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

Comparing AnyEvent/Changes (file contents):
Revision 1.99 by root, Sat May 24 23:56:26 2008 UTC vs.
Revision 1.350 by root, Mon Sep 28 17:30:53 2009 UTC

1Revision history for Perl extension AnyEvent. 1Revision history for Perl extension AnyEvent.
2 2
3 - AnyEvent:Handle::on_starttls/on_stoptls methods were broken
4 (reported by Torsten Foertsch).
5
65.2 Mon Sep 14 07:04:49 CEST 2009
7 - INCOMPATIBLE CHANGE: do no longer support register_read_type
8 and register_write_type in AnyEvent::Handle, instead support
9 package names (the facility was mostly abused).
10 - implement "packagename-as-read/write type" support in
11 AnyEvent::Handle.
12 - AnyEvent::Handle: new options "keepalive" and "oobinline".
13 - oobinline set by default to avoid security issues.
14 - the pure-perl event loop backend wrongly detected times() fallback
15 support (spotted by Pavel Boldin).
16
175.12 Tue Sep 1 20:26:50 CEST 2009
18 - be more lenient when parsing resolv.conf files, as some people
19 use hashmarks as comment indicator inside directives (reported
20 by Michael S. Fischer).
21 - use same set of warnings as common::sense 2.0.
22 - fix a potential 32 bit overflow issue due to perl having problems
23 with large hex constants in the Perl backend.
24 - do not use the slower 5.6 version fo the enc_name code in AnyEvent::DNS
25 by default.
26 - fix some prototypes.
27 - (optionally) require Async::Interrupt 1.02.
28 - replace opendns fallbacks by verizon/level3 public dns servers.
29
305.112 Fri Aug 21 13:59:04 CEST 2009
31 - AnyEvent::Handle could cause hard crashes in Net::SSLeay when
32 using starttls with an illegal/uncreatable tls context.
33 - port to broken win32 perls.
34
355.111 Sun Aug 16 18:44:36 CEST 2009
36 - perl errornously requires sockaddr_un structures returned by the
37 kernel to have a specific length. work around this bug by padding
38 sockaddr_un structures (found to fail on OS X, as perl only uses
39 the correct code for GNU/Linux, but likely to fail on other
40 OSes as well).
41
425.11 Wed Aug 12 17:49:37 CEST 2009
43 - fixed a bug in run_cmd with input from filename.
44 - create a json coder object when none is given in write direction
45 to reduce dependency on JSON.
46
475.1 Tue Aug 11 03:17:32 CEST 2009
48 - tcp_connect now reports the correct error in $!, not an unrelated
49 one (was broken since 4.91).
50 - AE::Handle did not properly clear rtimeout/wtimeout watchers
51 when retrying a tcp_connect.
52 - new functions: AnyEvent::Util::run_cmd and close_all_fds_except.
53 - fixed wrong documentation in AE::Socket.
54
555.01 Mon Aug 10 03:16:32 CEST 2009
56 - last release broke EV child watchers :/.
57
585.0 Sun Aug 9 17:32:09 CEST 2009
59 - officially introduce and document the AE API.
60 - lowering the timeout at runtime did not have immediate effect
61 in AnyEvent::Handle.
62 - AE::Handle now has separate and independent read and write timeout
63 settings.
64 - probe for Guard module when AnyEvent::Util is loaded, not at
65 runtime (Event doesn't like eval "use" from callbacks, and it avoids
66 exporting the probe function).
67 - Impl::IOAsync's io watchers did not use the filehandle emulation
68 code to work around it's watcher limits.
69 - work around IO::Async's problems with overloaded objects.
70 - try to disable wearnings inside IO::Async, it's just too buggy.
71 - do no longer use Event by default, only EV or Perl, Event is too buggy.
72 - hackishly provide fast AE:: interface when EV or Perl is the backend.
73 - inherit VERSION from AnyEvent::VERSION in most modules.
74
754.91 Thu Aug 6 15:42:45 CEST 2009
76 - AE::Handle::starttls could get out of sync when the read buffer
77 already contains some TLS handshake.
