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

Comparing AnyEvent/Changes (file contents):
Revision 1.200 by root, Mon Oct 27 02:58:19 2008 UTC vs.
Revision 1.558 by root, Wed Oct 31 15:42:06 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
10 - AnyEvent::Socket::inet_aton did not work when dns resolution was
11 used to find the addresses (analyzed and patch by Maxime Soulé).
12
137.02 Tue Aug 14 04:27:58 CEST 2012
14 - AnyEvent::Util::run_cmd could block indefinitely (analyzed and test
15 program by Yorhel).
16 - verified that AnyEvent::Socket follows RFC5952.
17 - try to parse "ADDR#PORT" in addition to "ADDR PORT".
18
197.01 Sun May 13 01:03:17 CEST 2012
20 - fail with EPROTO in AnyEvent::Handle wqhen TLS is requested but
21 not available, instead of throwing an exception.
22 - use File::Spec to get the tmpdir in t/*, to avoid needless
23 failures on (most, not mine :) windows boxes.
24 - new handle read types: tls_detect and tls_autostart.
25
267.0 Fri Apr 13 06:33:30 CEST 2012
27 - child watchers are broken in POE 1.352 (also many earlier
28 versions) and there seems to be no way to work around it, as POE
29 itself is inherently racy. Document this regression and add a delay
30 in t/68_poe_03_child.t for the time being.
31 - new module AnyEvent::IO, that is a frontend to either a pure-perl
32 synchronous I/O implementation (AnyEvent::IO::Perl), or to the
33 asynchronous I/O extension IO::AIO (AnyEvent::IO::IOAIO).
34 - load /etc/hosts only when DNS has no answer.
35 - stat /etc/hosts on every access and reload it if it changed.
36 - load /etc/hosts via AnyEvent::IO - potentially asynchronous.
37 - fix a buggy croak in the dh parameter reading in AnyEvent::TLS.
38 - AnyEvent::Log log_to_file and log_to_path now use AnyEvent::IO. As
39 a side effect, they now use true appending as opposed to libc
40 appending, and the file might not have been opened when the function
41 returns.
42 - the default logging level was not properly documented in a variety of
43 places, this has hopefully been rectified.
44 - updated uts46data.pl for unicode 6.1.0.
45 - made log messages generated by AnyEvent submodules not
46 include the package name anymore, as it will be logged
47 by default already.
48 - upgrade to the trick used by common-sense 3.5 to work
49 around extra warning torture/breakage under perl 5.15.x.
50 - log messages by anyevent are now ucfirst, are usually full
51 sentences and do no longer include the package name.
52 - the storable read type would sometimes throw an exception instead
53 of causing EBADMSG (analyzed by Maxime Soulé).
54 - work around a bug in openssl 1.0.1 which enforces a minimum rsa
55 keysize (reported by Paul Howarth).
56 - documented Rocco fabricating statements about AnyEvent and me.
57
586.14 Tue Jan 31 20:00:24 CET 2012
59 - AnyEvent::Impl::Tk was broken due to a mysterious "die" inside,
60 probably an editing mistake (reported by Darin McBride).
61
626.13 Thu Jan 12 07:27:01 CET 2012
63 - AnyEvent::Util::fork_call checked for POSIX availability in the wrong way,
64 causing it to fail if POSIX wasn't loaded already (analysed by Rock Power).
65 - AnyEvent::Handle::push_read (line => did pass $1 directly, so
66 regex-matching inside the callback would change the parameter, despite
67 $1 being dynamically scoped per-block. this perl bug is now being
68 worked around (testcase by Cindy Wang).
69
706.12 Mon Dec 12 13:21:10 CET 2011
71 - $! was clobbered by subsequent calls in tcp_connect, due to
72 the postpone not saving/restoring it (analyzed by Richard Garnish).
73
746.11 Tue Nov 22 10:36:05 CET 2011
75 - Tk cannot create windows when tainted unless you set a title, so
76 set a dummy title for AnyEvent's dummy mainwindow. How dummy.
77 - escape any nonprintable/nonascii characters when stringifying
78 backtraces.
79 - log the reason loading an interface module fails to load at level debug.
80
816.1 Tue Oct 4 19:44:30 CEST 2011
82 - INCOMPATIBLE CHANGE: the default log level is now "4" (error
83 and above), and some messages inside AnyEvent have been elevated
84 to higher log levels to print by default.
85 - AnyEvent::log, unlike AnyEvent::Log::log, did not always exit on fatal.
86 - work around yet another signal race bug, newly introduced by POE. My
87 how that module sucks.
88 - add google public ipv6 dns as fallback servers, so AE::DNS can finally
89 support ipv6-only operation even with defaults (also use ::1 instead
90 of 127.0.0.1).
91 - give better guidelines on how to chose a logging level.
92 - AnyEvent::Log can now cap (limit) log message priority by context.
93 - AE::log will now save and restore $! and $@, so callers don't have to.
