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

Comparing AnyEvent/Changes (file contents):
Revision 1.160 by root, Wed Jul 2 12:27:42 2008 UTC vs.
Revision 1.446 by root, Mon Jul 4 21:14:42 2011 UTC

1Revision history for Perl extension AnyEvent. 1Revision history for Perl extension AnyEvent.
2 2
34.160 3TODO: document TCP_*** constants
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: break the API and put the rr ttl where it belongs.
9TODO: Kevin Ryde (3.8K) 1 inbox Glib child processes
10 - add support for (currently third-party) AnyEvent::Impl::FLTK.
11
125.34 Thu May 12 09:59:41 CEST 2011
13 - AE::TLS didn't support both cert_file and key at the same time,
14 due to some copy&paste bug (reported and analyzed by vti).
15 - improve AE::Handle timeout documentation slightly.
16
175.33 Mon Feb 7 21:51:41 CET 2011
18 - never appeared on CPAN, for unknown reasons.
19 - AnyEvent::Handle on_connect_error - do not try to destroy a handle
20 object when it's already gone (analyzed by Ryan Bullock).
21 - update idna algorithm(s) to UTS#46 version 6.0.0.
22 - preserve trailing dot in idn_to_ascii - testsuite says to preserve,
23 uts46 says it's an error, and testsuite makes more sense.
24 - remove leading dots in idn_to_ascii - testsuite says to remove,
25 uts46 says it's an error, and testsuite makes more sense.
26 - fixed a bug in idn ignored character processing.
27 - passes the uts46 6.0.0 testsuite except for three cases which
28 apparently are bugs in the testsuite itself.
29 - passes the uts46 6.0.1 testsuite except for eleven cases which
30 apparently are bugs in the testsuite itself.
31
325.31 Sun Jan 23 11:49:19 CET 2011
33 - work around a POE bug where POE would first poll for new events and
34 invoke handlers BEFORE handling the message queue.
35 - fix parsing of "domain" statement in resolv.conf (Johannes Plunien).
36 - stop_read now does just that in TLS mode. see its
37 documentation for caveats.
38 - update to common::sense 3.4 warning set (some warnings in perl impose
39 an arbitrary 32kb limit on regex matches, which affects
40 AnyEvent::Handle).
41 - AE::Handle had a broken rbuf_max method and missing wbuf_max method.
42 - new experimental AE::Handle::resettls method.
43
445.3 Fri Dec 31 05:48:13 CET 2010
45 - major IOAsync patch by Paul Evans - 0.33 is required, as the major
46 issues with IO::Async have been solved in that version, including
47 the availabilty of a default loop!
48 - implement new wbuf_max parameter for AnyEvent::Handle.
49 - added Cocoa::EventLoop backend by Daisuke Murase.
50 - the default testsuite now honors PERL_ANYEVENT_MODEL (Paul Evans).
51 - clarify handle on_prepare documentation.
52 - set CLOEXEC flag on parent-side fd's in AnyEvent::Util::run_cmd,
53 to avoid leaking them to any child processes.
54 - try to force files into the arch-dependend lib directory, for the
55 half-"turly obsessive compulsive person"(s) that depend on it.
56
575.29 Sun Dec 5 10:49:21 CET 2010
58 - convert EV backend to EV 4.00 API (so better upgrade EV too).
59 - AE::Handle: implement dynamic read_size adjustment, add
60 max_read_size parameter, reduce default read_size to 2048.
61 - add some FAQ entries for servers and/or on_eof vs. on_error.
62 - work around OS bugs (cygwin again) when lingering in AE::Handle.
63 - reduce memory usage slightly when connect was used in AE::Handle.
64 - wrok around more segfaults in Tk.
65 - document tls_ctx => undef as valid.
66 - detect WNOHANG value at compiletime.
67 - include some probably linux-only support for building AnyEvent
68 as part of the perl core.
69 - improve compatibility of t/08_idna.t to perl 5.8.
70 - make AnyEvent::Strict truly optional (in case somebody builds a
71 minimal perl...).
72
735.28 Wed Oct 13 04:14:23 CEST 2010
74 - due to a glitch, AnyEvent's internal getprotobyname wasn't
75 actually used in tcp_server.
76 - implement AnyEvent::CondVar->new method, which was documented but
77 missing.
