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

Comparing AnyEvent/Changes (file contents):
Revision 1.91 by root, Sat May 24 01:15:19 2008 UTC vs.
Revision 1.454 by root, Tue Aug 2 22:30:00 2011 UTC

1Revision history for Perl extension AnyEvent. 1Revision history for Perl extension AnyEvent.
2 2
3TODO: preferences, retry-callback, memleaks 3TODO: document TCP_*** constants
43.51 4TODO: catch AE::xxx for anyevent strict and anyevent debug
5TODO: anyevent::trace? or debug tracing all watchers?
6TODO: anyevent strict checking for $_ changes?
7
8TODO: Kevin Ryde (3.8K) 1 inbox Glib child processes
9 - INCOMPATIBLE CHANGE: AnyEvent::DNS resource records now include
10 the TTL as fourth element - this affects ->resolve directly,
11 and ->request and ->dns_unpack indirectly. Convenience functions
12 such as AnyEvent::DNS::{a,txt,mx,srv} are not affected.
13 - INCOMPATIBLE CHANGE: backend authors now should not implement
14 one_event or loop, but instead the AnyEvent::CondVar::_wait method.
15 - some read types (regex, netstring, json, storable) could
16 errornously keep an AE::Handle object alive on their own
17 (reported by Mohammad Toossi).
18 - add more workarounds around child watcher bugs in IO::Async
19 and (newly introduced) in POE.
20 - use glib child watchers and try to work around its limitations,
21 if possible (based on a patch by Kevin Ryde).
22 - add support for (currently third-party) AnyEvent::Impl::FLTK.
23 - many read types in AnyEvent::Handle would malfunction during
24 transitions to TLS because of a "delete $self->{rbuf}".
25 (reported by Gerald Galster).
26 - resolve AnyEvent::Util::guard at first-use-time, not at load time,
27 for further speed and memory savings when not used.
28 - improve condvar blocking wait performance (probably only noticable
29 with faster backends), also make it easier for event loops that
30 do not support blocking waits.
31
325.34 Thu May 12 09:59:41 CEST 2011
33 - AE::TLS didn't support both cert_file and key at the same time,
34 due to some copy&paste bug (reported and analyzed by vti).
35 - improve AE::Handle timeout documentation slightly.
36
375.33 Mon Feb 7 21:51:41 CET 2011
38 - never appeared on CPAN, for unknown reasons.
39 - AnyEvent::Handle on_connect_error - do not try to destroy a handle
40 object when it's already gone (analyzed by Ryan Bullock).
41 - update idna algorithm(s) to UTS#46 version 6.0.0.
42 - preserve trailing dot in idn_to_ascii - testsuite says to preserve,
43 uts46 says it's an error, and testsuite makes more sense.
44 - remove leading dots in idn_to_ascii - testsuite says to remove,
45 uts46 says it's an error, and testsuite makes more sense.
46 - fixed a bug in idn ignored character processing.
47 - passes the uts46 6.0.0 testsuite except for three cases which
48 apparently are bugs in the testsuite itself.
49 - passes the uts46 6.0.1 testsuite except for eleven cases which
50 apparently are bugs in the testsuite itself.
51
525.31 Sun Jan 23 11:49:19 CET 2011
53 - work around a POE bug where POE would first poll for new events and
54 invoke handlers BEFORE handling the message queue.
55 - fix parsing of "domain" statement in resolv.conf (Johannes Plunien).
56 - stop_read now does just that in TLS mode. see its
57 documentation for caveats.
58 - update to common::sense 3.4 warning set (some warnings in perl impose
59 an arbitrary 32kb limit on regex matches, which affects
60 AnyEvent::Handle).
61 - AE::Handle had a broken rbuf_max method and missing wbuf_max method.
62 - new experimental AE::Handle::resettls method.
63
645.3 Fri Dec 31 05:48:13 CET 2010
65 - major IOAsync patch by Paul Evans - 0.33 is required, as the major
66 issues with IO::Async have been solved in that version, including
67 the availabilty of a default loop!
68 - implement new wbuf_max parameter for AnyEvent::Handle.
69 - added Cocoa::EventLoop backend by Daisuke Murase.
70 - the default testsuite now honors PERL_ANYEVENT_MODEL (Paul Evans).
71 - clarify handle on_prepare documentation.
72 - set CLOEXEC flag on parent-side fd's in AnyEvent::Util::run_cmd,
73 to avoid leaking them to any child processes.
74 - try to force files into the arch-dependend lib directory, for the
75 half-"turly obsessive compulsive person"(s) that depend on it.
76
775.29 Sun Dec 5 10:49:21 CET 2010
78 - convert EV backend to EV 4.00 API (so better upgrade EV too).
79 - AE::Handle: implement dynamic read_size adjustment, add
80 max_read_size parameter, reduce default read_size to 2048.
81 - add some FAQ entries for servers and/or on_eof vs. on_error.
82 - work around OS bugs (cygwin again) when lingering in AE::Handle.
83 - reduce memory usage slightly when connect was used in AE::Handle.
84 - wrok around more segfaults in Tk.
85 - document tls_ctx => undef as valid.
86 - detect WNOHANG value at compiletime.
87 - include some probably linux-only support for building AnyEvent
88 as part of the perl core.
89 - improve compatibility of t/08_idna.t to perl 5.8.
90 - make AnyEvent::Strict truly optional (in case somebody builds a
91 minimal perl...).
