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

Comparing AnyEvent/Changes (file contents):
Revision 1.87 by root, Fri May 23 16:36:02 2008 UTC vs.
Revision 1.505 by root, Fri Sep 2 04:29:11 2011 UTC

1Revision history for Perl extension AnyEvent. 1Revision history for Perl extension AnyEvent.
2 2
33.51 3TODO: document TCP_*** constants
4
5TOOD: stop internals abusing
6 - INCOMPATIBLE CHANGE: the default log level is now "3" (critical
7 and above), and some emssages inside AnyEvent have bene elevated
8 to critical, so.
9 - add google public ipv6 dns as flalback servers, so AE::DNS can finally
10 support ipv6-only operation even with defaults (also use ::1 instead
11 of 127.0.0.1).
12 - give guidelines on how to chose a logging level.
13 - AE::Strict now checks that registered fds are still valid at regular
14 intervals.
15 - increase signal latency test timer considerably for poor little osx.
16 - mention some debugging aids in the tutorial.
17
186.02 Fri Aug 26 20:08:31 CEST 2011
19 - log_to_syslog now accepts facility strings.
20 - log_to_syslog would not allow facilities - "$facility|$level"
21 *literally* means that string. sheesh.
22 - the CPAN parser got confused and thought we provide Sys::Syslog, try
23 to work around.
24 - renamed AnyEvent::Impl::FLTK2 to FLTK.
25
266.01 Fri Aug 26 07:04:11 CEST 2011
27 - INCOMPATIBLE CHANGE: AnyEvent::DNS resource records now include
28 the TTL as fourth element - this affects ->resolve, ->request
29 and ->dns_unpack indirectly. Convenience functions such as
30 AnyEvent::DNS::{a,txt,mx,srv} are not affected, neither is
31 AnyEvent::Socket.
32 - INCOMPATIBLE CHANGE: backend authors now should not implement
33 one_event or loop, but instead the AnyEvent::CondVar::_wait and _poll
34 methods.
35 - INCOMPATIBLE CHANGE: verbosity levels have been redefined, see
36 AnyEvent (PERL_ANYEVENT_VERBOSE) and AnyEvent::Log for details.
37 - DEPRECIATION: please change
38 "use AnyEvent::Impl::Perl" to "use AnyEvent::Loop" and
39 "AnyEvent::Impl::Perl::loop" to "AnyEvent::Loop::run"
40 in your programs, for future compatibility.
41 - *blush* idle watcher emulation was borked.
42 - AnyEvent::Strict now checks the AE::xxx functions as well.
43 - querying too long domain names would barf deep inside AnyEvent::DNS - since
44 this is hard to avoid for applications, AnyEvent:.DNS::request now checks for
45 for too long domain names explicitly and passes an undef to the callback
46 instead (reported by James Bromberger).
47 - make AnyEvent::Util more robust against EINTR (analyzed by Andrew Suffield).
48 - new functions: AnyEvent::postpone, AnyEvent::log.
49 - new module: AnyEvent::Log, for simple logging needs.
50 - new env variable, PERL_ANYEVENT_LOG, for fine-grained logging config.
51 - new env variable, PERL_ANYEVENT_DEBUG_SHELL, to automatically start
52 a debug shell.
53 - new env variable, PERL_ANYEVENT_DEBUG_WRAP, to automatically enable
54 watcher instrumentation/wrapping (see AnyEvent::Debug::wrap).
55 - new env variable, PERL_ANYEVENT_HOSTS, to override the place where
56 F</etc/hosts> can be found.
57 - an empty PERL_ANYEVENT_RESOLV_CONF now selects the default
58 configuration.
59 - env variables can now be specified using AE_xyz instead of
60 PERL_ANYEVENT_xyz, subject to some rules.
61 - lots of new AnyEvent::Debug functionality.
62 - AnyEvent::Strict now checks for modifications $_ by the callback
63 (by setting it to readonly while executing the callback, if possible).
64 - IO::Async sometimes cancels the wrong timer when you call cancel_timer.
65 Invest extra overhead to work around this peculiar behaviour.
66 - split the pure perl backend into a loop module and an impl
67 module (AnyEvent::Loop and AnyEvent::Impl::Perl).
68 - some read types (regex, netstring, json, storable) could
69 errornously keep an AE::Handle object alive on their own
70 (reported by Mohammad Toossi).
71 - AnyEvent::Socket::resolve_sockaddr and all functions using it now
72 supports /etc/hosts overrides.
73 - add more workarounds around child watcher bugs in IO::Async
74 and (newly introduced) in POE.
75 - use glib child watchers and try to work around its limitations,
76 if possible (based on a patch by Kevin Ryde).
77 - add support for FLTK (fltk2), via AnyEvent::Impl::FLTK2.
78 - many read types in AnyEvent::Handle would malfunction during
79 transitions to TLS because of a "delete $self->{rbuf}".
80 (reported by Gerald Galster).
81 - improve condvar blocking wait performance (probably only noticable
82 with faster backends), also make it easier for event loops that
83 do not support blocking waits.
84 - PERL_ANYEVENT_MODEL now also allows full module names, although
85 the usefulness of this feature is not yet clear.
86 - add tests for all supported event loops, conditional on PERL_ANYEVENT_LOOP_TESTS.
87 - add t/80_ssltext.t.
88 - no longer use AUTOLOAD for the initial loading (for no very good reason).
89 - tcp_server now tries to unlink unix domain sockets when it gets destroyed.
90 - do not run event loop detection multiple times when programs call their cached
91 copies of AnyEvent::detect.
92 - parse_hostport now accepts absolute paths als unix domain sockets.
93 - suppress the idiotic warning inside POE by patching the POE::Kernel
94 run flags instead of asking the user to jump through contortions.
95 - add Task::Weaken as dependency (as recommends only at this time).
96 - replace PF_UNSPEC by 0, for good old minix.
