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

Comparing AnyEvent/Changes (file contents):
Revision 1.133 by root, Thu May 29 22:34:51 2008 UTC vs.
Revision 1.507 by root, Mon Sep 5 07:21:53 2011 UTC

1Revision history for Perl extension AnyEvent. 1Revision history for Perl extension AnyEvent.
2 2
3TODO: document TCP_*** constants
4
5TODO: log message level redistribute?
6TOOD: stop internals abusing
7TODO: AE::logger
8TODO: add debug/trace logging to some modules
9TODO: poe child watcher hang?
10TODO: $autload model kill?
11 - INCOMPATIBLE CHANGE: the default log level is now "4" (error
12 and above), and some messages inside AnyEvent have been elevated
13 to higher log levels to print by default.
14 - add google public ipv6 dns as fallback servers, so AE::DNS can finally
15 support ipv6-only operation even with defaults (also use ::1 instead
16 of 127.0.0.1).
17 - give better guidelines on how to chose a logging level.
18 - AE::Strict now checks that registered fds are still valid at regular
19 intervals.
20 - AnyEvent::log, unlike AnyEvent::Log::log, did not always exit on fatal.
21 - skip frequently called 'require' calls if the module appears loaded.
22 - increase signal latency test timer considerably for poor little osx.
23 - mention some debugging aids in the tutorial.
24
256.02 Fri Aug 26 20:08:31 CEST 2011
26 - log_to_syslog now accepts facility strings.
27 - log_to_syslog would not allow facilities - "$facility|$level"
28 *literally* means that string. sheesh.
29 - the CPAN parser got confused and thought we provide Sys::Syslog, try
30 to work around.
31 - renamed AnyEvent::Impl::FLTK2 to FLTK.
32
336.01 Fri Aug 26 07:04:11 CEST 2011
34 - INCOMPATIBLE CHANGE: AnyEvent::DNS resource records now include
35 the TTL as fourth element - this affects ->resolve, ->request
36 and ->dns_unpack indirectly. Convenience functions such as
37 AnyEvent::DNS::{a,txt,mx,srv} are not affected, neither is
38 AnyEvent::Socket.
39 - INCOMPATIBLE CHANGE: backend authors now should not implement
40 one_event or loop, but instead the AnyEvent::CondVar::_wait and _poll
41 methods.
42 - INCOMPATIBLE CHANGE: verbosity levels have been redefined, see
43 AnyEvent (PERL_ANYEVENT_VERBOSE) and AnyEvent::Log for details.
44 - DEPRECIATION: please change
45 "use AnyEvent::Impl::Perl" to "use AnyEvent::Loop" and
46 "AnyEvent::Impl::Perl::loop" to "AnyEvent::Loop::run"
47 in your programs, for future compatibility.
48 - *blush* idle watcher emulation was borked.
49 - AnyEvent::Strict now checks the AE::xxx functions as well.
50 - querying too long domain names would barf deep inside AnyEvent::DNS - since
51 this is hard to avoid for applications, AnyEvent:.DNS::request now checks for
52 for too long domain names explicitly and passes an undef to the callback
53 instead (reported by James Bromberger).
54 - make AnyEvent::Util more robust against EINTR (analyzed by Andrew Suffield).
55 - new functions: AnyEvent::postpone, AnyEvent::log.
56 - new module: AnyEvent::Log, for simple logging needs.
57 - new env variable, PERL_ANYEVENT_LOG, for fine-grained logging config.
58 - new env variable, PERL_ANYEVENT_DEBUG_SHELL, to automatically start
59 a debug shell.
60 - new env variable, PERL_ANYEVENT_DEBUG_WRAP, to automatically enable
61 watcher instrumentation/wrapping (see AnyEvent::Debug::wrap).
62 - new env variable, PERL_ANYEVENT_HOSTS, to override the place where
63 F</etc/hosts> can be found.
64 - an empty PERL_ANYEVENT_RESOLV_CONF now selects the default
65 configuration.
66 - env variables can now be specified using AE_xyz instead of
67 PERL_ANYEVENT_xyz, subject to some rules.
68 - lots of new AnyEvent::Debug functionality.
69 - AnyEvent::Strict now checks for modifications $_ by the callback
70 (by setting it to readonly while executing the callback, if possible).
71 - IO::Async sometimes cancels the wrong timer when you call cancel_timer.
72 Invest extra overhead to work around this peculiar behaviour.
73 - split the pure perl backend into a loop module and an impl
74 module (AnyEvent::Loop and AnyEvent::Impl::Perl).
75 - some read types (regex, netstring, json, storable) could
76 errornously keep an AE::Handle object alive on their own
77 (reported by Mohammad Toossi).
78 - AnyEvent::Socket::resolve_sockaddr and all functions using it now
79 supports /etc/hosts overrides.
80 - add more workarounds around child watcher bugs in IO::Async
81 and (newly introduced) in POE.
82 - use glib child watchers and try to work around its limitations,
83 if possible (based on a patch by Kevin Ryde).
84 - add support for FLTK (fltk2), via AnyEvent::Impl::FLTK2.
85 - many read types in AnyEvent::Handle would malfunction during
86 transitions to TLS because of a "delete $self->{rbuf}".
87 (reported by Gerald Galster).
88 - improve condvar blocking wait performance (probably only noticable
89 with faster backends), also make it easier for event loops that
90 do not support blocking waits.
91 - PERL_ANYEVENT_MODEL now also allows full module names, although
92 the usefulness of this feature is not yet clear.
