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

Comparing AnyEvent/Changes (file contents):
Revision 1.200 by root, Mon Oct 27 02:58:19 2008 UTC vs.
Revision 1.452 by root, Tue Aug 2 20:02:44 2011 UTC

1Revision history for Perl extension AnyEvent. 1Revision history for Perl extension AnyEvent.
2 2
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: 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 - some read types (regex, netstring, json, storable) could
14 errornously keep an AE::Handle object alive on their own
15 (reported by Mohammad Toossi).
16 - add support for (currently third-party) AnyEvent::Impl::FLTK.
17 - many read types in AnyEvent::Handle would malfunction during
18 transitions to TLS because of a "delete $self->{rbuf}".
19 (reported by Gerald Galster).
20 - resolve AnyEvent::Util::guard at first-use-time, not at load time,
21 for further speed and memory savings when not used.
22 - improve condvar blocking wait performance (probably only noticable
23 with faster backends), also make it easier for event loops that
24 do not support blockign waits.
25
265.34 Thu May 12 09:59:41 CEST 2011
27 - AE::TLS didn't support both cert_file and key at the same time,
28 due to some copy&paste bug (reported and analyzed by vti).
29 - improve AE::Handle timeout documentation slightly.
30
315.33 Mon Feb 7 21:51:41 CET 2011
32 - never appeared on CPAN, for unknown reasons.
33 - AnyEvent::Handle on_connect_error - do not try to destroy a handle
34 object when it's already gone (analyzed by Ryan Bullock).
35 - update idna algorithm(s) to UTS#46 version 6.0.0.
36 - preserve trailing dot in idn_to_ascii - testsuite says to preserve,
37 uts46 says it's an error, and testsuite makes more sense.
38 - remove leading dots in idn_to_ascii - testsuite says to remove,
39 uts46 says it's an error, and testsuite makes more sense.
40 - fixed a bug in idn ignored character processing.
41 - passes the uts46 6.0.0 testsuite except for three cases which
42 apparently are bugs in the testsuite itself.
43 - passes the uts46 6.0.1 testsuite except for eleven cases which
44 apparently are bugs in the testsuite itself.
45
465.31 Sun Jan 23 11:49:19 CET 2011
47 - work around a POE bug where POE would first poll for new events and
48 invoke handlers BEFORE handling the message queue.
49 - fix parsing of "domain" statement in resolv.conf (Johannes Plunien).
50 - stop_read now does just that in TLS mode. see its
51 documentation for caveats.
52 - update to common::sense 3.4 warning set (some warnings in perl impose
53 an arbitrary 32kb limit on regex matches, which affects
54 AnyEvent::Handle).
55 - AE::Handle had a broken rbuf_max method and missing wbuf_max method.
56 - new experimental AE::Handle::resettls method.
57
585.3 Fri Dec 31 05:48:13 CET 2010
59 - major IOAsync patch by Paul Evans - 0.33 is required, as the major
60 issues with IO::Async have been solved in that version, including
61 the availabilty of a default loop!
62 - implement new wbuf_max parameter for AnyEvent::Handle.
63 - added Cocoa::EventLoop backend by Daisuke Murase.
64 - the default testsuite now honors PERL_ANYEVENT_MODEL (Paul Evans).
65 - clarify handle on_prepare documentation.
66 - set CLOEXEC flag on parent-side fd's in AnyEvent::Util::run_cmd,
67 to avoid leaking them to any child processes.
68 - try to force files into the arch-dependend lib directory, for the
69 half-"turly obsessive compulsive person"(s) that depend on it.
70
715.29 Sun Dec 5 10:49:21 CET 2010
72 - convert EV backend to EV 4.00 API (so better upgrade EV too).
73 - AE::Handle: implement dynamic read_size adjustment, add
74 max_read_size parameter, reduce default read_size to 2048.
75 - add some FAQ entries for servers and/or on_eof vs. on_error.
76 - work around OS bugs (cygwin again) when lingering in AE::Handle.
77 - reduce memory usage slightly when connect was used in AE::Handle.
78 - wrok around more segfaults in Tk.