78 - AE::Handle did not properly free the TLS session, causing bigger
79 memory leaks in Net::SSLeay than expected (reported by toaster).
80 - AE::Socket::tcp_connect will now artificially delay invoking
81 the callback to avoid returning after invoking the callback.
82 - convert many internal watcher uses to AE API.
83
844.9 Sat Aug 1 11:07:01 CEST 2009
85 - Glib silently fails when registering a timeout with
86 a negative value, so avoid that.
87 - call condvar callback immediately when it is set
88 after the condition is already signalled.
89 - check rbuf_max condition only after trying to consume
90 data.
91 - normalise signal numbers to names when using the ae's
92 signal handling, but do not document this (yet).
93 - pure perl signal emulation did not properly set nonblocking
94 mode on the signal pipe on win32 (this is unlikely to have caused
95 any issues).
96 - new module: AnyEvent::Debug.
97 - AnyEvent::Strict now checks that a signal was specified
98 by name (not name or number).
99 - reduce memory footprint in typical cases by ~50kb by
100 only compiling signal code when necessary.
101 - add AnyEvent::Handle->rbuf_max.
102 - grab the AE namespace for future new API, implement stubs for
103 said future/alternative API.
104 - new function: AnyEvent::Socket::format_hostport.
105
1064.881 Tue Jul 28 12:51:53 CEST 2009
107 - work around a bug in local in pre-5.10 perls, causing
108 AnyEvent::Handle to recurse when it should not
109 (analyzed by elmex).
110
1114.88 Tue Jul 28 04:04:37 CEST 2009
112 - re-bless the handle into a dummy package after calling
113 AnyEvent::Handle::destroy, so the user does not need to check
114 for errors after every push_write etc.
115 - do not attempt to run t/02_signals.t on obviously broken
116 platforms.
117
1184.87 Sun Jul 26 02:06:16 CEST 2009
119 - do not attempt to linger when there is no longer a valid fh.
120 - enforce tls mode to be either accept or connect, do not simply
121 segfault in Net::SSLeay.
122 - AnyEvent::Handle can now call tcp_connect itself (new parameters
123 connect, on_prepare, on_connect and on_connect_error). Updated
124 tutorial accordingly.
125 - add AnyEvent::Impl::Irssi backend.
126
1274.86 Mon Jul 20 23:52:29 CEST 2009
128 - since the verbose warning is not enough, explicitly document
129 that versions before 1.33 of Net::SSLeay are not secure.
130 - work around signal handling races in Event and (...) Event::Lib.
131 - try to align signal-race timer to full-second boundaries.
132 - work around Tk not liking negative timeouts.
133 - don't complain of different grades of environmental unfriendlyness
134 in IO::Async.
135
1364.85 Sat Jul 18 06:16:14 CEST 2009
137 - nail the signal race problem in perl once and for all
138 (see $AnyEvent::MAX_SIGNAL_LATENCY).
139 - take advantage of Async::Interrupt if it is available.
140 - load Time::HiRes and Guard modules on demand only.
141 - add optional/recommended modules section to AnyEvent
142 documentation.
143 - reduce memory usage considerably (and reduce startup penalty)
144 by not using "strict", "warnings" and "overload" modules.
145 - work around buggy windows/openbsd perls and provide EBADMSG
146 and EPROTO ourselves when missing.
147 - improve perl 5.6 compatibility of the core event loop.
148 - made Net::SSLeay version 1.33 a soft requirement.
149
1504.83 Fri Jul 17 16:56:26 CEST 2009
151 - implement AnyEvent::Socket::getprotobyname.
152 - AnyEvent::CondVar's will now detect recursive blocking
153 waits and will croak, as too many people fall into
154 this trap.
155 - AnyEvent::Handle will now call ->destroy on itself after
156 executing the on_error callback, instead of doing some
157 half-baked internal shutdown, for fatal errors.
158 - clarify on_eof behaviour w.r.t. the read queue and
159 on_read callbacks.
