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

Comparing AnyEvent/Changes (file contents):
Revision 1.216 by root, Sun Feb 8 00:20:53 2009 UTC vs.
Revision 1.560 by root, Wed Nov 14 22:51:45 2012 UTC

1Revision history for Perl extension AnyEvent. 1Revision history for Perl extension AnyEvent.
2 2
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: leka on guard undef resolve (sten)
11 - AnyEvent::Socket::inet_aton did not work when dns resolution was
12 used to find the addresses (analyzed and patch by Maxime Soulé).
13 - fix a memory leak in the /etc/hosts lookup code when hosts don't
14 resolve and are not in hosts (analyzed by sten).
15
167.02 Tue Aug 14 04:27:58 CEST 2012
17 - AnyEvent::Util::run_cmd could block indefinitely (analyzed and test
18 program by Yorhel).
19 - verified that AnyEvent::Socket follows RFC5952.
20 - try to parse "ADDR#PORT" in addition to "ADDR PORT".
21
227.01 Sun May 13 01:03:17 CEST 2012
23 - fail with EPROTO in AnyEvent::Handle wqhen TLS is requested but
24 not available, instead of throwing an exception.
25 - use File::Spec to get the tmpdir in t/*, to avoid needless
26 failures on (most, not mine :) windows boxes.
27 - new handle read types: tls_detect and tls_autostart.
28
297.0 Fri Apr 13 06:33:30 CEST 2012
30 - child watchers are broken in POE 1.352 (also many earlier
31 versions) and there seems to be no way to work around it, as POE
32 itself is inherently racy. Document this regression and add a delay
33 in t/68_poe_03_child.t for the time being.
34 - new module AnyEvent::IO, that is a frontend to either a pure-perl
35 synchronous I/O implementation (AnyEvent::IO::Perl), or to the
36 asynchronous I/O extension IO::AIO (AnyEvent::IO::IOAIO).
37 - load /etc/hosts only when DNS has no answer.
38 - stat /etc/hosts on every access and reload it if it changed.
39 - load /etc/hosts via AnyEvent::IO - potentially asynchronous.
40 - fix a buggy croak in the dh parameter reading in AnyEvent::TLS.
41 - AnyEvent::Log log_to_file and log_to_path now use AnyEvent::IO. As
42 a side effect, they now use true appending as opposed to libc
43 appending, and the file might not have been opened when the function
44 returns.
45 - the default logging level was not properly documented in a variety of
46 places, this has hopefully been rectified.
47 - updated uts46data.pl for unicode 6.1.0.
48 - made log messages generated by AnyEvent submodules not
49 include the package name anymore, as it will be logged
50 by default already.
51 - upgrade to the trick used by common-sense 3.5 to work
52 around extra warning torture/breakage under perl 5.15.x.
53 - log messages by anyevent are now ucfirst, are usually full
54 sentences and do no longer include the package name.
55 - the storable read type would sometimes throw an exception instead
56 of causing EBADMSG (analyzed by Maxime Soulé).
57 - work around a bug in openssl 1.0.1 which enforces a minimum rsa
58 keysize (reported by Paul Howarth).
59 - documented Rocco fabricating statements about AnyEvent and me.
60
616.14 Tue Jan 31 20:00:24 CET 2012
62 - AnyEvent::Impl::Tk was broken due to a mysterious "die" inside,
63 probably an editing mistake (reported by Darin McBride).
64
656.13 Thu Jan 12 07:27:01 CET 2012
66 - AnyEvent::Util::fork_call checked for POSIX availability in the wrong way,
67 causing it to fail if POSIX wasn't loaded already (analysed by Rock Power).
68 - AnyEvent::Handle::push_read (line => did pass $1 directly, so
69 regex-matching inside the callback would change the parameter, despite
70 $1 being dynamically scoped per-block. this perl bug is now being
71 worked around (testcase by Cindy Wang).
72
736.12 Mon Dec 12 13:21:10 CET 2011
74 - $! was clobbered by subsequent calls in tcp_connect, due to
75 the postpone not saving/restoring it (analyzed by Richard Garnish).
76
776.11 Tue Nov 22 10:36:05 CET 2011
78 - Tk cannot create windows when tainted unless you set a title, so
79 set a dummy title for AnyEvent's dummy mainwindow. How dummy.
80 - escape any nonprintable/nonascii characters when stringifying
81 backtraces.
82 - log the reason loading an interface module fails to load at level debug.
83
846.1 Tue Oct 4 19:44:30 CEST 2011
85 - INCOMPATIBLE CHANGE: the default log level is now "4" (error
86 and above), and some messages inside AnyEvent have been elevated
87 to higher log levels to print by default.
88 - AnyEvent::log, unlike AnyEvent::Log::log, did not always exit on fatal.
