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

Comparing AnyEvent/Changes (file contents):
Revision 1.242 by root, Mon Jun 22 11:57:05 2009 UTC vs.
Revision 1.565 by root, Mon Jan 14 17:57:09 2013 UTC

1Revision history for Perl extension AnyEvent. 1Revision history for Perl extension AnyEvent.
2 2
3TODO: document ->loop, and maybe one_event. 3TODO: document TCP_*** constants
4TODO: iolae.pl 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: maybe always try to downgrade in push_write.
11 - AnyEvent::Debug shell can now run coro shell commands, if
12 available.
13 - t/63* tests were wrongly in MANIFEST.
14
157.04 Thu Nov 15 02:23:58 CET 2012
16 - debugging code left in.
17
187.03 Thu Nov 15 02:16:11 CET 2012
19 - AnyEvent::Socket::inet_aton did not work when dns resolution was
20 used to find the addresses (analyzed and patch by Maxime Soulé).
21 - fix a memory leak in the /etc/hosts lookup code when hosts don't
22 resolve and are not in hosts (analyzed by sten).
23
247.02 Tue Aug 14 04:27:58 CEST 2012
25 - AnyEvent::Util::run_cmd could block indefinitely (analyzed and test
26 program by Yorhel).
27 - verified that AnyEvent::Socket follows RFC5952.
28 - try to parse "ADDR#PORT" in addition to "ADDR PORT".
29
307.01 Sun May 13 01:03:17 CEST 2012
31 - fail with EPROTO in AnyEvent::Handle wqhen TLS is requested but
32 not available, instead of throwing an exception.
33 - use File::Spec to get the tmpdir in t/*, to avoid needless
34 failures on (most, not mine :) windows boxes.
35 - new handle read types: tls_detect and tls_autostart.
36
377.0 Fri Apr 13 06:33:30 CEST 2012
38 - child watchers are broken in POE 1.352 (also many earlier
39 versions) and there seems to be no way to work around it, as POE
40 itself is inherently racy. Document this regression and add a delay
41 in t/68_poe_03_child.t for the time being.
42 - new module AnyEvent::IO, that is a frontend to either a pure-perl
43 synchronous I/O implementation (AnyEvent::IO::Perl), or to the
44 asynchronous I/O extension IO::AIO (AnyEvent::IO::IOAIO).
45 - load /etc/hosts only when DNS has no answer.
46 - stat /etc/hosts on every access and reload it if it changed.
47 - load /etc/hosts via AnyEvent::IO - potentially asynchronous.
48 - fix a buggy croak in the dh parameter reading in AnyEvent::TLS.
49 - AnyEvent::Log log_to_file and log_to_path now use AnyEvent::IO. As
50 a side effect, they now use true appending as opposed to libc
51 appending, and the file might not have been opened when the function
52 returns.
53 - the default logging level was not properly documented in a variety of
54 places, this has hopefully been rectified.
55 - updated uts46data.pl for unicode 6.1.0.
56 - made log messages generated by AnyEvent submodules not
57 include the package name anymore, as it will be logged
58 by default already.
59 - upgrade to the trick used by common-sense 3.5 to work
60 around extra warning torture/breakage under perl 5.15.x.
61 - log messages by anyevent are now ucfirst, are usually full
62 sentences and do no longer include the package name.
63 - the storable read type would sometimes throw an exception instead
64 of causing EBADMSG (analyzed by Maxime Soulé).
65 - work around a bug in openssl 1.0.1 which enforces a minimum rsa
66 keysize (reported by Paul Howarth).
67 - documented Rocco fabricating statements about AnyEvent and me.
68
696.14 Tue Jan 31 20:00:24 CET 2012
70 - AnyEvent::Impl::Tk was broken due to a mysterious "die" inside,
71 probably an editing mistake (reported by Darin McBride).
72
736.13 Thu Jan 12 07:27:01 CET 2012
74 - AnyEvent::Util::fork_call checked for POSIX availability in the wrong way,
75 causing it to fail if POSIX wasn't loaded already (analysed by Rock Power).
76 - AnyEvent::Handle::push_read (line => did pass $1 directly, so
77 regex-matching inside the callback would change the parameter, despite
78 $1 being dynamically scoped per-block. this perl bug is now being
79 worked around (testcase by Cindy Wang).
80
816.12 Mon Dec 12 13:21:10 CET 2011
82 - $! was clobbered by subsequent calls in tcp_connect, due to
83 the postpone not saving/restoring it (analyzed by Richard Garnish).
