simple logging needs.
165 - new env variable, PERL_ANYEVENT_LOG, for fine-grained logging config.
166 - new env variable, PERL_ANYEVENT_DEBUG_SHELL, to automatically start
167 a debug shell.
168 - new env variable, PERL_ANYEVENT_DEBUG_WRAP, to automatically enable
169 watcher instrumentation/wrapping (see AnyEvent::Debug::wrap).
170 - new env variable, PERL_ANYEVENT_HOSTS, to override the place where
171 /etc/hosts can be found.
172 - an empty PERL_ANYEVENT_RESOLV_CONF now selects the default
173 configuration.
174 - env variables can now be specified using AE_xyz instead of
175 PERL_ANYEVENT_xyz, subject to some rules.
176 - lots of new AnyEvent::Debug functionality.
177 - AnyEvent::Strict now checks for modifications of $_ by the callback
178 (by setting it to readonly while executing the callback, if possible).
179 - IO::Async sometimes cancels the wrong timer when you call cancel_timer.
180 Invest extra overhead to work around this peculiar behaviour.
181 - split the pure perl backend into a loop module and an impl
182 module (AnyEvent::Loop and AnyEvent::Impl::Perl).
183 - some read types (regex, netstring, json, storable) could
184 errornously keep an AE::Handle object alive on their own
185 (reported by Mohammad Toossi).
186 - AnyEvent::Socket::resolve_sockaddr and all functions using it now
187 supports /etc/hosts overrides.
188 - add more workarounds around child watcher bugs in IO::Async
189 and (newly introduced) in POE.
190 - use glib child watchers and try to work around its limitations,
191 if possible (based on a patch by Kevin Ryde).
192 - add support for FLTK (fltk2), via AnyEvent::Impl::FLTK2.
193 - many read types in AnyEvent::Handle would malfunction during
194 transitions to TLS because of a "delete $self->{rbuf}".
195 (reported by Gerald Galster).
196 - improve condvar blocking wait performance (probably only noticable
197 with faster backends), also make it easier for event loops that
198 do not support blocking waits.
199 - PERL_ANYEVENT_MODEL now also allows full module names, although
200 the usefulness of this feature is not yet clear.
201 - add tests for all supported event loops, conditional on PERL_ANYEVENT_LOOP_TESTS.
202 - add t/80_ssltext.t.
203 - no longer use AUTOLOAD for the initial loading (for no very good reason).
204 - tcp_server now tries to unlink unix domain sockets when it gets destroyed.
205 - do not run event loop detection multiple times when programs call their cached
206 copies of AnyEvent::detect.
207 - parse_hostport now accepts absolute paths als unix domain sockets.
208 - suppress the idiotic warning inside POE by patching the POE::Kernel
209 run flags instead of asking the user to jump through contortions.
210 - add Task::Weaken as dependency (as recommends only at this time).
211 - replace PF_UNSPEC by 0, for good old minix.
212
2136.0 (not officially released, but was on CPAN for about 15 minutes :/)
214
2155.34 Thu May 12 09:59:41 CEST 2011
216 - AE::TLS didn't support both cert_file and key at the same time,
217 due to some copy&paste bug (reported and analyzed by vti).
218 - improve AE::Handle timeout documentation slightly.
219
2205.33 Mon Feb 7 21:51:41 CET 2011
221 - never appeared on CPAN, for unknown reasons.
222 - AnyEvent::Handle on_connect_error - do not try to destroy a handle
223 object when it's already gone (analyzed by Ryan Bullock).
224 - update idna algorithm(s) to UTS#46 version 6.0.0.
225 - preserve trailing dot in idn_to_ascii - testsuite says to preserve,
226 uts46 says it's an error, and testsuite makes more sense.
227 - remove leading dots in idn_to_ascii - testsuite says to remove,
228 uts46 says it's an error, and testsuite makes more sense.
229 - fixed a bug in idn ignored character processing.
230 - passes the uts46 6.0.0 testsuite except for three cases which
231 apparently are bugs in the testsuite itself.
232 - passes the uts46 6.0.1 testsuite except for eleven cases which
233 apparently are bugs in the testsuite itself.
234
2355.31 Sun Jan 23 11:49:19 CET 2011
236 - work around a POE bug where POE would first poll for new events and
237 invoke handlers BEFORE handling the message queue.
238 - fix parsing of "domain" statement in resolv.conf (Johannes Plunien).
239 - stop_read now does just that in TLS mode. see its
240 documentation for caveats.
241 - update to common::sense 3.4 warning set (some warnings in perl impose
242 an arbitrary 32kb limit on regex matches, which affects
243 AnyEvent::Handle).
244 - AE::Handle had a broken rbuf_max method and missing wbuf_max method.
245 - new experimental AE::Handle::resettls method.
246
2475.3 Fri Dec 31 05:48:13 CET 2010
248 - major IOAsync patch by Paul Evans - 0.33 is required, as the major
249 issues with IO::Async have been solved in that version, including
250 the availabilty of a default loop!
251 - implement new wbuf_max parameter for AnyEvent::Handle.
252 - added Cocoa::EventLoop backend by Daisuke Murase.
253 - the default testsuite now honors PERL_ANYEVENT_MODEL (Paul Evans).
254 - clarify handle on_prepare documentation.
255 - set CLOEXEC flag on parent-side fd's in AnyEvent::Util::run_cmd,
256 to avoid leaking them to any child processes.
257 - try to force files into the arch-dependend lib directory, for the
258 half-"turly obsessive compulsive person"(s) that depend on it.