79 - document tls_ctx => undef as valid.
80 - detect WNOHANG value at compiletime.
81 - include some probably linux-only support for building AnyEvent
82 as part of the perl core.
83 - improve compatibility of t/08_idna.t to perl 5.8.
84 - make AnyEvent::Strict truly optional (in case somebody builds a
85 minimal perl...).
86
875.28 Wed Oct 13 04:14:23 CEST 2010
88 - due to a glitch, AnyEvent's internal getprotobyname wasn't
89 actually used in tcp_server.
90 - implement AnyEvent::CondVar->new method, which was documented but
91 missing.
92 - support autoloading of read types in unshift_read.
93 - AnyEvent::Strict: do the test for invalid fh arguments differently,
94 resulting in more informative error messages.
95 - new AnyEvent::FAQ manpage.
96 - clarify that you can actually call push_read etc. in on_read.
97 - put keys and certs in /tmp, not into the current directory,
98 and document this in the SECURITY CONSIDERATIONS sections
99 of AnyEvent::TLS.
100 - support ";" as resolv.conf comment character in AnyEvent::DNS, as
101 requested by Ron Isaacson.
102 - document $AnyEvent::DNS::RESOLVER variable.
103 - incorporated major typo patches by Abhijit Menon-Sen.
104 - AnyEvent::Handle now croaks when negative timeouts are passed.
105 - add a shitload of TCP_xxx sockopt name constants to AnyEvent::Util,
106 undocumented, but who knows when they might come in handy.
107
1085.271 Tue Jun 8 12:05:46 CEST 2010
109 - backport to perl 5.8.x.
110
1115.27 Sun Jun 6 12:12:05 CEST 2010
112 - postpone differently in AnyEvent::Socket now, as
113 when not, canceling the connection attempt might fail
114 (found by Felix Antonius Wilhelm Ostmann).
115 - explicitly check for non-stream sockets in AE::Handle, too many
116 clueless people fell into the trap of this somehow working.
117 - simplified and reworked the "OTHER MODULES" section.
118 - better/more condvar examples.
119
1205.261 Wed Apr 28 16:13:36 CEST 2010
121 - AF_INET6 was not properly used from Socket6 during configuration
122 time (found by Andy Grundman).
123
1245.26 Mon Apr 12 04:49:35 CEST 2010
125 - don't generate spurious readyness notifications when select
126 returns EINTR in the pure perl backend (analysed and testcase
127 by Bas Denissen).
128 - give same examples in SYNOPSIS sections of AnyEvent and AE.
129 - provider faster implementations for AE::time/now/now_update for
130 EV and Perl backends.
131 - provide faster AE::cv and AE::time implementations for all
132 backends.
133 - no longer support lower/mixed-case signal names (this was
134 never documented nor universally supported).
135 - some more memory µ-optimisations, and somewhat less messy
136 code to cope with both AnyEvent and AE APIs.
137
1385.251 Sat Mar 13 00:58:21 CET 2010
139 - make test failed if it was run before make install on perl <=
140 5.8.8, and older versions of perl make our live very difficult,
141 so write constants.pl during Makefile.PL time.
142
1435.25 Sat Mar 13 00:23:14 CET 2010
144 - fix a race condition in AnyEvent::Handle that would cause
145 a "bio_read: unsupported method" error (in _another_
146 ssl connection) after user code threw an exception.
147 - added AnyEvent::Handle->destroyed method.
148 - speed up AnyEvent::Socket::format_address and ::format_ipv6.
149 - the AnyEvent::Util::fh_nonblocking function worked only
150 by lucky accident on win32.
151 - smaller and faster AnyEvent::Util::fh_nonblocking.
152 - when the (required!) Time::HiRes module is missing, AnyEvent
153 did not fall back to built-in time properly.
154 - do not load Fcntl at runtime, saving memory and loading time.
155 - precompile a number of constants and use them instead of runtime
156 detection and eval.
157 - free detection code after detection and similar memory optimisations.
158 - Perl backend timer interval best effort drift has been improved
159 (same algorithm as EV).
160 - update unicode idna mapping table.