84
856.11 Tue Nov 22 10:36:05 CET 2011
86 - Tk cannot create windows when tainted unless you set a title, so
87 set a dummy title for AnyEvent's dummy mainwindow. How dummy.
88 - escape any nonprintable/nonascii characters when stringifying
89 backtraces.
90 - log the reason loading an interface module fails to load at level debug.
91
926.1 Tue Oct 4 19:44:30 CEST 2011
93 - INCOMPATIBLE CHANGE: the default log level is now "4" (error
94 and above), and some messages inside AnyEvent have been elevated
95 to higher log levels to print by default.
96 - AnyEvent::log, unlike AnyEvent::Log::log, did not always exit on fatal.
97 - work around yet another signal race bug, newly introduced by POE. My
98 how that module sucks.
99 - add google public ipv6 dns as fallback servers, so AE::DNS can finally
100 support ipv6-only operation even with defaults (also use ::1 instead
101 of 127.0.0.1).
102 - give better guidelines on how to chose a logging level.
103 - AnyEvent::Log can now cap (limit) log message priority by context.
104 - AE::log will now save and restore $! and $@, so callers don't have to.
105 - AE::Strict now checks that registered fds are still valid at regular
106 intervals.
107 - skip frequently called 'require' calls if the module appears loaded.
108 - increase signal latency test timer considerably for poor little osx,
109 which can't be bothered with basic posix functionality such as signal
110 delivery.
111 - mention some debugging aids in the tutorial.
112 - new env variable PERL_ANYEVENT_MAX_SIGNAL_LATENCY.
113 - private (%namespace) log contexts in PERL_ANYEVENT_LOG now have the name
114 of the context as title.
115 - AnyEvent::Strict would not restore read-write access to $_ when a callback
116 died - this is a user bug, but since AE::Strict exists to help dbeugging...
117 - detect the extremely evil IO::Async::Loop::AnyEvent module and refuse
118 to work with it - the author is well aware of the issues but refuses
119 to fix them. This extreme step has been taken because the module actively
120 attacks the goals of AnyEvent and makes it harder for module authors to use
121 AnyEvent.
122
1236.02 Fri Aug 26 20:08:31 CEST 2011
124 - log_to_syslog now accepts facility strings.
125 - log_to_syslog would not allow facilities - "$facility|$level"
126 *literally* means that string. sheesh.
127 - the CPAN parser got confused and thought we provide Sys::Syslog, try
128 to work around.
129 - renamed AnyEvent::Impl::FLTK2 to FLTK.
130
1316.01 Fri Aug 26 07:04:11 CEST 2011
132 - INCOMPATIBLE CHANGE: AnyEvent::DNS resource records now include
133 the TTL as fourth element - this affects ->resolve, ->request
134 and ->dns_unpack indirectly. Convenience functions such as
135 AnyEvent::DNS::{a,txt,mx,srv} are not affected, neither is
136 AnyEvent::Socket.
137 - INCOMPATIBLE CHANGE: backend authors now should not implement
138 one_event or loop, but instead the AnyEvent::CondVar::_wait and _poll
139 methods.
140 - INCOMPATIBLE CHANGE: verbosity levels have been redefined, see
141 AnyEvent (PERL_ANYEVENT_VERBOSE) and AnyEvent::Log for details.
142 - DEPRECIATION: please change
143 "use AnyEvent::Impl::Perl" to "use AnyEvent::Loop" and
144 "AnyEvent::Impl::Perl::loop" to "AnyEvent::Loop::run"
145 in your programs, for future compatibility.
146 - *blush* idle watcher emulation was borked.
147 - AnyEvent::Strict now checks the AE::xxx functions as well.
148 - querying too long domain names would barf deep inside AnyEvent::DNS - since
149 this is hard to avoid for applications, AnyEvent:.DNS::request now checks for
150 for too long domain names explicitly and passes an undef to the callback
151 instead (reported by James Bromberger).
152 - make AnyEvent::Util more robust against EINTR (analyzed by Andrew Suffield).
153 - new functions: AnyEvent::postpone, AnyEvent::log.
154 - new module: AnyEvent::Log, for simple logging needs.
155 - new env variable, PERL_ANYEVENT_LOG, for fine-grained logging config.
156 - new env variable, PERL_ANYEVENT_DEBUG_SHELL, to automatically start
157 a debug shell.
158 - new env variable, PERL_ANYEVENT_DEBUG_WRAP, to automatically enable
159 watcher instrumentation/wrapping (see AnyEvent::Debug::wrap).
