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

Comparing AnyEvent/Changes (file contents):
Revision 1.77 by root, Sat May 17 19:05:51 2008 UTC vs.
Revision 1.337 by root, Sun Aug 16 16:54:51 2009 UTC

1Revision history for Perl extension AnyEvent. 1Revision history for Perl extension AnyEvent.
2 2
35.111 Sun Aug 16 18:44:36 CEST 2009
4 - perl errornously requires sockaddr_un structures returned by the
5 kernel to have a specific length. work around this bug by padding
6 sockaddr_un structures (found to fail on OS X, as perl only uses
7 the correct code for GNU/Linux, but likely to fail on other
8 OSes as well).
9
105.11 Wed Aug 12 17:49:37 CEST 2009
11 - fixed a bug in run_cmd with input from filename.
12 - create a json coder object when none is given in write direction
13 to reduce dependency on JSON.
14
155.1 Tue Aug 11 03:17:32 CEST 2009
16 - tcp_connect now reports the correct error in $!, not an unrelated
17 one (was broken since 4.91).
18 - AE::Handle did not properly clear rtimeout/wtimeout watchers
19 when retrying a tcp_connect.
20 - new functions: AnyEvent::Util::run_cmd and close_all_fds_except.
21 - fixed wrong documentation in AE::Socket.
22
235.01 Mon Aug 10 03:16:32 CEST 2009
24 - last release broke EV child watchers :/.
25
265.0 Sun Aug 9 17:32:09 CEST 2009
27 - officially introduce and document the AE API.
28 - lowering the timeout at runtime did not have immediate effect
29 in AnyEvent::Handle.
30 - AE::Handle now has separate and independent read and write timeout
31 settings.
32 - probe for Guard module when AnyEvent::Util is loaded, not at
33 runtime (Event doesn't like eval "use" from callbacks, and it avoids
34 exporting the probe function).
35 - Impl::IOAsync's io watchers did not use the filehandle emulation
36 code to work around it's watcher limits.
37 - work around IO::Async's problems with overloaded objects.
38 - try to disable wearnings inside IO::Async, it's just too buggy.
39 - do no longer use Event by default, only EV or Perl, Event is too buggy.
40 - hackishly provide fast AE:: interface when EV or Perl is the backend.
41 - inherit VERSION from AnyEvent::VERSION in most modules.
42
434.91 Thu Aug 6 15:42:45 CEST 2009
44 - AE::Handle::starttls could get out of sync when the read buffer
45 already contains some TLS handshake.
46 - AE::Handle did not properly free the TLS session, causing bigger
47 memory leaks in Net::SSLeay than expected (reported by toaster).
48 - AE::Socket::tcp_connect will now artificially delay invoking
49 the callback to avoid returning after invoking the callback.
50 - convert many internal watcher uses to AE API.
51
524.9 Sat Aug 1 11:07:01 CEST 2009
53 - Glib silently fails when registering a timeout with
54 a negative value, so avoid that.
55 - call condvar callback immediately when it is set
56 after the condition is already signalled.
57 - check rbuf_max condition only after trying to consume
58 data.
59 - normalise signal numbers to names when using the ae's
60 signal handling, but do not document this (yet).
61 - pure perl signal emulation did not properly set nonblocking
62 mode on the signal pipe on win32 (this is unlikely to have caused
63 any issues).
64 - new module: AnyEvent::Debug.
65 - AnyEvent::Strict now checks that a signal was specified
66 by name (not name or number).
67 - reduce memory footprint in typical cases by ~50kb by
68 only compiling signal code when necessary.
69 - add AnyEvent::Handle->rbuf_max.
70 - grab the AE namespace for future new API, implement stubs for
71 said future/alternative API.
72 - new function: AnyEvent::Socket::format_hostport.
73
744.881 Tue Jul 28 12:51:53 CEST 2009
75 - work around a bug in local in pre-5.10 perls, causing
76 AnyEvent::Handle to recurse when it should not
77 (analyzed by elmex).
78
794.88 Tue Jul 28 04:04:37 CEST 2009
80 - re-bless the handle into a dummy package after calling
81 AnyEvent::Handle::destroy, so the user does not need to check
82 for errors after every push_write etc.
83 - do not attempt to run t/02_signals.t on obviously broken
84 platforms.
85
864.87 Sun Jul 26 02:06:16 CEST 2009
87 - do not attempt to linger when there is no longer a valid fh.