89 - work around yet another signal race bug, newly introduced by POE. My
90 how that module sucks.
91 - add google public ipv6 dns as fallback servers, so AE::DNS can finally
92 support ipv6-only operation even with defaults (also use ::1 instead
93 of 127.0.0.1).
94 - give better guidelines on how to chose a logging level.
95 - AnyEvent::Log can now cap (limit) log message priority by context.
96 - AE::log will now save and restore $! and $@, so callers don't have to.
97 - AE::Strict now checks that registered fds are still valid at regular
98 intervals.
99 - skip frequently called 'require' calls if the module appears loaded.
100 - increase signal latency test timer considerably for poor little osx,
101 which can't be bothered with basic posix functionality such as signal
102 delivery.
103 - mention some debugging aids in the tutorial.
104 - new env variable PERL_ANYEVENT_MAX_SIGNAL_LATENCY.
105 - private (%namespace) log contexts in PERL_ANYEVENT_LOG now have the name
106 of the context as title.
107 - AnyEvent::Strict would not restore read-write access to $_ when a callback
108 died - this is a user bug, but since AE::Strict exists to help dbeugging...
109 - detect the extremely evil IO::Async::Loop::AnyEvent module and refuse
110 to work with it - the author is well aware of the issues but refuses
111 to fix them. This extreme step has been taken because the module actively
112 attacks the goals of AnyEvent and makes it harder for module authors to use
113 AnyEvent.
114
1156.02 Fri Aug 26 20:08:31 CEST 2011
116 - log_to_syslog now accepts facility strings.
117 - log_to_syslog would not allow facilities - "$facility|$level"
118 *literally* means that string. sheesh.
119 - the CPAN parser got confused and thought we provide Sys::Syslog, try
120 to work around.
121 - renamed AnyEvent::Impl::FLTK2 to FLTK.
122
1236.01 Fri Aug 26 07:04:11 CEST 2011
124 - INCOMPATIBLE CHANGE: AnyEvent::DNS resource records now include
125 the TTL as fourth element - this affects ->resolve, ->request
126 and ->dns_unpack indirectly. Convenience functions such as
127 AnyEvent::DNS::{a,txt,mx,srv} are not affected, neither is
128 AnyEvent::Socket.
129 - INCOMPATIBLE CHANGE: backend authors now should not implement
130 one_event or loop, but instead the AnyEvent::CondVar::_wait and _poll
131 methods.
132 - INCOMPATIBLE CHANGE: verbosity levels have been redefined, see
133 AnyEvent (PERL_ANYEVENT_VERBOSE) and AnyEvent::Log for details.
134 - DEPRECIATION: please change
135 "use AnyEvent::Impl::Perl" to "use AnyEvent::Loop" and
136 "AnyEvent::Impl::Perl::loop" to "AnyEvent::Loop::run"
137 in your programs, for future compatibility.
138 - *blush* idle watcher emulation was borked.
139 - AnyEvent::Strict now checks the AE::xxx functions as well.
140 - querying too long domain names would barf deep inside AnyEvent::DNS - since
141 this is hard to avoid for applications, AnyEvent:.DNS::request now checks for
142 for too long domain names explicitly and passes an undef to the callback
143 instead (reported by James Bromberger).
144 - make AnyEvent::Util more robust against EINTR (analyzed by Andrew Suffield).
145 - new functions: AnyEvent::postpone, AnyEvent::log.
146 - new module: AnyEvent::Log, for simple logging needs.
147 - new env variable, PERL_ANYEVENT_LOG, for fine-grained logging config.
148 - new env variable, PERL_ANYEVENT_DEBUG_SHELL, to automatically start
149 a debug shell.
150 - new env variable, PERL_ANYEVENT_DEBUG_WRAP, to automatically enable
151 watcher instrumentation/wrapping (see AnyEvent::Debug::wrap).
152 - new env variable, PERL_ANYEVENT_HOSTS, to override the place where
153 /etc/hosts can be found.
154 - an empty PERL_ANYEVENT_RESOLV_CONF now selects the default
155 configuration.
156 - env variables can now be specified using AE_xyz instead of
157 PERL_ANYEVENT_xyz, subject to some rules.
158 - lots of new AnyEvent::Debug functionality.
159 - AnyEvent::Strict now checks for modifications of $_ by the callback
160 (by setting it to readonly while executing the callback, if possible).
161 - IO::Async sometimes cancels the wrong timer when you call cancel_timer.
162 Invest extra overhead to work around this peculiar behaviour.
163 - split the pure perl backend into a loop module and an impl
164 module (AnyEvent::Loop and AnyEvent::Impl::Perl).
165 - some read types (regex, netstring, json, storable) could
166 errornously keep an AE::Handle object alive on their own
167 (reported by Mohammad Toossi).