93 - add tests for all supported event loops, conditional on PERL_ANYEVENT_LOOP_TESTS.
94 - add t/80_ssltext.t.
95 - no longer use AUTOLOAD for the initial loading (for no very good reason).
96 - tcp_server now tries to unlink unix domain sockets when it gets destroyed.
97 - do not run event loop detection multiple times when programs call their cached
98 copies of AnyEvent::detect.
99 - parse_hostport now accepts absolute paths als unix domain sockets.
100 - suppress the idiotic warning inside POE by patching the POE::Kernel
101 run flags instead of asking the user to jump through contortions.
102 - add Task::Weaken as dependency (as recommends only at this time).
103 - replace PF_UNSPEC by 0, for good old minix.
104
1056.0 (not officially released, but was on CPAN for about 15 minutes :/)
106
1075.34 Thu May 12 09:59:41 CEST 2011
108 - AE::TLS didn't support both cert_file and key at the same time,
109 due to some copy&paste bug (reported and analyzed by vti).
110 - improve AE::Handle timeout documentation slightly.
111
1125.33 Mon Feb 7 21:51:41 CET 2011
113 - never appeared on CPAN, for unknown reasons.
114 - AnyEvent::Handle on_connect_error - do not try to destroy a handle
115 object when it's already gone (analyzed by Ryan Bullock).
116 - update idna algorithm(s) to UTS#46 version 6.0.0.
117 - preserve trailing dot in idn_to_ascii - testsuite says to preserve,
118 uts46 says it's an error, and testsuite makes more sense.
119 - remove leading dots in idn_to_ascii - testsuite says to remove,
120 uts46 says it's an error, and testsuite makes more sense.
121 - fixed a bug in idn ignored character processing.
122 - passes the uts46 6.0.0 testsuite except for three cases which
123 apparently are bugs in the testsuite itself.
124 - passes the uts46 6.0.1 testsuite except for eleven cases which
125 apparently are bugs in the testsuite itself.
126
1275.31 Sun Jan 23 11:49:19 CET 2011
128 - work around a POE bug where POE would first poll for new events and
129 invoke handlers BEFORE handling the message queue.
130 - fix parsing of "domain" statement in resolv.conf (Johannes Plunien).
131 - stop_read now does just that in TLS mode. see its
132 documentation for caveats.
133 - update to common::sense 3.4 warning set (some warnings in perl impose
134 an arbitrary 32kb limit on regex matches, which affects
135 AnyEvent::Handle).
136 - AE::Handle had a broken rbuf_max method and missing wbuf_max method.
137 - new experimental AE::Handle::resettls method.
138
1395.3 Fri Dec 31 05:48:13 CET 2010
140 - major IOAsync patch by Paul Evans - 0.33 is required, as the major
141 issues with IO::Async have been solved in that version, including
142 the availabilty of a default loop!
143 - implement new wbuf_max parameter for AnyEvent::Handle.
144 - added Cocoa::EventLoop backend by Daisuke Murase.
145 - the default testsuite now honors PERL_ANYEVENT_MODEL (Paul Evans).
146 - clarify handle on_prepare documentation.
147 - set CLOEXEC flag on parent-side fd's in AnyEvent::Util::run_cmd,
148 to avoid leaking them to any child processes.
149 - try to force files into the arch-dependend lib directory, for the
150 half-"turly obsessive compulsive person"(s) that depend on it.
151
1525.29 Sun Dec 5 10:49:21 CET 2010
153 - convert EV backend to EV 4.00 API (so better upgrade EV too).
154 - AE::Handle: implement dynamic read_size adjustment, add
155 max_read_size parameter, reduce default read_size to 2048.
156 - add some FAQ entries for servers and/or on_eof vs. on_error.
157 - work around OS bugs (cygwin again) when lingering in AE::Handle.
158 - reduce memory usage slightly when connect was used in AE::Handle.
159 - wrok around more segfaults in Tk.
160 - document tls_ctx => undef as valid.
161 - detect WNOHANG value at compiletime.
162 - include some probably linux-only support for building AnyEvent
163 as part of the perl core.
164 - improve compatibility of t/08_idna.t to perl 5.8.
165 - make AnyEvent::Strict truly optional (in case somebody builds a
166 minimal perl...).
167
1685.28 Wed Oct 13 04:14:23 CEST 2010
169 - due to a glitch, AnyEvent's internal getprotobyname wasn't
170 actually used in tcp_server.
171 - implement AnyEvent::CondVar->new method, which was documented but
172 missing.
173 - support autoloading of read types in unshift_read.
174 - AnyEvent::Strict: do the test for invalid fh arguments differently,
175 resulting in more informative error messages.
176 - new AnyEvent::FAQ manpage.
177 - clarify that you can actually call push_read etc. in on_read.
178 - put keys and certs in /tmp, not into the current directory,
179 and document this in the SECURITY CONSIDERATIONS sections
180 of AnyEvent::TLS.
181 - support ";" as resolv.conf comment character in AnyEvent::DNS, as
182 requested by Ron Isaacson.
183 - document $AnyEvent::DNS::RESOLVER variable.
184 - incorporated major typo patches by Abhijit Menon-Sen.
185 - AnyEvent::Handle now croaks when negative timeouts are passed.
186 - add a shitload of TCP_xxx sockopt name constants to AnyEvent::Util,
187 undocumented, but who knows when they might come in handy.
188
1895.271 Tue Jun 8 12:05:46 CEST 2010
190 - backport to perl 5.8.x.