78 - support autoloading of read types in unshift_read.
79 - AnyEvent::Strict: do the test for invalid fh arguments differently,
80 resulting in more informative error messages.
81 - new AnyEvent::FAQ manpage.
82 - clarify that you can actually call push_read etc. in on_read.
83 - put keys and certs in /tmp, not into the current directory,
84 and document this in the SECURITY CONSIDERATIONS sections
85 of AnyEvent::TLS.
86 - support ";" as resolv.conf comment character in AnyEvent::DNS, as
87 requested by Ron Isaacson.
88 - document $AnyEvent::DNS::RESOLVER variable.
89 - incorporated major typo patches by Abhijit Menon-Sen.
90 - AnyEvent::Handle now croaks when negative timeouts are passed.
91 - add a shitload of TCP_xxx sockopt name constants to AnyEvent::Util,
92 undocumented, but who knows when they might come in handy.
93
945.271 Tue Jun 8 12:05:46 CEST 2010
95 - backport to perl 5.8.x.
96
975.27 Sun Jun 6 12:12:05 CEST 2010
98 - postpone differently in AnyEvent::Socket now, as
99 when not, canceling the connection attempt might fail
100 (found by Felix Antonius Wilhelm Ostmann).
101 - explicitly check for non-stream sockets in AE::Handle, too many
102 clueless people fell into the trap of this somehow working.
103 - simplified and reworked the "OTHER MODULES" section.
104 - better/more condvar examples.
105
1065.261 Wed Apr 28 16:13:36 CEST 2010
107 - AF_INET6 was not properly used from Socket6 during configuration
108 time (found by Andy Grundman).
109
1105.26 Mon Apr 12 04:49:35 CEST 2010
111 - don't generate spurious readyness notifications when select
112 returns EINTR in the pure perl backend (analysed and testcase
113 by Bas Denissen).
114 - give same examples in SYNOPSIS sections of AnyEvent and AE.
115 - provider faster implementations for AE::time/now/now_update for
116 EV and Perl backends.
117 - provide faster AE::cv and AE::time implementations for all
118 backends.
119 - no longer support lower/mixed-case signal names (this was
120 never documented nor universally supported).
121 - some more memory µ-optimisations, and somewhat less messy
122 code to cope with both AnyEvent and AE APIs.
123
1245.251 Sat Mar 13 00:58:21 CET 2010
125 - make test failed if it was run before make install on perl <=
126 5.8.8, and older versions of perl make our live very difficult,
127 so write constants.pl during Makefile.PL time.
128
1295.25 Sat Mar 13 00:23:14 CET 2010
130 - fix a race condition in AnyEvent::Handle that would cause
131 a "bio_read: unsupported method" error (in _another_
132 ssl connection) after user code threw an exception.
133 - added AnyEvent::Handle->destroyed method.
134 - speed up AnyEvent::Socket::format_address and ::format_ipv6.
135 - the AnyEvent::Util::fh_nonblocking function worked only
136 by lucky accident on win32.
137 - smaller and faster AnyEvent::Util::fh_nonblocking.
138 - when the (required!) Time::HiRes module is missing, AnyEvent
139 did not fall back to built-in time properly.
140 - do not load Fcntl at runtime, saving memory and loading time.
141 - precompile a number of constants and use them instead of runtime
142 detection and eval.
143 - free detection code after detection and similar memory optimisations.
144 - Perl backend timer interval best effort drift has been improved
145 (same algorithm as EV).
146 - update unicode idna mapping table.
147
1485.24 Tue Jan 5 11:39:43 CET 2010
149 - cygwin never reports errors from failed connects. 1.5 just gives
150 you continous readyness and EAGAIN, 1.7 is even more broken and
151 just hangs. work around both issues in a cygwin-specific hack.
152 - improve idle watcher documentation slightly.
153
1545.23 Sun Dec 20 23:48:00 CET 2009
155 - support IDNs in resolve_sockaddr, and therefore in tcp_connect.
156 - implement punycode_encode/decode, idn_nameprep,
157 idn_to_ascii and idn_to_unicode operations in AnyEvent::Util.
158 - provide $AE::VERSION.
159 - removed traces of "no strict 'refs'".
160
1615.22 Sat Dec 5 03:51:13 CET 2009
162 - downgrade-or-fail in AnyEvent::Handle::push_write, to
163 diagnose encoding failures earlier and more succinctly.