161
1625.24 Tue Jan 5 11:39:43 CET 2010
163 - cygwin never reports errors from failed connects. 1.5 just gives
164 you continous readyness and EAGAIN, 1.7 is even more broken and
165 just hangs. work around both issues in a cygwin-specific hack.
166 - improve idle watcher documentation slightly.
167
1685.23 Sun Dec 20 23:48:00 CET 2009
169 - support IDNs in resolve_sockaddr, and therefore in tcp_connect.
170 - implement punycode_encode/decode, idn_nameprep,
171 idn_to_ascii and idn_to_unicode operations in AnyEvent::Util.
172 - provide $AE::VERSION.
173 - removed traces of "no strict 'refs'".
174
1755.22 Sat Dec 5 03:51:13 CET 2009
176 - downgrade-or-fail in AnyEvent::Handle::push_write, to
177 diagnose encoding failures earlier and more succinctly.
178 (this works around bugs in perl, throwing away encoding info
179 when passing scalar data to extensions).
180 - add more examples to AnyEvent::Socket manpage.
181 - upgrade internal warning set to the same as common::sense 2.03.
182 - use pack "n/a*" for pre-5.8.9 perl compatibility in AnyEvent::DNS
183 (John Beppu).
184 - AnyEvent::Socket::inet_aton now properly supports ipv6, as documented.
185 - add google public dns servers to fallback server set and make sure
186 we load-balance properly between the three sets. also add all
187 fallback dns servers, not just a random one, to each dns config.
188
1895.21 Thu Nov 19 02:48:47 CET 2009
190 - fix a problem where socket constants were called with parameters
191 (spotted by David Friedland).
192 - fork_call never use'd POSIX (reported by Daisuke Maki).
193 - improve perl 5.6 compatibility further (but it still won't work
194 unless you rip out everything but the core).
195 - prefer Net::DNS::Resolver over ipconfig, if installed, on win32.
196 uses 10MB of RAM, but doesn't open a console window. *sigh*.
197
1985.202 Wed Oct 14 22:35:44 CEST 2009
199 - AnyEvent::DNS would unexpectedly clobber $_ under windows
200 (analysed by Matthias Waldorf).
201 - AnyEvent::Handle::run_cmd can now pass the PID of the
202 newly-created process, which is much less useful than it might
203 sound (based on patch by Yann Kerherve).
204
2055.201 Tue Sep 29 12:09:25 CEST 2009
206 - AnyEvent:Handle::on_starttls/on_stoptls methods were broken
207 (reported by Torsten Foertsch).
208 - common::sense 2.0 could cause tcp_server to throw an exception
209 (analysed by elmex).
210
2115.2 Mon Sep 14 07:04:49 CEST 2009
212 - INCOMPATIBLE CHANGE: do no longer support register_read_type
213 and register_write_type in AnyEvent::Handle, instead support
214 package names (the facility was mostly abused).
215 - implement "packagename-as-read/write type" support in
216 AnyEvent::Handle.
217 - AnyEvent::Handle: new options "keepalive" and "oobinline".
218 - oobinline set by default to avoid security issues.
219 - the pure-perl event loop backend wrongly detected times() fallback
220 support (spotted by Pavel Boldin).
221
2225.12 Tue Sep 1 20:26:50 CEST 2009
223 - be more lenient when parsing resolv.conf files, as some people
224 use hashmarks as comment indicator inside directives (reported
225 by Michael S. Fischer).
226 - use same set of warnings as common::sense 2.0.
227 - fix a potential 32 bit overflow issue due to perl having problems
228 with large hex constants in the Perl backend.
229 - do not use the slower 5.6 version fo the enc_name code in AnyEvent::DNS
230 by default.
231 - fix some prototypes.
232 - (optionally) require Async::Interrupt 1.02.
233 - replace opendns fallbacks by verizon/level3 public dns servers.
234
2355.112 Fri Aug 21 13:59:04 CEST 2009
236 - AnyEvent::Handle could cause hard crashes in Net::SSLeay when
237 using starttls with an illegal/uncreatable tls context.
238 - port to broken win32 perls.