97
986.0 (not officially released, but was on CPAN for about 15 minutes :/)
99
1005.34 Thu May 12 09:59:41 CEST 2011
101 - AE::TLS didn't support both cert_file and key at the same time,
102 due to some copy&paste bug (reported and analyzed by vti).
103 - improve AE::Handle timeout documentation slightly.
104
1055.33 Mon Feb 7 21:51:41 CET 2011
106 - never appeared on CPAN, for unknown reasons.
107 - AnyEvent::Handle on_connect_error - do not try to destroy a handle
108 object when it's already gone (analyzed by Ryan Bullock).
109 - update idna algorithm(s) to UTS#46 version 6.0.0.
110 - preserve trailing dot in idn_to_ascii - testsuite says to preserve,
111 uts46 says it's an error, and testsuite makes more sense.
112 - remove leading dots in idn_to_ascii - testsuite says to remove,
113 uts46 says it's an error, and testsuite makes more sense.
114 - fixed a bug in idn ignored character processing.
115 - passes the uts46 6.0.0 testsuite except for three cases which
116 apparently are bugs in the testsuite itself.
117 - passes the uts46 6.0.1 testsuite except for eleven cases which
118 apparently are bugs in the testsuite itself.
119
1205.31 Sun Jan 23 11:49:19 CET 2011
121 - work around a POE bug where POE would first poll for new events and
122 invoke handlers BEFORE handling the message queue.
123 - fix parsing of "domain" statement in resolv.conf (Johannes Plunien).
124 - stop_read now does just that in TLS mode. see its
125 documentation for caveats.
126 - update to common::sense 3.4 warning set (some warnings in perl impose
127 an arbitrary 32kb limit on regex matches, which affects
128 AnyEvent::Handle).
129 - AE::Handle had a broken rbuf_max method and missing wbuf_max method.
130 - new experimental AE::Handle::resettls method.
131
1325.3 Fri Dec 31 05:48:13 CET 2010
133 - major IOAsync patch by Paul Evans - 0.33 is required, as the major
134 issues with IO::Async have been solved in that version, including
135 the availabilty of a default loop!
136 - implement new wbuf_max parameter for AnyEvent::Handle.
137 - added Cocoa::EventLoop backend by Daisuke Murase.
138 - the default testsuite now honors PERL_ANYEVENT_MODEL (Paul Evans).
139 - clarify handle on_prepare documentation.
140 - set CLOEXEC flag on parent-side fd's in AnyEvent::Util::run_cmd,
141 to avoid leaking them to any child processes.
142 - try to force files into the arch-dependend lib directory, for the
143 half-"turly obsessive compulsive person"(s) that depend on it.
144
1455.29 Sun Dec 5 10:49:21 CET 2010
146 - convert EV backend to EV 4.00 API (so better upgrade EV too).
147 - AE::Handle: implement dynamic read_size adjustment, add
148 max_read_size parameter, reduce default read_size to 2048.
149 - add some FAQ entries for servers and/or on_eof vs. on_error.
150 - work around OS bugs (cygwin again) when lingering in AE::Handle.
151 - reduce memory usage slightly when connect was used in AE::Handle.
152 - wrok around more segfaults in Tk.
153 - document tls_ctx => undef as valid.
154 - detect WNOHANG value at compiletime.
155 - include some probably linux-only support for building AnyEvent
156 as part of the perl core.
157 - improve compatibility of t/08_idna.t to perl 5.8.
158 - make AnyEvent::Strict truly optional (in case somebody builds a
159 minimal perl...).
160
1615.28 Wed Oct 13 04:14:23 CEST 2010
162 - due to a glitch, AnyEvent's internal getprotobyname wasn't
163 actually used in tcp_server.
164 - implement AnyEvent::CondVar->new method, which was documented but
165 missing.
166 - support autoloading of read types in unshift_read.
167 - AnyEvent::Strict: do the test for invalid fh arguments differently,
168 resulting in more informative error messages.
169 - new AnyEvent::FAQ manpage.
170 - clarify that you can actually call push_read etc. in on_read.
171 - put keys and certs in /tmp, not into the current directory,
172 and document this in the SECURITY CONSIDERATIONS sections
173 of AnyEvent::TLS.
174 - support ";" as resolv.conf comment character in AnyEvent::DNS, as
175 requested by Ron Isaacson.
176 - document $AnyEvent::DNS::RESOLVER variable.
177 - incorporated major typo patches by Abhijit Menon-Sen.
178 - AnyEvent::Handle now croaks when negative timeouts are passed.
179 - add a shitload of TCP_xxx sockopt name constants to AnyEvent::Util,
180 undocumented, but who knows when they might come in handy.
181
1825.271 Tue Jun 8 12:05:46 CEST 2010
183 - backport to perl 5.8.x.
184
1855.27 Sun Jun 6 12:12:05 CEST 2010
186 - postpone differently in AnyEvent::Socket now, as
187 when not, canceling the connection attempt might fail
188 (found by Felix Antonius Wilhelm Ostmann).
189 - explicitly check for non-stream sockets in AE::Handle, too many
190 clueless people fell into the trap of this somehow working.
191 - simplified and reworked the "OTHER MODULES" section.
192 - better/more condvar examples.
193
1945.261 Wed Apr 28 16:13:36 CEST 2010
195 - AF_INET6 was not properly used from Socket6 during configuration
196 time (found by Andy Grundman).
197
1985.26 Mon Apr 12 04:49:35 CEST 2010
199 - don't generate spurious readyness notifications when select
200 returns EINTR in the pure perl backend (analysed and testcase
201 by Bas Denissen).
202 - give same examples in SYNOPSIS sections of AnyEvent and AE.
203 - provider faster implementations for AE::time/now/now_update for
204 EV and Perl backends.
205 - provide faster AE::cv and AE::time implementations for all
206 backends.
207 - no longer support lower/mixed-case signal names (this was
208 never documented nor universally supported).