164 (this works around bugs in perl, throwing away encoding info
165 when passing scalar data to extensions).
166 - add more examples to AnyEvent::Socket manpage.
167 - upgrade internal warning set to the same as common::sense 2.03.
168 - use pack "n/a*" for pre-5.8.9 perl compatibility in AnyEvent::DNS
169 (John Beppu).
170 - AnyEvent::Socket::inet_aton now properly supports ipv6, as documented.
171 - add google public dns servers to fallback server set and make sure
172 we load-balance properly between the three sets. also add all
173 fallback dns servers, not just a random one, to each dns config.
174
1755.21 Thu Nov 19 02:48:47 CET 2009
176 - fix a problem where socket constants were called with parameters
177 (spotted by David Friedland).
178 - fork_call never use'd POSIX (reported by Daisuke Maki).
179 - improve perl 5.6 compatibility further (but it still won't work
180 unless you rip out everything but the core).
181 - prefer Net::DNS::Resolver over ipconfig, if installed, on win32.
182 uses 10MB of RAM, but doesn't open a console window. *sigh*.
183
1845.202 Wed Oct 14 22:35:44 CEST 2009
185 - AnyEvent::DNS would unexpectedly clobber $_ under windows
186 (analysed by Matthias Waldorf).
187 - AnyEvent::Handle::run_cmd can now pass the PID of the
188 newly-created process, which is much less useful than it might
189 sound (based on patch by Yann Kerherve).
190
1915.201 Tue Sep 29 12:09:25 CEST 2009
192 - AnyEvent:Handle::on_starttls/on_stoptls methods were broken
193 (reported by Torsten Foertsch).
194 - common::sense 2.0 could cause tcp_server to throw an exception
195 (analysed by elmex).
196
1975.2 Mon Sep 14 07:04:49 CEST 2009
198 - INCOMPATIBLE CHANGE: do no longer support register_read_type
199 and register_write_type in AnyEvent::Handle, instead support
200 package names (the facility was mostly abused).
201 - implement "packagename-as-read/write type" support in
202 AnyEvent::Handle.
203 - AnyEvent::Handle: new options "keepalive" and "oobinline".
204 - oobinline set by default to avoid security issues.
205 - the pure-perl event loop backend wrongly detected times() fallback
206 support (spotted by Pavel Boldin).
207
2085.12 Tue Sep 1 20:26:50 CEST 2009
209 - be more lenient when parsing resolv.conf files, as some people
210 use hashmarks as comment indicator inside directives (reported
211 by Michael S. Fischer).
212 - use same set of warnings as common::sense 2.0.
213 - fix a potential 32 bit overflow issue due to perl having problems
214 with large hex constants in the Perl backend.
215 - do not use the slower 5.6 version fo the enc_name code in AnyEvent::DNS
216 by default.
217 - fix some prototypes.
218 - (optionally) require Async::Interrupt 1.02.
219 - replace opendns fallbacks by verizon/level3 public dns servers.
220
2215.112 Fri Aug 21 13:59:04 CEST 2009
222 - AnyEvent::Handle could cause hard crashes in Net::SSLeay when
223 using starttls with an illegal/uncreatable tls context.
224 - port to broken win32 perls.
225
2265.111 Sun Aug 16 18:44:36 CEST 2009
227 - perl errornously requires sockaddr_un structures returned by the
228 kernel to have a specific length. work around this bug by padding
229 sockaddr_un structures (found to fail on OS X, as perl only uses
230 the correct code for GNU/Linux, but likely to fail on other
231 OSes as well).
232
2335.11 Wed Aug 12 17:49:37 CEST 2009
234 - fixed a bug in run_cmd with input from filename.
235 - create a json coder object when none is given in write direction
236 to reduce dependency on JSON.
237
2385.1 Tue Aug 11 03:17:32 CEST 2009
239 - tcp_connect now reports the correct error in $!, not an unrelated
240 one (was broken since 4.91).
241 - AE::Handle did not properly clear rtimeout/wtimeout watchers
242 when retrying a tcp_connect.
243 - new functions: AnyEvent::Util::run_cmd and close_all_fds_except.
244 - fixed wrong documentation in AE::Socket.
245
2465.01 Mon Aug 10 03:16:32 CEST 2009
247 - last release broke EV child watchers :/.