239
2405.111 Sun Aug 16 18:44:36 CEST 2009
241 - perl errornously requires sockaddr_un structures returned by the
242 kernel to have a specific length. work around this bug by padding
243 sockaddr_un structures (found to fail on OS X, as perl only uses
244 the correct code for GNU/Linux, but likely to fail on other
245 OSes as well).
246
2475.11 Wed Aug 12 17:49:37 CEST 2009
248 - fixed a bug in run_cmd with input from filename.
249 - create a json coder object when none is given in write direction
250 to reduce dependency on JSON.
251
2525.1 Tue Aug 11 03:17:32 CEST 2009
253 - tcp_connect now reports the correct error in $!, not an unrelated
254 one (was broken since 4.91).
255 - AE::Handle did not properly clear rtimeout/wtimeout watchers
256 when retrying a tcp_connect.
257 - new functions: AnyEvent::Util::run_cmd and close_all_fds_except.
258 - fixed wrong documentation in AE::Socket.
259
2605.01 Mon Aug 10 03:16:32 CEST 2009
261 - last release broke EV child watchers :/.
262
2635.0 Sun Aug 9 17:32:09 CEST 2009
264 - officially introduce and document the AE API.
265 - lowering the timeout at runtime did not have immediate effect
266 in AnyEvent::Handle.
267 - AE::Handle now has separate and independent read and write timeout
268 settings.
269 - probe for Guard module when AnyEvent::Util is loaded, not at
270 runtime (Event doesn't like eval "use" from callbacks, and it avoids
271 exporting the probe function).
272 - Impl::IOAsync's io watchers did not use the filehandle emulation
273 code to work around it's watcher limits.
274 - work around IO::Async's problems with overloaded objects.
275 - try to disable wearnings inside IO::Async, it's just too buggy.
276 - do no longer use Event by default, only EV or Perl, Event is too buggy.
277 - hackishly provide fast AE:: interface when EV or Perl is the backend.
278 - inherit VERSION from AnyEvent::VERSION in most modules.
279
2804.91 Thu Aug 6 15:42:45 CEST 2009
281 - AE::Handle::starttls could get out of sync when the read buffer
282 already contains some TLS handshake.
283 - AE::Handle did not properly free the TLS session, causing bigger
284 memory leaks in Net::SSLeay than expected (reported by toaster).
285 - AE::Socket::tcp_connect will now artificially delay invoking
286 the callback to avoid returning after invoking the callback.
287 - convert many internal watcher uses to AE API.
288
2894.9 Sat Aug 1 11:07:01 CEST 2009
290 - Glib silently fails when registering a timeout with
291 a negative value, so avoid that.
292 - call condvar callback immediately when it is set
293 after the condition is already signalled.
294 - check rbuf_max condition only after trying to consume
295 data.
296 - normalise signal numbers to names when using the ae's
297 signal handling, but do not document this (yet).
298 - pure perl signal emulation did not properly set nonblocking
299 mode on the signal pipe on win32 (this is unlikely to have caused
300 any issues).
301 - new module: AnyEvent::Debug.
302 - AnyEvent::Strict now checks that a signal was specified
303 by name (not name or number).
304 - reduce memory footprint in typical cases by ~50kb by
305 only compiling signal code when necessary.
306 - add AnyEvent::Handle->rbuf_max.
307 - grab the AE namespace for future new API, implement stubs for
308 said future/alternative API.
309 - new function: AnyEvent::Socket::format_hostport.
310
3114.881 Tue Jul 28 12:51:53 CEST 2009
312 - work around a bug in local in pre-5.10 perls, causing
313 AnyEvent::Handle to recurse when it should not
314 (analysed by elmex).
315
3164.88 Tue Jul 28 04:04:37 CEST 2009
317 - re-bless the handle into a dummy package after calling
318 AnyEvent::Handle::destroy, so the user does not need to check
319 for errors after every push_write etc.
320 - do not attempt to run t/02_signals.t on obviously broken
321 platforms.
322
3234.87 Sun Jul 26 02:06:16 CEST 2009
324 - do not attempt to linger when there is no longer a valid fh.