259
2605.29 Sun Dec 5 10:49:21 CET 2010
261 - convert EV backend to EV 4.00 API (so better upgrade EV too).
262 - AE::Handle: implement dynamic read_size adjustment, add
263 max_read_size parameter, reduce default read_size to 2048.
264 - add some FAQ entries for servers and/or on_eof vs. on_error.
265 - work around OS bugs (cygwin again) when lingering in AE::Handle.
266 - reduce memory usage slightly when connect was used in AE::Handle.
267 - wrok around more segfaults in Tk.
268 - document tls_ctx => undef as valid.
269 - detect WNOHANG value at compiletime.
270 - include some probably linux-only support for building AnyEvent
271 as part of the perl core.
272 - improve compatibility of t/08_idna.t to perl 5.8.
273 - make AnyEvent::Strict truly optional (in case somebody builds a
274 minimal perl...).
275
2765.28 Wed Oct 13 04:14:23 CEST 2010
277 - due to a glitch, AnyEvent's internal getprotobyname wasn't
278 actually used in tcp_server.
279 - implement AnyEvent::CondVar->new method, which was documented but
280 missing.
281 - support autoloading of read types in unshift_read.
282 - AnyEvent::Strict: do the test for invalid fh arguments differently,
283 resulting in more informative error messages.
284 - new AnyEvent::FAQ manpage.
285 - clarify that you can actually call push_read etc. in on_read.
286 - put keys and certs in /tmp, not into the current directory,
287 and document this in the SECURITY CONSIDERATIONS sections
288 of AnyEvent::TLS.
289 - support ";" as resolv.conf comment character in AnyEvent::DNS, as
290 requested by Ron Isaacson.
291 - document $AnyEvent::DNS::RESOLVER variable.
292 - incorporated major typo patches by Abhijit Menon-Sen.
293 - AnyEvent::Handle now croaks when negative timeouts are passed.
294 - add a shitload of TCP_xxx sockopt name constants to AnyEvent::Util,
295 undocumented, but who knows when they might come in handy.
296
2975.271 Tue Jun 8 12:05:46 CEST 2010
298 - backport to perl 5.8.x.
299
3005.27 Sun Jun 6 12:12:05 CEST 2010
301 - postpone differently in AnyEvent::Socket now, as
302 when not, canceling the connection attempt might fail
303 (found by Felix Antonius Wilhelm Ostmann).
304 - explicitly check for non-stream sockets in AE::Handle, too many
305 clueless people fell into the trap of this somehow working.
306 - simplified and reworked the "OTHER MODULES" section.
307 - better/more condvar examples.
308
3095.261 Wed Apr 28 16:13:36 CEST 2010
310 - AF_INET6 was not properly used from Socket6 during configuration
311 time (found by Andy Grundman).
312
3135.26 Mon Apr 12 04:49:35 CEST 2010
314 - don't generate spurious readyness notifications when select
315 returns EINTR in the pure perl backend (analysed and testcase
316 by Bas Denissen).
317 - give same examples in SYNOPSIS sections of AnyEvent and AE.
318 - provider faster implementations for AE::time/now/now_update for
319 EV and Perl backends.
320 - provide faster AE::cv and AE::time implementations for all
321 backends.
322 - no longer support lower/mixed-case signal names (this was
323 never documented nor universally supported).
324 - some more memory µ-optimisations, and somewhat less messy
325 code to cope with both AnyEvent and AE APIs.
326
3275.251 Sat Mar 13 00:58:21 CET 2010
328 - make test failed if it was run before make install on perl <=
329 5.8.8, and older versions of perl make our live very difficult,
330 so write constants.pl during Makefile.PL time.
331
3325.25 Sat Mar 13 00:23:14 CET 2010
333 - fix a race condition in AnyEvent::Handle that would cause
334 a "bio_read: unsupported method" error (in _another_
335 ssl connection) after user code threw an exception.
336 - added AnyEvent::Handle->destroyed method.
337 - speed up AnyEvent::Socket::format_address and ::format_ipv6.
338 - the AnyEvent::Util::fh_nonblocking function worked only
339 by lucky accident on win32.
340 - smaller and faster AnyEvent::Util::fh_nonblocking.
341 - when the (required!) Time::HiRes module is missing, AnyEvent
342 did not fall back to built-in time properly.
343 - do not load Fcntl at runtime, saving memory and loading time.
344 - precompile a number of constants and use them instead of runtime
345 detection and eval.
346 - free detection code after detection and similar memory optimisations.
347 - Perl backend timer interval best effort drift has been improved
348 (same algorithm as EV).
349 - update unicode idna mapping table.
350
3515.24 Tue Jan 5 11:39:43 CET 2010
352 - cygwin never reports errors from failed connects. 1.5 just gives
353 you continous readyness and EAGAIN, 1.7 is even more broken and
354 just hangs. work around both issues in a cygwin-specific hack.
355 - improve idle watcher documentation slightly.
356
3575.23 Sun Dec 20 23:48:00 CET 2009
358 - support IDNs in resolve_sockaddr, and therefore in tcp_connect.
359 - implement punycode_encode/decode, idn_nameprep,
360 idn_to_ascii and idn_to_unicode operations in AnyEvent::Util.
361 - provide $AE::VERSION.
362 - removed traces of "no strict 'refs'".
363
3645.22 Sat Dec 5 03:51:13 CET 2009
365 - downgrade-or-fail in AnyEvent::Handle::push_write, to
366 diagnose encoding failures earlier and more succinctly.