94 - AE::Strict now checks that registered fds are still valid at regular
95 intervals.
96 - skip frequently called 'require' calls if the module appears loaded.
97 - increase signal latency test timer considerably for poor little osx,
98 which can't be bothered with basic posix functionality such as signal
99 delivery.
100 - mention some debugging aids in the tutorial.
101 - new env variable PERL_ANYEVENT_MAX_SIGNAL_LATENCY.
102 - private (%namespace) log contexts in PERL_ANYEVENT_LOG now have the name
103 of the context as title.
104 - AnyEvent::Strict would not restore read-write access to $_ when a callback
105 died - this is a user bug, but since AE::Strict exists to help dbeugging...
106 - detect the extremely evil IO::Async::Loop::AnyEvent module and refuse
107 to work with it - the author is well aware of the issues but refuses
108 to fix them. This extreme step has been taken because the module actively
109 attacks the goals of AnyEvent and makes it harder for module authors to use
110 AnyEvent.
111
1126.02 Fri Aug 26 20:08:31 CEST 2011
113 - log_to_syslog now accepts facility strings.
114 - log_to_syslog would not allow facilities - "$facility|$level"
115 *literally* means that string. sheesh.
116 - the CPAN parser got confused and thought we provide Sys::Syslog, try
117 to work around.
118 - renamed AnyEvent::Impl::FLTK2 to FLTK.
119
1206.01 Fri Aug 26 07:04:11 CEST 2011
121 - INCOMPATIBLE CHANGE: AnyEvent::DNS resource records now include
122 the TTL as fourth element - this affects ->resolve, ->request
123 and ->dns_unpack indirectly. Convenience functions such as
124 AnyEvent::DNS::{a,txt,mx,srv} are not affected, neither is
125 AnyEvent::Socket.
126 - INCOMPATIBLE CHANGE: backend authors now should not implement
127 one_event or loop, but instead the AnyEvent::CondVar::_wait and _poll
128 methods.
129 - INCOMPATIBLE CHANGE: verbosity levels have been redefined, see
130 AnyEvent (PERL_ANYEVENT_VERBOSE) and AnyEvent::Log for details.
131 - DEPRECIATION: please change
132 "use AnyEvent::Impl::Perl" to "use AnyEvent::Loop" and
133 "AnyEvent::Impl::Perl::loop" to "AnyEvent::Loop::run"
134 in your programs, for future compatibility.
135 - *blush* idle watcher emulation was borked.
136 - AnyEvent::Strict now checks the AE::xxx functions as well.
137 - querying too long domain names would barf deep inside AnyEvent::DNS - since
138 this is hard to avoid for applications, AnyEvent:.DNS::request now checks for
139 for too long domain names explicitly and passes an undef to the callback
140 instead (reported by James Bromberger).
141 - make AnyEvent::Util more robust against EINTR (analyzed by Andrew Suffield).
142 - new functions: AnyEvent::postpone, AnyEvent::log.
143 - new module: AnyEvent::Log, for simple logging needs.
144 - new env variable, PERL_ANYEVENT_LOG, for fine-grained logging config.
145 - new env variable, PERL_ANYEVENT_DEBUG_SHELL, to automatically start
146 a debug shell.
147 - new env variable, PERL_ANYEVENT_DEBUG_WRAP, to automatically enable
148 watcher instrumentation/wrapping (see AnyEvent::Debug::wrap).
149 - new env variable, PERL_ANYEVENT_HOSTS, to override the place where
150 /etc/hosts can be found.
151 - an empty PERL_ANYEVENT_RESOLV_CONF now selects the default
152 configuration.
153 - env variables can now be specified using AE_xyz instead of
154 PERL_ANYEVENT_xyz, subject to some rules.
155 - lots of new AnyEvent::Debug functionality.
156 - AnyEvent::Strict now checks for modifications of $_ by the callback
157 (by setting it to readonly while executing the callback, if possible).
158 - IO::Async sometimes cancels the wrong timer when you call cancel_timer.
159 Invest extra overhead to work around this peculiar behaviour.
160 - split the pure perl backend into a loop module and an impl
161 module (AnyEvent::Loop and AnyEvent::Impl::Perl).
162 - some read types (regex, netstring, json, storable) could
163 errornously keep an AE::Handle object alive on their own
164 (reported by Mohammad Toossi).
165 - AnyEvent::Socket::resolve_sockaddr and all functions using it now
166 supports /etc/hosts overrides.
167 - add more workarounds around child watcher bugs in IO::Async
168 and (newly introduced) in POE.
169 - use glib child watchers and try to work around its limitations,
170 if possible (based on a patch by Kevin Ryde).
171 - add support for FLTK (fltk2), via AnyEvent::Impl::FLTK2.
172 - many read types in AnyEvent::Handle would malfunction during
173 transitions to TLS because of a "delete $self->{rbuf}".
174 (reported by Gerald Galster).