248
2495.0 Sun Aug 9 17:32:09 CEST 2009
250 - officially introduce and document the AE API.
251 - lowering the timeout at runtime did not have immediate effect
252 in AnyEvent::Handle.
253 - AE::Handle now has separate and independent read and write timeout
254 settings.
255 - probe for Guard module when AnyEvent::Util is loaded, not at
256 runtime (Event doesn't like eval "use" from callbacks, and it avoids
257 exporting the probe function).
258 - Impl::IOAsync's io watchers did not use the filehandle emulation
259 code to work around it's watcher limits.
260 - work around IO::Async's problems with overloaded objects.
261 - try to disable wearnings inside IO::Async, it's just too buggy.
262 - do no longer use Event by default, only EV or Perl, Event is too buggy.
263 - hackishly provide fast AE:: interface when EV or Perl is the backend.
264 - inherit VERSION from AnyEvent::VERSION in most modules.
265
2664.91 Thu Aug 6 15:42:45 CEST 2009
267 - AE::Handle::starttls could get out of sync when the read buffer
268 already contains some TLS handshake.
269 - AE::Handle did not properly free the TLS session, causing bigger
270 memory leaks in Net::SSLeay than expected (reported by toaster).
271 - AE::Socket::tcp_connect will now artificially delay invoking
272 the callback to avoid returning after invoking the callback.
273 - convert many internal watcher uses to AE API.
274
2754.9 Sat Aug 1 11:07:01 CEST 2009
276 - Glib silently fails when registering a timeout with
277 a negative value, so avoid that.
278 - call condvar callback immediately when it is set
279 after the condition is already signalled.
280 - check rbuf_max condition only after trying to consume
281 data.
282 - normalise signal numbers to names when using the ae's
283 signal handling, but do not document this (yet).
284 - pure perl signal emulation did not properly set nonblocking
285 mode on the signal pipe on win32 (this is unlikely to have caused
286 any issues).
287 - new module: AnyEvent::Debug.
288 - AnyEvent::Strict now checks that a signal was specified
289 by name (not name or number).
290 - reduce memory footprint in typical cases by ~50kb by
291 only compiling signal code when necessary.
292 - add AnyEvent::Handle->rbuf_max.
293 - grab the AE namespace for future new API, implement stubs for
294 said future/alternative API.
295 - new function: AnyEvent::Socket::format_hostport.
296
2974.881 Tue Jul 28 12:51:53 CEST 2009
298 - work around a bug in local in pre-5.10 perls, causing
299 AnyEvent::Handle to recurse when it should not
300 (analysed by elmex).
301
3024.88 Tue Jul 28 04:04:37 CEST 2009
303 - re-bless the handle into a dummy package after calling
304 AnyEvent::Handle::destroy, so the user does not need to check
305 for errors after every push_write etc.
306 - do not attempt to run t/02_signals.t on obviously broken
307 platforms.
308
3094.87 Sun Jul 26 02:06:16 CEST 2009
310 - do not attempt to linger when there is no longer a valid fh.
311 - enforce tls mode to be either accept or connect, do not simply
312 segfault in Net::SSLeay.
313 - AnyEvent::Handle can now call tcp_connect itself (new parameters
314 connect, on_prepare, on_connect and on_connect_error). Updated
315 tutorial accordingly.
316 - add AnyEvent::Impl::Irssi backend.
317
3184.86 Mon Jul 20 23:52:29 CEST 2009
319 - since the verbose warning is not enough, explicitly document
320 that versions before 1.33 of Net::SSLeay are not secure.
321 - work around signal handling races in Event and (...) Event::Lib.
322 - try to align signal-race timer to full-second boundaries.
323 - work around Tk not liking negative timeouts.
324 - don't complain of different grades of environmental unfriendlyness
325 in IO::Async.
326
3274.85 Sat Jul 18 06:16:14 CEST 2009
328 - nail the signal race problem in perl once and for all
329 (see $AnyEvent::MAX_SIGNAL_LATENCY).
330 - take advantage of Async::Interrupt if it is available.
331 - load Time::HiRes and Guard modules on demand only.
332 - add optional/recommended modules section to AnyEvent
333 documentation.
334 - reduce memory usage considerably (and reduce startup penalty)
335 by not using "strict", "warnings" and "overload" modules.