367 (this works around bugs in perl, throwing away encoding info
368 when passing scalar data to extensions).
369 - add more examples to AnyEvent::Socket manpage.
370 - upgrade internal warning set to the same as common::sense 2.03.
371 - use pack "n/a*" for pre-5.8.9 perl compatibility in AnyEvent::DNS
372 (John Beppu).
373 - AnyEvent::Socket::inet_aton now properly supports ipv6, as documented.
374 - add google public dns servers to fallback server set and make sure
375 we load-balance properly between the three sets. also add all
376 fallback dns servers, not just a random one, to each dns config.
377
3785.21 Thu Nov 19 02:48:47 CET 2009
379 - fix a problem where socket constants were called with parameters
380 (spotted by David Friedland).
381 - fork_call never use'd POSIX (reported by Daisuke Maki).
382 - improve perl 5.6 compatibility further (but it still won't work
383 unless you rip out everything but the core).
384 - prefer Net::DNS::Resolver over ipconfig, if installed, on win32.
385 uses 10MB of RAM, but doesn't open a console window. *sigh*.
386
3875.202 Wed Oct 14 22:35:44 CEST 2009
388 - AnyEvent::DNS would unexpectedly clobber $_ under windows
389 (analysed by Matthias Waldorf).
390 - AnyEvent::Handle::run_cmd can now pass the PID of the
391 newly-created process, which is much less useful than it might
392 sound (based on patch by Yann Kerherve).
393
3945.201 Tue Sep 29 12:09:25 CEST 2009
395 - AnyEvent:Handle::on_starttls/on_stoptls methods were broken
396 (reported by Torsten Foertsch).
397 - common::sense 2.0 could cause tcp_server to throw an exception
398 (analysed by elmex).
399
4005.2 Mon Sep 14 07:04:49 CEST 2009
401 - INCOMPATIBLE CHANGE: do no longer support register_read_type
402 and register_write_type in AnyEvent::Handle, instead support
403 package names (the facility was mostly abused).
404 - implement "packagename-as-read/write type" support in
405 AnyEvent::Handle.
406 - AnyEvent::Handle: new options "keepalive" and "oobinline".
407 - oobinline set by default to avoid security issues.
408 - the pure-perl event loop backend wrongly detected times() fallback
409 support (spotted by Pavel Boldin).
410
4115.12 Tue Sep 1 20:26:50 CEST 2009
412 - be more lenient when parsing resolv.conf files, as some people
413 use hashmarks as comment indicator inside directives (reported
414 by Michael S. Fischer).
415 - use same set of warnings as common::sense 2.0.
416 - fix a potential 32 bit overflow issue due to perl having problems
417 with large hex constants in the Perl backend.
418 - do not use the slower 5.6 version fo the enc_name code in AnyEvent::DNS
419 by default.
420 - fix some prototypes.
421 - (optionally) require Async::Interrupt 1.02.
422 - replace opendns fallbacks by verizon/level3 public dns servers.
423
4245.112 Fri Aug 21 13:59:04 CEST 2009
425 - AnyEvent::Handle could cause hard crashes in Net::SSLeay when
426 using starttls with an illegal/uncreatable tls context.
427 - port to broken win32 perls.
428
4295.111 Sun Aug 16 18:44:36 CEST 2009
430 - perl errornously requires sockaddr_un structures returned by the
431 kernel to have a specific length. work around this bug by padding
432 sockaddr_un structures (found to fail on OS X, as perl only uses
433 the correct code for GNU/Linux, but likely to fail on other
434 OSes as well).
435
4365.11 Wed Aug 12 17:49:37 CEST 2009
437 - fixed a bug in run_cmd with input from filename.
438 - create a json coder object when none is given in write direction
439 to reduce dependency on JSON.
440
4415.1 Tue Aug 11 03:17:32 CEST 2009
442 - tcp_connect now reports the correct error in $!, not an unrelated
443 one (was broken since 4.91).
444 - AE::Handle did not properly clear rtimeout/wtimeout watchers
445 when retrying a tcp_connect.
446 - new functions: AnyEvent::Util::run_cmd and close_all_fds_except.
447 - fixed wrong documentation in AE::Socket.
448
4495.01 Mon Aug 10 03:16:32 CEST 2009
450 - last release broke EV child watchers :/.
451
4525.0 Sun Aug 9 17:32:09 CEST 2009
453 - officially introduce and document the AE API.
454 - lowering the timeout at runtime did not have immediate effect
455 in AnyEvent::Handle.
456 - AE::Handle now has separate and independent read and write timeout
457 settings.
458 - probe for Guard module when AnyEvent::Util is loaded, not at
459 runtime (Event doesn't like eval "use" from callbacks, and it avoids
460 exporting the probe function).
461 - Impl::IOAsync's io watchers did not use the filehandle emulation
462 code to work around it's watcher limits.
463 - work around IO::Async's problems with overloaded objects.
464 - try to disable wearnings inside IO::Async, it's just too buggy.
465 - do no longer use Event by default, only EV or Perl, Event is too buggy.
466 - hackishly provide fast AE:: interface when EV or Perl is the backend.
467 - inherit VERSION from AnyEvent::VERSION in most modules.
468
4694.91 Thu Aug 6 15:42:45 CEST 2009
470 - AE::Handle::starttls could get out of sync when the read buffer
471 already contains some TLS handshake.
472 - AE::Handle did not properly free the TLS session, causing bigger
473 memory leaks in Net::SSLeay than expected (reported by toaster).