92
935.28 Wed Oct 13 04:14:23 CEST 2010
94 - due to a glitch, AnyEvent's internal getprotobyname wasn't
95 actually used in tcp_server.
96 - implement AnyEvent::CondVar->new method, which was documented but
97 missing.
98 - support autoloading of read types in unshift_read.
99 - AnyEvent::Strict: do the test for invalid fh arguments differently,
100 resulting in more informative error messages.
101 - new AnyEvent::FAQ manpage.
102 - clarify that you can actually call push_read etc. in on_read.
103 - put keys and certs in /tmp, not into the current directory,
104 and document this in the SECURITY CONSIDERATIONS sections
105 of AnyEvent::TLS.
106 - support ";" as resolv.conf comment character in AnyEvent::DNS, as
107 requested by Ron Isaacson.
108 - document $AnyEvent::DNS::RESOLVER variable.
109 - incorporated major typo patches by Abhijit Menon-Sen.
110 - AnyEvent::Handle now croaks when negative timeouts are passed.
111 - add a shitload of TCP_xxx sockopt name constants to AnyEvent::Util,
112 undocumented, but who knows when they might come in handy.
113
1145.271 Tue Jun 8 12:05:46 CEST 2010
115 - backport to perl 5.8.x.
116
1175.27 Sun Jun 6 12:12:05 CEST 2010
118 - postpone differently in AnyEvent::Socket now, as
119 when not, canceling the connection attempt might fail
120 (found by Felix Antonius Wilhelm Ostmann).
121 - explicitly check for non-stream sockets in AE::Handle, too many
122 clueless people fell into the trap of this somehow working.
123 - simplified and reworked the "OTHER MODULES" section.
124 - better/more condvar examples.
125
1265.261 Wed Apr 28 16:13:36 CEST 2010
127 - AF_INET6 was not properly used from Socket6 during configuration
128 time (found by Andy Grundman).
129
1305.26 Mon Apr 12 04:49:35 CEST 2010
131 - don't generate spurious readyness notifications when select
132 returns EINTR in the pure perl backend (analysed and testcase
133 by Bas Denissen).
134 - give same examples in SYNOPSIS sections of AnyEvent and AE.
135 - provider faster implementations for AE::time/now/now_update for
136 EV and Perl backends.
137 - provide faster AE::cv and AE::time implementations for all
138 backends.
139 - no longer support lower/mixed-case signal names (this was
140 never documented nor universally supported).
141 - some more memory µ-optimisations, and somewhat less messy
142 code to cope with both AnyEvent and AE APIs.
143
1445.251 Sat Mar 13 00:58:21 CET 2010
145 - make test failed if it was run before make install on perl <=
146 5.8.8, and older versions of perl make our live very difficult,
147 so write constants.pl during Makefile.PL time.
148
1495.25 Sat Mar 13 00:23:14 CET 2010
150 - fix a race condition in AnyEvent::Handle that would cause
151 a "bio_read: unsupported method" error (in _another_
152 ssl connection) after user code threw an exception.
153 - added AnyEvent::Handle->destroyed method.
154 - speed up AnyEvent::Socket::format_address and ::format_ipv6.
155 - the AnyEvent::Util::fh_nonblocking function worked only
156 by lucky accident on win32.
157 - smaller and faster AnyEvent::Util::fh_nonblocking.
158 - when the (required!) Time::HiRes module is missing, AnyEvent
159 did not fall back to built-in time properly.
160 - do not load Fcntl at runtime, saving memory and loading time.
161 - precompile a number of constants and use them instead of runtime
162 detection and eval.
163 - free detection code after detection and similar memory optimisations.
164 - Perl backend timer interval best effort drift has been improved
165 (same algorithm as EV).
166 - update unicode idna mapping table.
167
1685.24 Tue Jan 5 11:39:43 CET 2010
169 - cygwin never reports errors from failed connects. 1.5 just gives
170 you continous readyness and EAGAIN, 1.7 is even more broken and
171 just hangs. work around both issues in a cygwin-specific hack.
172 - improve idle watcher documentation slightly.
173
1745.23 Sun Dec 20 23:48:00 CET 2009
175 - support IDNs in resolve_sockaddr, and therefore in tcp_connect.
176 - implement punycode_encode/decode, idn_nameprep,
177 idn_to_ascii and idn_to_unicode operations in AnyEvent::Util.
178 - provide $AE::VERSION.
179 - removed traces of "no strict 'refs'".
180
1815.22 Sat Dec 5 03:51:13 CET 2009
182 - downgrade-or-fail in AnyEvent::Handle::push_write, to
183 diagnose encoding failures earlier and more succinctly.
184 (this works around bugs in perl, throwing away encoding info
185 when passing scalar data to extensions).
186 - add more examples to AnyEvent::Socket manpage.
187 - upgrade internal warning set to the same as common::sense 2.03.
188 - use pack "n/a*" for pre-5.8.9 perl compatibility in AnyEvent::DNS
189 (John Beppu).
190 - AnyEvent::Socket::inet_aton now properly supports ipv6, as documented.
191 - add google public dns servers to fallback server set and make sure
192 we load-balance properly between the three sets. also add all
193 fallback dns servers, not just a random one, to each dns config.
194
1955.21 Thu Nov 19 02:48:47 CET 2009
196 - fix a problem where socket constants were called with parameters
197 (spotted by David Friedland).
198 - fork_call never use'd POSIX (reported by Daisuke Maki).