160 - ignore some possible spurious wake-ups in tcp_connect.
161
1624.82 Sat Jul 11 00:34:55 CEST 2009
163 - POE and Event backends didn't accept some callable objects as
164 callbacks.
165 - use Config module instead of POSIX module to detect signal names
166 in AnyEvent::Strict and AnyEvent::Impl::EventLib, as the POSIX
167 module doesn't even have all POSIX signals :/.
168 - use more workarounds around the many refcnt/corruption bugs in
169 Event::Lib.
170 - work around a race condition in perl's select, causing t/03_child.t
171 to rarely fail.
172
1734.81 Thu Jul 9 10:30:30 CEST 2009
174 - AnyEvent::Handle didn't properly diagnose write errors
175 (it expected -1 from syswrite, how lame... :).
176 - support file descriptors in addition to file handles
177 in AnyEvent->io.
178 - new env variables: PERL_ANYEVENT_RESOLV_CONF,
179 PERL_ANYEVENT_MAX_OUTSTANDING_DNS, PERL_ANYEVENT_CA_FILE
180 and PERL_ANYEVENT_CA_PATH.
181 - provide a sensible synopsis section for AnyEvent::TLS.
182 - add a "supported backends" section to the manpage.
183 - added simple io watcher test to testsuite, using a
184 portable_socketpair.
185 - tried to improve the stability of the Event::Lib backend,
186 YMMV.
187
1884.8 Mon Jul 6 23:45:16 CEST 2009
189 - AnyEvent::DNS did not properly follow CNAME records with
190 uppercase targets.
191 - AnyEvent::DNS would errornously return AAAA records
192 with v4 mapped addresses (a faulty record) as ipv4 addresses,
193 causing AnyEvent::Socket to throw an exception.
194 - added new module AnyEvent::TLS for easier SSL/TLS context
195 creation, with many options including hostname verification,
196 secure default configuration, lots of documentation and,
197 predefined diffie-hellman keys for perfect forward security
198 and much more. get it while it's still fresh!
199 - use AnyEvent::TLS in AnyEvent::Handle for context management.
200 - load AnyEvent::Handle only on demand in AnyEvent::DNS,
201 so AnyEvent::Socket users have smaller memory footprint
202 in the common case.
203 - add AnyEvent::Handle->push_shutdown method.
204 - add an additional error message parameter to AnyEvent::Handle's
205 on_error callback (for TLS, $! is still available).
206 - add AnyEvent::Handle on_starttls/on_stoptls callbacks.
207 - make AnyEvent::Handle more robust against early conenction
208 failures (during new), and return C<undef> in that case
209 from the constructor.
210 - AnyEvent::Handle will now try to load only JSON::XS first,
211 then fall back to JSON.
212 - format_ipv4/format_ipv6 are now exported by default, for symmetry,
213 and because it was documented that way.
214
2154.451 Fri Jul 3 00:28:58 CEST 2009
216 - do not clear rbuf when shutting down an AnyEvent::Handle
217 object - doing so breaks AnyEvent::HTTP.
218
2194.45 Mon Jun 29 22:59:26 CEST 2009
220 - a write error could cause AnyEvent::Handle to create
221 an I/O watcher with an undefined $fh.
222 - special-case mapped ipv4 addresses in both
223 AnyEvent::Socket::format_address and parse_address,
224 to treat them just like ipv4 addresses.
225 - updated and overhauled the AnyEvent::Intro doc.
226 - implement AnyEvent::Socket::format_ipv4/ipv6.
227 - slightly speed up portable_pipe/socketpair.
228 - expand condvar begin/end documentation.
229
2304.42 Fri Jun 26 08:32:18 CEST 2009
231 - preliminary and neccesarily incomplete support for IO::Async.
232 - reset SIGCHLD to DEFAULT when AnyEvent is loaded, in
233 case it was set to IGNORE, to ensure we can catch child statuses
234 even when the calling env acts stupidly.
235 - updated benchmarks with IO::Async, which performs very well.