175 - improve condvar blocking wait performance (probably only noticable
176 with faster backends), also make it easier for event loops that
177 do not support blocking waits.
178 - PERL_ANYEVENT_MODEL now also allows full module names, although
179 the usefulness of this feature is not yet clear.
180 - add tests for all supported event loops, conditional on PERL_ANYEVENT_LOOP_TESTS.
181 - add t/80_ssltext.t.
182 - no longer use AUTOLOAD for the initial loading (for no very good reason).
183 - tcp_server now tries to unlink unix domain sockets when it gets destroyed.
184 - do not run event loop detection multiple times when programs call their cached
185 copies of AnyEvent::detect.
186 - parse_hostport now accepts absolute paths als unix domain sockets.
187 - suppress the idiotic warning inside POE by patching the POE::Kernel
188 run flags instead of asking the user to jump through contortions.
189 - add Task::Weaken as dependency (as recommends only at this time).
190 - replace PF_UNSPEC by 0, for good old minix.
191
1926.0 (not officially released, but was on CPAN for about 15 minutes :/)
193
1945.34 Thu May 12 09:59:41 CEST 2011
195 - AE::TLS didn't support both cert_file and key at the same time,
196 due to some copy&paste bug (reported and analyzed by vti).
197 - improve AE::Handle timeout documentation slightly.
198
1995.33 Mon Feb 7 21:51:41 CET 2011
200 - never appeared on CPAN, for unknown reasons.
201 - AnyEvent::Handle on_connect_error - do not try to destroy a handle
202 object when it's already gone (analyzed by Ryan Bullock).
203 - update idna algorithm(s) to UTS#46 version 6.0.0.
204 - preserve trailing dot in idn_to_ascii - testsuite says to preserve,
205 uts46 says it's an error, and testsuite makes more sense.
206 - remove leading dots in idn_to_ascii - testsuite says to remove,
207 uts46 says it's an error, and testsuite makes more sense.
208 - fixed a bug in idn ignored character processing.
209 - passes the uts46 6.0.0 testsuite except for three cases which
210 apparently are bugs in the testsuite itself.
211 - passes the uts46 6.0.1 testsuite except for eleven cases which
212 apparently are bugs in the testsuite itself.
213
2145.31 Sun Jan 23 11:49:19 CET 2011
215 - work around a POE bug where POE would first poll for new events and
216 invoke handlers BEFORE handling the message queue.
217 - fix parsing of "domain" statement in resolv.conf (Johannes Plunien).
218 - stop_read now does just that in TLS mode. see its
219 documentation for caveats.
220 - update to common::sense 3.4 warning set (some warnings in perl impose
221 an arbitrary 32kb limit on regex matches, which affects
222 AnyEvent::Handle).
223 - AE::Handle had a broken rbuf_max method and missing wbuf_max method.
224 - new experimental AE::Handle::resettls method.
225
2265.3 Fri Dec 31 05:48:13 CET 2010
227 - major IOAsync patch by Paul Evans - 0.33 is required, as the major
228 issues with IO::Async have been solved in that version, including
229 the availabilty of a default loop!
230 - implement new wbuf_max parameter for AnyEvent::Handle.
231 - added Cocoa::EventLoop backend by Daisuke Murase.
232 - the default testsuite now honors PERL_ANYEVENT_MODEL (Paul Evans).
233 - clarify handle on_prepare documentation.
234 - set CLOEXEC flag on parent-side fd's in AnyEvent::Util::run_cmd,
235 to avoid leaking them to any child processes.
236 - try to force files into the arch-dependend lib directory, for the
237 half-"turly obsessive compulsive person"(s) that depend on it.
238
2395.29 Sun Dec 5 10:49:21 CET 2010
240 - convert EV backend to EV 4.00 API (so better upgrade EV too).
241 - AE::Handle: implement dynamic read_size adjustment, add
242 max_read_size parameter, reduce default read_size to 2048.
243 - add some FAQ entries for servers and/or on_eof vs. on_error.
244 - work around OS bugs (cygwin again) when lingering in AE::Handle.
245 - reduce memory usage slightly when connect was used in AE::Handle.
246 - wrok around more segfaults in Tk.
247 - document tls_ctx => undef as valid.
248 - detect WNOHANG value at compiletime.
249 - include some probably linux-only support for building AnyEvent
250 as part of the perl core.
251 - improve compatibility of t/08_idna.t to perl 5.8.
252 - make AnyEvent::Strict truly optional (in case somebody builds a
253 minimal perl...).
254
2555.28 Wed Oct 13 04:14:23 CEST 2010
256 - due to a glitch, AnyEvent's internal getprotobyname wasn't
257 actually used in tcp_server.
258 - implement AnyEvent::CondVar->new method, which was documented but
259 missing.
260 - support autoloading of read types in unshift_read.
261 - AnyEvent::Strict: do the test for invalid fh arguments differently,
262 resulting in more informative error messages.
263 - new AnyEvent::FAQ manpage.