199 - improve perl 5.6 compatibility further (but it still won't work
200 unless you rip out everything but the core).
201 - prefer Net::DNS::Resolver over ipconfig, if installed, on win32.
202 uses 10MB of RAM, but doesn't open a console window. *sigh*.
203
2045.202 Wed Oct 14 22:35:44 CEST 2009
205 - AnyEvent::DNS would unexpectedly clobber $_ under windows
206 (analysed by Matthias Waldorf).
207 - AnyEvent::Handle::run_cmd can now pass the PID of the
208 newly-created process, which is much less useful than it might
209 sound (based on patch by Yann Kerherve).
210
2115.201 Tue Sep 29 12:09:25 CEST 2009
212 - AnyEvent:Handle::on_starttls/on_stoptls methods were broken
213 (reported by Torsten Foertsch).
214 - common::sense 2.0 could cause tcp_server to throw an exception
215 (analysed by elmex).
216
2175.2 Mon Sep 14 07:04:49 CEST 2009
218 - INCOMPATIBLE CHANGE: do no longer support register_read_type
219 and register_write_type in AnyEvent::Handle, instead support
220 package names (the facility was mostly abused).
221 - implement "packagename-as-read/write type" support in
222 AnyEvent::Handle.
223 - AnyEvent::Handle: new options "keepalive" and "oobinline".
224 - oobinline set by default to avoid security issues.
225 - the pure-perl event loop backend wrongly detected times() fallback
226 support (spotted by Pavel Boldin).
227
2285.12 Tue Sep 1 20:26:50 CEST 2009
229 - be more lenient when parsing resolv.conf files, as some people
230 use hashmarks as comment indicator inside directives (reported
231 by Michael S. Fischer).
232 - use same set of warnings as common::sense 2.0.
233 - fix a potential 32 bit overflow issue due to perl having problems
234 with large hex constants in the Perl backend.
235 - do not use the slower 5.6 version fo the enc_name code in AnyEvent::DNS
236 by default.
237 - fix some prototypes.
238 - (optionally) require Async::Interrupt 1.02.
239 - replace opendns fallbacks by verizon/level3 public dns servers.
240
2415.112 Fri Aug 21 13:59:04 CEST 2009
242 - AnyEvent::Handle could cause hard crashes in Net::SSLeay when
243 using starttls with an illegal/uncreatable tls context.
244 - port to broken win32 perls.
245
2465.111 Sun Aug 16 18:44:36 CEST 2009
247 - perl errornously requires sockaddr_un structures returned by the
248 kernel to have a specific length. work around this bug by padding
249 sockaddr_un structures (found to fail on OS X, as perl only uses
250 the correct code for GNU/Linux, but likely to fail on other
251 OSes as well).
252
2535.11 Wed Aug 12 17:49:37 CEST 2009
254 - fixed a bug in run_cmd with input from filename.
255 - create a json coder object when none is given in write direction
256 to reduce dependency on JSON.
257
2585.1 Tue Aug 11 03:17:32 CEST 2009
259 - tcp_connect now reports the correct error in $!, not an unrelated
260 one (was broken since 4.91).
261 - AE::Handle did not properly clear rtimeout/wtimeout watchers
262 when retrying a tcp_connect.
263 - new functions: AnyEvent::Util::run_cmd and close_all_fds_except.
264 - fixed wrong documentation in AE::Socket.
265
2665.01 Mon Aug 10 03:16:32 CEST 2009
267 - last release broke EV child watchers :/.
268
2695.0 Sun Aug 9 17:32:09 CEST 2009
270 - officially introduce and document the AE API.
271 - lowering the timeout at runtime did not have immediate effect
272 in AnyEvent::Handle.
273 - AE::Handle now has separate and independent read and write timeout
274 settings.
275 - probe for Guard module when AnyEvent::Util is loaded, not at
276 runtime (Event doesn't like eval "use" from callbacks, and it avoids
277 exporting the probe function).
278 - Impl::IOAsync's io watchers did not use the filehandle emulation
279 code to work around it's watcher limits.
280 - work around IO::Async's problems with overloaded objects.
281 - try to disable wearnings inside IO::Async, it's just too buggy.
282 - do no longer use Event by default, only EV or Perl, Event is too buggy.
283 - hackishly provide fast AE:: interface when EV or Perl is the backend.
284 - inherit VERSION from AnyEvent::VERSION in most modules.
285
2864.91 Thu Aug 6 15:42:45 CEST 2009
287 - AE::Handle::starttls could get out of sync when the read buffer
288 already contains some TLS handshake.
289 - AE::Handle did not properly free the TLS session, causing bigger
290 memory leaks in Net::SSLeay than expected (reported by toaster).
291 - AE::Socket::tcp_connect will now artificially delay invoking
292 the callback to avoid returning after invoking the callback.
293 - convert many internal watcher uses to AE API.
294
2954.9 Sat Aug 1 11:07:01 CEST 2009
296 - Glib silently fails when registering a timeout with
297 a negative value, so avoid that.
298 - call condvar callback immediately when it is set
299 after the condition is already signalled.
300 - check rbuf_max condition only after trying to consume
301 data.
302 - normalise signal numbers to names when using the ae's
303 signal handling, but do not document this (yet).
304 - pure perl signal emulation did not properly set nonblocking
305 mode on the signal pipe on win32 (this is unlikely to have caused
306 any issues).