160 - new env variable, PERL_ANYEVENT_HOSTS, to override the place where
161 /etc/hosts can be found.
162 - an empty PERL_ANYEVENT_RESOLV_CONF now selects the default
163 configuration.
164 - env variables can now be specified using AE_xyz instead of
165 PERL_ANYEVENT_xyz, subject to some rules.
166 - lots of new AnyEvent::Debug functionality.
167 - AnyEvent::Strict now checks for modifications of $_ by the callback
168 (by setting it to readonly while executing the callback, if possible).
169 - IO::Async sometimes cancels the wrong timer when you call cancel_timer.
170 Invest extra overhead to work around this peculiar behaviour.
171 - split the pure perl backend into a loop module and an impl
172 module (AnyEvent::Loop and AnyEvent::Impl::Perl).
173 - some read types (regex, netstring, json, storable) could
174 errornously keep an AE::Handle object alive on their own
175 (reported by Mohammad Toossi).
176 - AnyEvent::Socket::resolve_sockaddr and all functions using it now
177 supports /etc/hosts overrides.
178 - add more workarounds around child watcher bugs in IO::Async
179 and (newly introduced) in POE.
180 - use glib child watchers and try to work around its limitations,
181 if possible (based on a patch by Kevin Ryde).
182 - add support for FLTK (fltk2), via AnyEvent::Impl::FLTK2.
183 - many read types in AnyEvent::Handle would malfunction during
184 transitions to TLS because of a "delete $self->{rbuf}".
185 (reported by Gerald Galster).
186 - improve condvar blocking wait performance (probably only noticable
187 with faster backends), also make it easier for event loops that
188 do not support blocking waits.
189 - PERL_ANYEVENT_MODEL now also allows full module names, although
190 the usefulness of this feature is not yet clear.
191 - add tests for all supported event loops, conditional on PERL_ANYEVENT_LOOP_TESTS.
192 - add t/80_ssltext.t.
193 - no longer use AUTOLOAD for the initial loading (for no very good reason).
194 - tcp_server now tries to unlink unix domain sockets when it gets destroyed.
195 - do not run event loop detection multiple times when programs call their cached
196 copies of AnyEvent::detect.
197 - parse_hostport now accepts absolute paths als unix domain sockets.
198 - suppress the idiotic warning inside POE by patching the POE::Kernel
199 run flags instead of asking the user to jump through contortions.
200 - add Task::Weaken as dependency (as recommends only at this time).
201 - replace PF_UNSPEC by 0, for good old minix.
202
2036.0 (not officially released, but was on CPAN for about 15 minutes :/)
204
2055.34 Thu May 12 09:59:41 CEST 2011
206 - AE::TLS didn't support both cert_file and key at the same time,
207 due to some copy&paste bug (reported and analyzed by vti).
208 - improve AE::Handle timeout documentation slightly.
209
2105.33 Mon Feb 7 21:51:41 CET 2011
211 - never appeared on CPAN, for unknown reasons.
212 - AnyEvent::Handle on_connect_error - do not try to destroy a handle
213 object when it's already gone (analyzed by Ryan Bullock).
214 - update idna algorithm(s) to UTS#46 version 6.0.0.
215 - preserve trailing dot in idn_to_ascii - testsuite says to preserve,
216 uts46 says it's an error, and testsuite makes more sense.
217 - remove leading dots in idn_to_ascii - testsuite says to remove,
218 uts46 says it's an error, and testsuite makes more sense.
219 - fixed a bug in idn ignored character processing.
220 - passes the uts46 6.0.0 testsuite except for three cases which
221 apparently are bugs in the testsuite itself.
222 - passes the uts46 6.0.1 testsuite except for eleven cases which
223 apparently are bugs in the testsuite itself.
224
2255.31 Sun Jan 23 11:49:19 CET 2011
226 - work around a POE bug where POE would first poll for new events and
227 invoke handlers BEFORE handling the message queue.
228 - fix parsing of "domain" statement in resolv.conf (Johannes Plunien).
229 - stop_read now does just that in TLS mode. see its
230 documentation for caveats.
231 - update to common::sense 3.4 warning set (some warnings in perl impose
232 an arbitrary 32kb limit on regex matches, which affects
233 AnyEvent::Handle).
234 - AE::Handle had a broken rbuf_max method and missing wbuf_max method.
235 - new experimental AE::Handle::resettls method.