264 - clarify that you can actually call push_read etc. in on_read.
265 - put keys and certs in /tmp, not into the current directory,
266 and document this in the SECURITY CONSIDERATIONS sections
267 of AnyEvent::TLS.
268 - support ";" as resolv.conf comment character in AnyEvent::DNS, as
269 requested by Ron Isaacson.
270 - document $AnyEvent::DNS::RESOLVER variable.
271 - incorporated major typo patches by Abhijit Menon-Sen.
272 - AnyEvent::Handle now croaks when negative timeouts are passed.
273 - add a shitload of TCP_xxx sockopt name constants to AnyEvent::Util,
274 undocumented, but who knows when they might come in handy.
275
2765.271 Tue Jun 8 12:05:46 CEST 2010
277 - backport to perl 5.8.x.
278
2795.27 Sun Jun 6 12:12:05 CEST 2010
280 - postpone differently in AnyEvent::Socket now, as
281 when not, canceling the connection attempt might fail
282 (found by Felix Antonius Wilhelm Ostmann).
283 - explicitly check for non-stream sockets in AE::Handle, too many
284 clueless people fell into the trap of this somehow working.
285 - simplified and reworked the "OTHER MODULES" section.
286 - better/more condvar examples.
287
2885.261 Wed Apr 28 16:13:36 CEST 2010
289 - AF_INET6 was not properly used from Socket6 during configuration
290 time (found by Andy Grundman).
291
2925.26 Mon Apr 12 04:49:35 CEST 2010
293 - don't generate spurious readyness notifications when select
294 returns EINTR in the pure perl backend (analysed and testcase
295 by Bas Denissen).
296 - give same examples in SYNOPSIS sections of AnyEvent and AE.
297 - provider faster implementations for AE::time/now/now_update for
298 EV and Perl backends.
299 - provide faster AE::cv and AE::time implementations for all
300 backends.
301 - no longer support lower/mixed-case signal names (this was
302 never documented nor universally supported).
303 - some more memory µ-optimisations, and somewhat less messy
304 code to cope with both AnyEvent and AE APIs.
305
3065.251 Sat Mar 13 00:58:21 CET 2010
307 - make test failed if it was run before make install on perl <=
308 5.8.8, and older versions of perl make our live very difficult,
309 so write constants.pl during Makefile.PL time.
310
3115.25 Sat Mar 13 00:23:14 CET 2010
312 - fix a race condition in AnyEvent::Handle that would cause
313 a "bio_read: unsupported method" error (in _another_
314 ssl connection) after user code threw an exception.
315 - added AnyEvent::Handle->destroyed method.
316 - speed up AnyEvent::Socket::format_address and ::format_ipv6.
317 - the AnyEvent::Util::fh_nonblocking function worked only
318 by lucky accident on win32.
319 - smaller and faster AnyEvent::Util::fh_nonblocking.
320 - when the (required!) Time::HiRes module is missing, AnyEvent
321 did not fall back to built-in time properly.
322 - do not load Fcntl at runtime, saving memory and loading time.
323 - precompile a number of constants and use them instead of runtime
324 detection and eval.
325 - free detection code after detection and similar memory optimisations.
326 - Perl backend timer interval best effort drift has been improved
327 (same algorithm as EV).
328 - update unicode idna mapping table.
329
3305.24 Tue Jan 5 11:39:43 CET 2010
331 - cygwin never reports errors from failed connects. 1.5 just gives
332 you continous readyness and EAGAIN, 1.7 is even more broken and
333 just hangs. work around both issues in a cygwin-specific hack.
334 - improve idle watcher documentation slightly.
335
3365.23 Sun Dec 20 23:48:00 CET 2009
337 - support IDNs in resolve_sockaddr, and therefore in tcp_connect.
338 - implement punycode_encode/decode, idn_nameprep,
339 idn_to_ascii and idn_to_unicode operations in AnyEvent::Util.
340 - provide $AE::VERSION.
341 - removed traces of "no strict 'refs'".
342
3435.22 Sat Dec 5 03:51:13 CET 2009
344 - downgrade-or-fail in AnyEvent::Handle::push_write, to
345 diagnose encoding failures earlier and more succinctly.
346 (this works around bugs in perl, throwing away encoding info
347 when passing scalar data to extensions).
348 - add more examples to AnyEvent::Socket manpage.
349 - upgrade internal warning set to the same as common::sense 2.03.
350 - use pack "n/a*" for pre-5.8.9 perl compatibility in AnyEvent::DNS
351 (John Beppu).
352 - AnyEvent::Socket::inet_aton now properly supports ipv6, as documented.
353 - add google public dns servers to fallback server set and make sure
354 we load-balance properly between the three sets. also add all
355 fallback dns servers, not just a random one, to each dns config.
356
3575.21 Thu Nov 19 02:48:47 CET 2009
358 - fix a problem where socket constants were called with parameters
359 (spotted by David Friedland).