474 - AE::Socket::tcp_connect will now artificially delay invoking
475 the callback to avoid returning after invoking the callback.
476 - convert many internal watcher uses to AE API.
477
4784.9 Sat Aug 1 11:07:01 CEST 2009
479 - Glib silently fails when registering a timeout with
480 a negative value, so avoid that.
481 - call condvar callback immediately when it is set
482 after the condition is already signalled.
483 - check rbuf_max condition only after trying to consume
484 data.
485 - normalise signal numbers to names when using the ae's
486 signal handling, but do not document this (yet).
487 - pure perl signal emulation did not properly set nonblocking
488 mode on the signal pipe on win32 (this is unlikely to have caused
489 any issues).
490 - new module: AnyEvent::Debug.
491 - AnyEvent::Strict now checks that a signal was specified
492 by name (not name or number).
493 - reduce memory footprint in typical cases by ~50kb by
494 only compiling signal code when necessary.
495 - add AnyEvent::Handle->rbuf_max.
496 - grab the AE namespace for future new API, implement stubs for
497 said future/alternative API.
498 - new function: AnyEvent::Socket::format_hostport.
499
5004.881 Tue Jul 28 12:51:53 CEST 2009
501 - work around a bug in local in pre-5.10 perls, causing
502 AnyEvent::Handle to recurse when it should not
503 (analysed by elmex).
504
5054.88 Tue Jul 28 04:04:37 CEST 2009
506 - re-bless the handle into a dummy package after calling
507 AnyEvent::Handle::destroy, so the user does not need to check
508 for errors after every push_write etc.
509 - do not attempt to run t/02_signals.t on obviously broken
510 platforms.
511
5124.87 Sun Jul 26 02:06:16 CEST 2009
513 - do not attempt to linger when there is no longer a valid fh.
514 - enforce tls mode to be either accept or connect, do not simply
515 segfault in Net::SSLeay.
516 - AnyEvent::Handle can now call tcp_connect itself (new parameters
517 connect, on_prepare, on_connect and on_connect_error). Updated
518 tutorial accordingly.
519 - add AnyEvent::Impl::Irssi backend.
520
5214.86 Mon Jul 20 23:52:29 CEST 2009
522 - since the verbose warning is not enough, explicitly document
523 that versions before 1.33 of Net::SSLeay are not secure.
524 - work around signal handling races in Event and (...) Event::Lib.
525 - try to align signal-race timer to full-second boundaries.
526 - work around Tk not liking negative timeouts.
527 - don't complain of different grades of environmental unfriendlyness
528 in IO::Async.
529
5304.85 Sat Jul 18 06:16:14 CEST 2009
531 - nail the signal race problem in perl once and for all
532 (see $AnyEvent::MAX_SIGNAL_LATENCY).
533 - take advantage of Async::Interrupt if it is available.
534 - load Time::HiRes and Guard modules on demand only.
535 - add optional/recommended modules section to AnyEvent
536 documentation.
537 - reduce memory usage considerably (and reduce startup penalty)
538 by not using "strict", "warnings" and "overload" modules.
539 - work around buggy windows/openbsd perls and provide EBADMSG
540 and EPROTO ourselves when missing.
541 - improve perl 5.6 compatibility of the core event loop.
542 - made Net::SSLeay version 1.33 a soft requirement.
543
5444.83 Fri Jul 17 16:56:26 CEST 2009
545 - implement AnyEvent::Socket::getprotobyname.
546 - AnyEvent::CondVar's will now detect recursive blocking
547 waits and will croak, as too many people fall into
548 this trap.
549 - AnyEvent::Handle will now call ->destroy on itself after
550 executing the on_error callback, instead of doing some
551 half-baked internal shutdown, for fatal errors.
552 - clarify on_eof behaviour w.r.t. the read queue and
553 on_read callbacks.
554 - ignore some possible spurious wake-ups in tcp_connect.
555
5564.82 Sat Jul 11 00:34:55 CEST 2009
557 - POE and Event backends didn't accept some callable objects as
558 callbacks.
559 - use Config module instead of POSIX module to detect signal names
560 in AnyEvent::Strict and AnyEvent::Impl::EventLib, as the POSIX
561 module doesn't even have all POSIX signals :/.
562 - use more workarounds around the many refcnt/corruption bugs in
563 Event::Lib.
564 - work around a race condition in perl's select, causing t/03_child.t
565 to rarely fail.
566
5674.81 Thu Jul 9 10:30:30 CEST 2009
568 - AnyEvent::Handle didn't properly diagnose write errors
569 (it expected -1 from syswrite, how lame... :).
570 - support file descriptors in addition to file handles
571 in AnyEvent->io.
572 - new env variables: PERL_ANYEVENT_RESOLV_CONF,
573 PERL_ANYEVENT_MAX_OUTSTANDING_DNS, PERL_ANYEVENT_CA_FILE
574 and PERL_ANYEVENT_CA_PATH.
575 - provide a sensible synopsis section for AnyEvent::TLS.
576 - add a "supported backends" section to the manpage.
577 - added simple io watcher test to testsuite, using a
578 portable_socketpair.
579 - tried to improve the stability of the Event::Lib backend,
580 YMMV.
581
5824.8 Mon Jul 6 23:45:16 CEST 2009
583 - AnyEvent::DNS did not properly follow CNAME records with
584 uppercase targets.
585 - AnyEvent::DNS would errornously return AAAA records
586 with v4 mapped addresses (a faulty record) as ipv4 addresses,
587 causing AnyEvent::Socket to throw an exception.