168 - AnyEvent::Socket::resolve_sockaddr and all functions using it now
169 supports /etc/hosts overrides.
170 - add more workarounds around child watcher bugs in IO::Async
171 and (newly introduced) in POE.
172 - use glib child watchers and try to work around its limitations,
173 if possible (based on a patch by Kevin Ryde).
174 - add support for FLTK (fltk2), via AnyEvent::Impl::FLTK2.
175 - many read types in AnyEvent::Handle would malfunction during
176 transitions to TLS because of a "delete $self->{rbuf}".
177 (reported by Gerald Galster).
178 - improve condvar blocking wait performance (probably only noticable
179 with faster backends), also make it easier for event loops that
180 do not support blocking waits.
181 - PERL_ANYEVENT_MODEL now also allows full module names, although
182 the usefulness of this feature is not yet clear.
183 - add tests for all supported event loops, conditional on PERL_ANYEVENT_LOOP_TESTS.
184 - add t/80_ssltext.t.
185 - no longer use AUTOLOAD for the initial loading (for no very good reason).
186 - tcp_server now tries to unlink unix domain sockets when it gets destroyed.
187 - do not run event loop detection multiple times when programs call their cached
188 copies of AnyEvent::detect.
189 - parse_hostport now accepts absolute paths als unix domain sockets.
190 - suppress the idiotic warning inside POE by patching the POE::Kernel
191 run flags instead of asking the user to jump through contortions.
192 - add Task::Weaken as dependency (as recommends only at this time).
193 - replace PF_UNSPEC by 0, for good old minix.
194
1956.0 (not officially released, but was on CPAN for about 15 minutes :/)
196
1975.34 Thu May 12 09:59:41 CEST 2011
198 - AE::TLS didn't support both cert_file and key at the same time,
199 due to some copy&paste bug (reported and analyzed by vti).
200 - improve AE::Handle timeout documentation slightly.
201
2025.33 Mon Feb 7 21:51:41 CET 2011
203 - never appeared on CPAN, for unknown reasons.
204 - AnyEvent::Handle on_connect_error - do not try to destroy a handle
205 object when it's already gone (analyzed by Ryan Bullock).
206 - update idna algorithm(s) to UTS#46 version 6.0.0.
207 - preserve trailing dot in idn_to_ascii - testsuite says to preserve,
208 uts46 says it's an error, and testsuite makes more sense.
209 - remove leading dots in idn_to_ascii - testsuite says to remove,
210 uts46 says it's an error, and testsuite makes more sense.
211 - fixed a bug in idn ignored character processing.
212 - passes the uts46 6.0.0 testsuite except for three cases which
213 apparently are bugs in the testsuite itself.
214 - passes the uts46 6.0.1 testsuite except for eleven cases which
215 apparently are bugs in the testsuite itself.
216
2175.31 Sun Jan 23 11:49:19 CET 2011
218 - work around a POE bug where POE would first poll for new events and
219 invoke handlers BEFORE handling the message queue.
220 - fix parsing of "domain" statement in resolv.conf (Johannes Plunien).
221 - stop_read now does just that in TLS mode. see its
222 documentation for caveats.
223 - update to common::sense 3.4 warning set (some warnings in perl impose
224 an arbitrary 32kb limit on regex matches, which affects
225 AnyEvent::Handle).
226 - AE::Handle had a broken rbuf_max method and missing wbuf_max method.
227 - new experimental AE::Handle::resettls method.
228
2295.3 Fri Dec 31 05:48:13 CET 2010
230 - major IOAsync patch by Paul Evans - 0.33 is required, as the major
231 issues with IO::Async have been solved in that version, including
232 the availabilty of a default loop!
233 - implement new wbuf_max parameter for AnyEvent::Handle.
234 - added Cocoa::EventLoop backend by Daisuke Murase.
235 - the default testsuite now honors PERL_ANYEVENT_MODEL (Paul Evans).
236 - clarify handle on_prepare documentation.
237 - set CLOEXEC flag on parent-side fd's in AnyEvent::Util::run_cmd,
238 to avoid leaking them to any child processes.
239 - try to force files into the arch-dependend lib directory, for the
240 half-"turly obsessive compulsive person"(s) that depend on it.
241
2425.29 Sun Dec 5 10:49:21 CET 2010
243 - convert EV backend to EV 4.00 API (so better upgrade EV too).
244 - AE::Handle: implement dynamic read_size adjustment, add
245 max_read_size parameter, reduce default read_size to 2048.
246 - add some FAQ entries for servers and/or on_eof vs. on_error.
247 - work around OS bugs (cygwin again) when lingering in AE::Handle.
248 - reduce memory usage slightly when connect was used in AE::Handle.
249 - wrok around more segfaults in Tk.