236
2374.412 Wed Jun 24 01:35:57 CEST 2009
238 - support an "untaint" attribute for AnyEvent::DNS
239 and set it on the default resolver.
240 - implement and document AnyEvent::Impl::Perl::loop.
241 - remove all anyevent-env variables from %ENV
242 when running in tainted mode.
243 - mention and extend the IO::Lambda benchmark.
244
2454.411 Sun Jun 7 18:48:02 CEST 2009
246 - do not try to use F_SETFD on windows.
247
2484.41 Thu May 14 06:40:11 CEST 2009
249 - work around issues in older perls (5.8.0?) when
250 a signal handler is deleted from the %SIG hash.
251 - use POSIX::_exit in child test, to avoid running
252 destructors.
253 - speed up CHLD handling by relying on SIGCHLD being
254 synchronously delivered, even when we roll our own
255 implementation.
256 - AnyEvent::DNS: add the "dname" resource record name
257 for cosmetic reasons.
258
2594.4 Sun Apr 26 20:12:33 CEST 2009
260 - implemented idle watchers, where applicable.
261 - AnyEvent->time died when Event backend was in use.
262 - fix a memleak in the Tk backend.
263 - sped up Tk timer handling.
264 - clip negative "after" values to 0 in AnyEvent::Impl::Event
265 to avoid spamming out warning messages.
266 - fix Qt timers without interval.
267 - avoid Qt zero-timeout special case, in old libqt's.
268
2694.352 Mon Apr 20 16:31:11 CEST 2009
270 - fix AnyEvent::Strict error messages for child watchers.
271 - fix/update Makefile.PL "recommends" versions.
272 - add AnyEvent->now_update.
273
2744.351 Sat Apr 11 07:56:14 CEST 2009
275 - actually make the signal pipe work under win32.
276 - localise $! in signal handler to avoid changing $!, although
277 perl itself does not seem to save/restore errno either.
278 - set the cloexec flag on the signal pipe (normally set by perl too).
279
2804.35 Fri Mar 27 11:48:20 CET 2009
281 - event models relying on AnyEvent's signal watcher emulation did
282 invoke the callback asynchronously, contradicting documentation
283 and causing signals to get lost (this includes AnyEvent's own
284 event loop). AnyEvent now uses the standard pipe trick to make
285 callback execution synchronous to the event loop.
286 - AnyEvent::Handle didn't free TLS context data on DESTROY
287 (patch by Pavel Shaydo).
288 - work around the ever-incompatibly-changing API of MakeMaker.
289 - document that changing global variables without restoring them
290 is a bad idea in Perl (noted by Adam Rosenstein).
291 - AnyEvent::Strict now barfs if ->io is passed a file.
292
2934.34 Thu Feb 12 18:32:45 CET 2009
294 - separately buffer TLS read data, as otherwise the read queue could
295 deadlock as receiving data is not expected while draining the
296 read queue (which cna only happen with TLS).
297 - raise EBADMSG error on JSON decoding errors.
298 - fix some minor manpage bugs (reported by Maximilian Gaß).
299 - speed up select bitmask parsing quite a bit in the pure perl
300 backend.
301 - use CORE::select instead of just select in AnyEvent::Impl::Perl.
302
3034.331 Tue Jan 6 21:07:25 CET 2009
304 - socketpair fails on many vista machines because vista has
305 completely broken accept/getpeername and getsockname functions,
306 so we provide our own socketpair emulation that kind of works
307 (AnyEvent::Util::portable_pipe).
308 - new function: AnyEvent::Util::portable_socketpair.
309 - take advantage of the Guard module if it exists.
310
3114.33 Fri Nov 21 02:35:40 CET 2008
312 - AnyEvent::Strict did errornously flag a fileno of 0 as
313 illegal.
314 - reduce memory usage and slightly speed up the pure perl backend
315 by only storing the file descriptor, not the file handle.
316 - add missing autocork method to AnyEvent::Handle (reported
317 by Adam Rosenstein).
318 - AnyEvent::DNS->resolve errornously documented an $rcode result
319 argument, but there isn't (spotted by Henrik Krohns).