336 - work around buggy windows/openbsd perls and provide EBADMSG
337 and EPROTO ourselves when missing.
338 - improve perl 5.6 compatibility of the core event loop.
339 - made Net::SSLeay version 1.33 a soft requirement.
340
3414.83 Fri Jul 17 16:56:26 CEST 2009
342 - implement AnyEvent::Socket::getprotobyname.
343 - AnyEvent::CondVar's will now detect recursive blocking
344 waits and will croak, as too many people fall into
345 this trap.
346 - AnyEvent::Handle will now call ->destroy on itself after
347 executing the on_error callback, instead of doing some
348 half-baked internal shutdown, for fatal errors.
349 - clarify on_eof behaviour w.r.t. the read queue and
350 on_read callbacks.
351 - ignore some possible spurious wake-ups in tcp_connect.
352
3534.82 Sat Jul 11 00:34:55 CEST 2009
354 - POE and Event backends didn't accept some callable objects as
355 callbacks.
356 - use Config module instead of POSIX module to detect signal names
357 in AnyEvent::Strict and AnyEvent::Impl::EventLib, as the POSIX
358 module doesn't even have all POSIX signals :/.
359 - use more workarounds around the many refcnt/corruption bugs in
360 Event::Lib.
361 - work around a race condition in perl's select, causing t/03_child.t
362 to rarely fail.
363
3644.81 Thu Jul 9 10:30:30 CEST 2009
365 - AnyEvent::Handle didn't properly diagnose write errors
366 (it expected -1 from syswrite, how lame... :).
367 - support file descriptors in addition to file handles
368 in AnyEvent->io.
369 - new env variables: PERL_ANYEVENT_RESOLV_CONF,
370 PERL_ANYEVENT_MAX_OUTSTANDING_DNS, PERL_ANYEVENT_CA_FILE
371 and PERL_ANYEVENT_CA_PATH.
372 - provide a sensible synopsis section for AnyEvent::TLS.
373 - add a "supported backends" section to the manpage.
374 - added simple io watcher test to testsuite, using a
375 portable_socketpair.
376 - tried to improve the stability of the Event::Lib backend,
377 YMMV.
378
3794.8 Mon Jul 6 23:45:16 CEST 2009
380 - AnyEvent::DNS did not properly follow CNAME records with
381 uppercase targets.
382 - AnyEvent::DNS would errornously return AAAA records
383 with v4 mapped addresses (a faulty record) as ipv4 addresses,
384 causing AnyEvent::Socket to throw an exception.
385 - added new module AnyEvent::TLS for easier SSL/TLS context
386 creation, with many options including hostname verification,
387 secure default configuration, lots of documentation and,
388 predefined diffie-hellman keys for perfect forward security
389 and much more. get it while it's still fresh!
390 - use AnyEvent::TLS in AnyEvent::Handle for context management.
391 - load AnyEvent::Handle only on demand in AnyEvent::DNS,
392 so AnyEvent::Socket users have smaller memory footprint
393 in the common case.
394 - add AnyEvent::Handle->push_shutdown method.
395 - add an additional error message parameter to AnyEvent::Handle's
396 on_error callback (for TLS, $! is still available).
397 - add AnyEvent::Handle on_starttls/on_stoptls callbacks.
398 - make AnyEvent::Handle more robust against early conenction
399 failures (during new), and return C<undef> in that case
400 from the constructor.
401 - AnyEvent::Handle will now try to load only JSON::XS first,
402 then fall back to JSON.
403 - format_ipv4/format_ipv6 are now exported by default, for symmetry,
404 and because it was documented that way.
405
4064.451 Fri Jul 3 00:28:58 CEST 2009
407 - do not clear rbuf when shutting down an AnyEvent::Handle
408 object - doing so breaks AnyEvent::HTTP.
409
4104.45 Mon Jun 29 22:59:26 CEST 2009
411 - a write error could cause AnyEvent::Handle to create
412 an I/O watcher with an undefined $fh.
413 - special-case mapped ipv4 addresses in both
414 AnyEvent::Socket::format_address and parse_address,
415 to treat them just like ipv4 addresses.
416 - updated and overhauled the AnyEvent::Intro doc.
417 - implement AnyEvent::Socket::format_ipv4/ipv6.
418 - slightly speed up portable_pipe/socketpair.