250 - document tls_ctx => undef as valid.
251 - detect WNOHANG value at compiletime.
252 - include some probably linux-only support for building AnyEvent
253 as part of the perl core.
254 - improve compatibility of t/08_idna.t to perl 5.8.
255 - make AnyEvent::Strict truly optional (in case somebody builds a
256 minimal perl...).
257
2585.28 Wed Oct 13 04:14:23 CEST 2010
259 - due to a glitch, AnyEvent's internal getprotobyname wasn't
260 actually used in tcp_server.
261 - implement AnyEvent::CondVar->new method, which was documented but
262 missing.
263 - support autoloading of read types in unshift_read.
264 - AnyEvent::Strict: do the test for invalid fh arguments differently,
265 resulting in more informative error messages.
266 - new AnyEvent::FAQ manpage.
267 - clarify that you can actually call push_read etc. in on_read.
268 - put keys and certs in /tmp, not into the current directory,
269 and document this in the SECURITY CONSIDERATIONS sections
270 of AnyEvent::TLS.
271 - support ";" as resolv.conf comment character in AnyEvent::DNS, as
272 requested by Ron Isaacson.
273 - document $AnyEvent::DNS::RESOLVER variable.
274 - incorporated major typo patches by Abhijit Menon-Sen.
275 - AnyEvent::Handle now croaks when negative timeouts are passed.
276 - add a shitload of TCP_xxx sockopt name constants to AnyEvent::Util,
277 undocumented, but who knows when they might come in handy.
278
2795.271 Tue Jun 8 12:05:46 CEST 2010
280 - backport to perl 5.8.x.
281
2825.27 Sun Jun 6 12:12:05 CEST 2010
283 - postpone differently in AnyEvent::Socket now, as
284 when not, canceling the connection attempt might fail
285 (found by Felix Antonius Wilhelm Ostmann).
286 - explicitly check for non-stream sockets in AE::Handle, too many
287 clueless people fell into the trap of this somehow working.
288 - simplified and reworked the "OTHER MODULES" section.
289 - better/more condvar examples.
290
2915.261 Wed Apr 28 16:13:36 CEST 2010
292 - AF_INET6 was not properly used from Socket6 during configuration
293 time (found by Andy Grundman).
294
2955.26 Mon Apr 12 04:49:35 CEST 2010
296 - don't generate spurious readyness notifications when select
297 returns EINTR in the pure perl backend (analysed and testcase
298 by Bas Denissen).
299 - give same examples in SYNOPSIS sections of AnyEvent and AE.
300 - provider faster implementations for AE::time/now/now_update for
301 EV and Perl backends.
302 - provide faster AE::cv and AE::time implementations for all
303 backends.
304 - no longer support lower/mixed-case signal names (this was
305 never documented nor universally supported).
306 - some more memory µ-optimisations, and somewhat less messy
307 code to cope with both AnyEvent and AE APIs.
308
3095.251 Sat Mar 13 00:58:21 CET 2010
310 - make test failed if it was run before make install on perl <=
311 5.8.8, and older versions of perl make our live very difficult,
312 so write constants.pl during Makefile.PL time.
313
3145.25 Sat Mar 13 00:23:14 CET 2010
315 - fix a race condition in AnyEvent::Handle that would cause
316 a "bio_read: unsupported method" error (in _another_
317 ssl connection) after user code threw an exception.
318 - added AnyEvent::Handle->destroyed method.
319 - speed up AnyEvent::Socket::format_address and ::format_ipv6.
320 - the AnyEvent::Util::fh_nonblocking function worked only
321 by lucky accident on win32.
322 - smaller and faster AnyEvent::Util::fh_nonblocking.
323 - when the (required!) Time::HiRes module is missing, AnyEvent
324 did not fall back to built-in time properly.
325 - do not load Fcntl at runtime, saving memory and loading time.
326 - precompile a number of constants and use them instead of runtime
327 detection and eval.
328 - free detection code after detection and similar memory optimisations.
329 - Perl backend timer interval best effort drift has been improved
330 (same algorithm as EV).
331 - update unicode idna mapping table.
332
3335.24 Tue Jan 5 11:39:43 CET 2010
334 - cygwin never reports errors from failed connects. 1.5 just gives
335 you continous readyness and EAGAIN, 1.7 is even more broken and
336 just hangs. work around both issues in a cygwin-specific hack.
337 - improve idle watcher documentation slightly.
338
3395.23 Sun Dec 20 23:48:00 CET 2009
340 - support IDNs in resolve_sockaddr, and therefore in tcp_connect.
341 - implement punycode_encode/decode, idn_nameprep,
342 idn_to_ascii and idn_to_unicode operations in AnyEvent::Util.
343 - provide $AE::VERSION.
344 - removed traces of "no strict 'refs'".