209 - some more memory µ-optimisations, and somewhat less messy
210 code to cope with both AnyEvent and AE APIs.
211
2125.251 Sat Mar 13 00:58:21 CET 2010
213 - make test failed if it was run before make install on perl <=
214 5.8.8, and older versions of perl make our live very difficult,
215 so write constants.pl during Makefile.PL time.
216
2175.25 Sat Mar 13 00:23:14 CET 2010
218 - fix a race condition in AnyEvent::Handle that would cause
219 a "bio_read: unsupported method" error (in _another_
220 ssl connection) after user code threw an exception.
221 - added AnyEvent::Handle->destroyed method.
222 - speed up AnyEvent::Socket::format_address and ::format_ipv6.
223 - the AnyEvent::Util::fh_nonblocking function worked only
224 by lucky accident on win32.
225 - smaller and faster AnyEvent::Util::fh_nonblocking.
226 - when the (required!) Time::HiRes module is missing, AnyEvent
227 did not fall back to built-in time properly.
228 - do not load Fcntl at runtime, saving memory and loading time.
229 - precompile a number of constants and use them instead of runtime
230 detection and eval.
231 - free detection code after detection and similar memory optimisations.
232 - Perl backend timer interval best effort drift has been improved
233 (same algorithm as EV).
234 - update unicode idna mapping table.
235
2365.24 Tue Jan 5 11:39:43 CET 2010
237 - cygwin never reports errors from failed connects. 1.5 just gives
238 you continous readyness and EAGAIN, 1.7 is even more broken and
239 just hangs. work around both issues in a cygwin-specific hack.
240 - improve idle watcher documentation slightly.
241
2425.23 Sun Dec 20 23:48:00 CET 2009
243 - support IDNs in resolve_sockaddr, and therefore in tcp_connect.
244 - implement punycode_encode/decode, idn_nameprep,
245 idn_to_ascii and idn_to_unicode operations in AnyEvent::Util.
246 - provide $AE::VERSION.
247 - removed traces of "no strict 'refs'".
248
2495.22 Sat Dec 5 03:51:13 CET 2009
250 - downgrade-or-fail in AnyEvent::Handle::push_write, to
251 diagnose encoding failures earlier and more succinctly.
252 (this works around bugs in perl, throwing away encoding info
253 when passing scalar data to extensions).
254 - add more examples to AnyEvent::Socket manpage.
255 - upgrade internal warning set to the same as common::sense 2.03.
256 - use pack "n/a*" for pre-5.8.9 perl compatibility in AnyEvent::DNS
257 (John Beppu).
258 - AnyEvent::Socket::inet_aton now properly supports ipv6, as documented.
259 - add google public dns servers to fallback server set and make sure
260 we load-balance properly between the three sets. also add all
261 fallback dns servers, not just a random one, to each dns config.
262
2635.21 Thu Nov 19 02:48:47 CET 2009
264 - fix a problem where socket constants were called with parameters
265 (spotted by David Friedland).
266 - fork_call never use'd POSIX (reported by Daisuke Maki).
267 - improve perl 5.6 compatibility further (but it still won't work
268 unless you rip out everything but the core).
269 - prefer Net::DNS::Resolver over ipconfig, if installed, on win32.
270 uses 10MB of RAM, but doesn't open a console window. *sigh*.
271
2725.202 Wed Oct 14 22:35:44 CEST 2009
273 - AnyEvent::DNS would unexpectedly clobber $_ under windows
274 (analysed by Matthias Waldorf).
275 - AnyEvent::Handle::run_cmd can now pass the PID of the
276 newly-created process, which is much less useful than it might
277 sound (based on patch by Yann Kerherve).
278
2795.201 Tue Sep 29 12:09:25 CEST 2009
280 - AnyEvent:Handle::on_starttls/on_stoptls methods were broken
281 (reported by Torsten Foertsch).
282 - common::sense 2.0 could cause tcp_server to throw an exception
283 (analysed by elmex).
284
2855.2 Mon Sep 14 07:04:49 CEST 2009
286 - INCOMPATIBLE CHANGE: do no longer support register_read_type
287 and register_write_type in AnyEvent::Handle, instead support
288 package names (the facility was mostly abused).
289 - implement "packagename-as-read/write type" support in
290 AnyEvent::Handle.
291 - AnyEvent::Handle: new options "keepalive" and "oobinline".
292 - oobinline set by default to avoid security issues.
293 - the pure-perl event loop backend wrongly detected times() fallback
294 support (spotted by Pavel Boldin).
295
2965.12 Tue Sep 1 20:26:50 CEST 2009
297 - be more lenient when parsing resolv.conf files, as some people
298 use hashmarks as comment indicator inside directives (reported
299 by Michael S. Fischer).
300 - use same set of warnings as common::sense 2.0.
301 - fix a potential 32 bit overflow issue due to perl having problems
302 with large hex constants in the Perl backend.
303 - do not use the slower 5.6 version fo the enc_name code in AnyEvent::DNS
304 by default.
305 - fix some prototypes.
306 - (optionally) require Async::Interrupt 1.02.
307 - replace opendns fallbacks by verizon/level3 public dns servers.
308
3095.112 Fri Aug 21 13:59:04 CEST 2009
310 - AnyEvent::Handle could cause hard crashes in Net::SSLeay when
311 using starttls with an illegal/uncreatable tls context.
312 - port to broken win32 perls.
313
3145.111 Sun Aug 16 18:44:36 CEST 2009
315 - perl errornously requires sockaddr_un structures returned by the
316 kernel to have a specific length. work around this bug by padding
317 sockaddr_un structures (found to fail on OS X, as perl only uses
318 the correct code for GNU/Linux, but likely to fail on other
319 OSes as well).
320
3215.11 Wed Aug 12 17:49:37 CEST 2009
322 - fixed a bug in run_cmd with input from filename.