88 - enforce tls mode to be either accept or connect, do not simply
89 segfault in Net::SSLeay.
90 - AnyEvent::Handle can now call tcp_connect itself (new parameters
91 connect, on_prepare, on_connect and on_connect_error). Updated
92 tutorial accordingly.
93 - add AnyEvent::Impl::Irssi backend.
94
954.86 Mon Jul 20 23:52:29 CEST 2009
96 - since the verbose warning is not enough, explicitly document
97 that versions before 1.33 of Net::SSLeay are not secure.
98 - work around signal handling races in Event and (...) Event::Lib.
99 - try to align signal-race timer to full-second boundaries.
100 - work around Tk not liking negative timeouts.
101 - don't complain of different grades of environmental unfriendlyness
102 in IO::Async.
103
1044.85 Sat Jul 18 06:16:14 CEST 2009
105 - nail the signal race problem in perl once and for all
106 (see $AnyEvent::MAX_SIGNAL_LATENCY).
107 - take advantage of Async::Interrupt if it is available.
108 - load Time::HiRes and Guard modules on demand only.
109 - add optional/recommended modules section to AnyEvent
110 documentation.
111 - reduce memory usage considerably (and reduce startup penalty)
112 by not using "strict", "warnings" and "overload" modules.
113 - work around buggy windows/openbsd perls and provide EBADMSG
114 and EPROTO ourselves when missing.
115 - improve perl 5.6 compatibility of the core event loop.
116 - made Net::SSLeay version 1.33 a soft requirement.
117
1184.83 Fri Jul 17 16:56:26 CEST 2009
119 - implement AnyEvent::Socket::getprotobyname.
120 - AnyEvent::CondVar's will now detect recursive blocking
121 waits and will croak, as too many people fall into
122 this trap.
123 - AnyEvent::Handle will now call ->destroy on itself after
124 executing the on_error callback, instead of doing some
125 half-baked internal shutdown, for fatal errors.
126 - clarify on_eof behaviour w.r.t. the read queue and
127 on_read callbacks.
128 - ignore some possible spurious wake-ups in tcp_connect.
129
1304.82 Sat Jul 11 00:34:55 CEST 2009
131 - POE and Event backends didn't accept some callable objects as
132 callbacks.
133 - use Config module instead of POSIX module to detect signal names
134 in AnyEvent::Strict and AnyEvent::Impl::EventLib, as the POSIX
135 module doesn't even have all POSIX signals :/.
136 - use more workarounds around the many refcnt/corruption bugs in
137 Event::Lib.
138 - work around a race condition in perl's select, causing t/03_child.t
139 to rarely fail.
140
1414.81 Thu Jul 9 10:30:30 CEST 2009
142 - AnyEvent::Handle didn't properly diagnose write errors
143 (it expected -1 from syswrite, how lame... :).
144 - support file descriptors in addition to file handles
145 in AnyEvent->io.
146 - new env variables: PERL_ANYEVENT_RESOLV_CONF,
147 PERL_ANYEVENT_MAX_OUTSTANDING_DNS, PERL_ANYEVENT_CA_FILE
148 and PERL_ANYEVENT_CA_PATH.
149 - provide a sensible synopsis section for AnyEvent::TLS.
150 - add a "supported backends" section to the manpage.
151 - added simple io watcher test to testsuite, using a
152 portable_socketpair.
153 - tried to improve the stability of the Event::Lib backend,
154 YMMV.
155
1564.8 Mon Jul 6 23:45:16 CEST 2009
157 - AnyEvent::DNS did not properly follow CNAME records with
158 uppercase targets.
159 - AnyEvent::DNS would errornously return AAAA records
160 with v4 mapped addresses (a faulty record) as ipv4 addresses,
161 causing AnyEvent::Socket to throw an exception.
162 - added new module AnyEvent::TLS for easier SSL/TLS context
163 creation, with many options including hostname verification,
164 secure default configuration, lots of documentation and,
165 predefined diffie-hellman keys for perfect forward security
166 and much more. get it while it's still fresh!
167 - use AnyEvent::TLS in AnyEvent::Handle for context management.
168 - load AnyEvent::Handle only on demand in AnyEvent::DNS,
169 so AnyEvent::Socket users have smaller memory footprint
170 in the common case.
171 - add AnyEvent::Handle->push_shutdown method.
172 - add an additional error message parameter to AnyEvent::Handle's
173 on_error callback (for TLS, $! is still available).