325 - enforce tls mode to be either accept or connect, do not simply
326 segfault in Net::SSLeay.
327 - AnyEvent::Handle can now call tcp_connect itself (new parameters
328 connect, on_prepare, on_connect and on_connect_error). Updated
329 tutorial accordingly.
330 - add AnyEvent::Impl::Irssi backend.
331
3324.86 Mon Jul 20 23:52:29 CEST 2009
333 - since the verbose warning is not enough, explicitly document
334 that versions before 1.33 of Net::SSLeay are not secure.
335 - work around signal handling races in Event and (...) Event::Lib.
336 - try to align signal-race timer to full-second boundaries.
337 - work around Tk not liking negative timeouts.
338 - don't complain of different grades of environmental unfriendlyness
339 in IO::Async.
340
3414.85 Sat Jul 18 06:16:14 CEST 2009
342 - nail the signal race problem in perl once and for all
343 (see $AnyEvent::MAX_SIGNAL_LATENCY).
344 - take advantage of Async::Interrupt if it is available.
345 - load Time::HiRes and Guard modules on demand only.
346 - add optional/recommended modules section to AnyEvent
347 documentation.
348 - reduce memory usage considerably (and reduce startup penalty)
349 by not using "strict", "warnings" and "overload" modules.
350 - work around buggy windows/openbsd perls and provide EBADMSG
351 and EPROTO ourselves when missing.
352 - improve perl 5.6 compatibility of the core event loop.
353 - made Net::SSLeay version 1.33 a soft requirement.
354
3554.83 Fri Jul 17 16:56:26 CEST 2009
356 - implement AnyEvent::Socket::getprotobyname.
357 - AnyEvent::CondVar's will now detect recursive blocking
358 waits and will croak, as too many people fall into
359 this trap.
360 - AnyEvent::Handle will now call ->destroy on itself after
361 executing the on_error callback, instead of doing some
362 half-baked internal shutdown, for fatal errors.
363 - clarify on_eof behaviour w.r.t. the read queue and
364 on_read callbacks.
365 - ignore some possible spurious wake-ups in tcp_connect.
366
3674.82 Sat Jul 11 00:34:55 CEST 2009
368 - POE and Event backends didn't accept some callable objects as
369 callbacks.
370 - use Config module instead of POSIX module to detect signal names
371 in AnyEvent::Strict and AnyEvent::Impl::EventLib, as the POSIX
372 module doesn't even have all POSIX signals :/.
373 - use more workarounds around the many refcnt/corruption bugs in
374 Event::Lib.
375 - work around a race condition in perl's select, causing t/03_child.t
376 to rarely fail.
377
3784.81 Thu Jul 9 10:30:30 CEST 2009
379 - AnyEvent::Handle didn't properly diagnose write errors
380 (it expected -1 from syswrite, how lame... :).
381 - support file descriptors in addition to file handles
382 in AnyEvent->io.
383 - new env variables: PERL_ANYEVENT_RESOLV_CONF,
384 PERL_ANYEVENT_MAX_OUTSTANDING_DNS, PERL_ANYEVENT_CA_FILE
385 and PERL_ANYEVENT_CA_PATH.
386 - provide a sensible synopsis section for AnyEvent::TLS.
387 - add a "supported backends" section to the manpage.
388 - added simple io watcher test to testsuite, using a
389 portable_socketpair.
390 - tried to improve the stability of the Event::Lib backend,
391 YMMV.
392
3934.8 Mon Jul 6 23:45:16 CEST 2009
394 - AnyEvent::DNS did not properly follow CNAME records with
395 uppercase targets.
396 - AnyEvent::DNS would errornously return AAAA records
397 with v4 mapped addresses (a faulty record) as ipv4 addresses,
398 causing AnyEvent::Socket to throw an exception.
399 - added new module AnyEvent::TLS for easier SSL/TLS context
400 creation, with many options including hostname verification,
401 secure default configuration, lots of documentation and,
402 predefined diffie-hellman keys for perfect forward security
403 and much more. get it while it's still fresh!
404 - use AnyEvent::TLS in AnyEvent::Handle for context management.