323 - create a json coder object when none is given in write direction
324 to reduce dependency on JSON.
325
3265.1 Tue Aug 11 03:17:32 CEST 2009
327 - tcp_connect now reports the correct error in $!, not an unrelated
328 one (was broken since 4.91).
329 - AE::Handle did not properly clear rtimeout/wtimeout watchers
330 when retrying a tcp_connect.
331 - new functions: AnyEvent::Util::run_cmd and close_all_fds_except.
332 - fixed wrong documentation in AE::Socket.
333
3345.01 Mon Aug 10 03:16:32 CEST 2009
335 - last release broke EV child watchers :/.
336
3375.0 Sun Aug 9 17:32:09 CEST 2009
338 - officially introduce and document the AE API.
339 - lowering the timeout at runtime did not have immediate effect
340 in AnyEvent::Handle.
341 - AE::Handle now has separate and independent read and write timeout
342 settings.
343 - probe for Guard module when AnyEvent::Util is loaded, not at
344 runtime (Event doesn't like eval "use" from callbacks, and it avoids
345 exporting the probe function).
346 - Impl::IOAsync's io watchers did not use the filehandle emulation
347 code to work around it's watcher limits.
348 - work around IO::Async's problems with overloaded objects.
349 - try to disable wearnings inside IO::Async, it's just too buggy.
350 - do no longer use Event by default, only EV or Perl, Event is too buggy.
351 - hackishly provide fast AE:: interface when EV or Perl is the backend.
352 - inherit VERSION from AnyEvent::VERSION in most modules.
353
3544.91 Thu Aug 6 15:42:45 CEST 2009
355 - AE::Handle::starttls could get out of sync when the read buffer
356 already contains some TLS handshake.
357 - AE::Handle did not properly free the TLS session, causing bigger
358 memory leaks in Net::SSLeay than expected (reported by toaster).
359 - AE::Socket::tcp_connect will now artificially delay invoking
360 the callback to avoid returning after invoking the callback.
361 - convert many internal watcher uses to AE API.
362
3634.9 Sat Aug 1 11:07:01 CEST 2009
364 - Glib silently fails when registering a timeout with
365 a negative value, so avoid that.
366 - call condvar callback immediately when it is set
367 after the condition is already signalled.
368 - check rbuf_max condition only after trying to consume
369 data.
370 - normalise signal numbers to names when using the ae's
371 signal handling, but do not document this (yet).
372 - pure perl signal emulation did not properly set nonblocking
373 mode on the signal pipe on win32 (this is unlikely to have caused
374 any issues).
375 - new module: AnyEvent::Debug.
376 - AnyEvent::Strict now checks that a signal was specified
377 by name (not name or number).
378 - reduce memory footprint in typical cases by ~50kb by
379 only compiling signal code when necessary.
380 - add AnyEvent::Handle->rbuf_max.
381 - grab the AE namespace for future new API, implement stubs for
382 said future/alternative API.
383 - new function: AnyEvent::Socket::format_hostport.
384
3854.881 Tue Jul 28 12:51:53 CEST 2009
386 - work around a bug in local in pre-5.10 perls, causing
387 AnyEvent::Handle to recurse when it should not
388 (analysed by elmex).
389
3904.88 Tue Jul 28 04:04:37 CEST 2009
391 - re-bless the handle into a dummy package after calling
392 AnyEvent::Handle::destroy, so the user does not need to check
393 for errors after every push_write etc.
394 - do not attempt to run t/02_signals.t on obviously broken
395 platforms.
396
3974.87 Sun Jul 26 02:06:16 CEST 2009
398 - do not attempt to linger when there is no longer a valid fh.
399 - enforce tls mode to be either accept or connect, do not simply
400 segfault in Net::SSLeay.
401 - AnyEvent::Handle can now call tcp_connect itself (new parameters
402 connect, on_prepare, on_connect and on_connect_error). Updated
403 tutorial accordingly.
404 - add AnyEvent::Impl::Irssi backend.
405
4064.86 Mon Jul 20 23:52:29 CEST 2009
407 - since the verbose warning is not enough, explicitly document
408 that versions before 1.33 of Net::SSLeay are not secure.
409 - work around signal handling races in Event and (...) Event::Lib.
410 - try to align signal-race timer to full-second boundaries.
411 - work around Tk not liking negative timeouts.
412 - don't complain of different grades of environmental unfriendlyness
413 in IO::Async.
414
4154.85 Sat Jul 18 06:16:14 CEST 2009
416 - nail the signal race problem in perl once and for all
417 (see $AnyEvent::MAX_SIGNAL_LATENCY).
418 - take advantage of Async::Interrupt if it is available.
419 - load Time::HiRes and Guard modules on demand only.
420 - add optional/recommended modules section to AnyEvent
421 documentation.
422 - reduce memory usage considerably (and reduce startup penalty)
423 by not using "strict", "warnings" and "overload" modules.
424 - work around buggy windows/openbsd perls and provide EBADMSG
425 and EPROTO ourselves when missing.
426 - improve perl 5.6 compatibility of the core event loop.
427 - made Net::SSLeay version 1.33 a soft requirement.
428
4294.83 Fri Jul 17 16:56:26 CEST 2009
430 - implement AnyEvent::Socket::getprotobyname.
431 - AnyEvent::CondVar's will now detect recursive blocking
432 waits and will croak, as too many people fall into
433 this trap.
434 - AnyEvent::Handle will now call ->destroy on itself after
435 executing the on_error callback, instead of doing some
436 half-baked internal shutdown, for fatal errors.
437 - clarify on_eof behaviour w.r.t. the read queue and
438 on_read callbacks.
439 - ignore some possible spurious wake-ups in tcp_connect.
440
4414.82 Sat Jul 11 00:34:55 CEST 2009
442 - POE and Event backends didn't accept some callable objects as
443 callbacks.