320 - the naked truth about Tk - it's basically unmaintained.
321
3224.32 Mon Nov 3 22:46:32 CET 2008
323 - fix AnyEvent::Socket::resolve_sockaddr to properly support
324 unix sockets again and choose a proper default for
325 the socket type.
326 - fix call to resolve_sockaddr in tcp_connect.
327
3284.31 Thu Oct 30 04:41:48 CET 2008
329 - implemented AnyEvent::Handle->destroy method.
330 - hint about unexpected effects in TLS mode.
331 - speed up AnyEvent::Impl::Perl by using more arrays
332 instead of hashes without the slightest loss of readability :)
333 - work around a perl argument refcounting bug.
334 - with some perl patches applied, AnyEvent now seems leak-free.
335
3364.3 Fri Oct 3 09:18:43 CEST 2008
337 - AnyEvent will now install a no-op signal handler for SIGPIPE,
338 unless one has been installed already.
339 - warn about not loaidng AnyEvent::Impl::POE early enough
340 (patch by Adam Rosenstein).
341 - fixed a great number of bugs and corner cases in AnyEvent::Handle:
342 - fix a bug in where in SSL connect mode, the
343 client would first wait for some data by the server and
344 otherwise hang. (reported and analyzed in an absolutely
345 exemplary manner by Adam Rosenstein).
346 - fix a bug in where SSL EOF would not be treated
347 as stream EOF, putting the connection into a hung state.
348 (reported and analyzed in an absolutely exemplary manner by Adam
349 Rosenstein).
350 - fix a potential segfault when the TLS context
351 would go missing in a read callback (e.g. due to stoptls) -
352 Net::SSLeay of course makes no type checking whatsoever.
353 - AnyEvent::Handle will not stop the read watcher in TLS mode.
354 - AnyEvent::Handle->stoptls will send a TLS close notify
355 instead of simply destroying the stream now.
356 - fix a bug where large blocks of data written in TLS mode
357 would not be sent unless triggered by receives.
358 - on_drain will now take the tls write buffer into account.
359 - SSL operations have been streamlined a bit and should be faster
360 now, more to come, though.
361 - added a FAQ section, will document nontrivial issues in there.
362
3634.234 Mon Sep 29 04:08:13 CEST 2008
364 - fix child watcher documentation: only child exits will be reported,
365 no trace events.
366 - mention SIGPIPE in AnyEvent::Handle.
367 - perl backend will now use POSIX::times as monotonic clock
368 when available and the monotonic clock option is not.
369 - run even without Time::HiRes in some cases now, but do not
370 provide sub-second accuracy (all *supported* perl versions have
371 Time::HiRes).
372
3734.233 Fri Aug 22 01:48:05 CEST 2008
374 - fix a bug in the testsuite, causing a hang.
375 - clarified lots of AnyEvent::Handle constructor arguments.
376
3774.232 Thu Aug 21 20:44:25 CEST 2008
378 - fix a bug in AnyEvent::Handle that could cause two fatal error
379 callbacks on EOF, which caused AnyEvent::HTTP to signal success twice
380 (testcase by Leon Brocard).
381 - clarified on_eof/on_error documentation a bit.
382
3834.231 Tue Jul 29 13:12:15 CEST 2008
384 - remove some debugging code left in AnyEvent::Util::fork_call
385 (and no, it's impossible to implement with the broken windows
386 perls without resource leaks or worse).
387
3884.23 Tue Jul 29 12:19:59 CEST 2008
389 - document the first parameter passed to condvar callbacks to be
390 the callback.
391 - add AnyEvent::Socket::{ntoa,aton} aliases.
392 - optimize the AE::Handle->push_read (line) for the default
393 eol marker.
394 - optimize push_read (packstring|storable) for small packets.
395 - invoke on_error callback when no on_eof callback is set.
396 - fix a bug in push_read (storable) of unknown impact.
397
3984.22 Sun Jul 20 16:34:13 CEST 2008
399 - new function AnyEvent::Socket::parse_hostport.