191
1925.27 Sun Jun 6 12:12:05 CEST 2010
193 - postpone differently in AnyEvent::Socket now, as
194 when not, canceling the connection attempt might fail
195 (found by Felix Antonius Wilhelm Ostmann).
196 - explicitly check for non-stream sockets in AE::Handle, too many
197 clueless people fell into the trap of this somehow working.
198 - simplified and reworked the "OTHER MODULES" section.
199 - better/more condvar examples.
200
2015.261 Wed Apr 28 16:13:36 CEST 2010
202 - AF_INET6 was not properly used from Socket6 during configuration
203 time (found by Andy Grundman).
204
2055.26 Mon Apr 12 04:49:35 CEST 2010
206 - don't generate spurious readyness notifications when select
207 returns EINTR in the pure perl backend (analysed and testcase
208 by Bas Denissen).
209 - give same examples in SYNOPSIS sections of AnyEvent and AE.
210 - provider faster implementations for AE::time/now/now_update for
211 EV and Perl backends.
212 - provide faster AE::cv and AE::time implementations for all
213 backends.
214 - no longer support lower/mixed-case signal names (this was
215 never documented nor universally supported).
216 - some more memory µ-optimisations, and somewhat less messy
217 code to cope with both AnyEvent and AE APIs.
218
2195.251 Sat Mar 13 00:58:21 CET 2010
220 - make test failed if it was run before make install on perl <=
221 5.8.8, and older versions of perl make our live very difficult,
222 so write constants.pl during Makefile.PL time.
223
2245.25 Sat Mar 13 00:23:14 CET 2010
225 - fix a race condition in AnyEvent::Handle that would cause
226 a "bio_read: unsupported method" error (in _another_
227 ssl connection) after user code threw an exception.
228 - added AnyEvent::Handle->destroyed method.
229 - speed up AnyEvent::Socket::format_address and ::format_ipv6.
230 - the AnyEvent::Util::fh_nonblocking function worked only
231 by lucky accident on win32.
232 - smaller and faster AnyEvent::Util::fh_nonblocking.
233 - when the (required!) Time::HiRes module is missing, AnyEvent
234 did not fall back to built-in time properly.
235 - do not load Fcntl at runtime, saving memory and loading time.
236 - precompile a number of constants and use them instead of runtime
237 detection and eval.
238 - free detection code after detection and similar memory optimisations.
239 - Perl backend timer interval best effort drift has been improved
240 (same algorithm as EV).
241 - update unicode idna mapping table.
242
2435.24 Tue Jan 5 11:39:43 CET 2010
244 - cygwin never reports errors from failed connects. 1.5 just gives
245 you continous readyness and EAGAIN, 1.7 is even more broken and
246 just hangs. work around both issues in a cygwin-specific hack.
247 - improve idle watcher documentation slightly.
248
2495.23 Sun Dec 20 23:48:00 CET 2009
250 - support IDNs in resolve_sockaddr, and therefore in tcp_connect.
251 - implement punycode_encode/decode, idn_nameprep,
252 idn_to_ascii and idn_to_unicode operations in AnyEvent::Util.
253 - provide $AE::VERSION.
254 - removed traces of "no strict 'refs'".
255
2565.22 Sat Dec 5 03:51:13 CET 2009
257 - downgrade-or-fail in AnyEvent::Handle::push_write, to
258 diagnose encoding failures earlier and more succinctly.
259 (this works around bugs in perl, throwing away encoding info
260 when passing scalar data to extensions).
261 - add more examples to AnyEvent::Socket manpage.
262 - upgrade internal warning set to the same as common::sense 2.03.
263 - use pack "n/a*" for pre-5.8.9 perl compatibility in AnyEvent::DNS
264 (John Beppu).
265 - AnyEvent::Socket::inet_aton now properly supports ipv6, as documented.
266 - add google public dns servers to fallback server set and make sure
267 we load-balance properly between the three sets. also add all
268 fallback dns servers, not just a random one, to each dns config.
269
2705.21 Thu Nov 19 02:48:47 CET 2009
271 - fix a problem where socket constants were called with parameters
272 (spotted by David Friedland).
273 - fork_call never use'd POSIX (reported by Daisuke Maki).
274 - improve perl 5.6 compatibility further (but it still won't work
275 unless you rip out everything but the core).
276 - prefer Net::DNS::Resolver over ipconfig, if installed, on win32.
277 uses 10MB of RAM, but doesn't open a console window. *sigh*.
278
2795.202 Wed Oct 14 22:35:44 CEST 2009
280 - AnyEvent::DNS would unexpectedly clobber $_ under windows
281 (analysed by Matthias Waldorf).
282 - AnyEvent::Handle::run_cmd can now pass the PID of the
283 newly-created process, which is much less useful than it might
284 sound (based on patch by Yann Kerherve).
285
2865.201 Tue Sep 29 12:09:25 CEST 2009
287 - AnyEvent:Handle::on_starttls/on_stoptls methods were broken
288 (reported by Torsten Foertsch).
289 - common::sense 2.0 could cause tcp_server to throw an exception
290 (analysed by elmex).
291
2925.2 Mon Sep 14 07:04:49 CEST 2009
293 - INCOMPATIBLE CHANGE: do no longer support register_read_type
294 and register_write_type in AnyEvent::Handle, instead support
295 package names (the facility was mostly abused).
296 - implement "packagename-as-read/write type" support in
297 AnyEvent::Handle.
298 - AnyEvent::Handle: new options "keepalive" and "oobinline".