444 - use Config module instead of POSIX module to detect signal names
445 in AnyEvent::Strict and AnyEvent::Impl::EventLib, as the POSIX
446 module doesn't even have all POSIX signals :/.
447 - use more workarounds around the many refcnt/corruption bugs in
448 Event::Lib.
449 - work around a race condition in perl's select, causing t/03_child.t
450 to rarely fail.
451
4524.81 Thu Jul 9 10:30:30 CEST 2009
453 - AnyEvent::Handle didn't properly diagnose write errors
454 (it expected -1 from syswrite, how lame... :).
455 - support file descriptors in addition to file handles
456 in AnyEvent->io.
457 - new env variables: PERL_ANYEVENT_RESOLV_CONF,
458 PERL_ANYEVENT_MAX_OUTSTANDING_DNS, PERL_ANYEVENT_CA_FILE
459 and PERL_ANYEVENT_CA_PATH.
460 - provide a sensible synopsis section for AnyEvent::TLS.
461 - add a "supported backends" section to the manpage.
462 - added simple io watcher test to testsuite, using a
463 portable_socketpair.
464 - tried to improve the stability of the Event::Lib backend,
465 YMMV.
466
4674.8 Mon Jul 6 23:45:16 CEST 2009
468 - AnyEvent::DNS did not properly follow CNAME records with
469 uppercase targets.
470 - AnyEvent::DNS would errornously return AAAA records
471 with v4 mapped addresses (a faulty record) as ipv4 addresses,
472 causing AnyEvent::Socket to throw an exception.
473 - added new module AnyEvent::TLS for easier SSL/TLS context
474 creation, with many options including hostname verification,
475 secure default configuration, lots of documentation and,
476 predefined diffie-hellman keys for perfect forward security
477 and much more. get it while it's still fresh!
478 - use AnyEvent::TLS in AnyEvent::Handle for context management.
479 - load AnyEvent::Handle only on demand in AnyEvent::DNS,
480 so AnyEvent::Socket users have smaller memory footprint
481 in the common case.
482 - add AnyEvent::Handle->push_shutdown method.
483 - add an additional error message parameter to AnyEvent::Handle's
484 on_error callback (for TLS, $! is still available).
485 - add AnyEvent::Handle on_starttls/on_stoptls callbacks.
486 - make AnyEvent::Handle more robust against early conenction
487 failures (during new), and return C<undef> in that case
488 from the constructor.
489 - AnyEvent::Handle will now try to load only JSON::XS first,
490 then fall back to JSON.
491 - format_ipv4/format_ipv6 are now exported by default, for symmetry,
492 and because it was documented that way.
493
4944.451 Fri Jul 3 00:28:58 CEST 2009
495 - do not clear rbuf when shutting down an AnyEvent::Handle
496 object - doing so breaks AnyEvent::HTTP.
497
4984.45 Mon Jun 29 22:59:26 CEST 2009
499 - a write error could cause AnyEvent::Handle to create
500 an I/O watcher with an undefined $fh.
501 - special-case mapped ipv4 addresses in both
502 AnyEvent::Socket::format_address and parse_address,
503 to treat them just like ipv4 addresses.
504 - updated and overhauled the AnyEvent::Intro doc.
505 - implement AnyEvent::Socket::format_ipv4/ipv6.
506 - slightly speed up portable_pipe/socketpair.
507 - expand condvar begin/end documentation.
508
5094.42 Fri Jun 26 08:32:18 CEST 2009
510 - preliminary and neccesarily incomplete support for IO::Async.
511 - reset SIGCHLD to DEFAULT when AnyEvent is loaded, in
512 case it was set to IGNORE, to ensure we can catch child statuses
513 even when the calling env acts stupidly.
514 - updated benchmarks with IO::Async, which performs very well.
515
5164.412 Wed Jun 24 01:35:57 CEST 2009
517 - support an "untaint" attribute for AnyEvent::DNS
518 and set it on the default resolver.
519 - implement and document AnyEvent::Impl::Perl::loop.
520 - remove all anyevent-env variables from %ENV
521 when running in tainted mode.
522 - mention and extend the IO::Lambda benchmark.
523
5244.411 Sun Jun 7 18:48:02 CEST 2009
525 - do not try to use F_SETFD on windows.
526
5274.41 Thu May 14 06:40:11 CEST 2009
528 - work around issues in older perls (5.8.0?) when
529 a signal handler is deleted from the %SIG hash.
530 - use POSIX::_exit in child test, to avoid running
531 destructors.
532 - speed up CHLD handling by relying on SIGCHLD being
533 synchronously delivered, even when we roll our own
534 implementation.
535 - AnyEvent::DNS: add the "dname" resource record name
536 for cosmetic reasons.
537
5384.4 Sun Apr 26 20:12:33 CEST 2009
539 - implemented idle watchers, where applicable.
540 - AnyEvent->time died when Event backend was in use.
541 - fix a memleak in the Tk backend.
542 - sped up Tk timer handling.
543 - clip negative "after" values to 0 in AnyEvent::Impl::Event
544 to avoid spamming out warning messages.
545 - fix Qt timers without interval.
546 - avoid Qt zero-timeout special case, in old libqt's.
547
5484.352 Mon Apr 20 16:31:11 CEST 2009
549 - fix AnyEvent::Strict error messages for child watchers.
550 - fix/update Makefile.PL "recommends" versions.
551 - add AnyEvent->now_update.
552
5534.351 Sat Apr 11 07:56:14 CEST 2009
554 - actually make the signal pipe work under win32.
555 - localise $! in signal handler to avoid changing $!, although
556 perl itself does not seem to save/restore errno either.
557 - set the cloexec flag on the signal pipe (normally set by perl too).
558
5594.35 Fri Mar 27 11:48:20 CET 2009
560 - event models relying on AnyEvent's signal watcher emulation did
561 invoke the callback asynchronously, contradicting documentation
562 and causing signals to get lost (this includes AnyEvent's own
563 event loop). AnyEvent now uses the standard pipe trick to make
564 callback execution synchronous to the event loop.