588 - added new module AnyEvent::TLS for easier SSL/TLS context
589 creation, with many options including hostname verification,
590 secure default configuration, lots of documentation and,
591 predefined diffie-hellman keys for perfect forward security
592 and much more. get it while it's still fresh!
593 - use AnyEvent::TLS in AnyEvent::Handle for context management.
594 - load AnyEvent::Handle only on demand in AnyEvent::DNS,
595 so AnyEvent::Socket users have smaller memory footprint
596 in the common case.
597 - add AnyEvent::Handle->push_shutdown method.
598 - add an additional error message parameter to AnyEvent::Handle's
599 on_error callback (for TLS, $! is still available).
600 - add AnyEvent::Handle on_starttls/on_stoptls callbacks.
601 - make AnyEvent::Handle more robust against early conenction
602 failures (during new), and return C<undef> in that case
603 from the constructor.
604 - AnyEvent::Handle will now try to load only JSON::XS first,
605 then fall back to JSON.
606 - format_ipv4/format_ipv6 are now exported by default, for symmetry,
607 and because it was documented that way.
608
6094.451 Fri Jul 3 00:28:58 CEST 2009
610 - do not clear rbuf when shutting down an AnyEvent::Handle
611 object - doing so breaks AnyEvent::HTTP.
612
6134.45 Mon Jun 29 22:59:26 CEST 2009
614 - a write error could cause AnyEvent::Handle to create
615 an I/O watcher with an undefined $fh.
616 - special-case mapped ipv4 addresses in both
617 AnyEvent::Socket::format_address and parse_address,
618 to treat them just like ipv4 addresses.
619 - updated and overhauled the AnyEvent::Intro doc.
620 - implement AnyEvent::Socket::format_ipv4/ipv6.
621 - slightly speed up portable_pipe/socketpair.
622 - expand condvar begin/end documentation.
623
6244.42 Fri Jun 26 08:32:18 CEST 2009
625 - preliminary and neccesarily incomplete support for IO::Async.
626 - reset SIGCHLD to DEFAULT when AnyEvent is loaded, in
627 case it was set to IGNORE, to ensure we can catch child statuses
628 even when the calling env acts stupidly.
629 - updated benchmarks with IO::Async, which performs very well.
630
6314.412 Wed Jun 24 01:35:57 CEST 2009
632 - support an "untaint" attribute for AnyEvent::DNS
633 and set it on the default resolver.
634 - implement and document AnyEvent::Impl::Perl::loop.
635 - remove all anyevent-env variables from %ENV
636 when running in tainted mode.
637 - mention and extend the IO::Lambda benchmark.
638
6394.411 Sun Jun 7 18:48:02 CEST 2009
640 - do not try to use F_SETFD on windows.
641
6424.41 Thu May 14 06:40:11 CEST 2009
643 - work around issues in older perls (5.8.0?) when
644 a signal handler is deleted from the %SIG hash.
645 - use POSIX::_exit in child test, to avoid running
646 destructors.
647 - speed up CHLD handling by relying on SIGCHLD being
648 synchronously delivered, even when we roll our own
649 implementation.
650 - AnyEvent::DNS: add the "dname" resource record name
651 for cosmetic reasons.
652
6534.4 Sun Apr 26 20:12:33 CEST 2009
654 - implemented idle watchers, where applicable.
655 - AnyEvent->time died when Event backend was in use.
656 - fix a memleak in the Tk backend.
657 - sped up Tk timer handling.
658 - clip negative "after" values to 0 in AnyEvent::Impl::Event
659 to avoid spamming out warning messages.
660 - fix Qt timers without interval.
661 - avoid Qt zero-timeout special case, in old libqt's.
662
6634.352 Mon Apr 20 16:31:11 CEST 2009
664 - fix AnyEvent::Strict error messages for child watchers.
665 - fix/update Makefile.PL "recommends" versions.
666 - add AnyEvent->now_update.
667
6684.351 Sat Apr 11 07:56:14 CEST 2009
669 - actually make the signal pipe work under win32.
670 - localise $! in signal handler to avoid changing $!, although
671 perl itself does not seem to save/restore errno either.
672 - set the cloexec flag on the signal pipe (normally set by perl too).
673
6744.35 Fri Mar 27 11:48:20 CET 2009
675 - event models relying on AnyEvent's signal watcher emulation did
676 invoke the callback asynchronously, contradicting documentation
677 and causing signals to get lost (this includes AnyEvent's own
678 event loop). AnyEvent now uses the standard pipe trick to make
679 callback execution synchronous to the event loop.
680 - AnyEvent::Handle didn't free TLS context data on DESTROY
681 (patch by Pavel Shaydo).
682 - work around the ever-incompatibly-changing API of MakeMaker.
683 - document that changing global variables without restoring them
684 is a bad idea in Perl (noted by Adam Rosenstein).
685 - AnyEvent::Strict now barfs if ->io is passed a file.
686
6874.34 Thu Feb 12 18:32:45 CET 2009
688 - separately buffer TLS read data, as otherwise the read queue could
689 deadlock as receiving data is not expected while draining the
690 read queue (which cna only happen with TLS).
691 - raise EBADMSG error on JSON decoding errors.
692 - fix some minor manpage bugs (reported by Maximilian Gaß).
693 - speed up select bitmask parsing quite a bit in the pure perl
694 backend.
695 - use CORE::select instead of just select in AnyEvent::Impl::Perl.