299 - oobinline set by default to avoid security issues.
300 - the pure-perl event loop backend wrongly detected times() fallback
301 support (spotted by Pavel Boldin).
302
3035.12 Tue Sep 1 20:26:50 CEST 2009
304 - be more lenient when parsing resolv.conf files, as some people
305 use hashmarks as comment indicator inside directives (reported
306 by Michael S. Fischer).
307 - use same set of warnings as common::sense 2.0.
308 - fix a potential 32 bit overflow issue due to perl having problems
309 with large hex constants in the Perl backend.
310 - do not use the slower 5.6 version fo the enc_name code in AnyEvent::DNS
311 by default.
312 - fix some prototypes.
313 - (optionally) require Async::Interrupt 1.02.
314 - replace opendns fallbacks by verizon/level3 public dns servers.
315
3165.112 Fri Aug 21 13:59:04 CEST 2009
317 - AnyEvent::Handle could cause hard crashes in Net::SSLeay when
318 using starttls with an illegal/uncreatable tls context.
319 - port to broken win32 perls.
320
3215.111 Sun Aug 16 18:44:36 CEST 2009
322 - perl errornously requires sockaddr_un structures returned by the
323 kernel to have a specific length. work around this bug by padding
324 sockaddr_un structures (found to fail on OS X, as perl only uses
325 the correct code for GNU/Linux, but likely to fail on other
326 OSes as well).
327
3285.11 Wed Aug 12 17:49:37 CEST 2009
329 - fixed a bug in run_cmd with input from filename.
330 - create a json coder object when none is given in write direction
331 to reduce dependency on JSON.
332
3335.1 Tue Aug 11 03:17:32 CEST 2009
334 - tcp_connect now reports the correct error in $!, not an unrelated
335 one (was broken since 4.91).
336 - AE::Handle did not properly clear rtimeout/wtimeout watchers
337 when retrying a tcp_connect.
338 - new functions: AnyEvent::Util::run_cmd and close_all_fds_except.
339 - fixed wrong documentation in AE::Socket.
340
3415.01 Mon Aug 10 03:16:32 CEST 2009
342 - last release broke EV child watchers :/.
343
3445.0 Sun Aug 9 17:32:09 CEST 2009
345 - officially introduce and document the AE API.
346 - lowering the timeout at runtime did not have immediate effect
347 in AnyEvent::Handle.
348 - AE::Handle now has separate and independent read and write timeout
349 settings.
350 - probe for Guard module when AnyEvent::Util is loaded, not at
351 runtime (Event doesn't like eval "use" from callbacks, and it avoids
352 exporting the probe function).
353 - Impl::IOAsync's io watchers did not use the filehandle emulation
354 code to work around it's watcher limits.
355 - work around IO::Async's problems with overloaded objects.
356 - try to disable wearnings inside IO::Async, it's just too buggy.
357 - do no longer use Event by default, only EV or Perl, Event is too buggy.
358 - hackishly provide fast AE:: interface when EV or Perl is the backend.
359 - inherit VERSION from AnyEvent::VERSION in most modules.
360
3614.91 Thu Aug 6 15:42:45 CEST 2009
362 - AE::Handle::starttls could get out of sync when the read buffer
363 already contains some TLS handshake.
364 - AE::Handle did not properly free the TLS session, causing bigger
365 memory leaks in Net::SSLeay than expected (reported by toaster).
366 - AE::Socket::tcp_connect will now artificially delay invoking
367 the callback to avoid returning after invoking the callback.
368 - convert many internal watcher uses to AE API.
369
3704.9 Sat Aug 1 11:07:01 CEST 2009
371 - Glib silently fails when registering a timeout with
372 a negative value, so avoid that.
373 - call condvar callback immediately when it is set
374 after the condition is already signalled.
375 - check rbuf_max condition only after trying to consume
376 data.
377 - normalise signal numbers to names when using the ae's
378 signal handling, but do not document this (yet).
379 - pure perl signal emulation did not properly set nonblocking
380 mode on the signal pipe on win32 (this is unlikely to have caused
381 any issues).
382 - new module: AnyEvent::Debug.
383 - AnyEvent::Strict now checks that a signal was specified
384 by name (not name or number).
385 - reduce memory footprint in typical cases by ~50kb by
386 only compiling signal code when necessary.
387 - add AnyEvent::Handle->rbuf_max.
388 - grab the AE namespace for future new API, implement stubs for
389 said future/alternative API.
390 - new function: AnyEvent::Socket::format_hostport.
391
3924.881 Tue Jul 28 12:51:53 CEST 2009
393 - work around a bug in local in pre-5.10 perls, causing
394 AnyEvent::Handle to recurse when it should not
395 (analysed by elmex).
396
3974.88 Tue Jul 28 04:04:37 CEST 2009
398 - re-bless the handle into a dummy package after calling
399 AnyEvent::Handle::destroy, so the user does not need to check
400 for errors after every push_write etc.
401 - do not attempt to run t/02_signals.t on obviously broken
402 platforms.
403
4044.87 Sun Jul 26 02:06:16 CEST 2009
405 - do not attempt to linger when there is no longer a valid fh.
406 - enforce tls mode to be either accept or connect, do not simply
407 segfault in Net::SSLeay.
408 - AnyEvent::Handle can now call tcp_connect itself (new parameters
409 connect, on_prepare, on_connect and on_connect_error). Updated
410 tutorial accordingly.
411 - add AnyEvent::Impl::Irssi backend.