307 - new module: AnyEvent::Debug.
308 - AnyEvent::Strict now checks that a signal was specified
309 by name (not name or number).
310 - reduce memory footprint in typical cases by ~50kb by
311 only compiling signal code when necessary.
312 - add AnyEvent::Handle->rbuf_max.
313 - grab the AE namespace for future new API, implement stubs for
314 said future/alternative API.
315 - new function: AnyEvent::Socket::format_hostport.
316
3174.881 Tue Jul 28 12:51:53 CEST 2009
318 - work around a bug in local in pre-5.10 perls, causing
319 AnyEvent::Handle to recurse when it should not
320 (analysed by elmex).
321
3224.88 Tue Jul 28 04:04:37 CEST 2009
323 - re-bless the handle into a dummy package after calling
324 AnyEvent::Handle::destroy, so the user does not need to check
325 for errors after every push_write etc.
326 - do not attempt to run t/02_signals.t on obviously broken
327 platforms.
328
3294.87 Sun Jul 26 02:06:16 CEST 2009
330 - do not attempt to linger when there is no longer a valid fh.
331 - enforce tls mode to be either accept or connect, do not simply
332 segfault in Net::SSLeay.
333 - AnyEvent::Handle can now call tcp_connect itself (new parameters
334 connect, on_prepare, on_connect and on_connect_error). Updated
335 tutorial accordingly.
336 - add AnyEvent::Impl::Irssi backend.
337
3384.86 Mon Jul 20 23:52:29 CEST 2009
339 - since the verbose warning is not enough, explicitly document
340 that versions before 1.33 of Net::SSLeay are not secure.
341 - work around signal handling races in Event and (...) Event::Lib.
342 - try to align signal-race timer to full-second boundaries.
343 - work around Tk not liking negative timeouts.
344 - don't complain of different grades of environmental unfriendlyness
345 in IO::Async.
346
3474.85 Sat Jul 18 06:16:14 CEST 2009
348 - nail the signal race problem in perl once and for all
349 (see $AnyEvent::MAX_SIGNAL_LATENCY).
350 - take advantage of Async::Interrupt if it is available.
351 - load Time::HiRes and Guard modules on demand only.
352 - add optional/recommended modules section to AnyEvent
353 documentation.
354 - reduce memory usage considerably (and reduce startup penalty)
355 by not using "strict", "warnings" and "overload" modules.
356 - work around buggy windows/openbsd perls and provide EBADMSG
357 and EPROTO ourselves when missing.
358 - improve perl 5.6 compatibility of the core event loop.
359 - made Net::SSLeay version 1.33 a soft requirement.
360
3614.83 Fri Jul 17 16:56:26 CEST 2009
362 - implement AnyEvent::Socket::getprotobyname.
363 - AnyEvent::CondVar's will now detect recursive blocking
364 waits and will croak, as too many people fall into
365 this trap.
366 - AnyEvent::Handle will now call ->destroy on itself after
367 executing the on_error callback, instead of doing some
368 half-baked internal shutdown, for fatal errors.
369 - clarify on_eof behaviour w.r.t. the read queue and
370 on_read callbacks.
371 - ignore some possible spurious wake-ups in tcp_connect.
372
3734.82 Sat Jul 11 00:34:55 CEST 2009
374 - POE and Event backends didn't accept some callable objects as
375 callbacks.
376 - use Config module instead of POSIX module to detect signal names
377 in AnyEvent::Strict and AnyEvent::Impl::EventLib, as the POSIX
378 module doesn't even have all POSIX signals :/.
379 - use more workarounds around the many refcnt/corruption bugs in
380 Event::Lib.
381 - work around a race condition in perl's select, causing t/03_child.t
382 to rarely fail.
383
3844.81 Thu Jul 9 10:30:30 CEST 2009
385 - AnyEvent::Handle didn't properly diagnose write errors
386 (it expected -1 from syswrite, how lame... :).
387 - support file descriptors in addition to file handles
388 in AnyEvent->io.
389 - new env variables: PERL_ANYEVENT_RESOLV_CONF,
390 PERL_ANYEVENT_MAX_OUTSTANDING_DNS, PERL_ANYEVENT_CA_FILE
391 and PERL_ANYEVENT_CA_PATH.
392 - provide a sensible synopsis section for AnyEvent::TLS.
393 - add a "supported backends" section to the manpage.
394 - added simple io watcher test to testsuite, using a
395 portable_socketpair.
396 - tried to improve the stability of the Event::Lib backend,
397 YMMV.
398
3994.8 Mon Jul 6 23:45:16 CEST 2009
400 - AnyEvent::DNS did not properly follow CNAME records with
401 uppercase targets.
402 - AnyEvent::DNS would errornously return AAAA records
403 with v4 mapped addresses (a faulty record) as ipv4 addresses,
404 causing AnyEvent::Socket to throw an exception.
405 - added new module AnyEvent::TLS for easier SSL/TLS context
406 creation, with many options including hostname verification,
407 secure default configuration, lots of documentation and,
408 predefined diffie-hellman keys for perfect forward security
409 and much more. get it while it's still fresh!
410 - use AnyEvent::TLS in AnyEvent::Handle for context management.
411 - load AnyEvent::Handle only on demand in AnyEvent::DNS,
412 so AnyEvent::Socket users have smaller memory footprint
413 in the common case.
414 - add AnyEvent::Handle->push_shutdown method.