696
6974.331 Tue Jan 6 21:07:25 CET 2009
698 - socketpair fails on many vista machines because vista has
699 completely broken accept/getpeername and getsockname functions,
700 so we provide our own socketpair emulation that kind of works
701 (AnyEvent::Util::portable_pipe).
702 - new function: AnyEvent::Util::portable_socketpair.
703 - take advantage of the Guard module if it exists.
704
7054.33 Fri Nov 21 02:35:40 CET 2008
706 - AnyEvent::Strict did errornously flag a fileno of 0 as
707 illegal.
708 - reduce memory usage and slightly speed up the pure perl backend
709 by only storing the file descriptor, not the file handle.
710 - add missing autocork method to AnyEvent::Handle (reported
711 by Adam Rosenstein).
712 - AnyEvent::DNS->resolve errornously documented an $rcode result
713 argument, but there isn't (spotted by Henrik Krohns).
714 - the naked truth about Tk - it's basically unmaintained.
715
7164.32 Mon Nov 3 22:46:32 CET 2008
717 - fix AnyEvent::Socket::resolve_sockaddr to properly support
718 unix sockets again and choose a proper default for
719 the socket type.
720 - fix call to resolve_sockaddr in tcp_connect.
721
7224.31 Thu Oct 30 04:41:48 CET 2008
723 - implemented AnyEvent::Handle->destroy method.
724 - hint about unexpected effects in TLS mode.
725 - speed up AnyEvent::Impl::Perl by using more arrays
726 instead of hashes without the slightest loss of readability :)
727 - work around a perl argument refcounting bug.
728 - with some perl patches applied, AnyEvent now seems leak-free.
729
7304.3 Fri Oct 3 09:18:43 CEST 2008
731 - AnyEvent will now install a no-op signal handler for SIGPIPE,
732 unless one has been installed already.
733 - warn about not loaidng AnyEvent::Impl::POE early enough
734 (patch by Adam Rosenstein).
735 - fixed a great number of bugs and corner cases in AnyEvent::Handle.
736 - fix a bug in where in SSL connect mode, the
737 client would first wait for some data by the server and
738 otherwise hang. (reported and analysed in an absolutely
739 exemplary manner by Adam Rosenstein).
740 - fix a bug in where SSL EOF would not be treated
741 as stream EOF, putting the connection into a hung state.
742 (reported and analysed in an absolutely exemplary manner by Adam
743 Rosenstein).
744 - fix a potential segfault when the TLS context
745 would go missing in a read callback (e.g. due to stoptls) -
746 Net::SSLeay of course makes no type checking whatsoever.
747 - AnyEvent::Handle will not stop the read watcher in TLS mode.
748 - AnyEvent::Handle->stoptls will send a TLS close notify
749 instead of simply destroying the stream now.
750 - fix a bug where large blocks of data written in TLS mode
751 would not be sent unless triggered by receives.
752 - on_drain will now take the tls write buffer into account.
753 - SSL operations have been streamlined a bit and should be faster
754 now, more to come, though.
755 - added a FAQ section, will document nontrivial issues in there.
756
7574.234 Mon Sep 29 04:08:13 CEST 2008
758 - fix child watcher documentation: only child exits will be reported,
759 no trace events.
760 - mention SIGPIPE in AnyEvent::Handle.
761 - perl backend will now use POSIX::times as monotonic clock
762 when available and the monotonic clock option is not.
763 - run even without Time::HiRes in some cases now, but do not
764 provide sub-second accuracy (all *supported* perl versions have
765 Time::HiRes).
766
7674.233 Fri Aug 22 01:48:05 CEST 2008
768 - fix a bug in the testsuite, causing a hang.
769 - clarified lots of AnyEvent::Handle constructor arguments.
770
7714.232 Thu Aug 21 20:44:25 CEST 2008
772 - fix a bug in AnyEvent::Handle that could cause two fatal error
773 callbacks on EOF, which caused AnyEvent::HTTP to signal success twice
774 (testcase by Leon Brocard).
775 - clarified on_eof/on_error documentation a bit.
776
7774.231 Tue Jul 29 13:12:15 CEST 2008
778 - remove some debugging code left in AnyEvent::Util::fork_call
779 (and no, it's impossible to implement with the broken windows
780 perls without resource leaks or worse).
781
7824.23 Tue Jul 29 12:19:59 CEST 2008
783 - document the first parameter passed to condvar callbacks to be
784 the callback.
785 - add AnyEvent::Socket::{ntoa,aton} aliases.
786 - optimize the AE::Handle->push_read (line) for the default
787 eol marker.
788 - optimize push_read (packstring|storable) for small packets.
789 - invoke on_error callback when no on_eof callback is set.
790 - fix a bug in push_read (storable) of unknown impact.
791
7924.22 Sun Jul 20 16:34:13 CEST 2008
793 - new function AnyEvent::Socket::parse_hostport.
794 - as the bulkheads at microsoft can't even get getprotobyname reliably
795 working on their shitty fucking broken os we need to hardcode
796 some common protocol numbers in AnyEvent::Socket. How can
797 people even bother with such a pile of shit as windows.
798
7994.21 Thu Jul 17 14:40:05 CEST 2008
800 - INCOMPATIBLE CHANGE: AnyEvent::Util::fork_call now has a prototype
801 which is incompatible to previous invocation syntax.
802 - work around more windows perl bugs in fork_call. I cannot imagine
803 why anybody would bother implementing fork in such an obviously
804 unusable way.