405 - load AnyEvent::Handle only on demand in AnyEvent::DNS,
406 so AnyEvent::Socket users have smaller memory footprint
407 in the common case.
408 - add AnyEvent::Handle->push_shutdown method.
409 - add an additional error message parameter to AnyEvent::Handle's
410 on_error callback (for TLS, $! is still available).
411 - add AnyEvent::Handle on_starttls/on_stoptls callbacks.
412 - make AnyEvent::Handle more robust against early conenction
413 failures (during new), and return C<undef> in that case
414 from the constructor.
415 - AnyEvent::Handle will now try to load only JSON::XS first,
416 then fall back to JSON.
417 - format_ipv4/format_ipv6 are now exported by default, for symmetry,
418 and because it was documented that way.
419
4204.451 Fri Jul 3 00:28:58 CEST 2009
421 - do not clear rbuf when shutting down an AnyEvent::Handle
422 object - doing so breaks AnyEvent::HTTP.
423
4244.45 Mon Jun 29 22:59:26 CEST 2009
425 - a write error could cause AnyEvent::Handle to create
426 an I/O watcher with an undefined $fh.
427 - special-case mapped ipv4 addresses in both
428 AnyEvent::Socket::format_address and parse_address,
429 to treat them just like ipv4 addresses.
430 - updated and overhauled the AnyEvent::Intro doc.
431 - implement AnyEvent::Socket::format_ipv4/ipv6.
432 - slightly speed up portable_pipe/socketpair.
433 - expand condvar begin/end documentation.
434
4354.42 Fri Jun 26 08:32:18 CEST 2009
436 - preliminary and neccesarily incomplete support for IO::Async.
437 - reset SIGCHLD to DEFAULT when AnyEvent is loaded, in
438 case it was set to IGNORE, to ensure we can catch child statuses
439 even when the calling env acts stupidly.
440 - updated benchmarks with IO::Async, which performs very well.
441
4424.412 Wed Jun 24 01:35:57 CEST 2009
443 - support an "untaint" attribute for AnyEvent::DNS
444 and set it on the default resolver.
445 - implement and document AnyEvent::Impl::Perl::loop.
446 - remove all anyevent-env variables from %ENV
447 when running in tainted mode.
448 - mention and extend the IO::Lambda benchmark.
449
4504.411 Sun Jun 7 18:48:02 CEST 2009
451 - do not try to use F_SETFD on windows.
452
4534.41 Thu May 14 06:40:11 CEST 2009
454 - work around issues in older perls (5.8.0?) when
455 a signal handler is deleted from the %SIG hash.
456 - use POSIX::_exit in child test, to avoid running
457 destructors.
458 - speed up CHLD handling by relying on SIGCHLD being
459 synchronously delivered, even when we roll our own
460 implementation.
461 - AnyEvent::DNS: add the "dname" resource record name
462 for cosmetic reasons.
463
4644.4 Sun Apr 26 20:12:33 CEST 2009
465 - implemented idle watchers, where applicable.
466 - AnyEvent->time died when Event backend was in use.
467 - fix a memleak in the Tk backend.
468 - sped up Tk timer handling.
469 - clip negative "after" values to 0 in AnyEvent::Impl::Event
470 to avoid spamming out warning messages.
471 - fix Qt timers without interval.
472 - avoid Qt zero-timeout special case, in old libqt's.
473
4744.352 Mon Apr 20 16:31:11 CEST 2009
475 - fix AnyEvent::Strict error messages for child watchers.
476 - fix/update Makefile.PL "recommends" versions.
477 - add AnyEvent->now_update.
478
4794.351 Sat Apr 11 07:56:14 CEST 2009
480 - actually make the signal pipe work under win32.
481 - localise $! in signal handler to avoid changing $!, although
482 perl itself does not seem to save/restore errno either.
483 - set the cloexec flag on the signal pipe (normally set by perl too).
484
4854.35 Fri Mar 27 11:48:20 CET 2009
486 - event models relying on AnyEvent's signal watcher emulation did
487 invoke the callback asynchronously, contradicting documentation
488 and causing signals to get lost (this includes AnyEvent's own
489 event loop). AnyEvent now uses the standard pipe trick to make
490 callback execution synchronous to the event loop.