419 - expand condvar begin/end documentation.
420
4214.42 Fri Jun 26 08:32:18 CEST 2009
422 - preliminary and neccesarily incomplete support for IO::Async.
423 - reset SIGCHLD to DEFAULT when AnyEvent is loaded, in
424 case it was set to IGNORE, to ensure we can catch child statuses
425 even when the calling env acts stupidly.
426 - updated benchmarks with IO::Async, which performs very well.
427
4284.412 Wed Jun 24 01:35:57 CEST 2009
429 - support an "untaint" attribute for AnyEvent::DNS
430 and set it on the default resolver.
431 - implement and document AnyEvent::Impl::Perl::loop.
432 - remove all anyevent-env variables from %ENV
433 when running in tainted mode.
434 - mention and extend the IO::Lambda benchmark.
435
4364.411 Sun Jun 7 18:48:02 CEST 2009
437 - do not try to use F_SETFD on windows.
438
4394.41 Thu May 14 06:40:11 CEST 2009
440 - work around issues in older perls (5.8.0?) when
441 a signal handler is deleted from the %SIG hash.
442 - use POSIX::_exit in child test, to avoid running
443 destructors.
444 - speed up CHLD handling by relying on SIGCHLD being
445 synchronously delivered, even when we roll our own
446 implementation.
447 - AnyEvent::DNS: add the "dname" resource record name
448 for cosmetic reasons.
449
4504.4 Sun Apr 26 20:12:33 CEST 2009
451 - implemented idle watchers, where applicable.
452 - AnyEvent->time died when Event backend was in use.
453 - fix a memleak in the Tk backend.
454 - sped up Tk timer handling.
455 - clip negative "after" values to 0 in AnyEvent::Impl::Event
456 to avoid spamming out warning messages.
457 - fix Qt timers without interval.
458 - avoid Qt zero-timeout special case, in old libqt's.
459
4604.352 Mon Apr 20 16:31:11 CEST 2009
461 - fix AnyEvent::Strict error messages for child watchers.
462 - fix/update Makefile.PL "recommends" versions.
463 - add AnyEvent->now_update.
464
4654.351 Sat Apr 11 07:56:14 CEST 2009
466 - actually make the signal pipe work under win32.
467 - localise $! in signal handler to avoid changing $!, although
468 perl itself does not seem to save/restore errno either.
469 - set the cloexec flag on the signal pipe (normally set by perl too).
470
4714.35 Fri Mar 27 11:48:20 CET 2009
472 - event models relying on AnyEvent's signal watcher emulation did
473 invoke the callback asynchronously, contradicting documentation
474 and causing signals to get lost (this includes AnyEvent's own
475 event loop). AnyEvent now uses the standard pipe trick to make
476 callback execution synchronous to the event loop.
477 - AnyEvent::Handle didn't free TLS context data on DESTROY
478 (patch by Pavel Shaydo).
479 - work around the ever-incompatibly-changing API of MakeMaker.
480 - document that changing global variables without restoring them
481 is a bad idea in Perl (noted by Adam Rosenstein).
482 - AnyEvent::Strict now barfs if ->io is passed a file.
483
4844.34 Thu Feb 12 18:32:45 CET 2009
485 - separately buffer TLS read data, as otherwise the read queue could
486 deadlock as receiving data is not expected while draining the
487 read queue (which cna only happen with TLS).
488 - raise EBADMSG error on JSON decoding errors.
489 - fix some minor manpage bugs (reported by Maximilian Gaß).
490 - speed up select bitmask parsing quite a bit in the pure perl
491 backend.
492 - use CORE::select instead of just select in AnyEvent::Impl::Perl.
493
4944.331 Tue Jan 6 21:07:25 CET 2009
495 - socketpair fails on many vista machines because vista has
496 completely broken accept/getpeername and getsockname functions,
497 so we provide our own socketpair emulation that kind of works
498 (AnyEvent::Util::portable_pipe).
499 - new function: AnyEvent::Util::portable_socketpair.
500 - take advantage of the Guard module if it exists.
501
5024.33 Fri Nov 21 02:35:40 CET 2008
503 - AnyEvent::Strict did errornously flag a fileno of 0 as
504 illegal.
505 - reduce memory usage and slightly speed up the pure perl backend
506 by only storing the file descriptor, not the file handle.