236
2375.3 Fri Dec 31 05:48:13 CET 2010
238 - major IOAsync patch by Paul Evans - 0.33 is required, as the major
239 issues with IO::Async have been solved in that version, including
240 the availabilty of a default loop!
241 - implement new wbuf_max parameter for AnyEvent::Handle.
242 - added Cocoa::EventLoop backend by Daisuke Murase.
243 - the default testsuite now honors PERL_ANYEVENT_MODEL (Paul Evans).
244 - clarify handle on_prepare documentation.
245 - set CLOEXEC flag on parent-side fd's in AnyEvent::Util::run_cmd,
246 to avoid leaking them to any child processes.
247 - try to force files into the arch-dependend lib directory, for the
248 half-"turly obsessive compulsive person"(s) that depend on it.
249
2505.29 Sun Dec 5 10:49:21 CET 2010
251 - convert EV backend to EV 4.00 API (so better upgrade EV too).
252 - AE::Handle: implement dynamic read_size adjustment, add
253 max_read_size parameter, reduce default read_size to 2048.
254 - add some FAQ entries for servers and/or on_eof vs. on_error.
255 - work around OS bugs (cygwin again) when lingering in AE::Handle.
256 - reduce memory usage slightly when connect was used in AE::Handle.
257 - wrok around more segfaults in Tk.
258 - document tls_ctx => undef as valid.
259 - detect WNOHANG value at compiletime.
260 - include some probably linux-only support for building AnyEvent
261 as part of the perl core.
262 - improve compatibility of t/08_idna.t to perl 5.8.
263 - make AnyEvent::Strict truly optional (in case somebody builds a
264 minimal perl...).
265
2665.28 Wed Oct 13 04:14:23 CEST 2010
267 - due to a glitch, AnyEvent's internal getprotobyname wasn't
268 actually used in tcp_server.
269 - implement AnyEvent::CondVar->new method, which was documented but
270 missing.
271 - support autoloading of read types in unshift_read.
272 - AnyEvent::Strict: do the test for invalid fh arguments differently,
273 resulting in more informative error messages.
274 - new AnyEvent::FAQ manpage.
275 - clarify that you can actually call push_read etc. in on_read.
276 - put keys and certs in /tmp, not into the current directory,
277 and document this in the SECURITY CONSIDERATIONS sections
278 of AnyEvent::TLS.
279 - support ";" as resolv.conf comment character in AnyEvent::DNS, as
280 requested by Ron Isaacson.
281 - document $AnyEvent::DNS::RESOLVER variable.
282 - incorporated major typo patches by Abhijit Menon-Sen.
283 - AnyEvent::Handle now croaks when negative timeouts are passed.
284 - add a shitload of TCP_xxx sockopt name constants to AnyEvent::Util,
285 undocumented, but who knows when they might come in handy.
286
2875.271 Tue Jun 8 12:05:46 CEST 2010
288 - backport to perl 5.8.x.
289
2905.27 Sun Jun 6 12:12:05 CEST 2010
291 - postpone differently in AnyEvent::Socket now, as
292 when not, canceling the connection attempt might fail
293 (found by Felix Antonius Wilhelm Ostmann).
294 - explicitly check for non-stream sockets in AE::Handle, too many
295 clueless people fell into the trap of this somehow working.
296 - simplified and reworked the "OTHER MODULES" section.
297 - better/more condvar examples.
298
2995.261 Wed Apr 28 16:13:36 CEST 2010
300 - AF_INET6 was not properly used from Socket6 during configuration
301 time (found by Andy Grundman).
302
3035.26 Mon Apr 12 04:49:35 CEST 2010
304 - don't generate spurious readyness notifications when select
305 returns EINTR in the pure perl backend (analysed and testcase
306 by Bas Denissen).
307 - give same examples in SYNOPSIS sections of AnyEvent and AE.
308 - provider faster implementations for AE::time/now/now_update for
309 EV and Perl backends.
310 - provide faster AE::cv and AE::time implementations for all
311 backends.
312 - no longer support lower/mixed-case signal names (this was
313 never documented nor universally supported).
314 - some more memory µ-optimisations, and somewhat less messy
315 code to cope with both AnyEvent and AE APIs.
316
3175.251 Sat Mar 13 00:58:21 CET 2010
318 - make test failed if it was run before make install on perl <=
319 5.8.8, and older versions of perl make our live very difficult,
320 so write constants.pl during Makefile.PL time.
321
3225.25 Sat Mar 13 00:23:14 CET 2010
323 - fix a race condition in AnyEvent::Handle that would cause
324 a "bio_read: unsupported method" error (in _another_
325 ssl connection) after user code threw an exception.