412
4134.86 Mon Jul 20 23:52:29 CEST 2009
414 - since the verbose warning is not enough, explicitly document
415 that versions before 1.33 of Net::SSLeay are not secure.
416 - work around signal handling races in Event and (...) Event::Lib.
417 - try to align signal-race timer to full-second boundaries.
418 - work around Tk not liking negative timeouts.
419 - don't complain of different grades of environmental unfriendlyness
420 in IO::Async.
421
4224.85 Sat Jul 18 06:16:14 CEST 2009
423 - nail the signal race problem in perl once and for all
424 (see $AnyEvent::MAX_SIGNAL_LATENCY).
425 - take advantage of Async::Interrupt if it is available.
426 - load Time::HiRes and Guard modules on demand only.
427 - add optional/recommended modules section to AnyEvent
428 documentation.
429 - reduce memory usage considerably (and reduce startup penalty)
430 by not using "strict", "warnings" and "overload" modules.
431 - work around buggy windows/openbsd perls and provide EBADMSG
432 and EPROTO ourselves when missing.
433 - improve perl 5.6 compatibility of the core event loop.
434 - made Net::SSLeay version 1.33 a soft requirement.
435
4364.83 Fri Jul 17 16:56:26 CEST 2009
437 - implement AnyEvent::Socket::getprotobyname.
438 - AnyEvent::CondVar's will now detect recursive blocking
439 waits and will croak, as too many people fall into
440 this trap.
441 - AnyEvent::Handle will now call ->destroy on itself after
442 executing the on_error callback, instead of doing some
443 half-baked internal shutdown, for fatal errors.
444 - clarify on_eof behaviour w.r.t. the read queue and
445 on_read callbacks.
446 - ignore some possible spurious wake-ups in tcp_connect.
447
4484.82 Sat Jul 11 00:34:55 CEST 2009
449 - POE and Event backends didn't accept some callable objects as
450 callbacks.
451 - use Config module instead of POSIX module to detect signal names
452 in AnyEvent::Strict and AnyEvent::Impl::EventLib, as the POSIX
453 module doesn't even have all POSIX signals :/.
454 - use more workarounds around the many refcnt/corruption bugs in
455 Event::Lib.
456 - work around a race condition in perl's select, causing t/03_child.t
457 to rarely fail.
458
4594.81 Thu Jul 9 10:30:30 CEST 2009
460 - AnyEvent::Handle didn't properly diagnose write errors
461 (it expected -1 from syswrite, how lame... :).
462 - support file descriptors in addition to file handles
463 in AnyEvent->io.
464 - new env variables: PERL_ANYEVENT_RESOLV_CONF,
465 PERL_ANYEVENT_MAX_OUTSTANDING_DNS, PERL_ANYEVENT_CA_FILE
466 and PERL_ANYEVENT_CA_PATH.
467 - provide a sensible synopsis section for AnyEvent::TLS.
468 - add a "supported backends" section to the manpage.
469 - added simple io watcher test to testsuite, using a
470 portable_socketpair.
471 - tried to improve the stability of the Event::Lib backend,
472 YMMV.
473
4744.8 Mon Jul 6 23:45:16 CEST 2009
475 - AnyEvent::DNS did not properly follow CNAME records with
476 uppercase targets.
477 - AnyEvent::DNS would errornously return AAAA records
478 with v4 mapped addresses (a faulty record) as ipv4 addresses,
479 causing AnyEvent::Socket to throw an exception.
480 - added new module AnyEvent::TLS for easier SSL/TLS context
481 creation, with many options including hostname verification,
482 secure default configuration, lots of documentation and,
483 predefined diffie-hellman keys for perfect forward security
484 and much more. get it while it's still fresh!
485 - use AnyEvent::TLS in AnyEvent::Handle for context management.
486 - load AnyEvent::Handle only on demand in AnyEvent::DNS,
487 so AnyEvent::Socket users have smaller memory footprint
488 in the common case.
489 - add AnyEvent::Handle->push_shutdown method.
490 - add an additional error message parameter to AnyEvent::Handle's
491 on_error callback (for TLS, $! is still available).
492 - add AnyEvent::Handle on_starttls/on_stoptls callbacks.
493 - make AnyEvent::Handle more robust against early conenction
494 failures (during new), and return C<undef> in that case
495 from the constructor.
496 - AnyEvent::Handle will now try to load only JSON::XS first,
497 then fall back to JSON.
498 - format_ipv4/format_ipv6 are now exported by default, for symmetry,
499 and because it was documented that way.
500
5014.451 Fri Jul 3 00:28:58 CEST 2009
502 - do not clear rbuf when shutting down an AnyEvent::Handle
503 object - doing so breaks AnyEvent::HTTP.
504
5054.45 Mon Jun 29 22:59:26 CEST 2009
506 - a write error could cause AnyEvent::Handle to create
507 an I/O watcher with an undefined $fh.
508 - special-case mapped ipv4 addresses in both
509 AnyEvent::Socket::format_address and parse_address,
510 to treat them just like ipv4 addresses.
511 - updated and overhauled the AnyEvent::Intro doc.
512 - implement AnyEvent::Socket::format_ipv4/ipv6.
513 - slightly speed up portable_pipe/socketpair.
514 - expand condvar begin/end documentation.
515
5164.42 Fri Jun 26 08:32:18 CEST 2009
517 - preliminary and neccesarily incomplete support for IO::Async.
518 - reset SIGCHLD to DEFAULT when AnyEvent is loaded, in
519 case it was set to IGNORE, to ensure we can catch child statuses
520 even when the calling env acts stupidly.