360 - fork_call never use'd POSIX (reported by Daisuke Maki).
361 - improve perl 5.6 compatibility further (but it still won't work
362 unless you rip out everything but the core).
363 - prefer Net::DNS::Resolver over ipconfig, if installed, on win32.
364 uses 10MB of RAM, but doesn't open a console window. *sigh*.
365
3665.202 Wed Oct 14 22:35:44 CEST 2009
367 - AnyEvent::DNS would unexpectedly clobber $_ under windows
368 (analysed by Matthias Waldorf).
369 - AnyEvent::Handle::run_cmd can now pass the PID of the
370 newly-created process, which is much less useful than it might
371 sound (based on patch by Yann Kerherve).
372
3735.201 Tue Sep 29 12:09:25 CEST 2009
374 - AnyEvent:Handle::on_starttls/on_stoptls methods were broken
375 (reported by Torsten Foertsch).
376 - common::sense 2.0 could cause tcp_server to throw an exception
377 (analysed by elmex).
378
3795.2 Mon Sep 14 07:04:49 CEST 2009
380 - INCOMPATIBLE CHANGE: do no longer support register_read_type
381 and register_write_type in AnyEvent::Handle, instead support
382 package names (the facility was mostly abused).
383 - implement "packagename-as-read/write type" support in
384 AnyEvent::Handle.
385 - AnyEvent::Handle: new options "keepalive" and "oobinline".
386 - oobinline set by default to avoid security issues.
387 - the pure-perl event loop backend wrongly detected times() fallback
388 support (spotted by Pavel Boldin).
389
3905.12 Tue Sep 1 20:26:50 CEST 2009
391 - be more lenient when parsing resolv.conf files, as some people
392 use hashmarks as comment indicator inside directives (reported
393 by Michael S. Fischer).
394 - use same set of warnings as common::sense 2.0.
395 - fix a potential 32 bit overflow issue due to perl having problems
396 with large hex constants in the Perl backend.
397 - do not use the slower 5.6 version fo the enc_name code in AnyEvent::DNS
398 by default.
399 - fix some prototypes.
400 - (optionally) require Async::Interrupt 1.02.
401 - replace opendns fallbacks by verizon/level3 public dns servers.
402
4035.112 Fri Aug 21 13:59:04 CEST 2009
404 - AnyEvent::Handle could cause hard crashes in Net::SSLeay when
405 using starttls with an illegal/uncreatable tls context.
406 - port to broken win32 perls.
407
4085.111 Sun Aug 16 18:44:36 CEST 2009
409 - perl errornously requires sockaddr_un structures returned by the
410 kernel to have a specific length. work around this bug by padding
411 sockaddr_un structures (found to fail on OS X, as perl only uses
412 the correct code for GNU/Linux, but likely to fail on other
413 OSes as well).
414
4155.11 Wed Aug 12 17:49:37 CEST 2009
416 - fixed a bug in run_cmd with input from filename.
417 - create a json coder object when none is given in write direction
418 to reduce dependency on JSON.
419
4205.1 Tue Aug 11 03:17:32 CEST 2009
421 - tcp_connect now reports the correct error in $!, not an unrelated
422 one (was broken since 4.91).
423 - AE::Handle did not properly clear rtimeout/wtimeout watchers
424 when retrying a tcp_connect.
425 - new functions: AnyEvent::Util::run_cmd and close_all_fds_except.
426 - fixed wrong documentation in AE::Socket.
427
4285.01 Mon Aug 10 03:16:32 CEST 2009
429 - last release broke EV child watchers :/.
430
4315.0 Sun Aug 9 17:32:09 CEST 2009
432 - officially introduce and document the AE API.
433 - lowering the timeout at runtime did not have immediate effect
434 in AnyEvent::Handle.
435 - AE::Handle now has separate and independent read and write timeout
436 settings.
437 - probe for Guard module when AnyEvent::Util is loaded, not at
438 runtime (Event doesn't like eval "use" from callbacks, and it avoids
439 exporting the probe function).
440 - Impl::IOAsync's io watchers did not use the filehandle emulation
441 code to work around it's watcher limits.
442 - work around IO::Async's problems with overloaded objects.
443 - try to disable wearnings inside IO::Async, it's just too buggy.
444 - do no longer use Event by default, only EV or Perl, Event is too buggy.
445 - hackishly provide fast AE:: interface when EV or Perl is the backend.
446 - inherit VERSION from AnyEvent::VERSION in most modules.
447
4484.91 Thu Aug 6 15:42:45 CEST 2009
449 - AE::Handle::starttls could get out of sync when the read buffer
450 already contains some TLS handshake.
451 - AE::Handle did not properly free the TLS session, causing bigger
452 memory leaks in Net::SSLeay than expected (reported by toaster).
453 - AE::Socket::tcp_connect will now artificially delay invoking
454 the callback to avoid returning after invoking the callback.
455 - convert many internal watcher uses to AE API.