326 - added AnyEvent::Handle->destroyed method.
327 - speed up AnyEvent::Socket::format_address and ::format_ipv6.
328 - the AnyEvent::Util::fh_nonblocking function worked only
329 by lucky accident on win32.
330 - smaller and faster AnyEvent::Util::fh_nonblocking.
331 - when the (required!) Time::HiRes module is missing, AnyEvent
332 did not fall back to built-in time properly.
333 - do not load Fcntl at runtime, saving memory and loading time.
334 - precompile a number of constants and use them instead of runtime
335 detection and eval.
336 - free detection code after detection and similar memory optimisations.
337 - Perl backend timer interval best effort drift has been improved
338 (same algorithm as EV).
339 - update unicode idna mapping table.
340
3415.24 Tue Jan 5 11:39:43 CET 2010
342 - cygwin never reports errors from failed connects. 1.5 just gives
343 you continous readyness and EAGAIN, 1.7 is even more broken and
344 just hangs. work around both issues in a cygwin-specific hack.
345 - improve idle watcher documentation slightly.
346
3475.23 Sun Dec 20 23:48:00 CET 2009
348 - support IDNs in resolve_sockaddr, and therefore in tcp_connect.
349 - implement punycode_encode/decode, idn_nameprep,
350 idn_to_ascii and idn_to_unicode operations in AnyEvent::Util.
351 - provide $AE::VERSION.
352 - removed traces of "no strict 'refs'".
353
3545.22 Sat Dec 5 03:51:13 CET 2009
355 - downgrade-or-fail in AnyEvent::Handle::push_write, to
356 diagnose encoding failures earlier and more succinctly.
357 (this works around bugs in perl, throwing away encoding info
358 when passing scalar data to extensions).
359 - add more examples to AnyEvent::Socket manpage.
360 - upgrade internal warning set to the same as common::sense 2.03.
361 - use pack "n/a*" for pre-5.8.9 perl compatibility in AnyEvent::DNS
362 (John Beppu).
363 - AnyEvent::Socket::inet_aton now properly supports ipv6, as documented.
364 - add google public dns servers to fallback server set and make sure
365 we load-balance properly between the three sets. also add all
366 fallback dns servers, not just a random one, to each dns config.
367
3685.21 Thu Nov 19 02:48:47 CET 2009
369 - fix a problem where socket constants were called with parameters
370 (spotted by David Friedland).
371 - fork_call never use'd POSIX (reported by Daisuke Maki).
372 - improve perl 5.6 compatibility further (but it still won't work
373 unless you rip out everything but the core).
374 - prefer Net::DNS::Resolver over ipconfig, if installed, on win32.
375 uses 10MB of RAM, but doesn't open a console window. *sigh*.
376
3775.202 Wed Oct 14 22:35:44 CEST 2009
378 - AnyEvent::DNS would unexpectedly clobber $_ under windows
379 (analysed by Matthias Waldorf).
380 - AnyEvent::Handle::run_cmd can now pass the PID of the
381 newly-created process, which is much less useful than it might
382 sound (based on patch by Yann Kerherve).
383
3845.201 Tue Sep 29 12:09:25 CEST 2009
385 - AnyEvent:Handle::on_starttls/on_stoptls methods were broken
386 (reported by Torsten Foertsch).
387 - common::sense 2.0 could cause tcp_server to throw an exception
388 (analysed by elmex).
389
3905.2 Mon Sep 14 07:04:49 CEST 2009
391 - INCOMPATIBLE CHANGE: do no longer support register_read_type
392 and register_write_type in AnyEvent::Handle, instead support
393 package names (the facility was mostly abused).
394 - implement "packagename-as-read/write type" support in
395 AnyEvent::Handle.
396 - AnyEvent::Handle: new options "keepalive" and "oobinline".
397 - oobinline set by default to avoid security issues.
398 - the pure-perl event loop backend wrongly detected times() fallback
399 support (spotted by Pavel Boldin).
400
4015.12 Tue Sep 1 20:26:50 CEST 2009
402 - be more lenient when parsing resolv.conf files, as some people
403 use hashmarks as comment indicator inside directives (reported
404 by Michael S. Fischer).
405 - use same set of warnings as common::sense 2.0.
406 - fix a potential 32 bit overflow issue due to perl having problems
407 with large hex constants in the Perl backend.
408 - do not use the slower 5.6 version fo the enc_name code in AnyEvent::DNS
409 by default.
410 - fix some prototypes.