507 - add missing autocork method to AnyEvent::Handle (reported
508 by Adam Rosenstein).
509 - AnyEvent::DNS->resolve errornously documented an $rcode result
510 argument, but there isn't (spotted by Henrik Krohns).
511 - the naked truth about Tk - it's basically unmaintained.
512
5134.32 Mon Nov 3 22:46:32 CET 2008
514 - fix AnyEvent::Socket::resolve_sockaddr to properly support
515 unix sockets again and choose a proper default for
516 the socket type.
517 - fix call to resolve_sockaddr in tcp_connect.
518
5194.31 Thu Oct 30 04:41:48 CET 2008
520 - implemented AnyEvent::Handle->destroy method.
521 - hint about unexpected effects in TLS mode.
522 - speed up AnyEvent::Impl::Perl by using more arrays
523 instead of hashes without the slightest loss of readability :)
524 - work around a perl argument refcounting bug.
525 - with some perl patches applied, AnyEvent now seems leak-free.
526
5274.3 Fri Oct 3 09:18:43 CEST 2008
528 - AnyEvent will now install a no-op signal handler for SIGPIPE,
529 unless one has been installed already.
530 - warn about not loaidng AnyEvent::Impl::POE early enough
531 (patch by Adam Rosenstein).
532 - fixed a great number of bugs and corner cases in AnyEvent::Handle.
533 - fix a bug in where in SSL connect mode, the
534 client would first wait for some data by the server and
535 otherwise hang. (reported and analysed in an absolutely
536 exemplary manner by Adam Rosenstein).
537 - fix a bug in where SSL EOF would not be treated
538 as stream EOF, putting the connection into a hung state.
539 (reported and analysed in an absolutely exemplary manner by Adam
540 Rosenstein).
541 - fix a potential segfault when the TLS context
542 would go missing in a read callback (e.g. due to stoptls) -
543 Net::SSLeay of course makes no type checking whatsoever.
544 - AnyEvent::Handle will not stop the read watcher in TLS mode.
545 - AnyEvent::Handle->stoptls will send a TLS close notify
546 instead of simply destroying the stream now.
547 - fix a bug where large blocks of data written in TLS mode
548 would not be sent unless triggered by receives.
549 - on_drain will now take the tls write buffer into account.
550 - SSL operations have been streamlined a bit and should be faster
551 now, more to come, though.
552 - added a FAQ section, will document nontrivial issues in there.
553
5544.234 Mon Sep 29 04:08:13 CEST 2008
555 - fix child watcher documentation: only child exits will be reported,
556 no trace events.
557 - mention SIGPIPE in AnyEvent::Handle.
558 - perl backend will now use POSIX::times as monotonic clock
559 when available and the monotonic clock option is not.
560 - run even without Time::HiRes in some cases now, but do not
561 provide sub-second accuracy (all *supported* perl versions have
562 Time::HiRes).
563
5644.233 Fri Aug 22 01:48:05 CEST 2008
565 - fix a bug in the testsuite, causing a hang.
566 - clarified lots of AnyEvent::Handle constructor arguments.
567
5684.232 Thu Aug 21 20:44:25 CEST 2008
569 - fix a bug in AnyEvent::Handle that could cause two fatal error
570 callbacks on EOF, which caused AnyEvent::HTTP to signal success twice
571 (testcase by Leon Brocard).
572 - clarified on_eof/on_error documentation a bit.
573
5744.231 Tue Jul 29 13:12:15 CEST 2008
575 - remove some debugging code left in AnyEvent::Util::fork_call
576 (and no, it's impossible to implement with the broken windows
577 perls without resource leaks or worse).
578
5794.23 Tue Jul 29 12:19:59 CEST 2008
580 - document the first parameter passed to condvar callbacks to be
581 the callback.
582 - add AnyEvent::Socket::{ntoa,aton} aliases.
583 - optimize the AE::Handle->push_read (line) for the default
584 eol marker.
585 - optimize push_read (packstring|storable) for small packets.
586 - invoke on_error callback when no on_eof callback is set.
587 - fix a bug in push_read (storable) of unknown impact.
588
5894.22 Sun Jul 20 16:34:13 CEST 2008
590 - new function AnyEvent::Socket::parse_hostport.