456
4574.9 Sat Aug 1 11:07:01 CEST 2009
458 - Glib silently fails when registering a timeout with
459 a negative value, so avoid that.
460 - call condvar callback immediately when it is set
461 after the condition is already signalled.
462 - check rbuf_max condition only after trying to consume
463 data.
464 - normalise signal numbers to names when using the ae's
465 signal handling, but do not document this (yet).
466 - pure perl signal emulation did not properly set nonblocking
467 mode on the signal pipe on win32 (this is unlikely to have caused
468 any issues).
469 - new module: AnyEvent::Debug.
470 - AnyEvent::Strict now checks that a signal was specified
471 by name (not name or number).
472 - reduce memory footprint in typical cases by ~50kb by
473 only compiling signal code when necessary.
474 - add AnyEvent::Handle->rbuf_max.
475 - grab the AE namespace for future new API, implement stubs for
476 said future/alternative API.
477 - new function: AnyEvent::Socket::format_hostport.
478
4794.881 Tue Jul 28 12:51:53 CEST 2009
480 - work around a bug in local in pre-5.10 perls, causing
481 AnyEvent::Handle to recurse when it should not
482 (analysed by elmex).
483
4844.88 Tue Jul 28 04:04:37 CEST 2009
485 - re-bless the handle into a dummy package after calling
486 AnyEvent::Handle::destroy, so the user does not need to check
487 for errors after every push_write etc.
488 - do not attempt to run t/02_signals.t on obviously broken
489 platforms.
490
4914.87 Sun Jul 26 02:06:16 CEST 2009
492 - do not attempt to linger when there is no longer a valid fh.
493 - enforce tls mode to be either accept or connect, do not simply
494 segfault in Net::SSLeay.
495 - AnyEvent::Handle can now call tcp_connect itself (new parameters
496 connect, on_prepare, on_connect and on_connect_error). Updated
497 tutorial accordingly.
498 - add AnyEvent::Impl::Irssi backend.
499
5004.86 Mon Jul 20 23:52:29 CEST 2009
501 - since the verbose warning is not enough, explicitly document
502 that versions before 1.33 of Net::SSLeay are not secure.
503 - work around signal handling races in Event and (...) Event::Lib.
504 - try to align signal-race timer to full-second boundaries.
505 - work around Tk not liking negative timeouts.
506 - don't complain of different grades of environmental unfriendlyness
507 in IO::Async.
508
5094.85 Sat Jul 18 06:16:14 CEST 2009
510 - nail the signal race problem in perl once and for all
511 (see $AnyEvent::MAX_SIGNAL_LATENCY).
512 - take advantage of Async::Interrupt if it is available.
513 - load Time::HiRes and Guard modules on demand only.
514 - add optional/recommended modules section to AnyEvent
515 documentation.
516 - reduce memory usage considerably (and reduce startup penalty)
517 by not using "strict", "warnings" and "overload" modules.
518 - work around buggy windows/openbsd perls and provide EBADMSG
519 and EPROTO ourselves when missing.
520 - improve perl 5.6 compatibility of the core event loop.
521 - made Net::SSLeay version 1.33 a soft requirement.
522
5234.83 Fri Jul 17 16:56:26 CEST 2009
524 - implement AnyEvent::Socket::getprotobyname.
525 - AnyEvent::CondVar's will now detect recursive blocking
526 waits and will croak, as too many people fall into
527 this trap.
528 - AnyEvent::Handle will now call ->destroy on itself after
529 executing the on_error callback, instead of doing some
530 half-baked internal shutdown, for fatal errors.
531 - clarify on_eof behaviour w.r.t. the read queue and
532 on_read callbacks.
533 - ignore some possible spurious wake-ups in tcp_connect.
534
5354.82 Sat Jul 11 00:34:55 CEST 2009
536 - POE and Event backends didn't accept some callable objects as
537 callbacks.
538 - use Config module instead of POSIX module to detect signal names
539 in AnyEvent::Strict and AnyEvent::Impl::EventLib, as the POSIX
540 module doesn't even have all POSIX signals :/.
541 - use more workarounds around the many refcnt/corruption bugs in
542 Event::Lib.
543 - work around a race condition in perl's select, causing t/03_child.t
544 to rarely fail.
545
5464.81 Thu Jul 9 10:30:30 CEST 2009
547 - AnyEvent::Handle didn't properly diagnose write errors
548 (it expected -1 from syswrite, how lame... :).
549 - support file descriptors in addition to file handles
550 in AnyEvent->io.
551 - new env variables: PERL_ANYEVENT_RESOLV_CONF,
552 PERL_ANYEVENT_MAX_OUTSTANDING_DNS, PERL_ANYEVENT_CA_FILE
553 and PERL_ANYEVENT_CA_PATH.
554 - provide a sensible synopsis section for AnyEvent::TLS.
555 - add a "supported backends" section to the manpage.