415 - add an additional error message parameter to AnyEvent::Handle's
416 on_error callback (for TLS, $! is still available).
417 - add AnyEvent::Handle on_starttls/on_stoptls callbacks.
418 - make AnyEvent::Handle more robust against early conenction
419 failures (during new), and return C<undef> in that case
420 from the constructor.
421 - AnyEvent::Handle will now try to load only JSON::XS first,
422 then fall back to JSON.
423 - format_ipv4/format_ipv6 are now exported by default, for symmetry,
424 and because it was documented that way.
425
4264.451 Fri Jul 3 00:28:58 CEST 2009
427 - do not clear rbuf when shutting down an AnyEvent::Handle
428 object - doing so breaks AnyEvent::HTTP.
429
4304.45 Mon Jun 29 22:59:26 CEST 2009
431 - a write error could cause AnyEvent::Handle to create
432 an I/O watcher with an undefined $fh.
433 - special-case mapped ipv4 addresses in both
434 AnyEvent::Socket::format_address and parse_address,
435 to treat them just like ipv4 addresses.
436 - updated and overhauled the AnyEvent::Intro doc.
437 - implement AnyEvent::Socket::format_ipv4/ipv6.
438 - slightly speed up portable_pipe/socketpair.
439 - expand condvar begin/end documentation.
440
4414.42 Fri Jun 26 08:32:18 CEST 2009
442 - preliminary and neccesarily incomplete support for IO::Async.
443 - reset SIGCHLD to DEFAULT when AnyEvent is loaded, in
444 case it was set to IGNORE, to ensure we can catch child statuses
445 even when the calling env acts stupidly.
446 - updated benchmarks with IO::Async, which performs very well.
447
4484.412 Wed Jun 24 01:35:57 CEST 2009
449 - support an "untaint" attribute for AnyEvent::DNS
450 and set it on the default resolver.
451 - implement and document AnyEvent::Impl::Perl::loop.
452 - remove all anyevent-env variables from %ENV
453 when running in tainted mode.
454 - mention and extend the IO::Lambda benchmark.
455
4564.411 Sun Jun 7 18:48:02 CEST 2009
457 - do not try to use F_SETFD on windows.
458
4594.41 Thu May 14 06:40:11 CEST 2009
460 - work around issues in older perls (5.8.0?) when
461 a signal handler is deleted from the %SIG hash.
462 - use POSIX::_exit in child test, to avoid running
463 destructors.
464 - speed up CHLD handling by relying on SIGCHLD being
465 synchronously delivered, even when we roll our own
466 implementation.
467 - AnyEvent::DNS: add the "dname" resource record name
468 for cosmetic reasons.
469
4704.4 Sun Apr 26 20:12:33 CEST 2009
471 - implemented idle watchers, where applicable.
472 - AnyEvent->time died when Event backend was in use.
473 - fix a memleak in the Tk backend.
474 - sped up Tk timer handling.
475 - clip negative "after" values to 0 in AnyEvent::Impl::Event
476 to avoid spamming out warning messages.
477 - fix Qt timers without interval.
478 - avoid Qt zero-timeout special case, in old libqt's.
479
4804.352 Mon Apr 20 16:31:11 CEST 2009
481 - fix AnyEvent::Strict error messages for child watchers.
482 - fix/update Makefile.PL "recommends" versions.
483 - add AnyEvent->now_update.
484
4854.351 Sat Apr 11 07:56:14 CEST 2009
486 - actually make the signal pipe work under win32.
487 - localise $! in signal handler to avoid changing $!, although
488 perl itself does not seem to save/restore errno either.
489 - set the cloexec flag on the signal pipe (normally set by perl too).
490
4914.35 Fri Mar 27 11:48:20 CET 2009
492 - event models relying on AnyEvent's signal watcher emulation did
493 invoke the callback asynchronously, contradicting documentation
494 and causing signals to get lost (this includes AnyEvent's own
495 event loop). AnyEvent now uses the standard pipe trick to make
496 callback execution synchronous to the event loop.
497 - AnyEvent::Handle didn't free TLS context data on DESTROY
498 (patch by Pavel Shaydo).
499 - work around the ever-incompatibly-changing API of MakeMaker.
500 - document that changing global variables without restoring them
501 is a bad idea in Perl (noted by Adam Rosenstein).
502 - AnyEvent::Strict now barfs if ->io is passed a file.
503
5044.34 Thu Feb 12 18:32:45 CET 2009
505 - separately buffer TLS read data, as otherwise the read queue could
506 deadlock as receiving data is not expected while draining the
507 read queue (which cna only happen with TLS).
508 - raise EBADMSG error on JSON decoding errors.
509 - fix some minor manpage bugs (reported by Maximilian Gaß).
510 - speed up select bitmask parsing quite a bit in the pure perl
511 backend.
512 - use CORE::select instead of just select in AnyEvent::Impl::Perl.
513
5144.331 Tue Jan 6 21:07:25 CET 2009
515 - socketpair fails on many vista machines because vista has
516 completely broken accept/getpeername and getsockname functions,
517 so we provide our own socketpair emulation that kind of works
518 (AnyEvent::Util::portable_pipe).
519 - new function: AnyEvent::Util::portable_socketpair.
520 - take advantage of the Guard module if it exists.
521
5224.33 Fri Nov 21 02:35:40 CET 2008
523 - AnyEvent::Strict did errornously flag a fileno of 0 as
524 illegal.