411 - (optionally) require Async::Interrupt 1.02.
412 - replace opendns fallbacks by verizon/level3 public dns servers.
413
4145.112 Fri Aug 21 13:59:04 CEST 2009
415 - AnyEvent::Handle could cause hard crashes in Net::SSLeay when
416 using starttls with an illegal/uncreatable tls context.
417 - port to broken win32 perls.
418
4195.111 Sun Aug 16 18:44:36 CEST 2009
420 - perl errornously requires sockaddr_un structures returned by the
421 kernel to have a specific length. work around this bug by padding
422 sockaddr_un structures (found to fail on OS X, as perl only uses
423 the correct code for GNU/Linux, but likely to fail on other
424 OSes as well).
425
4265.11 Wed Aug 12 17:49:37 CEST 2009
427 - fixed a bug in run_cmd with input from filename.
428 - create a json coder object when none is given in write direction
429 to reduce dependency on JSON.
430
4315.1 Tue Aug 11 03:17:32 CEST 2009
432 - tcp_connect now reports the correct error in $!, not an unrelated
433 one (was broken since 4.91).
434 - AE::Handle did not properly clear rtimeout/wtimeout watchers
435 when retrying a tcp_connect.
436 - new functions: AnyEvent::Util::run_cmd and close_all_fds_except.
437 - fixed wrong documentation in AE::Socket.
438
4395.01 Mon Aug 10 03:16:32 CEST 2009
440 - last release broke EV child watchers :/.
441
4425.0 Sun Aug 9 17:32:09 CEST 2009
443 - officially introduce and document the AE API.
444 - lowering the timeout at runtime did not have immediate effect
445 in AnyEvent::Handle.
446 - AE::Handle now has separate and independent read and write timeout
447 settings.
448 - probe for Guard module when AnyEvent::Util is loaded, not at
449 runtime (Event doesn't like eval "use" from callbacks, and it avoids
450 exporting the probe function).
451 - Impl::IOAsync's io watchers did not use the filehandle emulation
452 code to work around it's watcher limits.
453 - work around IO::Async's problems with overloaded objects.
454 - try to disable wearnings inside IO::Async, it's just too buggy.
455 - do no longer use Event by default, only EV or Perl, Event is too buggy.
456 - hackishly provide fast AE:: interface when EV or Perl is the backend.
457 - inherit VERSION from AnyEvent::VERSION in most modules.
458
4594.91 Thu Aug 6 15:42:45 CEST 2009
460 - AE::Handle::starttls could get out of sync when the read buffer
461 already contains some TLS handshake.
462 - AE::Handle did not properly free the TLS session, causing bigger
463 memory leaks in Net::SSLeay than expected (reported by toaster).
464 - AE::Socket::tcp_connect will now artificially delay invoking
465 the callback to avoid returning after invoking the callback.
466 - convert many internal watcher uses to AE API.
467
4684.9 Sat Aug 1 11:07:01 CEST 2009
469 - Glib silently fails when registering a timeout with
470 a negative value, so avoid that.
471 - call condvar callback immediately when it is set
472 after the condition is already signalled.
473 - check rbuf_max condition only after trying to consume
474 data.
475 - normalise signal numbers to names when using the ae's
476 signal handling, but do not document this (yet).
477 - pure perl signal emulation did not properly set nonblocking
478 mode on the signal pipe on win32 (this is unlikely to have caused
479 any issues).
480 - new module: AnyEvent::Debug.
481 - AnyEvent::Strict now checks that a signal was specified
482 by name (not name or number).
483 - reduce memory footprint in typical cases by ~50kb by
484 only compiling signal code when necessary.
485 - add AnyEvent::Handle->rbuf_max.
486 - grab the AE namespace for future new API, implement stubs for
487 said future/alternative API.
488 - new function: AnyEvent::Socket::format_hostport.
489
4904.881 Tue Jul 28 12:51:53 CEST 2009
491 - work around a bug in local in pre-5.10 perls, causing
492 AnyEvent::Handle to recurse when it should not
493 (analysed by elmex).
494
4954.88 Tue Jul 28 04:04:37 CEST 2009
496 - re-bless the handle into a dummy package after calling
497 AnyEvent::Handle::destroy, so the user does not need to check
498 for errors after every push_write etc.
499 - do not attempt to run t/02_signals.t on obviously broken
500 platforms.
501
5024.87 Sun Jul 26 02:06:16 CEST 2009
503 - do not attempt to linger when there is no longer a valid fh.