400 - as the bulkheads at microsoft can't even get getprotobyname reliably
401 working on their shitty fucking broken os we need to hardcode
402 some common protocol numbers in AnyEvent::Socket. How can
403 people even bother with such a pile of shit as windows.
404
4054.21 Thu Jul 17 14:40:05 CEST 2008
406 - INCOMPATIBLE CHANGE: AnyEvent::Util::fork_call now has a prototype
407 which is incompatible to previous invocation syntax.
408 - work around more windows perl bugs in fork_call. I cannot imagine
409 why anybody would bother implementing fork in such an obviously
410 unusable way.
411 - avoid calling $do_retry if the request has already finished in
412 AnyEvent::DNS, thus avoiding an exception (reported by Anatoly
413 K. Sharifulin).
414
4154.2 Sat Jul 12 22:42:11 CEST 2008
416 - support an "interval" argument for repeating timers.
417 - fix a bug in the pure perl event loop that caused timers to
418 be delayed under some circumstances.
419 - implement optional argument checking via PERL_ANYEVENT_STRICT.
420 - not importing anything from Event causes it to malfunction
421 and not use Time::HiRes, work around this bug.
422 - more bug workarounds against the endlessly broken Tk module,
423 seems to work now on 32 bit machines, but less so on 64 bit.
424 - minor optimisations applied to most backends.
425
4264.161 Fri Jul 4 14:24:48 CEST 2008
427 - fixed the t/handle/02_write.t, it read after write, with the
428 assumption that all data could be written before read
429 (reported by Jim Razmus).
430
4314.160 Thu Jul 3 04:02:21 CEST 2008
432 - re-registering signal watchers after unregistering one did
433 not work if a backend falls back on AnyEvent's default
434 implementation (also affected child watchers).
435 - new AnyEvent::Handle options: autocork and no_delay.
436
4374.152 Sun Jun 22 14:15:44 CEST 2008
438 - allow for 32-bit perls that implement shifts differently
439 on different architectures in parse_ipv4 (reported and
440 analysed by Keiichi DAIBA).
441
4424.151 Fri Jun 6 17:34:24 CEST 2008
443 - make sure specifying _only_ on_read and never pushing reads
444 works.
445
4464.15 Fri Jun 6 13:00:46 CEST 2008
447 - the pure perl backend would keep some watchers alive when more than
448 one watcher was registered for the same fd.
449 - new "packstring" and "storable" read and write types
450 for AnyEvent::Handle.
451 - allow on_eof handler to be called after on_error with EPIPE returns.
452 - do not immediately call on_read callback in handle constructor.
453
4544.14 Thu Jun 5 20:29:31 CEST 2008
455 - Fixed a bug in DNS SRV priority sorting.
456 - AnyEvent::Util::guard now reports runtime errors while
457 executing the guard block as warnings.
458 - handle 0-byte-reads just before EOF correctly in AnyEvent::Handle.
459
4604.13 Thu Jun 5 00:47:59 CEST 2008
461 - AnyEvent::DNS only followed cname chains with length 2,
462 contrary to documentation. bumped it up to 10, thanks to
463 microsoft, the current broken-dns-config-king.
464 - AnyEvent::DNS didn't check for socket return status
465 but instead relied on perl not creating filehandles in that
466 case - too bad it gives you a bogus file handle (reported
467 and analysed by Vladimir Timofeev).
468 - fix queue management logic in AnyEvent::Handle:
469 when on_read was registered, the queue was empty
470 and no progress could be made AnyEvent::Handle would
471 enter an endless loop.
472 - correctly start reading again when the handle became
473 busy again after idling.
474 - correctly treat tls shut-downs as EOF condition
475 (for the time being, would be nice to have a callback for that).
476 - correctly call eof callback under all conditions
477 (hopefully).
478 - the timeout callback did not expect that $self can go
479 away any time.
480
4814.12 Tue Jun 3 10:58:04 CEST 2008
482 - include AnyEvent::Intro, a tutorial for anyevent,
483 anyevent::socket and anyevent::handle.