805 - avoid calling $do_retry if the request has already finished in
806 AnyEvent::DNS, thus avoiding an exception (reported by Anatoly
807 K. Sharifulin).
808
8094.2 Sat Jul 12 22:42:11 CEST 2008
810 - support an "interval" argument for repeating timers.
811 - fix a bug in the pure perl event loop that caused timers to
812 be delayed under some circumstances.
813 - implement optional argument checking via PERL_ANYEVENT_STRICT.
814 - not importing anything from Event causes it to malfunction
815 and not use Time::HiRes, work around this bug.
816 - more bug workarounds against the endlessly broken Tk module,
817 seems to work now on 32 bit machines, but less so on 64 bit.
818 - minor optimisations applied to most backends.
819
8204.161 Fri Jul 4 14:24:48 CEST 2008
821 - fixed the t/handle/02_write.t, it read after write, with the
822 assumption that all data could be written before read
823 (reported by Jim Razmus).
824
8254.160 Thu Jul 3 04:02:21 CEST 2008
826 - re-registering signal watchers after unregistering one did
827 not work if a backend falls back on AnyEvent's default
828 implementation (also affected child watchers).
829 - new AnyEvent::Handle options: autocork and no_delay.
830
8314.152 Sun Jun 22 14:15:44 CEST 2008
832 - allow for 32-bit perls that implement shifts differently
833 on different architectures in parse_ipv4 (reported and
834 analysed by Keiichi DAIBA).
835
8364.151 Fri Jun 6 17:34:24 CEST 2008
837 - make sure specifying _only_ on_read and never pushing reads
838 works.
839
8404.15 Fri Jun 6 13:00:46 CEST 2008
841 - the pure perl backend would keep some watchers alive when more than
842 one watcher was registered for the same fd.
843 - new "packstring" and "storable" read and write types
844 for AnyEvent::Handle.
845 - allow on_eof handler to be called after on_error with EPIPE returns.
846 - do not immediately call on_read callback in handle constructor.
847
8484.14 Thu Jun 5 20:29:31 CEST 2008
849 - Fixed a bug in DNS SRV priority sorting.
850 - AnyEvent::Util::guard now reports runtime errors while
851 executing the guard block as warnings.
852 - handle 0-byte-reads just before EOF correctly in AnyEvent::Handle.
853
8544.13 Thu Jun 5 00:47:59 CEST 2008
855 - AnyEvent::DNS only followed cname chains with length 2,
856 contrary to documentation. bumped it up to 10, thanks to
857 microsoft, the current broken-dns-config-king.
858 - AnyEvent::DNS didn't check for socket return status
859 but instead relied on perl not creating filehandles in that
860 case - too bad it gives you a bogus file handle (reported
861 and analysed by Vladimir Timofeev).
862 - fix queue management logic in AnyEvent::Handle:
863 when on_read was registered, the queue was empty
864 and no progress could be made AnyEvent::Handle would
865 enter an endless loop.
866 - correctly start reading again when the handle became
867 busy again after idling.
868 - correctly treat tls shut-downs as EOF condition
869 (for the time being, would be nice to have a callback for that).
870 - correctly call eof callback under all conditions
871 (hopefully).
872 - the timeout callback did not expect that $self can go
873 away any time.
874
8754.12 Tue Jun 3 10:58:04 CEST 2008
876 - include AnyEvent::Intro, a tutorial for anyevent,
877 anyevent::socket and anyevent::handle.
878 - allow more options in on_error.
879
8804.11 Fri May 30 23:42:25 CEST 2008
881 - INCOMPATIBLE CHANGE: replace ptr by real PTR lookup, provide
882 reverse_lookup and reverse_verify to replace it, support
883 v4mapped and v4compat addresses.
884 - provide more documentation for the resolver class.
885 - really replace longest run of :0: by :: in format_address,
886 also properly convert :: and ::1 again.
887 - support NAPTR record name and decode it.
888 - implement random weight sampling for SRV records, as per
889 rfc 2782.
890 - correctly abort on srv-record targets of ".".
891 - added AnyEvent::DNS::wait_for_slot.
892 - in the unlikely event of a virtual circuit connection
893 being invalidated by a delayed udp reply, AnyEvent::DNS
894 could die. this has been fixed.
895 - plug a probable memleak in the DNS vc code.
896 - use configured timeout also for VC request phase.
897 - implement timeout and max_outstanding methods
898 for resolver class.
899 - update version numbers in all modules.
900
9014.1 Thu May 29 05:45:40 CEST 2008
902 - INCOMPATIBLE CHANGE: renamed xxx_ip functions to xxx_address
903 which mirrors their purpose better (old names still available
904 for a while). Also moved AnyEvent::DNS::addr to
905 AnyEvent::Socket::resolve_sockaddr.
906 - implement AnyEvent->time and AnyEvent->now.
907 - fix IPv6 support in pack_sockaddr.
908 - officially un-experimentalise all newly introduced code.
909 - support unix domain sockets everywhere by specifying
910 a host of "unix/" and the pathname as service.
911 - implement an activity timeout in AnyEvent::Handle.
912 - added a regex read type to AnyEvent::Handle.
913 - added a json read/write type to AnyEvent::Handle.
914 - always croak in on_error in AnyEvent::Handle.
915 - document how to subclass AnyEvent::Handle.
916 - implement AnyEvent::Util::fork_call.
917 - add support for IPv6 nameservers and nameserver statements.
918 - work around _yet_ _another_ windows perl bug in where empty
919 select masks cause errors in the pure perl backend (this
920 workaround cannot be implemented for the other event loops,
921 but EV also works around this).