521 - updated benchmarks with IO::Async, which performs very well.
522
5234.412 Wed Jun 24 01:35:57 CEST 2009
524 - support an "untaint" attribute for AnyEvent::DNS
525 and set it on the default resolver.
526 - implement and document AnyEvent::Impl::Perl::loop.
527 - remove all anyevent-env variables from %ENV
528 when running in tainted mode.
529 - mention and extend the IO::Lambda benchmark.
530
5314.411 Sun Jun 7 18:48:02 CEST 2009
532 - do not try to use F_SETFD on windows.
533
5344.41 Thu May 14 06:40:11 CEST 2009
535 - work around issues in older perls (5.8.0?) when
536 a signal handler is deleted from the %SIG hash.
537 - use POSIX::_exit in child test, to avoid running
538 destructors.
539 - speed up CHLD handling by relying on SIGCHLD being
540 synchronously delivered, even when we roll our own
541 implementation.
542 - AnyEvent::DNS: add the "dname" resource record name
543 for cosmetic reasons.
544
5454.4 Sun Apr 26 20:12:33 CEST 2009
546 - implemented idle watchers, where applicable.
547 - AnyEvent->time died when Event backend was in use.
548 - fix a memleak in the Tk backend.
549 - sped up Tk timer handling.
550 - clip negative "after" values to 0 in AnyEvent::Impl::Event
551 to avoid spamming out warning messages.
552 - fix Qt timers without interval.
553 - avoid Qt zero-timeout special case, in old libqt's.
554
5554.352 Mon Apr 20 16:31:11 CEST 2009
556 - fix AnyEvent::Strict error messages for child watchers.
557 - fix/update Makefile.PL "recommends" versions.
558 - add AnyEvent->now_update.
559
5604.351 Sat Apr 11 07:56:14 CEST 2009
561 - actually make the signal pipe work under win32.
562 - localise $! in signal handler to avoid changing $!, although
563 perl itself does not seem to save/restore errno either.
564 - set the cloexec flag on the signal pipe (normally set by perl too).
565
5664.35 Fri Mar 27 11:48:20 CET 2009
567 - event models relying on AnyEvent's signal watcher emulation did
568 invoke the callback asynchronously, contradicting documentation
569 and causing signals to get lost (this includes AnyEvent's own
570 event loop). AnyEvent now uses the standard pipe trick to make
571 callback execution synchronous to the event loop.
572 - AnyEvent::Handle didn't free TLS context data on DESTROY
573 (patch by Pavel Shaydo).
574 - work around the ever-incompatibly-changing API of MakeMaker.
575 - document that changing global variables without restoring them
576 is a bad idea in Perl (noted by Adam Rosenstein).
577 - AnyEvent::Strict now barfs if ->io is passed a file.
578
5794.34 Thu Feb 12 18:32:45 CET 2009
580 - separately buffer TLS read data, as otherwise the read queue could
581 deadlock as receiving data is not expected while draining the
582 read queue (which cna only happen with TLS).
583 - raise EBADMSG error on JSON decoding errors.
584 - fix some minor manpage bugs (reported by Maximilian Gaß).
585 - speed up select bitmask parsing quite a bit in the pure perl
586 backend.
587 - use CORE::select instead of just select in AnyEvent::Impl::Perl.
588
5894.331 Tue Jan 6 21:07:25 CET 2009
590 - socketpair fails on many vista machines because vista has
591 completely broken accept/getpeername and getsockname functions,
592 so we provide our own socketpair emulation that kind of works
593 (AnyEvent::Util::portable_pipe).
594 - new function: AnyEvent::Util::portable_socketpair.
595 - take advantage of the Guard module if it exists.
596
5974.33 Fri Nov 21 02:35:40 CET 2008
598 - AnyEvent::Strict did errornously flag a fileno of 0 as
599 illegal.
600 - reduce memory usage and slightly speed up the pure perl backend
601 by only storing the file descriptor, not the file handle.
602 - add missing autocork method to AnyEvent::Handle (reported
603 by Adam Rosenstein).
604 - AnyEvent::DNS->resolve errornously documented an $rcode result
605 argument, but there isn't (spotted by Henrik Krohns).
606 - the naked truth about Tk - it's basically unmaintained.
607
6084.32 Mon Nov 3 22:46:32 CET 2008
609 - fix AnyEvent::Socket::resolve_sockaddr to properly support
610 unix sockets again and choose a proper default for
611 the socket type.
612 - fix call to resolve_sockaddr in tcp_connect.
613
6144.31 Thu Oct 30 04:41:48 CET 2008
615 - implemented AnyEvent::Handle->destroy method.
616 - hint about unexpected effects in TLS mode.
617 - speed up AnyEvent::Impl::Perl by using more arrays
618 instead of hashes without the slightest loss of readability :)
619 - work around a perl argument refcounting bug.
620 - with some perl patches applied, AnyEvent now seems leak-free.
621
6224.3 Fri Oct 3 09:18:43 CEST 2008
623 - AnyEvent will now install a no-op signal handler for SIGPIPE,
624 unless one has been installed already.
625 - warn about not loaidng AnyEvent::Impl::POE early enough
626 (patch by Adam Rosenstein).
627 - fixed a great number of bugs and corner cases in AnyEvent::Handle.
628 - fix a bug in where in SSL connect mode, the
629 client would first wait for some data by the server and
630 otherwise hang. (reported and analysed in an absolutely
631 exemplary manner by Adam Rosenstein).