345
3465.22 Sat Dec 5 03:51:13 CET 2009
347 - downgrade-or-fail in AnyEvent::Handle::push_write, to
348 diagnose encoding failures earlier and more succinctly.
349 (this works around bugs in perl, throwing away encoding info
350 when passing scalar data to extensions).
351 - add more examples to AnyEvent::Socket manpage.
352 - upgrade internal warning set to the same as common::sense 2.03.
353 - use pack "n/a*" for pre-5.8.9 perl compatibility in AnyEvent::DNS
354 (John Beppu).
355 - AnyEvent::Socket::inet_aton now properly supports ipv6, as documented.
356 - add google public dns servers to fallback server set and make sure
357 we load-balance properly between the three sets. also add all
358 fallback dns servers, not just a random one, to each dns config.
359
3605.21 Thu Nov 19 02:48:47 CET 2009
361 - fix a problem where socket constants were called with parameters
362 (spotted by David Friedland).
363 - fork_call never use'd POSIX (reported by Daisuke Maki).
364 - improve perl 5.6 compatibility further (but it still won't work
365 unless you rip out everything but the core).
366 - prefer Net::DNS::Resolver over ipconfig, if installed, on win32.
367 uses 10MB of RAM, but doesn't open a console window. *sigh*.
368
3695.202 Wed Oct 14 22:35:44 CEST 2009
370 - AnyEvent::DNS would unexpectedly clobber $_ under windows
371 (analysed by Matthias Waldorf).
372 - AnyEvent::Handle::run_cmd can now pass the PID of the
373 newly-created process, which is much less useful than it might
374 sound (based on patch by Yann Kerherve).
375
3765.201 Tue Sep 29 12:09:25 CEST 2009
377 - AnyEvent:Handle::on_starttls/on_stoptls methods were broken
378 (reported by Torsten Foertsch).
379 - common::sense 2.0 could cause tcp_server to throw an exception
380 (analysed by elmex).
381
3825.2 Mon Sep 14 07:04:49 CEST 2009
383 - INCOMPATIBLE CHANGE: do no longer support register_read_type
384 and register_write_type in AnyEvent::Handle, instead support
385 package names (the facility was mostly abused).
386 - implement "packagename-as-read/write type" support in
387 AnyEvent::Handle.
388 - AnyEvent::Handle: new options "keepalive" and "oobinline".
389 - oobinline set by default to avoid security issues.
390 - the pure-perl event loop backend wrongly detected times() fallback
391 support (spotted by Pavel Boldin).
392
3935.12 Tue Sep 1 20:26:50 CEST 2009
394 - be more lenient when parsing resolv.conf files, as some people
395 use hashmarks as comment indicator inside directives (reported
396 by Michael S. Fischer).
397 - use same set of warnings as common::sense 2.0.
398 - fix a potential 32 bit overflow issue due to perl having problems
399 with large hex constants in the Perl backend.
400 - do not use the slower 5.6 version fo the enc_name code in AnyEvent::DNS
401 by default.
402 - fix some prototypes.
403 - (optionally) require Async::Interrupt 1.02.
404 - replace opendns fallbacks by verizon/level3 public dns servers.
405
4065.112 Fri Aug 21 13:59:04 CEST 2009
407 - AnyEvent::Handle could cause hard crashes in Net::SSLeay when
408 using starttls with an illegal/uncreatable tls context.
409 - port to broken win32 perls.
410
4115.111 Sun Aug 16 18:44:36 CEST 2009
412 - perl errornously requires sockaddr_un structures returned by the
413 kernel to have a specific length. work around this bug by padding
414 sockaddr_un structures (found to fail on OS X, as perl only uses
415 the correct code for GNU/Linux, but likely to fail on other
416 OSes as well).
417
4185.11 Wed Aug 12 17:49:37 CEST 2009
419 - fixed a bug in run_cmd with input from filename.
420 - create a json coder object when none is given in write direction
421 to reduce dependency on JSON.
422
4235.1 Tue Aug 11 03:17:32 CEST 2009
424 - tcp_connect now reports the correct error in $!, not an unrelated
425 one (was broken since 4.91).
426 - AE::Handle did not properly clear rtimeout/wtimeout watchers
427 when retrying a tcp_connect.
428 - new functions: AnyEvent::Util::run_cmd and close_all_fds_except.
429 - fixed wrong documentation in AE::Socket.
430
4315.01 Mon Aug 10 03:16:32 CEST 2009
432 - last release broke EV child watchers :/.
433
4345.0 Sun Aug 9 17:32:09 CEST 2009
435 - officially introduce and document the AE API.
436 - lowering the timeout at runtime did not have immediate effect
437 in AnyEvent::Handle.
438 - AE::Handle now has separate and independent read and write timeout
439 settings.