565 - AnyEvent::Handle didn't free TLS context data on DESTROY
566 (patch by Pavel Shaydo).
567 - work around the ever-incompatibly-changing API of MakeMaker.
568 - document that changing global variables without restoring them
569 is a bad idea in Perl (noted by Adam Rosenstein).
570 - AnyEvent::Strict now barfs if ->io is passed a file.
571
5724.34 Thu Feb 12 18:32:45 CET 2009
573 - separately buffer TLS read data, as otherwise the read queue could
574 deadlock as receiving data is not expected while draining the
575 read queue (which cna only happen with TLS).
576 - raise EBADMSG error on JSON decoding errors.
577 - fix some minor manpage bugs (reported by Maximilian Gaß).
578 - speed up select bitmask parsing quite a bit in the pure perl
579 backend.
580 - use CORE::select instead of just select in AnyEvent::Impl::Perl.
581
5824.331 Tue Jan 6 21:07:25 CET 2009
583 - socketpair fails on many vista machines because vista has
584 completely broken accept/getpeername and getsockname functions,
585 so we provide our own socketpair emulation that kind of works
586 (AnyEvent::Util::portable_pipe).
587 - new function: AnyEvent::Util::portable_socketpair.
588 - take advantage of the Guard module if it exists.
589
5904.33 Fri Nov 21 02:35:40 CET 2008
591 - AnyEvent::Strict did errornously flag a fileno of 0 as
592 illegal.
593 - reduce memory usage and slightly speed up the pure perl backend
594 by only storing the file descriptor, not the file handle.
595 - add missing autocork method to AnyEvent::Handle (reported
596 by Adam Rosenstein).
597 - AnyEvent::DNS->resolve errornously documented an $rcode result
598 argument, but there isn't (spotted by Henrik Krohns).
599 - the naked truth about Tk - it's basically unmaintained.
600
6014.32 Mon Nov 3 22:46:32 CET 2008
602 - fix AnyEvent::Socket::resolve_sockaddr to properly support
603 unix sockets again and choose a proper default for
604 the socket type.
605 - fix call to resolve_sockaddr in tcp_connect.
606
6074.31 Thu Oct 30 04:41:48 CET 2008
608 - implemented AnyEvent::Handle->destroy method.
609 - hint about unexpected effects in TLS mode.
610 - speed up AnyEvent::Impl::Perl by using more arrays
611 instead of hashes without the slightest loss of readability :)
612 - work around a perl argument refcounting bug.
613 - with some perl patches applied, AnyEvent now seems leak-free.
614
6154.3 Fri Oct 3 09:18:43 CEST 2008
616 - AnyEvent will now install a no-op signal handler for SIGPIPE,
617 unless one has been installed already.
618 - warn about not loaidng AnyEvent::Impl::POE early enough
619 (patch by Adam Rosenstein).
620 - fixed a great number of bugs and corner cases in AnyEvent::Handle.
621 - fix a bug in where in SSL connect mode, the
622 client would first wait for some data by the server and
623 otherwise hang. (reported and analysed in an absolutely
624 exemplary manner by Adam Rosenstein).
625 - fix a bug in where SSL EOF would not be treated
626 as stream EOF, putting the connection into a hung state.
627 (reported and analysed in an absolutely exemplary manner by Adam
628 Rosenstein).
629 - fix a potential segfault when the TLS context
630 would go missing in a read callback (e.g. due to stoptls) -
631 Net::SSLeay of course makes no type checking whatsoever.
632 - AnyEvent::Handle will not stop the read watcher in TLS mode.
633 - AnyEvent::Handle->stoptls will send a TLS close notify
634 instead of simply destroying the stream now.
635 - fix a bug where large blocks of data written in TLS mode
636 would not be sent unless triggered by receives.
637 - on_drain will now take the tls write buffer into account.
638 - SSL operations have been streamlined a bit and should be faster
639 now, more to come, though.
640 - added a FAQ section, will document nontrivial issues in there.
641
6424.234 Mon Sep 29 04:08:13 CEST 2008
643 - fix child watcher documentation: only child exits will be reported,
644 no trace events.
645 - mention SIGPIPE in AnyEvent::Handle.
646 - perl backend will now use POSIX::times as monotonic clock
647 when available and the monotonic clock option is not.
648 - run even without Time::HiRes in some cases now, but do not
649 provide sub-second accuracy (all *supported* perl versions have
650 Time::HiRes).
651
6524.233 Fri Aug 22 01:48:05 CEST 2008
653 - fix a bug in the testsuite, causing a hang.
654 - clarified lots of AnyEvent::Handle constructor arguments.
655
6564.232 Thu Aug 21 20:44:25 CEST 2008
657 - fix a bug in AnyEvent::Handle that could cause two fatal error
658 callbacks on EOF, which caused AnyEvent::HTTP to signal success twice
659 (testcase by Leon Brocard).
660 - clarified on_eof/on_error documentation a bit.
661
6624.231 Tue Jul 29 13:12:15 CEST 2008
663 - remove some debugging code left in AnyEvent::Util::fork_call
664 (and no, it's impossible to implement with the broken windows
665 perls without resource leaks or worse).
666
6674.23 Tue Jul 29 12:19:59 CEST 2008
668 - document the first parameter passed to condvar callbacks to be
669 the callback.
670 - add AnyEvent::Socket::{ntoa,aton} aliases.
671 - optimize the AE::Handle->push_read (line) for the default
672 eol marker.
673 - optimize push_read (packstring|storable) for small packets.
674 - invoke on_error callback when no on_eof callback is set.
675 - fix a bug in push_read (storable) of unknown impact.
676
6774.22 Sun Jul 20 16:34:13 CEST 2008
678 - new function AnyEvent::Socket::parse_hostport.