525 - reduce memory usage and slightly speed up the pure perl backend
526 by only storing the file descriptor, not the file handle.
527 - add missing autocork method to AnyEvent::Handle (reported
528 by Adam Rosenstein).
529 - AnyEvent::DNS->resolve errornously documented an $rcode result
530 argument, but there isn't (spotted by Henrik Krohns).
531 - the naked truth about Tk - it's basically unmaintained.
532
5334.32 Mon Nov 3 22:46:32 CET 2008
534 - fix AnyEvent::Socket::resolve_sockaddr to properly support
535 unix sockets again and choose a proper default for
536 the socket type.
537 - fix call to resolve_sockaddr in tcp_connect.
538
5394.31 Thu Oct 30 04:41:48 CET 2008
540 - implemented AnyEvent::Handle->destroy method.
541 - hint about unexpected effects in TLS mode.
542 - speed up AnyEvent::Impl::Perl by using more arrays
543 instead of hashes without the slightest loss of readability :)
544 - work around a perl argument refcounting bug.
545 - with some perl patches applied, AnyEvent now seems leak-free.
546
5474.3 Fri Oct 3 09:18:43 CEST 2008
548 - AnyEvent will now install a no-op signal handler for SIGPIPE,
549 unless one has been installed already.
550 - warn about not loaidng AnyEvent::Impl::POE early enough
551 (patch by Adam Rosenstein).
552 - fixed a great number of bugs and corner cases in AnyEvent::Handle.
553 - fix a bug in where in SSL connect mode, the
554 client would first wait for some data by the server and
555 otherwise hang. (reported and analysed in an absolutely
556 exemplary manner by Adam Rosenstein).
557 - fix a bug in where SSL EOF would not be treated
558 as stream EOF, putting the connection into a hung state.
559 (reported and analysed in an absolutely exemplary manner by Adam
560 Rosenstein).
561 - fix a potential segfault when the TLS context
562 would go missing in a read callback (e.g. due to stoptls) -
563 Net::SSLeay of course makes no type checking whatsoever.
564 - AnyEvent::Handle will not stop the read watcher in TLS mode.
565 - AnyEvent::Handle->stoptls will send a TLS close notify
566 instead of simply destroying the stream now.
567 - fix a bug where large blocks of data written in TLS mode
568 would not be sent unless triggered by receives.
569 - on_drain will now take the tls write buffer into account.
570 - SSL operations have been streamlined a bit and should be faster
571 now, more to come, though.
572 - added a FAQ section, will document nontrivial issues in there.
573
5744.234 Mon Sep 29 04:08:13 CEST 2008
575 - fix child watcher documentation: only child exits will be reported,
576 no trace events.
577 - mention SIGPIPE in AnyEvent::Handle.
578 - perl backend will now use POSIX::times as monotonic clock
579 when available and the monotonic clock option is not.
580 - run even without Time::HiRes in some cases now, but do not
581 provide sub-second accuracy (all *supported* perl versions have
582 Time::HiRes).
583
5844.233 Fri Aug 22 01:48:05 CEST 2008
585 - fix a bug in the testsuite, causing a hang.
586 - clarified lots of AnyEvent::Handle constructor arguments.
587
5884.232 Thu Aug 21 20:44:25 CEST 2008
589 - fix a bug in AnyEvent::Handle that could cause two fatal error
590 callbacks on EOF, which caused AnyEvent::HTTP to signal success twice
591 (testcase by Leon Brocard).
592 - clarified on_eof/on_error documentation a bit.
593
5944.231 Tue Jul 29 13:12:15 CEST 2008
595 - remove some debugging code left in AnyEvent::Util::fork_call
596 (and no, it's impossible to implement with the broken windows
597 perls without resource leaks or worse).
598
5994.23 Tue Jul 29 12:19:59 CEST 2008
600 - document the first parameter passed to condvar callbacks to be
601 the callback.
602 - add AnyEvent::Socket::{ntoa,aton} aliases.
603 - optimize the AE::Handle->push_read (line) for the default
604 eol marker.
605 - optimize push_read (packstring|storable) for small packets.
606 - invoke on_error callback when no on_eof callback is set.
607 - fix a bug in push_read (storable) of unknown impact.
608
6094.22 Sun Jul 20 16:34:13 CEST 2008
610 - new function AnyEvent::Socket::parse_hostport.
611 - as the bulkheads at microsoft can't even get getprotobyname reliably
612 working on their shitty fucking broken os we need to hardcode
613 some common protocol numbers in AnyEvent::Socket. How can
614 people even bother with such a pile of shit as windows.
615
6164.21 Thu Jul 17 14:40:05 CEST 2008
617 - INCOMPATIBLE CHANGE: AnyEvent::Util::fork_call now has a prototype
618 which is incompatible to previous invocation syntax.
619 - work around more windows perl bugs in fork_call. I cannot imagine
620 why anybody would bother implementing fork in such an obviously
621 unusable way.
622 - avoid calling $do_retry if the request has already finished in
623 AnyEvent::DNS, thus avoiding an exception (reported by Anatoly
624 K. Sharifulin).
625
6264.2 Sat Jul 12 22:42:11 CEST 2008
627 - support an "interval" argument for repeating timers.
628 - fix a bug in the pure perl event loop that caused timers to
629 be delayed under some circumstances.
630 - implement optional argument checking via PERL_ANYEVENT_STRICT.
631 - not importing anything from Event causes it to malfunction
632 and not use Time::HiRes, work around this bug.