484 - allow more options in on_error.
485
4864.11 Fri May 30 23:42:25 CEST 2008
487 - INCOMPATIBLE CHANGE: replace ptr by real PTR lookup, provide
488 reverse_lookup and reverse_verify to replace it, support
489 v4mapped and v4compat addresses.
490 - provide more documentation for the resolver class.
491 - really replace longest run of :0: by :: in format_address,
492 also properly convert :: and ::1 again.
493 - support NAPTR record name and decode it.
494 - implement random weight sampling for SRV records, as per
495 rfc 2782.
496 - correctly abort on srv-record targets of ".".
497 - added AnyEvent::DNS::wait_for_slot.
498 - in the unlikely event of a virtual circuit connection
499 being invalidated by a delayed udp reply, AnyEvent::DNS
500 could die. this has been fixed.
501 - plug a probable memleak in the DNS vc code.
502 - use configured timeout also for VC request phase.
503 - implement timeout and max_outstanding methods
504 for resolver class.
505 - update version numbers in all modules.
506
5074.1 Thu May 29 05:45:40 CEST 2008
508 - INCOMPATIBLE CHANGE: renamed xxx_ip functions to xxx_address
509 which mirrors their purpose better (old names still available
510 for a while). Also moved AnyEvent::DNS::addr to
511 AnyEvent::Socket::resolve_sockaddr.
512 - implement AnyEvent->time and AnyEvent->now.
513 - fix IPv6 support in pack_sockaddr.
514 - officially un-experimentalise all newly introduced code.
515 - support unix domain sockets everywhere by specifying
516 a host of "unix/" and the pathname as service.
517 - implement an activity timeout in AnyEvent::Handle.
518 - added a regex read type to AnyEvent::Handle.
519 - added a json read/write type to AnyEvent::Handle.
520 - always croak in on_error in AnyEvent::Handle.
521 - document how to subclass AnyEvent::Handle.
522 - implement AnyEvent::Util::fork_call.
523 - add support for IPv6 nameservers and nameserver statements.
524 - work around _yet_ _another_ windows perl bug in where empty
525 select masks cause errors in the pure perl backend (this
526 workaround cannot be implemented for the other event loops,
527 but EV also works around this).
528 - supply AnyEvent::Socket::address_type and make good use of it.
529 - clarify and fix the Handle documentation w.r.t $self vs. $handle
530 vs. callback arguments.
531 - add some recommends to the META.yml.
532
5334.05 Mon May 26 19:44:06 CEST 2008
534 - some platforms ignore the rfc and prepend an extra sa_len member to
535 the sockaddr structure, cater for those. also use sockaddr_family,
536 didn't know it was there, but comes in handy for the workaround.
537 - undo the import WIN32 hack.
538
5394.04 Mon May 26 08:03:31 CEST 2008
540 - try to work around yet another windows bug: failed connects are reported
541 as if it were out-of-band data. windows users: you suck. Only
542 EV, Glib, Event and the pure perl backend can handle this condition.
543 - optimize write algorithm in AnyEvent::Handle.
544 - properly parse PERL_ANYEVENT_PROTOCOLS and use the info correctly.
545 - tcp_connect now properly iterates through all targets.
546 - check wether IPv6 sockets can actually be created, otherwise, disable ipv6
547 support entirely. also hardcode some AF_INET6 constants for the
548 sake of perl 5.8.
549 - vastly improved nameserver/suffix detection algorithm for windows
550 (still a hack...)
551 - try to use a fallback dns server if no dns servers could be found.
552 - splatter around some local $SIG{__DIE__}'s to work around bad
553 die handlers.
554
5554.03 Sun May 25 05:05:57 CEST 2008
556 - work around perl bugs on windows, where perl returns undocumented
557 error codes for sysread, syswrite etc.
3 - fix AnyEvent::Handle::unshift_read to really unshift and not push. 558 - fix AnyEvent::Handle::unshift_read to really unshift and not push.