679 - as the bulkheads at microsoft can't even get getprotobyname reliably
680 working on their shitty fucking broken os we need to hardcode
681 some common protocol numbers in AnyEvent::Socket. How can
682 people even bother with such a pile of shit as windows.
683
6844.21 Thu Jul 17 14:40:05 CEST 2008
685 - INCOMPATIBLE CHANGE: AnyEvent::Util::fork_call now has a prototype
686 which is incompatible to previous invocation syntax.
687 - work around more windows perl bugs in fork_call. I cannot imagine
688 why anybody would bother implementing fork in such an obviously
689 unusable way.
690 - avoid calling $do_retry if the request has already finished in
691 AnyEvent::DNS, thus avoiding an exception (reported by Anatoly
692 K. Sharifulin).
693
6944.2 Sat Jul 12 22:42:11 CEST 2008
695 - support an "interval" argument for repeating timers.
696 - fix a bug in the pure perl event loop that caused timers to
697 be delayed under some circumstances.
698 - implement optional argument checking via PERL_ANYEVENT_STRICT.
699 - not importing anything from Event causes it to malfunction
700 and not use Time::HiRes, work around this bug.
701 - more bug workarounds against the endlessly broken Tk module,
702 seems to work now on 32 bit machines, but less so on 64 bit.
703 - minor optimisations applied to most backends.
704
7054.161 Fri Jul 4 14:24:48 CEST 2008
706 - fixed the t/handle/02_write.t, it read after write, with the
707 assumption that all data could be written before read
708 (reported by Jim Razmus).
709
7104.160 Thu Jul 3 04:02:21 CEST 2008
711 - re-registering signal watchers after unregistering one did
712 not work if a backend falls back on AnyEvent's default
713 implementation (also affected child watchers).
714 - new AnyEvent::Handle options: autocork and no_delay.
715
7164.152 Sun Jun 22 14:15:44 CEST 2008
717 - allow for 32-bit perls that implement shifts differently
718 on different architectures in parse_ipv4 (reported and
719 analysed by Keiichi DAIBA).
720
7214.151 Fri Jun 6 17:34:24 CEST 2008
722 - make sure specifying _only_ on_read and never pushing reads
723 works.
724
7254.15 Fri Jun 6 13:00:46 CEST 2008
726 - the pure perl backend would keep some watchers alive when more than
727 one watcher was registered for the same fd.
728 - new "packstring" and "storable" read and write types
729 for AnyEvent::Handle.
730 - allow on_eof handler to be called after on_error with EPIPE returns.
731 - do not immediately call on_read callback in handle constructor.
732
7334.14 Thu Jun 5 20:29:31 CEST 2008
734 - Fixed a bug in DNS SRV priority sorting.
735 - AnyEvent::Util::guard now reports runtime errors while
736 executing the guard block as warnings.
737 - handle 0-byte-reads just before EOF correctly in AnyEvent::Handle.
738
7394.13 Thu Jun 5 00:47:59 CEST 2008
740 - AnyEvent::DNS only followed cname chains with length 2,
741 contrary to documentation. bumped it up to 10, thanks to
742 microsoft, the current broken-dns-config-king.
743 - AnyEvent::DNS didn't check for socket return status
744 but instead relied on perl not creating filehandles in that
745 case - too bad it gives you a bogus file handle (reported
746 and analysed by Vladimir Timofeev).
747 - fix queue management logic in AnyEvent::Handle:
748 when on_read was registered, the queue was empty
749 and no progress could be made AnyEvent::Handle would
750 enter an endless loop.
751 - correctly start reading again when the handle became
752 busy again after idling.
753 - correctly treat tls shut-downs as EOF condition
754 (for the time being, would be nice to have a callback for that).
755 - correctly call eof callback under all conditions
756 (hopefully).
757 - the timeout callback did not expect that $self can go
758 away any time.
759
7604.12 Tue Jun 3 10:58:04 CEST 2008
761 - include AnyEvent::Intro, a tutorial for anyevent,
762 anyevent::socket and anyevent::handle.
763 - allow more options in on_error.
764
7654.11 Fri May 30 23:42:25 CEST 2008
766 - INCOMPATIBLE CHANGE: replace ptr by real PTR lookup, provide
767 reverse_lookup and reverse_verify to replace it, support
768 v4mapped and v4compat addresses.
769 - provide more documentation for the resolver class.
770 - really replace longest run of :0: by :: in format_address,
771 also properly convert :: and ::1 again.
772 - support NAPTR record name and decode it.
773 - implement random weight sampling for SRV records, as per
774 rfc 2782.
775 - correctly abort on srv-record targets of ".".
776 - added AnyEvent::DNS::wait_for_slot.
777 - in the unlikely event of a virtual circuit connection
778 being invalidated by a delayed udp reply, AnyEvent::DNS
779 could die. this has been fixed.
780 - plug a probable memleak in the DNS vc code.
781 - use configured timeout also for VC request phase.
782 - implement timeout and max_outstanding methods
783 for resolver class.
784 - update version numbers in all modules.
785
7864.1 Thu May 29 05:45:40 CEST 2008
787 - INCOMPATIBLE CHANGE: renamed xxx_ip functions to xxx_address
788 which mirrors their purpose better (old names still available
789 for a while). Also moved AnyEvent::DNS::addr to
790 AnyEvent::Socket::resolve_sockaddr.
791 - implement AnyEvent->time and AnyEvent->now.
792 - fix IPv6 support in pack_sockaddr.
793 - officially un-experimentalise all newly introduced code.
794 - support unix domain sockets everywhere by specifying
795 a host of "unix/" and the pathname as service.
796 - implement an activity timeout in AnyEvent::Handle.
797 - added a regex read type to AnyEvent::Handle.
798 - added a json read/write type to AnyEvent::Handle.
799 - always croak in on_error in AnyEvent::Handle.
800 - document how to subclass AnyEvent::Handle.
801 - implement AnyEvent::Util::fork_call.