440 - probe for Guard module when AnyEvent::Util is loaded, not at
441 runtime (Event doesn't like eval "use" from callbacks, and it avoids
442 exporting the probe function).
443 - Impl::IOAsync's io watchers did not use the filehandle emulation
444 code to work around it's watcher limits.
445 - work around IO::Async's problems with overloaded objects.
446 - try to disable wearnings inside IO::Async, it's just too buggy.
447 - do no longer use Event by default, only EV or Perl, Event is too buggy.
448 - hackishly provide fast AE:: interface when EV or Perl is the backend.
449 - inherit VERSION from AnyEvent::VERSION in most modules.
450
4514.91 Thu Aug 6 15:42:45 CEST 2009
452 - AE::Handle::starttls could get out of sync when the read buffer
453 already contains some TLS handshake.
454 - AE::Handle did not properly free the TLS session, causing bigger
455 memory leaks in Net::SSLeay than expected (reported by toaster).
456 - AE::Socket::tcp_connect will now artificially delay invoking
457 the callback to avoid returning after invoking the callback.
458 - convert many internal watcher uses to AE API.
459
4604.9 Sat Aug 1 11:07:01 CEST 2009
461 - Glib silently fails when registering a timeout with
462 a negative value, so avoid that.
463 - call condvar callback immediately when it is set
464 after the condition is already signalled.
465 - check rbuf_max condition only after trying to consume
466 data.
467 - normalise signal numbers to names when using the ae's
468 signal handling, but do not document this (yet).
469 - pure perl signal emulation did not properly set nonblocking
470 mode on the signal pipe on win32 (this is unlikely to have caused
471 any issues).
472 - new module: AnyEvent::Debug.
473 - AnyEvent::Strict now checks that a signal was specified
474 by name (not name or number).
475 - reduce memory footprint in typical cases by ~50kb by
476 only compiling signal code when necessary.
477 - add AnyEvent::Handle->rbuf_max.
478 - grab the AE namespace for future new API, implement stubs for
479 said future/alternative API.
480 - new function: AnyEvent::Socket::format_hostport.
481
4824.881 Tue Jul 28 12:51:53 CEST 2009
483 - work around a bug in local in pre-5.10 perls, causing
484 AnyEvent::Handle to recurse when it should not
485 (analysed by elmex).
486
4874.88 Tue Jul 28 04:04:37 CEST 2009
488 - re-bless the handle into a dummy package after calling
489 AnyEvent::Handle::destroy, so the user does not need to check
490 for errors after every push_write etc.
491 - do not attempt to run t/02_signals.t on obviously broken
492 platforms.
493
4944.87 Sun Jul 26 02:06:16 CEST 2009
495 - do not attempt to linger when there is no longer a valid fh.
496 - enforce tls mode to be either accept or connect, do not simply
497 segfault in Net::SSLeay.
498 - AnyEvent::Handle can now call tcp_connect itself (new parameters
499 connect, on_prepare, on_connect and on_connect_error). Updated
500 tutorial accordingly.
501 - add AnyEvent::Impl::Irssi backend.
502
5034.86 Mon Jul 20 23:52:29 CEST 2009
504 - since the verbose warning is not enough, explicitly document
505 that versions before 1.33 of Net::SSLeay are not secure.
506 - work around signal handling races in Event and (...) Event::Lib.
507 - try to align signal-race timer to full-second boundaries.
508 - work around Tk not liking negative timeouts.
509 - don't complain of different grades of environmental unfriendlyness
510 in IO::Async.
511
5124.85 Sat Jul 18 06:16:14 CEST 2009
513 - nail the signal race problem in perl once and for all
514 (see $AnyEvent::MAX_SIGNAL_LATENCY).
515 - take advantage of Async::Interrupt if it is available.
516 - load Time::HiRes and Guard modules on demand only.
517 - add optional/recommended modules section to AnyEvent
518 documentation.
519 - reduce memory usage considerably (and reduce startup penalty)
520 by not using "strict", "warnings" and "overload" modules.
521 - work around buggy windows/openbsd perls and provide EBADMSG
522 and EPROTO ourselves when missing.
523 - improve perl 5.6 compatibility of the core event loop.
524 - made Net::SSLeay version 1.33 a soft requirement.
525
5264.83 Fri Jul 17 16:56:26 CEST 2009
527 - implement AnyEvent::Socket::getprotobyname.
528 - AnyEvent::CondVar's will now detect recursive blocking
529 waits and will croak, as too many people fall into
530 this trap.
531 - AnyEvent::Handle will now call ->destroy on itself after
532 executing the on_error callback, instead of doing some
533 half-baked internal shutdown, for fatal errors.
534 - clarify on_eof behaviour w.r.t. the read queue and
535 on_read callbacks.