4 - AnyEvent::Handle could sometimes stumble over 0-byte writes. 559 - AnyEvent::Handle could sometimes stumble over 0-byte writes.
5 - croak, not die, in AnyEvent::Handle; 560 - fix unpack_sockaddr, now works for AF_INET6 sockaddr's.
561 - updated documentation.
562 - lots of workarounds for perl 5.8 quirks.
6 - implement a more flexible type system for both reads and writes 563 - implement a more flexible type system for both reads and writes
7 in AnyEvent::Handle. 564 in AnyEvent::Handle.
565 - support IPv6 in tcp_server.
566 - pass local hostname and port to the prepare callback in tcp_server.
8 - make types extendable in AnyEvent::Handle. 567 - make types extendable in AnyEvent::Handle.
568 - croak, not die, in AnyEvent::Handle;
9 569
104.0 Sat May 24 19:58:08 CEST 2008 5704.0 Sat May 24 19:58:08 CEST 2008
11 - added (experimental) AnyEvent::DNS module. 571 - added (experimental) AnyEvent::DNS module.
12 - added (experimental) AnyEvent::Socket module. 572 - added (experimental) AnyEvent::Socket module.
13 - added (experiemntal) support for TLS/SSL in AnyEvent::Handle. 573 - added (experiemntal) support for TLS/SSL in AnyEvent::Handle.
… …
19 - AnyEvent::Socket::inet_aton will use AnyEvent::DNS now and not fork anymore. 579 - AnyEvent::Socket::inet_aton will use AnyEvent::DNS now and not fork anymore.
20 - try very hard not to freeze on broken windows perls. 580 - try very hard not to freeze on broken windows perls.
21 - make on_eof optional in AnyEvent::Handle. 581 - make on_eof optional in AnyEvent::Handle.
22 - support service names in tcp_connect and tcp_server. 582 - support service names in tcp_connect and tcp_server.
23 - work around netbsd/openbsd bug causing unix sockets not to 583 - work around netbsd/openbsd bug causing unix sockets not to
24 become writable untill empty. 584 become writable until empty.
25 - only make a dummy read for error detection when getpeername 585 - only make a dummy read for error detection when getpeername
26 returns ENOTCONN. 586 returns ENOTCONN.
27 - defend AnyEvent::Util::inet_aton against Coro::LWP brutal 587 - defend AnyEvent::Util::inet_aton against Coro::LWP's brutal
28 override. 588 override.
29 - new AnyEvent::Guard::cancel. 589 - new AnyEvent::Guard::cancel.
30 590
313.5 Sat May 17 23:17:14 CEST 2008 5913.5 Sat May 17 23:17:14 CEST 2008
32 - gracefully deal with systems not defining CLOCK_MONOTONIC 592 - gracefully deal with systems not defining CLOCK_MONOTONIC
… …
46 606
473.41 Sun May 11 19:53:13 CEST 2008 6073.41 Sun May 11 19:53:13 CEST 2008
48 - work around a bug in perl 5.8.8's local. Reported by Yi Ma Mao. 608 - work around a bug in perl 5.8.8's local. Reported by Yi Ma Mao.
49 609
503.4 Sun May 11 00:29:25 CEST 2008 6103.4 Sun May 11 00:29:25 CEST 2008
51 - complete re-implementation of condvars with a hopefully much mroe useful 611 - complete re-implementation of condvars with a hopefully much more useful
52 API (backward compatible functions exist but are undocumented). 612 API (backward compatible functions exist but are undocumented).
53 - AnyEvent::Handle fully rewritten with a hopefully more powerful API 613 - AnyEvent::Handle fully rewritten with a hopefully more powerful API
54 (still subject to change). 614 (still subject to change).
55 - take advantage of CLOCK_MONOTONIC, if available, in AnyEvent::Impl::Perl. 615 - take advantage of CLOCK_MONOTONIC, if available, in AnyEvent::Impl::Perl.
56 - provide AnyEvent::post_detect and @AnyEvent::post_detect, which 616 - provide AnyEvent::post_detect and @AnyEvent::post_detect, which

Diff Legend

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