174 - add AnyEvent::Handle on_starttls/on_stoptls callbacks.
175 - make AnyEvent::Handle more robust against early conenction
176 failures (during new), and return C<undef> in that case
177 from the constructor.
178 - AnyEvent::Handle will now try to load only JSON::XS first,
179 then fall back to JSON.
180 - format_ipv4/format_ipv6 are now exported by default, for symmetry,
181 and because it was documented that way.
182
1834.451 Fri Jul 3 00:28:58 CEST 2009
184 - do not clear rbuf when shutting down an AnyEvent::Handle
185 object - doing so breaks AnyEvent::HTTP.
186
1874.45 Mon Jun 29 22:59:26 CEST 2009
188 - a write error could cause AnyEvent::Handle to create
189 an I/O watcher with an undefined $fh.
190 - special-case mapped ipv4 addresses in both
191 AnyEvent::Socket::format_address and parse_address,
192 to treat them just like ipv4 addresses.
193 - updated and overhauled the AnyEvent::Intro doc.
194 - implement AnyEvent::Socket::format_ipv4/ipv6.
195 - slightly speed up portable_pipe/socketpair.
196 - expand condvar begin/end documentation.
197
1984.42 Fri Jun 26 08:32:18 CEST 2009
199 - preliminary and neccesarily incomplete support for IO::Async.
200 - reset SIGCHLD to DEFAULT when AnyEvent is loaded, in
201 case it was set to IGNORE, to ensure we can catch child statuses
202 even when the calling env acts stupidly.
203 - updated benchmarks with IO::Async, which performs very well.
204
2054.412 Wed Jun 24 01:35:57 CEST 2009
206 - support an "untaint" attribute for AnyEvent::DNS
207 and set it on the default resolver.
208 - implement and document AnyEvent::Impl::Perl::loop.
209 - remove all anyevent-env variables from %ENV
210 when running in tainted mode.
211 - mention and extend the IO::Lambda benchmark.
212
2134.411 Sun Jun 7 18:48:02 CEST 2009
214 - do not try to use F_SETFD on windows.
215
2164.41 Thu May 14 06:40:11 CEST 2009
217 - work around issues in older perls (5.8.0?) when
218 a signal handler is deleted from the %SIG hash.
219 - use POSIX::_exit in child test, to avoid running
220 destructors.
221 - speed up CHLD handling by relying on SIGCHLD being
222 synchronously delivered, even when we roll our own
223 implementation.
224 - AnyEvent::DNS: add the "dname" resource record name
225 for cosmetic reasons.
226
2274.4 Sun Apr 26 20:12:33 CEST 2009
228 - implemented idle watchers, where applicable.
229 - AnyEvent->time died when Event backend was in use.
230 - fix a memleak in the Tk backend.
231 - sped up Tk timer handling.
232 - clip negative "after" values to 0 in AnyEvent::Impl::Event
233 to avoid spamming out warning messages.
234 - fix Qt timers without interval.
235 - avoid Qt zero-timeout special case, in old libqt's.
236
2374.352 Mon Apr 20 16:31:11 CEST 2009
238 - fix AnyEvent::Strict error messages for child watchers.
239 - fix/update Makefile.PL "recommends" versions.
240 - add AnyEvent->now_update.
241
2424.351 Sat Apr 11 07:56:14 CEST 2009
243 - actually make the signal pipe work under win32.
244 - localise $! in signal handler to avoid changing $!, although
245 perl itself does not seem to save/restore errno either.
246 - set the cloexec flag on the signal pipe (normally set by perl too).
247
2484.35 Fri Mar 27 11:48:20 CET 2009
249 - event models relying on AnyEvent's signal watcher emulation did
250 invoke the callback asynchronously, contradicting documentation
251 and causing signals to get lost (this includes AnyEvent's own
252 event loop). AnyEvent now uses the standard pipe trick to make
253 callback execution synchronous to the event loop.
254 - AnyEvent::Handle didn't free TLS context data on DESTROY
255 (patch by Pavel Shaydo).
256 - work around the ever-incompatibly-changing API of MakeMaker.
257 - document that changing global variables without restoring them
258 is a bad idea in Perl (noted by Adam Rosenstein).
259 - AnyEvent::Strict now barfs if ->io is passed a file.