591 - as the bulkheads at microsoft can't even get getprotobyname reliably
592 working on their shitty fucking broken os we need to hardcode
593 some common protocol numbers in AnyEvent::Socket. How can
594 people even bother with such a pile of shit as windows.
595
5964.21 Thu Jul 17 14:40:05 CEST 2008
597 - INCOMPATIBLE CHANGE: AnyEvent::Util::fork_call now has a prototype
598 which is incompatible to previous invocation syntax.
599 - work around more windows perl bugs in fork_call. I cannot imagine
600 why anybody would bother implementing fork in such an obviously
601 unusable way.
602 - avoid calling $do_retry if the request has already finished in
603 AnyEvent::DNS, thus avoiding an exception (reported by Anatoly
604 K. Sharifulin).
605
6064.2 Sat Jul 12 22:42:11 CEST 2008
607 - support an "interval" argument for repeating timers.
608 - fix a bug in the pure perl event loop that caused timers to
609 be delayed under some circumstances.
610 - implement optional argument checking via PERL_ANYEVENT_STRICT.
611 - not importing anything from Event causes it to malfunction
612 and not use Time::HiRes, work around this bug.
613 - more bug workarounds against the endlessly broken Tk module,
614 seems to work now on 32 bit machines, but less so on 64 bit.
615 - minor optimisations applied to most backends.
616
6174.161 Fri Jul 4 14:24:48 CEST 2008
618 - fixed the t/handle/02_write.t, it read after write, with the
619 assumption that all data could be written before read
620 (reported by Jim Razmus).
621
6224.160 Thu Jul 3 04:02:21 CEST 2008
4 - re-registering signal watchers after unregistering one did 623 - re-registering signal watchers after unregistering one did
5 not work if a backend falls back on AnyEvent's default 624 not work if a backend falls back on AnyEvent's default
6 implementation (also affected child watchers). 625 implementation (also affected child watchers).
7 - new AnyEvent::Handle options: autocork and no_delay. 626 - new AnyEvent::Handle options: autocork and no_delay.
8 627
… …
154 - support service names in tcp_connect and tcp_server. 773 - support service names in tcp_connect and tcp_server.
155 - work around netbsd/openbsd bug causing unix sockets not to 774 - work around netbsd/openbsd bug causing unix sockets not to
156 become writable until empty. 775 become writable until empty.
157 - only make a dummy read for error detection when getpeername 776 - only make a dummy read for error detection when getpeername
158 returns ENOTCONN. 777 returns ENOTCONN.
159 - defend AnyEvent::Util::inet_aton against Coro::LWP brutal 778 - defend AnyEvent::Util::inet_aton against Coro::LWP's brutal
160 override. 779 override.
161 - new AnyEvent::Guard::cancel. 780 - new AnyEvent::Guard::cancel.
162 781
1633.5 Sat May 17 23:17:14 CEST 2008 7823.5 Sat May 17 23:17:14 CEST 2008
164 - gracefully deal with systems not defining CLOCK_MONOTONIC 783 - gracefully deal with systems not defining CLOCK_MONOTONIC
… …
178 797
1793.41 Sun May 11 19:53:13 CEST 2008 7983.41 Sun May 11 19:53:13 CEST 2008
180 - work around a bug in perl 5.8.8's local. Reported by Yi Ma Mao. 799 - work around a bug in perl 5.8.8's local. Reported by Yi Ma Mao.
181 800
1823.4 Sun May 11 00:29:25 CEST 2008 8013.4 Sun May 11 00:29:25 CEST 2008
183 - complete re-implementation of condvars with a hopefully much mroe useful 802 - complete re-implementation of condvars with a hopefully much more useful
184 API (backward compatible functions exist but are undocumented). 803 API (backward compatible functions exist but are undocumented).
185 - AnyEvent::Handle fully rewritten with a hopefully more powerful API 804 - AnyEvent::Handle fully rewritten with a hopefully more powerful API
186 (still subject to change). 805 (still subject to change).
187 - take advantage of CLOCK_MONOTONIC, if available, in AnyEvent::Impl::Perl. 806 - take advantage of CLOCK_MONOTONIC, if available, in AnyEvent::Impl::Perl.
188 - provide AnyEvent::post_detect and @AnyEvent::post_detect, which 807 - provide AnyEvent::post_detect and @AnyEvent::post_detect, which

Diff Legend

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