802 - add support for IPv6 nameservers and nameserver statements.
803 - work around _yet_ _another_ windows perl bug in where empty
804 select masks cause errors in the pure perl backend (this
805 workaround cannot be implemented for the other event loops,
806 but EV also works around this).
807 - supply AnyEvent::Socket::address_type and make good use of it.
808 - clarify and fix the Handle documentation w.r.t $self vs. $handle
809 vs. callback arguments.
810 - add some recommends to the META.yml.
811
8124.05 Mon May 26 19:44:06 CEST 2008
813 - some platforms ignore the rfc and prepend an extra sa_len member to
814 the sockaddr structure, cater for those. also use sockaddr_family,
815 didn't know it was there, but comes in handy for the workaround.
816 - undo the import WIN32 hack.
817
8184.04 Mon May 26 08:03:31 CEST 2008
819 - try to work around yet another windows bug: failed connects are reported
820 as if it were out-of-band data. windows users: you suck. Only
821 EV, Glib, Event and the pure perl backend can handle this condition.
822 - optimize write algorithm in AnyEvent::Handle.
823 - properly parse PERL_ANYEVENT_PROTOCOLS and use the info correctly.
824 - tcp_connect now properly iterates through all targets.
825 - check wether IPv6 sockets can actually be created, otherwise, disable ipv6
826 support entirely. also hardcode some AF_INET6 constants for the
827 sake of perl 5.8.
828 - vastly improved nameserver/suffix detection algorithm for windows
829 (still a hack...)
830 - try to use a fallback dns server if no dns servers could be found.
831 - splatter around some local $SIG{__DIE__}'s to work around bad
832 die handlers.
833
8344.03 Sun May 25 05:05:57 CEST 2008
835 - work around perl bugs on windows, where perl returns undocumented
836 error codes for sysread, syswrite etc.
837 - fix AnyEvent::Handle::unshift_read to really unshift and not push.
838 - AnyEvent::Handle could sometimes stumble over 0-byte writes.
839 - fix unpack_sockaddr, now works for AF_INET6 sockaddr's.
840 - updated documentation.
841 - lots of workarounds for perl 5.8 quirks.
842 - implement a more flexible type system for both reads and writes
843 in AnyEvent::Handle.
844 - support IPv6 in tcp_server.
845 - pass local hostname and port to the prepare callback in tcp_server.
846 - make types extendable in AnyEvent::Handle.
847 - croak, not die, in AnyEvent::Handle;
848
8494.0 Sat May 24 19:58:08 CEST 2008
850 - added (experimental) AnyEvent::DNS module.
851 - added (experimental) AnyEvent::Socket module.
852 - added (experiemntal) support for TLS/SSL in AnyEvent::Handle.
853 - make condvar->end call ->send as documented.
854 - make condvar constructor support the documented cb => argument.
855 - overload condvars so you can use them instead of code references.
856 - support more of the POSIX forms instead of a pure dotted quad for
857 verbatim ip addresses.
858 - AnyEvent::Socket::inet_aton will use AnyEvent::DNS now and not fork anymore.
4 - try very hard not to freeze on broken windows perls. 859 - try very hard not to freeze on broken windows perls.
5 - added (experimental) AnyEvent::DNS module.
6 - make on_eof optional in AnyEvent::Handle. 860 - make on_eof optional in AnyEvent::Handle.
7 - support service names in tcp_connect and tcp_server. 861 - support service names in tcp_connect and tcp_server.
8 - work around netbsd/openbsd bug causing unix sockets not to 862 - work around netbsd/openbsd bug causing unix sockets not to
9 become writable untill empty. 863 become writable until empty.
10 - only make a dummy read for error detection when getpeername 864 - only make a dummy read for error detection when getpeername
11 returns ENOTCONN. 865 returns ENOTCONN.
12 - defend AnyEvent::Util::inet_aton against Coro::LWP brutal 866 - defend AnyEvent::Util::inet_aton against Coro::LWP's brutal
13 override. 867 override.
868 - new AnyEvent::Guard::cancel.
14 869
153.5 Sat May 17 23:17:14 CEST 2008 8703.5 Sat May 17 23:17:14 CEST 2008
16 - gracefully deal with systems not defining CLOCK_MONOTONIC 871 - gracefully deal with systems not defining CLOCK_MONOTONIC
17 (as opposed to not supporting the clock itself, which worked in 3.41). 872 (as opposed to not supporting the clock itself, which worked in 3.41).
18 - added AnyEvent::Util::tcp_connect and AnyEvent::Util::tcp_server 873 - added AnyEvent::Util::tcp_connect and AnyEvent::Util::tcp_server
… …
30 885
313.41 Sun May 11 19:53:13 CEST 2008 8863.41 Sun May 11 19:53:13 CEST 2008
32 - work around a bug in perl 5.8.8's local. Reported by Yi Ma Mao. 887 - work around a bug in perl 5.8.8's local. Reported by Yi Ma Mao.
33 888
343.4 Sun May 11 00:29:25 CEST 2008 8893.4 Sun May 11 00:29:25 CEST 2008
35 - complete re-implementation of condvars with a hopefully much mroe useful 890 - complete re-implementation of condvars with a hopefully much more useful
36 API (backward compatible functions exist but are undocumented). 891 API (backward compatible functions exist but are undocumented).
37 - AnyEvent::Handle fully rewritten with a hopefully more powerful API 892 - AnyEvent::Handle fully rewritten with a hopefully more powerful API
38 (still subject to change). 893 (still subject to change).
39 - take advantage of CLOCK_MONOTONIC, if available, in AnyEvent::Impl::Perl. 894 - take advantage of CLOCK_MONOTONIC, if available, in AnyEvent::Impl::Perl.
40 - provide AnyEvent::post_detect and @AnyEvent::post_detect, which 895 - provide AnyEvent::post_detect and @AnyEvent::post_detect, which

Diff Legend

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