260
2614.34 Thu Feb 12 18:32:45 CET 2009
262 - separately buffer TLS read data, as otherwise the read queue could
263 deadlock as receiving data is not expected while draining the
264 read queue (which cna only happen with TLS).
265 - raise EBADMSG error on JSON decoding errors.
266 - fix some minor manpage bugs (reported by Maximilian Gaß).
267 - speed up select bitmask parsing quite a bit in the pure perl
268 backend.
269 - use CORE::select instead of just select in AnyEvent::Impl::Perl.
270
2714.331 Tue Jan 6 21:07:25 CET 2009
272 - socketpair fails on many vista machines because vista has
273 completely broken accept/getpeername and getsockname functions,
274 so we provide our own socketpair emulation that kind of works
275 (AnyEvent::Util::portable_pipe).
276 - new function: AnyEvent::Util::portable_socketpair.
277 - take advantage of the Guard module if it exists.
278
2794.33 Fri Nov 21 02:35:40 CET 2008
280 - AnyEvent::Strict did errornously flag a fileno of 0 as
281 illegal.
282 - reduce memory usage and slightly speed up the pure perl backend
283 by only storing the file descriptor, not the file handle.
284 - add missing autocork method to AnyEvent::Handle (reported
285 by Adam Rosenstein).
286 - AnyEvent::DNS->resolve errornously documented an $rcode result
287 argument, but there isn't (spotted by Henrik Krohns).
288 - the naked truth about Tk - it's basically unmaintained.
289
2904.32 Mon Nov 3 22:46:32 CET 2008
291 - fix AnyEvent::Socket::resolve_sockaddr to properly support
292 unix sockets again and choose a proper default for
293 the socket type.
294 - fix call to resolve_sockaddr in tcp_connect.
295
2964.31 Thu Oct 30 04:41:48 CET 2008
297 - implemented AnyEvent::Handle->destroy method.
298 - hint about unexpected effects in TLS mode.
299 - speed up AnyEvent::Impl::Perl by using more arrays
300 instead of hashes without the slightest loss of readability :)
301 - work around a perl argument refcounting bug.
302 - with some perl patches applied, AnyEvent now seems leak-free.
303
3044.3 Fri Oct 3 09:18:43 CEST 2008
305 - AnyEvent will now install a no-op signal handler for SIGPIPE,
306 unless one has been installed already.
307 - warn about not loaidng AnyEvent::Impl::POE early enough
308 (patch by Adam Rosenstein).
309 - fixed a great number of bugs and corner cases in AnyEvent::Handle:
310 - fix a bug in where in SSL connect mode, the
311 client would first wait for some data by the server and
312 otherwise hang. (reported and analyzed in an absolutely
313 exemplary manner by Adam Rosenstein).
314 - fix a bug in where SSL EOF would not be treated
315 as stream EOF, putting the connection into a hung state.
316 (reported and analyzed in an absolutely exemplary manner by Adam
317 Rosenstein).
318 - fix a potential segfault when the TLS context
319 would go missing in a read callback (e.g. due to stoptls) -
320 Net::SSLeay of course makes no type checking whatsoever.
321 - AnyEvent::Handle will not stop the read watcher in TLS mode.
322 - AnyEvent::Handle->stoptls will send a TLS close notify
323 instead of simply destroying the stream now.
324 - fix a bug where large blocks of data written in TLS mode
325 would not be sent unless triggered by receives.
326 - on_drain will now take the tls write buffer into account.
327 - SSL operations have been streamlined a bit and should be faster
328 now, more to come, though.
329 - added a FAQ section, will document nontrivial issues in there.
330
3314.234 Mon Sep 29 04:08:13 CEST 2008
332 - fix child watcher documentation: only child exits will be reported,
333 no trace events.
334 - mention SIGPIPE in AnyEvent::Handle.
335 - perl backend will now use POSIX::times as monotonic clock
336 when available and the monotonic clock option is not.
337 - run even without Time::HiRes in some cases now, but do not
338 provide sub-second accuracy (all *supported* perl versions have
339 Time::HiRes).
340
3414.233 Fri Aug 22 01:48:05 CEST 2008
342 - fix a bug in the testsuite, causing a hang.
343 - clarified lots of AnyEvent::Handle constructor arguments.
344
3454.232 Thu Aug 21 20:44:25 CEST 2008
346 - fix a bug in AnyEvent::Handle that could cause two fatal error
347 callbacks on EOF, which caused AnyEvent::HTTP to signal success twice
348 (testcase by Leon Brocard).