536 - ignore some possible spurious wake-ups in tcp_connect.
537
5384.82 Sat Jul 11 00:34:55 CEST 2009
539 - POE and Event backends didn't accept some callable objects as
540 callbacks.
541 - use Config module instead of POSIX module to detect signal names
542 in AnyEvent::Strict and AnyEvent::Impl::EventLib, as the POSIX
543 module doesn't even have all POSIX signals :/.
544 - use more workarounds around the many refcnt/corruption bugs in
545 Event::Lib.
546 - work around a race condition in perl's select, causing t/03_child.t
547 to rarely fail.
548
5494.81 Thu Jul 9 10:30:30 CEST 2009
550 - AnyEvent::Handle didn't properly diagnose write errors
551 (it expected -1 from syswrite, how lame... :).
552 - support file descriptors in addition to file handles
553 in AnyEvent->io.
554 - new env variables: PERL_ANYEVENT_RESOLV_CONF,
555 PERL_ANYEVENT_MAX_OUTSTANDING_DNS, PERL_ANYEVENT_CA_FILE
556 and PERL_ANYEVENT_CA_PATH.
557 - provide a sensible synopsis section for AnyEvent::TLS.
558 - add a "supported backends" section to the manpage.
559 - added simple io watcher test to testsuite, using a
560 portable_socketpair.
561 - tried to improve the stability of the Event::Lib backend,
562 YMMV.
563
5644.8 Mon Jul 6 23:45:16 CEST 2009
565 - AnyEvent::DNS did not properly follow CNAME records with
566 uppercase targets.
567 - AnyEvent::DNS would errornously return AAAA records
568 with v4 mapped addresses (a faulty record) as ipv4 addresses,
569 causing AnyEvent::Socket to throw an exception.
570 - added new module AnyEvent::TLS for easier SSL/TLS context
571 creation, with many options including hostname verification,
572 secure default configuration, lots of documentation and,
573 predefined diffie-hellman keys for perfect forward security
574 and much more. get it while it's still fresh!
575 - use AnyEvent::TLS in AnyEvent::Handle for context management.
576 - load AnyEvent::Handle only on demand in AnyEvent::DNS,
577 so AnyEvent::Socket users have smaller memory footprint
578 in the common case.
579 - add AnyEvent::Handle->push_shutdown method.
580 - add an additional error message parameter to AnyEvent::Handle's
581 on_error callback (for TLS, $! is still available).
582 - add AnyEvent::Handle on_starttls/on_stoptls callbacks.
583 - make AnyEvent::Handle more robust against early conenction
584 failures (during new), and return C<undef> in that case
585 from the constructor.
586 - AnyEvent::Handle will now try to load only JSON::XS first,
587 then fall back to JSON.
588 - format_ipv4/format_ipv6 are now exported by default, for symmetry,
589 and because it was documented that way.
590
5914.451 Fri Jul 3 00:28:58 CEST 2009
592 - do not clear rbuf when shutting down an AnyEvent::Handle
593 object - doing so breaks AnyEvent::HTTP.
594
5954.45 Mon Jun 29 22:59:26 CEST 2009
596 - a write error could cause AnyEvent::Handle to create
597 an I/O watcher with an undefined $fh.
598 - special-case mapped ipv4 addresses in both
599 AnyEvent::Socket::format_address and parse_address,
600 to treat them just like ipv4 addresses.
601 - updated and overhauled the AnyEvent::Intro doc.
602 - implement AnyEvent::Socket::format_ipv4/ipv6.
603 - slightly speed up portable_pipe/socketpair.
604 - expand condvar begin/end documentation.
605
6064.42 Fri Jun 26 08:32:18 CEST 2009
607 - preliminary and neccesarily incomplete support for IO::Async.
608 - reset SIGCHLD to DEFAULT when AnyEvent is loaded, in
609 case it was set to IGNORE, to ensure we can catch child statuses
610 even when the calling env acts stupidly.
611 - updated benchmarks with IO::Async, which performs very well.
612
6134.412 Wed Jun 24 01:35:57 CEST 2009
614 - support an "untaint" attribute for AnyEvent::DNS
615 and set it on the default resolver.
616 - implement and document AnyEvent::Impl::Perl::loop.
617 - remove all anyevent-env variables from %ENV
618 when running in tainted mode.
619 - mention and extend the IO::Lambda benchmark.
620
6214.411 Sun Jun 7 18:48:02 CEST 2009
622 - do not try to use F_SETFD on windows.
623
6244.41 Thu May 14 06:40:11 CEST 2009
625 - work around issues in older perls (5.8.0?) when
626 a signal handler is deleted from the %SIG hash.
627 - use POSIX::_exit in child test, to avoid running
628 destructors.