632 - fix a bug in where SSL EOF would not be treated
633 as stream EOF, putting the connection into a hung state.
634 (reported and analysed in an absolutely exemplary manner by Adam
635 Rosenstein).
636 - fix a potential segfault when the TLS context
637 would go missing in a read callback (e.g. due to stoptls) -
638 Net::SSLeay of course makes no type checking whatsoever.
639 - AnyEvent::Handle will not stop the read watcher in TLS mode.
640 - AnyEvent::Handle->stoptls will send a TLS close notify
641 instead of simply destroying the stream now.
642 - fix a bug where large blocks of data written in TLS mode
643 would not be sent unless triggered by receives.
644 - on_drain will now take the tls write buffer into account.
645 - SSL operations have been streamlined a bit and should be faster
646 now, more to come, though.
647 - added a FAQ section, will document nontrivial issues in there.
648
6494.234 Mon Sep 29 04:08:13 CEST 2008
650 - fix child watcher documentation: only child exits will be reported,
651 no trace events.
652 - mention SIGPIPE in AnyEvent::Handle.
653 - perl backend will now use POSIX::times as monotonic clock
654 when available and the monotonic clock option is not.
655 - run even without Time::HiRes in some cases now, but do not
656 provide sub-second accuracy (all *supported* perl versions have
657 Time::HiRes).
658
6594.233 Fri Aug 22 01:48:05 CEST 2008
660 - fix a bug in the testsuite, causing a hang.
661 - clarified lots of AnyEvent::Handle constructor arguments.
662
6634.232 Thu Aug 21 20:44:25 CEST 2008
664 - fix a bug in AnyEvent::Handle that could cause two fatal error
665 callbacks on EOF, which caused AnyEvent::HTTP to signal success twice
666 (testcase by Leon Brocard).
667 - clarified on_eof/on_error documentation a bit.
668
6694.231 Tue Jul 29 13:12:15 CEST 2008
670 - remove some debugging code left in AnyEvent::Util::fork_call
671 (and no, it's impossible to implement with the broken windows
672 perls without resource leaks or worse).
673
6744.23 Tue Jul 29 12:19:59 CEST 2008
675 - document the first parameter passed to condvar callbacks to be
676 the callback.
677 - add AnyEvent::Socket::{ntoa,aton} aliases.
678 - optimize the AE::Handle->push_read (line) for the default
679 eol marker.
680 - optimize push_read (packstring|storable) for small packets.
681 - invoke on_error callback when no on_eof callback is set.
682 - fix a bug in push_read (storable) of unknown impact.
683
6844.22 Sun Jul 20 16:34:13 CEST 2008
685 - new function AnyEvent::Socket::parse_hostport.
686 - as the bulkheads at microsoft can't even get getprotobyname reliably
687 working on their shitty fucking broken os we need to hardcode
688 some common protocol numbers in AnyEvent::Socket. How can
689 people even bother with such a pile of shit as windows.
690
6914.21 Thu Jul 17 14:40:05 CEST 2008
692 - INCOMPATIBLE CHANGE: AnyEvent::Util::fork_call now has a prototype
693 which is incompatible to previous invocation syntax.
694 - work around more windows perl bugs in fork_call. I cannot imagine
695 why anybody would bother implementing fork in such an obviously
696 unusable way.
697 - avoid calling $do_retry if the request has already finished in
698 AnyEvent::DNS, thus avoiding an exception (reported by Anatoly
699 K. Sharifulin).
700
7014.2 Sat Jul 12 22:42:11 CEST 2008
702 - support an "interval" argument for repeating timers.
703 - fix a bug in the pure perl event loop that caused timers to
704 be delayed under some circumstances.
705 - implement optional argument checking via PERL_ANYEVENT_STRICT.
706 - not importing anything from Event causes it to malfunction
707 and not use Time::HiRes, work around this bug.
708 - more bug workarounds against the endlessly broken Tk module,
709 seems to work now on 32 bit machines, but less so on 64 bit.
710 - minor optimisations applied to most backends.
711
7124.161 Fri Jul 4 14:24:48 CEST 2008
713 - fixed the t/handle/02_write.t, it read after write, with the
714 assumption that all data could be written before read
715 (reported by Jim Razmus).
716
7174.160 Thu Jul 3 04:02:21 CEST 2008
718 - re-registering signal watchers after unregistering one did
719 not work if a backend falls back on AnyEvent's default
720 implementation (also affected child watchers).
721 - new AnyEvent::Handle options: autocork and no_delay.
722
7234.152 Sun Jun 22 14:15:44 CEST 2008
724 - allow for 32-bit perls that implement shifts differently
725 on different architectures in parse_ipv4 (reported and
726 analysed by Keiichi DAIBA).
727
7284.151 Fri Jun 6 17:34:24 CEST 2008
729 - make sure specifying _only_ on_read and never pushing reads
730 works.
731
7324.15 Fri Jun 6 13:00:46 CEST 2008
733 - the pure perl backend would keep some watchers alive when more than
734 one watcher was registered for the same fd.
735 - new "packstring" and "storable" read and write types
736 for AnyEvent::Handle.
737 - allow on_eof handler to be called after on_error with EPIPE returns.
738 - do not immediately call on_read callback in handle constructor.
739
7404.14 Thu Jun 5 20:29:31 CEST 2008
741 - Fixed a bug in DNS SRV priority sorting.
742 - AnyEvent::Util::guard now reports runtime errors while
743 executing the guard block as warnings.