556 - added simple io watcher test to testsuite, using a
557 portable_socketpair.
558 - tried to improve the stability of the Event::Lib backend,
559 YMMV.
560
5614.8 Mon Jul 6 23:45:16 CEST 2009
562 - AnyEvent::DNS did not properly follow CNAME records with
563 uppercase targets.
564 - AnyEvent::DNS would errornously return AAAA records
565 with v4 mapped addresses (a faulty record) as ipv4 addresses,
566 causing AnyEvent::Socket to throw an exception.
567 - added new module AnyEvent::TLS for easier SSL/TLS context
568 creation, with many options including hostname verification,
569 secure default configuration, lots of documentation and,
570 predefined diffie-hellman keys for perfect forward security
571 and much more. get it while it's still fresh!
572 - use AnyEvent::TLS in AnyEvent::Handle for context management.
573 - load AnyEvent::Handle only on demand in AnyEvent::DNS,
574 so AnyEvent::Socket users have smaller memory footprint
575 in the common case.
576 - add AnyEvent::Handle->push_shutdown method.
577 - add an additional error message parameter to AnyEvent::Handle's
578 on_error callback (for TLS, $! is still available).
579 - add AnyEvent::Handle on_starttls/on_stoptls callbacks.
580 - make AnyEvent::Handle more robust against early conenction
581 failures (during new), and return C<undef> in that case
582 from the constructor.
583 - AnyEvent::Handle will now try to load only JSON::XS first,
584 then fall back to JSON.
585 - format_ipv4/format_ipv6 are now exported by default, for symmetry,
586 and because it was documented that way.
587
5884.451 Fri Jul 3 00:28:58 CEST 2009
589 - do not clear rbuf when shutting down an AnyEvent::Handle
590 object - doing so breaks AnyEvent::HTTP.
591
5924.45 Mon Jun 29 22:59:26 CEST 2009
593 - a write error could cause AnyEvent::Handle to create
594 an I/O watcher with an undefined $fh.
595 - special-case mapped ipv4 addresses in both
596 AnyEvent::Socket::format_address and parse_address,
597 to treat them just like ipv4 addresses.
598 - updated and overhauled the AnyEvent::Intro doc.
599 - implement AnyEvent::Socket::format_ipv4/ipv6.
600 - slightly speed up portable_pipe/socketpair.
601 - expand condvar begin/end documentation.
602
6034.42 Fri Jun 26 08:32:18 CEST 2009
604 - preliminary and neccesarily incomplete support for IO::Async.
605 - reset SIGCHLD to DEFAULT when AnyEvent is loaded, in
606 case it was set to IGNORE, to ensure we can catch child statuses
607 even when the calling env acts stupidly.
608 - updated benchmarks with IO::Async, which performs very well.
609
6104.412 Wed Jun 24 01:35:57 CEST 2009
611 - support an "untaint" attribute for AnyEvent::DNS
612 and set it on the default resolver.
613 - implement and document AnyEvent::Impl::Perl::loop.
614 - remove all anyevent-env variables from %ENV
615 when running in tainted mode.
616 - mention and extend the IO::Lambda benchmark.
617
6184.411 Sun Jun 7 18:48:02 CEST 2009
619 - do not try to use F_SETFD on windows.
620
6214.41 Thu May 14 06:40:11 CEST 2009
622 - work around issues in older perls (5.8.0?) when
623 a signal handler is deleted from the %SIG hash.
624 - use POSIX::_exit in child test, to avoid running
625 destructors.
626 - speed up CHLD handling by relying on SIGCHLD being
627 synchronously delivered, even when we roll our own
628 implementation.
629 - AnyEvent::DNS: add the "dname" resource record name
630 for cosmetic reasons.
631
6324.4 Sun Apr 26 20:12:33 CEST 2009
633 - implemented idle watchers, where applicable.
634 - AnyEvent->time died when Event backend was in use.
635 - fix a memleak in the Tk backend.
636 - sped up Tk timer handling.
637 - clip negative "after" values to 0 in AnyEvent::Impl::Event
638 to avoid spamming out warning messages.
639 - fix Qt timers without interval.
640 - avoid Qt zero-timeout special case, in old libqt's.
641
6424.352 Mon Apr 20 16:31:11 CEST 2009
643 - fix AnyEvent::Strict error messages for child watchers.
644 - fix/update Makefile.PL "recommends" versions.
645 - add AnyEvent->now_update.
646
6474.351 Sat Apr 11 07:56:14 CEST 2009
648 - actually make the signal pipe work under win32.
649 - localise $! in signal handler to avoid changing $!, although
650 perl itself does not seem to save/restore errno either.
651 - set the cloexec flag on the signal pipe (normally set by perl too).
652
6534.35 Fri Mar 27 11:48:20 CET 2009
654 - event models relying on AnyEvent's signal watcher emulation did
655 invoke the callback asynchronously, contradicting documentation
656 and causing signals to get lost (this includes AnyEvent's own
657 event loop). AnyEvent now uses the standard pipe trick to make
658 callback execution synchronous to the event loop.