633 - more bug workarounds against the endlessly broken Tk module,
634 seems to work now on 32 bit machines, but less so on 64 bit.
635 - minor optimisations applied to most backends.
636
6374.161 Fri Jul 4 14:24:48 CEST 2008
638 - fixed the t/handle/02_write.t, it read after write, with the
639 assumption that all data could be written before read
640 (reported by Jim Razmus).
641
6424.160 Thu Jul 3 04:02:21 CEST 2008
643 - re-registering signal watchers after unregistering one did
644 not work if a backend falls back on AnyEvent's default
645 implementation (also affected child watchers).
646 - new AnyEvent::Handle options: autocork and no_delay.
647
6484.152 Sun Jun 22 14:15:44 CEST 2008
649 - allow for 32-bit perls that implement shifts differently
650 on different architectures in parse_ipv4 (reported and
651 analysed by Keiichi DAIBA).
652
6534.151 Fri Jun 6 17:34:24 CEST 2008
654 - make sure specifying _only_ on_read and never pushing reads
655 works.
656
6574.15 Fri Jun 6 13:00:46 CEST 2008
658 - the pure perl backend would keep some watchers alive when more than
659 one watcher was registered for the same fd.
660 - new "packstring" and "storable" read and write types
661 for AnyEvent::Handle.
662 - allow on_eof handler to be called after on_error with EPIPE returns.
663 - do not immediately call on_read callback in handle constructor.
664
6654.14 Thu Jun 5 20:29:31 CEST 2008
666 - Fixed a bug in DNS SRV priority sorting.
667 - AnyEvent::Util::guard now reports runtime errors while
668 executing the guard block as warnings.
669 - handle 0-byte-reads just before EOF correctly in AnyEvent::Handle.
670
6714.13 Thu Jun 5 00:47:59 CEST 2008
672 - AnyEvent::DNS only followed cname chains with length 2,
673 contrary to documentation. bumped it up to 10, thanks to
674 microsoft, the current broken-dns-config-king.
675 - AnyEvent::DNS didn't check for socket return status
676 but instead relied on perl not creating filehandles in that
677 case - too bad it gives you a bogus file handle (reported
678 and analysed by Vladimir Timofeev).
679 - fix queue management logic in AnyEvent::Handle:
680 when on_read was registered, the queue was empty
681 and no progress could be made AnyEvent::Handle would
682 enter an endless loop.
683 - correctly start reading again when the handle became
684 busy again after idling.
685 - correctly treat tls shut-downs as EOF condition
686 (for the time being, would be nice to have a callback for that).
687 - correctly call eof callback under all conditions
688 (hopefully).
689 - the timeout callback did not expect that $self can go
690 away any time.
691
6924.12 Tue Jun 3 10:58:04 CEST 2008
693 - include AnyEvent::Intro, a tutorial for anyevent,
694 anyevent::socket and anyevent::handle.
695 - allow more options in on_error.
696
6974.11 Fri May 30 23:42:25 CEST 2008
698 - INCOMPATIBLE CHANGE: replace ptr by real PTR lookup, provide
699 reverse_lookup and reverse_verify to replace it, support
700 v4mapped and v4compat addresses.
701 - provide more documentation for the resolver class.
702 - really replace longest run of :0: by :: in format_address,
703 also properly convert :: and ::1 again.
704 - support NAPTR record name and decode it.
705 - implement random weight sampling for SRV records, as per
706 rfc 2782.
707 - correctly abort on srv-record targets of ".".
708 - added AnyEvent::DNS::wait_for_slot.
709 - in the unlikely event of a virtual circuit connection
710 being invalidated by a delayed udp reply, AnyEvent::DNS
711 could die. this has been fixed.
712 - plug a probable memleak in the DNS vc code.
713 - use configured timeout also for VC request phase.
714 - implement timeout and max_outstanding methods
715 for resolver class.
716 - update version numbers in all modules.
717
7184.1 Thu May 29 05:45:40 CEST 2008
719 - INCOMPATIBLE CHANGE: renamed xxx_ip functions to xxx_address
720 which mirrors their purpose better (old names still available
721 for a while). Also moved AnyEvent::DNS::addr to
722 AnyEvent::Socket::resolve_sockaddr.
723 - implement AnyEvent->time and AnyEvent->now.
724 - fix IPv6 support in pack_sockaddr.
725 - officially un-experimentalise all newly introduced code.
726 - support unix domain sockets everywhere by specifying
727 a host of "unix/" and the pathname as service.
728 - implement an activity timeout in AnyEvent::Handle.
729 - added a regex read type to AnyEvent::Handle.
730 - added a json read/write type to AnyEvent::Handle.
731 - always croak in on_error in AnyEvent::Handle.
732 - document how to subclass AnyEvent::Handle.
733 - implement AnyEvent::Util::fork_call.
734 - add support for IPv6 nameservers and nameserver statements.
735 - work around _yet_ _another_ windows perl bug in where empty
736 select masks cause errors in the pure perl backend (this
737 workaround cannot be implemented for the other event loops,
738 but EV also works around this).
739 - supply AnyEvent::Socket::address_type and make good use of it.
740 - clarify and fix the Handle documentation w.r.t $self vs. $handle
741 vs. callback arguments.
742 - add some recommends to the META.yml.