504 - enforce tls mode to be either accept or connect, do not simply
505 segfault in Net::SSLeay.
506 - AnyEvent::Handle can now call tcp_connect itself (new parameters
507 connect, on_prepare, on_connect and on_connect_error). Updated
508 tutorial accordingly.
509 - add AnyEvent::Impl::Irssi backend.
510
5114.86 Mon Jul 20 23:52:29 CEST 2009
512 - since the verbose warning is not enough, explicitly document
513 that versions before 1.33 of Net::SSLeay are not secure.
514 - work around signal handling races in Event and (...) Event::Lib.
515 - try to align signal-race timer to full-second boundaries.
516 - work around Tk not liking negative timeouts.
517 - don't complain of different grades of environmental unfriendlyness
518 in IO::Async.
519
5204.85 Sat Jul 18 06:16:14 CEST 2009
521 - nail the signal race problem in perl once and for all
522 (see $AnyEvent::MAX_SIGNAL_LATENCY).
523 - take advantage of Async::Interrupt if it is available.
524 - load Time::HiRes and Guard modules on demand only.
525 - add optional/recommended modules section to AnyEvent
526 documentation.
527 - reduce memory usage considerably (and reduce startup penalty)
528 by not using "strict", "warnings" and "overload" modules.
529 - work around buggy windows/openbsd perls and provide EBADMSG
530 and EPROTO ourselves when missing.
531 - improve perl 5.6 compatibility of the core event loop.
532 - made Net::SSLeay version 1.33 a soft requirement.
533
5344.83 Fri Jul 17 16:56:26 CEST 2009
535 - implement AnyEvent::Socket::getprotobyname.
536 - AnyEvent::CondVar's will now detect recursive blocking
537 waits and will croak, as too many people fall into
538 this trap.
539 - AnyEvent::Handle will now call ->destroy on itself after
540 executing the on_error callback, instead of doing some
541 half-baked internal shutdown, for fatal errors.
542 - clarify on_eof behaviour w.r.t. the read queue and
543 on_read callbacks.
544 - ignore some possible spurious wake-ups in tcp_connect.
545
5464.82 Sat Jul 11 00:34:55 CEST 2009
547 - POE and Event backends didn't accept some callable objects as
548 callbacks.
549 - use Config module instead of POSIX module to detect signal names
550 in AnyEvent::Strict and AnyEvent::Impl::EventLib, as the POSIX
551 module doesn't even have all POSIX signals :/.
552 - use more workarounds around the many refcnt/corruption bugs in
553 Event::Lib.
554 - work around a race condition in perl's select, causing t/03_child.t
555 to rarely fail.
556
5574.81 Thu Jul 9 10:30:30 CEST 2009
558 - AnyEvent::Handle didn't properly diagnose write errors
559 (it expected -1 from syswrite, how lame... :).
560 - support file descriptors in addition to file handles
561 in AnyEvent->io.
562 - new env variables: PERL_ANYEVENT_RESOLV_CONF,
563 PERL_ANYEVENT_MAX_OUTSTANDING_DNS, PERL_ANYEVENT_CA_FILE
564 and PERL_ANYEVENT_CA_PATH.
565 - provide a sensible synopsis section for AnyEvent::TLS.
566 - add a "supported backends" section to the manpage.
567 - added simple io watcher test to testsuite, using a
568 portable_socketpair.
569 - tried to improve the stability of the Event::Lib backend,
570 YMMV.
571
5724.8 Mon Jul 6 23:45:16 CEST 2009
573 - AnyEvent::DNS did not properly follow CNAME records with
574 uppercase targets.
575 - AnyEvent::DNS would errornously return AAAA records
576 with v4 mapped addresses (a faulty record) as ipv4 addresses,
577 causing AnyEvent::Socket to throw an exception.
578 - added new module AnyEvent::TLS for easier SSL/TLS context
579 creation, with many options including hostname verification,
580 secure default configuration, lots of documentation and,
581 predefined diffie-hellman keys for perfect forward security
582 and much more. get it while it's still fresh!
583 - use AnyEvent::TLS in AnyEvent::Handle for context management.
584 - load AnyEvent::Handle only on demand in AnyEvent::DNS,
585 so AnyEvent::Socket users have smaller memory footprint
586 in the common case.
587 - add AnyEvent::Handle->push_shutdown method.
588 - add an additional error message parameter to AnyEvent::Handle's
589 on_error callback (for TLS, $! is still available).
590 - add AnyEvent::Handle on_starttls/on_stoptls callbacks.