349 - clarified on_eof/on_error documentation a bit.
350
3514.231 Tue Jul 29 13:12:15 CEST 2008
352 - remove some debugging code left in AnyEvent::Util::fork_call
353 (and no, it's impossible to implement with the broken windows
354 perls without resource leaks or worse).
355
3564.23 Tue Jul 29 12:19:59 CEST 2008
357 - document the first parameter passed to condvar callbacks to be
358 the callback.
359 - add AnyEvent::Socket::{ntoa,aton} aliases.
360 - optimize the AE::Handle->push_read (line) for the default
361 eol marker.
362 - optimize push_read (packstring|storable) for small packets.
363 - invoke on_error callback when no on_eof callback is set.
364 - fix a bug in push_read (storable) of unknown impact.
365
3664.22 Sun Jul 20 16:34:13 CEST 2008
367 - new function AnyEvent::Socket::parse_hostport.
368 - as the bulkheads at microsoft can't even get getprotobyname reliably
369 working on their shitty fucking broken os we need to hardcode
370 some common protocol numbers in AnyEvent::Socket. How can
371 people even bother with such a pile of shit as windows.
372
3734.21 Thu Jul 17 14:40:05 CEST 2008
374 - INCOMPATIBLE CHANGE: AnyEvent::Util::fork_call now has a prototype
375 which is incompatible to previous invocation syntax.
376 - work around more windows perl bugs in fork_call. I cannot imagine
377 why anybody would bother implementing fork in such an obviously
378 unusable way.
379 - avoid calling $do_retry if the request has already finished in
380 AnyEvent::DNS, thus avoiding an exception (reported by Anatoly
381 K. Sharifulin).
382
3834.2 Sat Jul 12 22:42:11 CEST 2008
384 - support an "interval" argument for repeating timers.
385 - fix a bug in the pure perl event loop that caused timers to
386 be delayed under some circumstances.
387 - implement optional argument checking via PERL_ANYEVENT_STRICT.
388 - not importing anything from Event causes it to malfunction
389 and not use Time::HiRes, work around this bug.
390 - more bug workarounds against the endlessly broken Tk module,
391 seems to work now on 32 bit machines, but less so on 64 bit.
392 - minor optimisations applied to most backends.
393
3944.161 Fri Jul 4 14:24:48 CEST 2008
395 - fixed the t/handle/02_write.t, it read after write, with the
396 assumption that all data could be written before read
397 (reported by Jim Razmus).
398
3994.160 Thu Jul 3 04:02:21 CEST 2008
400 - re-registering signal watchers after unregistering one did
401 not work if a backend falls back on AnyEvent's default
402 implementation (also affected child watchers).
403 - new AnyEvent::Handle options: autocork and no_delay.
404
4054.152 Sun Jun 22 14:15:44 CEST 2008
406 - allow for 32-bit perls that implement shifts differently
407 on different architectures in parse_ipv4 (reported and
408 analysed by Keiichi DAIBA).
409
4104.151 Fri Jun 6 17:34:24 CEST 2008
411 - make sure specifying _only_ on_read and never pushing reads
412 works.
413
4144.15 Fri Jun 6 13:00:46 CEST 2008
415 - the pure perl backend would keep some watchers alive when more than
416 one watcher was registered for the same fd.
417 - new "packstring" and "storable" read and write types
418 for AnyEvent::Handle.
419 - allow on_eof handler to be called after on_error with EPIPE returns.
420 - do not immediately call on_read callback in handle constructor.
421
4224.14 Thu Jun 5 20:29:31 CEST 2008
423 - Fixed a bug in DNS SRV priority sorting.
424 - AnyEvent::Util::guard now reports runtime errors while
425 executing the guard block as warnings.
426 - handle 0-byte-reads just before EOF correctly in AnyEvent::Handle.
427
4284.13 Thu Jun 5 00:47:59 CEST 2008
429 - AnyEvent::DNS only followed cname chains with length 2,
430 contrary to documentation. bumped it up to 10, thanks to
431 microsoft, the current broken-dns-config-king.
432 - AnyEvent::DNS didn't check for socket return status
433 but instead relied on perl not creating filehandles in that
434 case - too bad it gives you a bogus file handle (reported
435 and analysed by Vladimir Timofeev).
436 - fix queue management logic in AnyEvent::Handle:
437 when on_read was registered, the queue was empty
438 and no progress could be made AnyEvent::Handle would
439 enter an endless loop.