743
7444.05 Mon May 26 19:44:06 CEST 2008
745 - some platforms ignore the rfc and prepend an extra sa_len member to
746 the sockaddr structure, cater for those. also use sockaddr_family,
747 didn't know it was there, but comes in handy for the workaround.
748 - undo the import WIN32 hack.
749
7504.04 Mon May 26 08:03:31 CEST 2008
751 - try to work around yet another windows bug: failed connects are reported
752 as if it were out-of-band data. windows users: you suck. Only
753 EV, Glib, Event and the pure perl backend can handle this condition.
754 - optimize write algorithm in AnyEvent::Handle.
755 - properly parse PERL_ANYEVENT_PROTOCOLS and use the info correctly.
756 - tcp_connect now properly iterates through all targets.
757 - check wether IPv6 sockets can actually be created, otherwise, disable ipv6
758 support entirely. also hardcode some AF_INET6 constants for the
759 sake of perl 5.8.
760 - vastly improved nameserver/suffix detection algorithm for windows
761 (still a hack...)
762 - try to use a fallback dns server if no dns servers could be found.
763 - splatter around some local $SIG{__DIE__}'s to work around bad
764 die handlers.
765
7664.03 Sun May 25 05:05:57 CEST 2008
767 - work around perl bugs on windows, where perl returns undocumented
768 error codes for sysread, syswrite etc.
769 - fix AnyEvent::Handle::unshift_read to really unshift and not push.
770 - AnyEvent::Handle could sometimes stumble over 0-byte writes.
771 - fix unpack_sockaddr, now works for AF_INET6 sockaddr's.
772 - updated documentation.
773 - lots of workarounds for perl 5.8 quirks.
774 - implement a more flexible type system for both reads and writes
775 in AnyEvent::Handle.
776 - support IPv6 in tcp_server.
777 - pass local hostname and port to the prepare callback in tcp_server.
778 - make types extendable in AnyEvent::Handle.
779 - croak, not die, in AnyEvent::Handle;
780
7814.0 Sat May 24 19:58:08 CEST 2008
5 - added (experimental) AnyEvent::DNS module. 782 - added (experimental) AnyEvent::DNS module.
6 - added (experimental) AnyEvent::Socket module. 783 - added (experimental) AnyEvent::Socket module.
784 - added (experiemntal) support for TLS/SSL in AnyEvent::Handle.
7 - make condvar->end call ->send as documented. 785 - make condvar->end call ->send as documented.
8 - make condvar constructor support the documented cb => argument. 786 - make condvar constructor support the documented cb => argument.
787 - overload condvars so you can use them instead of code references.
9 - support more of the POSIX forms instead of a pure dooted quad for 788 - support more of the POSIX forms instead of a pure dotted quad for
10 verbatim ip addresses. 789 verbatim ip addresses.
11 - AnyEvent::Socket::inet_aton will use AnyEvent::DNS now and not fork anymore. 790 - AnyEvent::Socket::inet_aton will use AnyEvent::DNS now and not fork anymore.
12 - try very hard not to freeze on broken windows perls. 791 - try very hard not to freeze on broken windows perls.
13 - make on_eof optional in AnyEvent::Handle. 792 - make on_eof optional in AnyEvent::Handle.
14 - support service names in tcp_connect and tcp_server. 793 - support service names in tcp_connect and tcp_server.
15 - work around netbsd/openbsd bug causing unix sockets not to 794 - work around netbsd/openbsd bug causing unix sockets not to
16 become writable untill empty. 795 become writable until empty.
17 - only make a dummy read for error detection when getpeername 796 - only make a dummy read for error detection when getpeername
18 returns ENOTCONN. 797 returns ENOTCONN.
19 - defend AnyEvent::Util::inet_aton against Coro::LWP brutal 798 - defend AnyEvent::Util::inet_aton against Coro::LWP's brutal
20 override. 799 override.
21 - new AnyEvent::Guard::cancel. 800 - new AnyEvent::Guard::cancel.
22 801
233.5 Sat May 17 23:17:14 CEST 2008 8023.5 Sat May 17 23:17:14 CEST 2008
24 - gracefully deal with systems not defining CLOCK_MONOTONIC 803 - gracefully deal with systems not defining CLOCK_MONOTONIC
… …
38 817
393.41 Sun May 11 19:53:13 CEST 2008 8183.41 Sun May 11 19:53:13 CEST 2008
40 - work around a bug in perl 5.8.8's local. Reported by Yi Ma Mao. 819 - work around a bug in perl 5.8.8's local. Reported by Yi Ma Mao.
41 820
423.4 Sun May 11 00:29:25 CEST 2008 8213.4 Sun May 11 00:29:25 CEST 2008
43 - complete re-implementation of condvars with a hopefully much mroe useful 822 - complete re-implementation of condvars with a hopefully much more useful
44 API (backward compatible functions exist but are undocumented). 823 API (backward compatible functions exist but are undocumented).
45 - AnyEvent::Handle fully rewritten with a hopefully more powerful API 824 - AnyEvent::Handle fully rewritten with a hopefully more powerful API
46 (still subject to change). 825 (still subject to change).
47 - take advantage of CLOCK_MONOTONIC, if available, in AnyEvent::Impl::Perl. 826 - take advantage of CLOCK_MONOTONIC, if available, in AnyEvent::Impl::Perl.
48 - provide AnyEvent::post_detect and @AnyEvent::post_detect, which 827 - provide AnyEvent::post_detect and @AnyEvent::post_detect, which

Diff Legend

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