591 - make AnyEvent::Handle more robust against early conenction
592 failures (during new), and return C<undef> in that case
593 from the constructor.
594 - AnyEvent::Handle will now try to load only JSON::XS first,
595 then fall back to JSON.
596 - format_ipv4/format_ipv6 are now exported by default, for symmetry,
597 and because it was documented that way.
598
5994.451 Fri Jul 3 00:28:58 CEST 2009
600 - do not clear rbuf when shutting down an AnyEvent::Handle
601 object - doing so breaks AnyEvent::HTTP.
602
6034.45 Mon Jun 29 22:59:26 CEST 2009
604 - a write error could cause AnyEvent::Handle to create
605 an I/O watcher with an undefined $fh.
606 - special-case mapped ipv4 addresses in both
607 AnyEvent::Socket::format_address and parse_address,
608 to treat them just like ipv4 addresses.
609 - updated and overhauled the AnyEvent::Intro doc.
610 - implement AnyEvent::Socket::format_ipv4/ipv6.
611 - slightly speed up portable_pipe/socketpair.
612 - expand condvar begin/end documentation.
613
6144.42 Fri Jun 26 08:32:18 CEST 2009
615 - preliminary and neccesarily incomplete support for IO::Async.
616 - reset SIGCHLD to DEFAULT when AnyEvent is loaded, in
617 case it was set to IGNORE, to ensure we can catch child statuses
618 even when the calling env acts stupidly.
619 - updated benchmarks with IO::Async, which performs very well.
620
6214.412 Wed Jun 24 01:35:57 CEST 2009
5 - support an "untaint" attribute for AnyEvent::DNS 622 - support an "untaint" attribute for AnyEvent::DNS
6 and set it on the default resolver. 623 and set it on the default resolver.
7 - implement and document AnyEvent::Impl::Perl::loop. 624 - implement and document AnyEvent::Impl::Perl::loop.
8 - remove all anyevent-env variables from %ENV 625 - remove all anyevent-env variables from %ENV
9 when running in tainted mode. 626 when running in tainted mode.
627 - mention and extend the IO::Lambda benchmark.
10 628
114.411 Sun Jun 7 18:48:02 CEST 2009 6294.411 Sun Jun 7 18:48:02 CEST 2009
12 - do not try to use F_SETFD on windows. 630 - do not try to use F_SETFD on windows.
13 631
144.41 Thu May 14 06:40:11 CEST 2009 6324.41 Thu May 14 06:40:11 CEST 2009
… …
1024.3 Fri Oct 3 09:18:43 CEST 2008 7204.3 Fri Oct 3 09:18:43 CEST 2008
103 - AnyEvent will now install a no-op signal handler for SIGPIPE, 721 - AnyEvent will now install a no-op signal handler for SIGPIPE,
104 unless one has been installed already. 722 unless one has been installed already.
105 - warn about not loaidng AnyEvent::Impl::POE early enough 723 - warn about not loaidng AnyEvent::Impl::POE early enough
106 (patch by Adam Rosenstein). 724 (patch by Adam Rosenstein).
107 - fixed a great number of bugs and corner cases in AnyEvent::Handle: 725 - fixed a great number of bugs and corner cases in AnyEvent::Handle.
108 - fix a bug in where in SSL connect mode, the 726 - fix a bug in where in SSL connect mode, the
109 client would first wait for some data by the server and 727 client would first wait for some data by the server and
110 otherwise hang. (reported and analyzed in an absolutely 728 otherwise hang. (reported and analysed in an absolutely
111 exemplary manner by Adam Rosenstein). 729 exemplary manner by Adam Rosenstein).
112 - fix a bug in where SSL EOF would not be treated 730 - fix a bug in where SSL EOF would not be treated
113 as stream EOF, putting the connection into a hung state. 731 as stream EOF, putting the connection into a hung state.
114 (reported and analyzed in an absolutely exemplary manner by Adam 732 (reported and analysed in an absolutely exemplary manner by Adam
115 Rosenstein). 733 Rosenstein).
116 - fix a potential segfault when the TLS context 734 - fix a potential segfault when the TLS context
117 would go missing in a read callback (e.g. due to stoptls) - 735 would go missing in a read callback (e.g. due to stoptls) -
118 Net::SSLeay of course makes no type checking whatsoever. 736 Net::SSLeay of course makes no type checking whatsoever.
119 - AnyEvent::Handle will not stop the read watcher in TLS mode. 737 - AnyEvent::Handle will not stop the read watcher in TLS mode.

Diff Legend

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