629 - speed up CHLD handling by relying on SIGCHLD being
630 synchronously delivered, even when we roll our own
631 implementation.
632 - AnyEvent::DNS: add the "dname" resource record name
633 for cosmetic reasons.
634
6354.4 Sun Apr 26 20:12:33 CEST 2009
636 - implemented idle watchers, where applicable.
637 - AnyEvent->time died when Event backend was in use.
638 - fix a memleak in the Tk backend.
639 - sped up Tk timer handling.
640 - clip negative "after" values to 0 in AnyEvent::Impl::Event
641 to avoid spamming out warning messages.
642 - fix Qt timers without interval.
643 - avoid Qt zero-timeout special case, in old libqt's.
644
6454.352 Mon Apr 20 16:31:11 CEST 2009
646 - fix AnyEvent::Strict error messages for child watchers.
647 - fix/update Makefile.PL "recommends" versions.
648 - add AnyEvent->now_update.
649
6504.351 Sat Apr 11 07:56:14 CEST 2009
651 - actually make the signal pipe work under win32.
652 - localise $! in signal handler to avoid changing $!, although
653 perl itself does not seem to save/restore errno either.
654 - set the cloexec flag on the signal pipe (normally set by perl too).
655
6564.35 Fri Mar 27 11:48:20 CET 2009
657 - event models relying on AnyEvent's signal watcher emulation did
658 invoke the callback asynchronously, contradicting documentation
659 and causing signals to get lost (this includes AnyEvent's own
660 event loop). AnyEvent now uses the standard pipe trick to make
661 callback execution synchronous to the event loop.
662 - AnyEvent::Handle didn't free TLS context data on DESTROY
663 (patch by Pavel Shaydo).
664 - work around the ever-incompatibly-changing API of MakeMaker.
665 - document that changing global variables without restoring them
666 is a bad idea in Perl (noted by Adam Rosenstein).
667 - AnyEvent::Strict now barfs if ->io is passed a file.
668
6694.34 Thu Feb 12 18:32:45 CET 2009
670 - separately buffer TLS read data, as otherwise the read queue could
671 deadlock as receiving data is not expected while draining the
672 read queue (which cna only happen with TLS).
3 - raise EBADMSG error on JSON decoding errors. 673 - raise EBADMSG error on JSON decoding errors.
4 - fix some minor manpage bugs (reported by Maximilian Gaß). 674 - fix some minor manpage bugs (reported by Maximilian Gaß).
5 - speed up select bitmask parsing quite a bit in the pure perl 675 - speed up select bitmask parsing quite a bit in the pure perl
6 backend. 676 backend.
7 - use CORE::select instead of just select in AnyEvent::Impl::Perl. 677 - use CORE::select instead of just select in AnyEvent::Impl::Perl.
… …
424.3 Fri Oct 3 09:18:43 CEST 2008 7124.3 Fri Oct 3 09:18:43 CEST 2008
43 - AnyEvent will now install a no-op signal handler for SIGPIPE, 713 - AnyEvent will now install a no-op signal handler for SIGPIPE,
44 unless one has been installed already. 714 unless one has been installed already.
45 - warn about not loaidng AnyEvent::Impl::POE early enough 715 - warn about not loaidng AnyEvent::Impl::POE early enough
46 (patch by Adam Rosenstein). 716 (patch by Adam Rosenstein).
47 - fixed a great number of bugs and corner cases in AnyEvent::Handle: 717 - fixed a great number of bugs and corner cases in AnyEvent::Handle.
48 - fix a bug in where in SSL connect mode, the 718 - fix a bug in where in SSL connect mode, the
49 client would first wait for some data by the server and 719 client would first wait for some data by the server and
50 otherwise hang. (reported and analyzed in an absolutely 720 otherwise hang. (reported and analysed in an absolutely
51 exemplary manner by Adam Rosenstein). 721 exemplary manner by Adam Rosenstein).
52 - fix a bug in where SSL EOF would not be treated 722 - fix a bug in where SSL EOF would not be treated
53 as stream EOF, putting the connection into a hung state. 723 as stream EOF, putting the connection into a hung state.
54 (reported and analyzed in an absolutely exemplary manner by Adam 724 (reported and analysed in an absolutely exemplary manner by Adam
55 Rosenstein). 725 Rosenstein).
56 - fix a potential segfault when the TLS context 726 - fix a potential segfault when the TLS context
57 would go missing in a read callback (e.g. due to stoptls) - 727 would go missing in a read callback (e.g. due to stoptls) -
58 Net::SSLeay of course makes no type checking whatsoever. 728 Net::SSLeay of course makes no type checking whatsoever.
59 - AnyEvent::Handle will not stop the read watcher in TLS mode. 729 - AnyEvent::Handle will not stop the read watcher in TLS mode.

Diff Legend

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