744 - handle 0-byte-reads just before EOF correctly in AnyEvent::Handle.
745
7464.13 Thu Jun 5 00:47:59 CEST 2008
747 - AnyEvent::DNS only followed cname chains with length 2,
748 contrary to documentation. bumped it up to 10, thanks to
749 microsoft, the current broken-dns-config-king.
750 - AnyEvent::DNS didn't check for socket return status
751 but instead relied on perl not creating filehandles in that
752 case - too bad it gives you a bogus file handle (reported
753 and analysed by Vladimir Timofeev).
754 - fix queue management logic in AnyEvent::Handle:
755 when on_read was registered, the queue was empty
756 and no progress could be made AnyEvent::Handle would
757 enter an endless loop.
758 - correctly start reading again when the handle became
759 busy again after idling.
760 - correctly treat tls shut-downs as EOF condition
761 (for the time being, would be nice to have a callback for that).
762 - correctly call eof callback under all conditions
763 (hopefully).
764 - the timeout callback did not expect that $self can go
765 away any time.
766
7674.12 Tue Jun 3 10:58:04 CEST 2008
768 - include AnyEvent::Intro, a tutorial for anyevent,
769 anyevent::socket and anyevent::handle.
770 - allow more options in on_error.
771
7724.11 Fri May 30 23:42:25 CEST 2008
773 - INCOMPATIBLE CHANGE: replace ptr by real PTR lookup, provide
774 reverse_lookup and reverse_verify to replace it, support
775 v4mapped and v4compat addresses.
3 - provide more documentation for the resolver class. 776 - provide more documentation for the resolver class.
4 - really replace longest run of :0: by :: in format_address, 777 - really replace longest run of :0: by :: in format_address,
5 also properly convert :: and ::1 again. 778 also properly convert :: and ::1 again.
6 - support naptr name and decode it 779 - support NAPTR record name and decode it.
7 - implement random weight sampling for SRV records, as per 780 - implement random weight sampling for SRV records, as per
8 rfc 2782. 781 rfc 2782.
9 - replace ptr by real ptr lookup, provide reverse_lookup 782 - correctly abort on srv-record targets of ".".
10 and reverse_verify to replace it, support v4mapped and 783 - added AnyEvent::DNS::wait_for_slot.
11 v4compat addresses. 784 - in the unlikely event of a virtual circuit connection
785 being invalidated by a delayed udp reply, AnyEvent::DNS
786 could die. this has been fixed.
787 - plug a probable memleak in the DNS vc code.
788 - use configured timeout also for VC request phase.
789 - implement timeout and max_outstanding methods
790 for resolver class.
791 - update version numbers in all modules.
12 792
134.1 Thu May 29 05:45:40 CEST 2008 7934.1 Thu May 29 05:45:40 CEST 2008
14 - INCOMPATIBLE CHANGE: renamed xxx_ip functions to xxx_address 794 - INCOMPATIBLE CHANGE: renamed xxx_ip functions to xxx_address
15 which mirrors their purpose better (old names still available 795 which mirrors their purpose better (old names still available
16 for a while). Also moved AnyEvent::DNS::addr to 796 for a while). Also moved AnyEvent::DNS::addr to
… …
88 - support service names in tcp_connect and tcp_server. 868 - support service names in tcp_connect and tcp_server.
89 - work around netbsd/openbsd bug causing unix sockets not to 869 - work around netbsd/openbsd bug causing unix sockets not to
90 become writable until empty. 870 become writable until empty.
91 - only make a dummy read for error detection when getpeername 871 - only make a dummy read for error detection when getpeername
92 returns ENOTCONN. 872 returns ENOTCONN.
93 - defend AnyEvent::Util::inet_aton against Coro::LWP brutal 873 - defend AnyEvent::Util::inet_aton against Coro::LWP's brutal
94 override. 874 override.
95 - new AnyEvent::Guard::cancel. 875 - new AnyEvent::Guard::cancel.
96 876
973.5 Sat May 17 23:17:14 CEST 2008 8773.5 Sat May 17 23:17:14 CEST 2008
98 - gracefully deal with systems not defining CLOCK_MONOTONIC 878 - gracefully deal with systems not defining CLOCK_MONOTONIC
… …
112 892
1133.41 Sun May 11 19:53:13 CEST 2008 8933.41 Sun May 11 19:53:13 CEST 2008
114 - work around a bug in perl 5.8.8's local. Reported by Yi Ma Mao. 894 - work around a bug in perl 5.8.8's local. Reported by Yi Ma Mao.
115 895
1163.4 Sun May 11 00:29:25 CEST 2008 8963.4 Sun May 11 00:29:25 CEST 2008
117 - complete re-implementation of condvars with a hopefully much mroe useful 897 - complete re-implementation of condvars with a hopefully much more useful
118 API (backward compatible functions exist but are undocumented). 898 API (backward compatible functions exist but are undocumented).
119 - AnyEvent::Handle fully rewritten with a hopefully more powerful API 899 - AnyEvent::Handle fully rewritten with a hopefully more powerful API
120 (still subject to change). 900 (still subject to change).
121 - take advantage of CLOCK_MONOTONIC, if available, in AnyEvent::Impl::Perl. 901 - take advantage of CLOCK_MONOTONIC, if available, in AnyEvent::Impl::Perl.
122 - provide AnyEvent::post_detect and @AnyEvent::post_detect, which 902 - provide AnyEvent::post_detect and @AnyEvent::post_detect, which

Diff Legend

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