440 - correctly start reading again when the handle became
441 busy again after idling.
442 - correctly treat tls shut-downs as EOF condition
443 (for the time being, would be nice to have a callback for that).
444 - correctly call eof callback under all conditions
445 (hopefully).
446 - the timeout callback did not expect that $self can go
447 away any time.
448
4494.12 Tue Jun 3 10:58:04 CEST 2008
450 - include AnyEvent::Intro, a tutorial for anyevent,
451 anyevent::socket and anyevent::handle.
452 - allow more options in on_error.
453
4544.11 Fri May 30 23:42:25 CEST 2008
455 - INCOMPATIBLE CHANGE: replace ptr by real PTR lookup, provide
456 reverse_lookup and reverse_verify to replace it, support
457 v4mapped and v4compat addresses.
458 - provide more documentation for the resolver class.
459 - really replace longest run of :0: by :: in format_address,
460 also properly convert :: and ::1 again.
461 - support NAPTR record name and decode it.
462 - implement random weight sampling for SRV records, as per
463 rfc 2782.
464 - correctly abort on srv-record targets of ".".
465 - added AnyEvent::DNS::wait_for_slot.
466 - in the unlikely event of a virtual circuit connection
467 being invalidated by a delayed udp reply, AnyEvent::DNS
468 could die. this has been fixed.
469 - plug a probable memleak in the DNS vc code.
470 - use configured timeout also for VC request phase.
471 - implement timeout and max_outstanding methods
472 for resolver class.
473 - update version numbers in all modules.
474
4754.1 Thu May 29 05:45:40 CEST 2008
476 - INCOMPATIBLE CHANGE: renamed xxx_ip functions to xxx_address
477 which mirrors their purpose better (old names still available
478 for a while). Also moved AnyEvent::DNS::addr to
479 AnyEvent::Socket::resolve_sockaddr.
480 - implement AnyEvent->time and AnyEvent->now.
481 - fix IPv6 support in pack_sockaddr.
482 - officially un-experimentalise all newly introduced code.
483 - support unix domain sockets everywhere by specifying
484 a host of "unix/" and the pathname as service.
485 - implement an activity timeout in AnyEvent::Handle.
486 - added a regex read type to AnyEvent::Handle.
487 - added a json read/write type to AnyEvent::Handle.
488 - always croak in on_error in AnyEvent::Handle.
489 - document how to subclass AnyEvent::Handle.
490 - implement AnyEvent::Util::fork_call.
491 - add support for IPv6 nameservers and nameserver statements.
492 - work around _yet_ _another_ windows perl bug in where empty
493 select masks cause errors in the pure perl backend (this
494 workaround cannot be implemented for the other event loops,
495 but EV also works around this).
496 - supply AnyEvent::Socket::address_type and make good use of it.
497 - clarify and fix the Handle documentation w.r.t $self vs. $handle
498 vs. callback arguments.
499 - add some recommends to the META.yml.
500
5014.05 Mon May 26 19:44:06 CEST 2008
502 - some platforms ignore the rfc and prepend an extra sa_len member to
503 the sockaddr structure, cater for those. also use sockaddr_family,
504 didn't know it was there, but comes in handy for the workaround.
505 - undo the import WIN32 hack.
506
5074.04 Mon May 26 08:03:31 CEST 2008
508 - try to work around yet another windows bug: failed connects are reported
509 as if it were out-of-band data. windows users: you suck. Only
510 EV, Glib, Event and the pure perl backend can handle this condition.
511 - optimize write algorithm in AnyEvent::Handle.
512 - properly parse PERL_ANYEVENT_PROTOCOLS and use the info correctly.
513 - tcp_connect now properly iterates through all targets.
514 - check wether IPv6 sockets can actually be created, otherwise, disable ipv6
515 support entirely. also hardcode some AF_INET6 constants for the
516 sake of perl 5.8.
517 - vastly improved nameserver/suffix detection algorithm for windows
518 (still a hack...)
519 - try to use a fallback dns server if no dns servers could be found.
520 - splatter around some local $SIG{__DIE__}'s to work around bad
521 die handlers.
522
5234.03 Sun May 25 05:05:57 CEST 2008
524 - work around perl bugs on windows, where perl returns undocumented
525 error codes for sysread, syswrite etc.
526 - fix AnyEvent::Handle::unshift_read to really unshift and not push.