659 - AnyEvent::Handle didn't free TLS context data on DESTROY
660 (patch by Pavel Shaydo).
661 - work around the ever-incompatibly-changing API of MakeMaker.
662 - document that changing global variables without restoring them
663 is a bad idea in Perl (noted by Adam Rosenstein).
664 - AnyEvent::Strict now barfs if ->io is passed a file.
665
6664.34 Thu Feb 12 18:32:45 CET 2009
667 - separately buffer TLS read data, as otherwise the read queue could
668 deadlock as receiving data is not expected while draining the
669 read queue (which cna only happen with TLS).
670 - raise EBADMSG error on JSON decoding errors.
671 - fix some minor manpage bugs (reported by Maximilian Gaß).
672 - speed up select bitmask parsing quite a bit in the pure perl
673 backend.
674 - use CORE::select instead of just select in AnyEvent::Impl::Perl.
675
6764.331 Tue Jan 6 21:07:25 CET 2009
677 - socketpair fails on many vista machines because vista has
678 completely broken accept/getpeername and getsockname functions,
679 so we provide our own socketpair emulation that kind of works
680 (AnyEvent::Util::portable_pipe).
681 - new function: AnyEvent::Util::portable_socketpair.
682 - take advantage of the Guard module if it exists.
683
6844.33 Fri Nov 21 02:35:40 CET 2008
685 - AnyEvent::Strict did errornously flag a fileno of 0 as
686 illegal.
687 - reduce memory usage and slightly speed up the pure perl backend
688 by only storing the file descriptor, not the file handle.
689 - add missing autocork method to AnyEvent::Handle (reported
690 by Adam Rosenstein).
691 - AnyEvent::DNS->resolve errornously documented an $rcode result
692 argument, but there isn't (spotted by Henrik Krohns).
693 - the naked truth about Tk - it's basically unmaintained.
694
6954.32 Mon Nov 3 22:46:32 CET 2008
696 - fix AnyEvent::Socket::resolve_sockaddr to properly support
697 unix sockets again and choose a proper default for
698 the socket type.
699 - fix call to resolve_sockaddr in tcp_connect.
700
7014.31 Thu Oct 30 04:41:48 CET 2008
3 - implemented AnyEvent::Handle->destroy method. 702 - implemented AnyEvent::Handle->destroy method.
4 - hint about unexpected effects in tls mode. 703 - hint about unexpected effects in TLS mode.
5 - speed up AnyEvent::Impl::Perl by using more arrays 704 - speed up AnyEvent::Impl::Perl by using more arrays
6 instead of hashes without the slightest loss of readability :) 705 instead of hashes without the slightest loss of readability :)
706 - work around a perl argument refcounting bug.
707 - with some perl patches applied, AnyEvent now seems leak-free.
7 708
84.3 Fri Oct 3 09:18:43 CEST 2008 7094.3 Fri Oct 3 09:18:43 CEST 2008
9 - AnyEvent will now install a no-op signal handler for SIGPIPE, 710 - AnyEvent will now install a no-op signal handler for SIGPIPE,
10 unless one has been installed already. 711 unless one has been installed already.
11 - warn about not loaidng AnyEvent::Impl::POE early enough 712 - warn about not loaidng AnyEvent::Impl::POE early enough
12 (patch by Adam Rosenstein). 713 (patch by Adam Rosenstein).
13 - fixed a great number of bugs and corner cases in AnyEvent::Handle: 714 - fixed a great number of bugs and corner cases in AnyEvent::Handle.
14 - fix a bug in where in SSL connect mode, the 715 - fix a bug in where in SSL connect mode, the
15 client would first wait for some data by the server and 716 client would first wait for some data by the server and
16 otherwise hang. (reported and analyzed in an absolutely 717 otherwise hang. (reported and analysed in an absolutely
17 exemplary manner by Adam Rosenstein). 718 exemplary manner by Adam Rosenstein).
18 - fix a bug in where SSL EOF would not be treated 719 - fix a bug in where SSL EOF would not be treated
19 as stream EOF, putting the connection into a hung state. 720 as stream EOF, putting the connection into a hung state.
20 (reported and analyzed in an absolutely exemplary manner by Adam 721 (reported and analysed in an absolutely exemplary manner by Adam
21 Rosenstein). 722 Rosenstein).
22 - fix a potential segfault when the TLS context 723 - fix a potential segfault when the TLS context
23 would go missing in a read callback (e.g. due to stoptls) - 724 would go missing in a read callback (e.g. due to stoptls) -
24 Net::SSLeay of course makes no type checking whatsoever. 725 Net::SSLeay of course makes no type checking whatsoever.
25 - AnyEvent::Handle will not stop the read watcher in TLS mode. 726 - AnyEvent::Handle will not stop the read watcher in TLS mode.

Diff Legend

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