922 - supply AnyEvent::Socket::address_type and make good use of it.
923 - clarify and fix the Handle documentation w.r.t $self vs. $handle
924 vs. callback arguments.
925 - add some recommends to the META.yml.
926
9274.05 Mon May 26 19:44:06 CEST 2008
928 - some platforms ignore the rfc and prepend an extra sa_len member to
929 the sockaddr structure, cater for those. also use sockaddr_family,
930 didn't know it was there, but comes in handy for the workaround.
931 - undo the import WIN32 hack.
932
9334.04 Mon May 26 08:03:31 CEST 2008
934 - try to work around yet another windows bug: failed connects are reported
935 as if it were out-of-band data. windows users: you suck. Only
936 EV, Glib, Event and the pure perl backend can handle this condition.
937 - optimize write algorithm in AnyEvent::Handle.
938 - properly parse PERL_ANYEVENT_PROTOCOLS and use the info correctly.
939 - tcp_connect now properly iterates through all targets.
940 - check wether IPv6 sockets can actually be created, otherwise, disable ipv6
941 support entirely. also hardcode some AF_INET6 constants for the
942 sake of perl 5.8.
943 - vastly improved nameserver/suffix detection algorithm for windows
944 (still a hack...)
945 - try to use a fallback dns server if no dns servers could be found.
946 - splatter around some local $SIG{__DIE__}'s to work around bad
947 die handlers.
948
9494.03 Sun May 25 05:05:57 CEST 2008
950 - work around perl bugs on windows, where perl returns undocumented
951 error codes for sysread, syswrite etc.
952 - fix AnyEvent::Handle::unshift_read to really unshift and not push.
953 - AnyEvent::Handle could sometimes stumble over 0-byte writes.
954 - fix unpack_sockaddr, now works for AF_INET6 sockaddr's.
955 - updated documentation.
956 - lots of workarounds for perl 5.8 quirks.
957 - implement a more flexible type system for both reads and writes
958 in AnyEvent::Handle.
959 - support IPv6 in tcp_server.
960 - pass local hostname and port to the prepare callback in tcp_server.
961 - make types extendable in AnyEvent::Handle.
962 - croak, not die, in AnyEvent::Handle;
963
9644.0 Sat May 24 19:58:08 CEST 2008
965 - added (experimental) AnyEvent::DNS module.
966 - added (experimental) AnyEvent::Socket module.
967 - added (experiemntal) support for TLS/SSL in AnyEvent::Handle.
968 - make condvar->end call ->send as documented.
969 - make condvar constructor support the documented cb => argument.
970 - overload condvars so you can use them instead of code references.
971 - support more of the POSIX forms instead of a pure dotted quad for
972 verbatim ip addresses.
973 - AnyEvent::Socket::inet_aton will use AnyEvent::DNS now and not fork anymore.
974 - try very hard not to freeze on broken windows perls.
975 - make on_eof optional in AnyEvent::Handle.
976 - support service names in tcp_connect and tcp_server.
977 - work around netbsd/openbsd bug causing unix sockets not to
978 become writable until empty.
979 - only make a dummy read for error detection when getpeername
980 returns ENOTCONN.
981 - defend AnyEvent::Util::inet_aton against Coro::LWP's brutal
982 override.
983 - new AnyEvent::Guard::cancel.
984
9853.5 Sat May 17 23:17:14 CEST 2008
986 - gracefully deal with systems not defining CLOCK_MONOTONIC
987 (as opposed to not supporting the clock itself, which worked in 3.41).
988 - added AnyEvent::Util::tcp_connect and AnyEvent::Util::tcp_server
989 helpers, to replace the ill-designed AnyEvent::Socket class.
990 - added some examples (eg/connect, eg/handle).
991 - added AnyEvent::Util::guard.
992 - AnyEvent::Util now exports most of its functions.
993 - fixed a bug in *_read_chunk, where the callback was called with $self as
994 first and second argument.
995 - simplified and fixed regex handling in AnyEvent::Handle::*_read_line.
996 - fix manpage condvars examples to use new syntax.
997 - considerably improve inet_aton with EV::ADNS backend to support
998 cname chains (common dns configuration bug, e.g. with www.google.de)
999 and aaaa records if no a record could be found.
1000
10013.41 Sun May 11 19:53:13 CEST 2008
1002 - work around a bug in perl 5.8.8's local. Reported by Yi Ma Mao.
1003
10043.4 Sun May 11 00:29:25 CEST 2008
1005 - complete re-implementation of condvars with a hopefully much more useful
1006 API (backward compatible functions exist but are undocumented).
1007 - AnyEvent::Handle fully rewritten with a hopefully more powerful API
1008 (still subject to change).
4 - take advantage of CLOCK_MONOTONIC, if available, in AnyEvent::Impl::Perl. 1009 - take advantage of CLOCK_MONOTONIC, if available, in AnyEvent::Impl::Perl.
5 - provide AnyEvent::post_detect and @AnyEvent::post_detect, which 1010 - provide AnyEvent::post_detect and @AnyEvent::post_detect, which
6 allows module authors to avoid forcing event loop detection. 1011 allows module authors to avoid forcing event loop detection.
7 (used by Coro::AIO, Coro::BDB, Coro::AnyEvent for example). 1012 (used by Coro::AIO, Coro::BDB, Coro::AnyEvent for example).
8 - remove coro backends: Coro now provides generic support for AnyEvent 1013 - remove coro backends: Coro now provides generic support for AnyEvent

Diff Legend

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