527 - AnyEvent::Handle could sometimes stumble over 0-byte writes.
528 - fix unpack_sockaddr, now works for AF_INET6 sockaddr's.
529 - updated documentation.
530 - lots of workarounds for perl 5.8 quirks.
531 - implement a more flexible type system for both reads and writes
532 in AnyEvent::Handle.
533 - support IPv6 in tcp_server.
534 - pass local hostname and port to the prepare callback in tcp_server.
535 - make types extendable in AnyEvent::Handle.
536 - croak, not die, in AnyEvent::Handle;
537
5384.0 Sat May 24 19:58:08 CEST 2008
539 - added (experimental) AnyEvent::DNS module.
540 - added (experimental) AnyEvent::Socket module.
541 - added (experiemntal) support for TLS/SSL in AnyEvent::Handle.
542 - make condvar->end call ->send as documented.
543 - make condvar constructor support the documented cb => argument.
544 - overload condvars so you can use them instead of code references.
545 - support more of the POSIX forms instead of a pure dotted quad for
546 verbatim ip addresses.
547 - AnyEvent::Socket::inet_aton will use AnyEvent::DNS now and not fork anymore.
548 - try very hard not to freeze on broken windows perls.
549 - make on_eof optional in AnyEvent::Handle.
550 - support service names in tcp_connect and tcp_server.
551 - work around netbsd/openbsd bug causing unix sockets not to
552 become writable until empty.
553 - only make a dummy read for error detection when getpeername
554 returns ENOTCONN.
555 - defend AnyEvent::Util::inet_aton against Coro::LWP's brutal
556 override.
557 - new AnyEvent::Guard::cancel.
558
5593.5 Sat May 17 23:17:14 CEST 2008
560 - gracefully deal with systems not defining CLOCK_MONOTONIC
561 (as opposed to not supporting the clock itself, which worked in 3.41).
3 - added AnyEvent::Util::tcp_connect and AnyEvent::Util::tcp_server 562 - added AnyEvent::Util::tcp_connect and AnyEvent::Util::tcp_server
4 helpers, to replace the ill-designed AnyEvent::Socket class. 563 helpers, to replace the ill-designed AnyEvent::Socket class.
5 - added some examples (eg/connect, eg/handle). 564 - added some examples (eg/connect, eg/handle).
6 - added AnyEvent::Util::guard. 565 - added AnyEvent::Util::guard.
7 - AnyEvent::Util now exports most of its functions. 566 - AnyEvent::Util now exports most of its functions.
8 - fixed a bug in *_read_chunk, where the callback was called with $self as 567 - fixed a bug in *_read_chunk, where the callback was called with $self as
9 first and second argument. 568 first and second argument.
10 - simplified and fixed regex handling in AnyEvent::Handle::*_read_line. 569 - simplified and fixed regex handling in AnyEvent::Handle::*_read_line.
11 - fix manpage condvars examples to use new syntax. 570 - fix manpage condvars examples to use new syntax.
571 - considerably improve inet_aton with EV::ADNS backend to support
572 cname chains (common dns configuration bug, e.g. with www.google.de)
573 and aaaa records if no a record could be found.
12 574
133.41 Sun May 11 19:53:13 CEST 2008 5753.41 Sun May 11 19:53:13 CEST 2008
14 - work around a bug in perl 5.8.8's local. Reported by Yi Ma Mao. 576 - work around a bug in perl 5.8.8's local. Reported by Yi Ma Mao.
15 577
163.4 Sun May 11 00:29:25 CEST 2008 5783.4 Sun May 11 00:29:25 CEST 2008
17 - complete re-implementation of condvars with a hopefully much mroe useful 579 - complete re-implementation of condvars with a hopefully much more useful
18 API (backward compatible functions exist but are undocumented). 580 API (backward compatible functions exist but are undocumented).
19 - AnyEvent::Handle fully rewritten with a hopefully more powerful API 581 - AnyEvent::Handle fully rewritten with a hopefully more powerful API
20 (still subject to change). 582 (still subject to change).
21 - take advantage of CLOCK_MONOTONIC, if available, in AnyEvent::Impl::Perl. 583 - take advantage of CLOCK_MONOTONIC, if available, in AnyEvent::Impl::Perl.
22 - provide AnyEvent::post_detect and @AnyEvent::post_detect, which 584 - provide AnyEvent::post_detect and @AnyEvent::post_detect, which

Diff Legend

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