491 - AnyEvent::Handle didn't free TLS context data on DESTROY
492 (patch by Pavel Shaydo).
493 - work around the ever-incompatibly-changing API of MakeMaker.
494 - document that changing global variables without restoring them
495 is a bad idea in Perl (noted by Adam Rosenstein).
496 - AnyEvent::Strict now barfs if ->io is passed a file.
497
4984.34 Thu Feb 12 18:32:45 CET 2009
499 - separately buffer TLS read data, as otherwise the read queue could
500 deadlock as receiving data is not expected while draining the
501 read queue (which cna only happen with TLS).
502 - raise EBADMSG error on JSON decoding errors.
503 - fix some minor manpage bugs (reported by Maximilian Gaß).
504 - speed up select bitmask parsing quite a bit in the pure perl
505 backend.
506 - use CORE::select instead of just select in AnyEvent::Impl::Perl.
507
5084.331 Tue Jan 6 21:07:25 CET 2009
509 - socketpair fails on many vista machines because vista has
510 completely broken accept/getpeername and getsockname functions,
511 so we provide our own socketpair emulation that kind of works
512 (AnyEvent::Util::portable_pipe).
513 - new function: AnyEvent::Util::portable_socketpair.
514 - take advantage of the Guard module if it exists.
515
5164.33 Fri Nov 21 02:35:40 CET 2008
517 - AnyEvent::Strict did errornously flag a fileno of 0 as
518 illegal.
519 - reduce memory usage and slightly speed up the pure perl backend
520 by only storing the file descriptor, not the file handle.
521 - add missing autocork method to AnyEvent::Handle (reported
522 by Adam Rosenstein).
523 - AnyEvent::DNS->resolve errornously documented an $rcode result
524 argument, but there isn't (spotted by Henrik Krohns).
525 - the naked truth about Tk - it's basically unmaintained.
526
5274.32 Mon Nov 3 22:46:32 CET 2008
528 - fix AnyEvent::Socket::resolve_sockaddr to properly support
529 unix sockets again and choose a proper default for
530 the socket type.
531 - fix call to resolve_sockaddr in tcp_connect.
532
5334.31 Thu Oct 30 04:41:48 CET 2008
3 - implemented AnyEvent::Handle->destroy method. 534 - implemented AnyEvent::Handle->destroy method.
4 - hint about unexpected effects in tls mode. 535 - hint about unexpected effects in TLS mode.
5 - speed up AnyEvent::Impl::Perl by using more arrays 536 - speed up AnyEvent::Impl::Perl by using more arrays
6 instead of hashes without the slightest loss of readability :) 537 instead of hashes without the slightest loss of readability :)
538 - work around a perl argument refcounting bug.
539 - with some perl patches applied, AnyEvent now seems leak-free.
7 540
84.3 Fri Oct 3 09:18:43 CEST 2008 5414.3 Fri Oct 3 09:18:43 CEST 2008
9 - AnyEvent will now install a no-op signal handler for SIGPIPE, 542 - AnyEvent will now install a no-op signal handler for SIGPIPE,
10 unless one has been installed already. 543 unless one has been installed already.
11 - warn about not loaidng AnyEvent::Impl::POE early enough 544 - warn about not loaidng AnyEvent::Impl::POE early enough
12 (patch by Adam Rosenstein). 545 (patch by Adam Rosenstein).
13 - fixed a great number of bugs and corner cases in AnyEvent::Handle: 546 - fixed a great number of bugs and corner cases in AnyEvent::Handle.
14 - fix a bug in where in SSL connect mode, the 547 - fix a bug in where in SSL connect mode, the
15 client would first wait for some data by the server and 548 client would first wait for some data by the server and
16 otherwise hang. (reported and analyzed in an absolutely 549 otherwise hang. (reported and analysed in an absolutely
17 exemplary manner by Adam Rosenstein). 550 exemplary manner by Adam Rosenstein).
18 - fix a bug in where SSL EOF would not be treated 551 - fix a bug in where SSL EOF would not be treated
19 as stream EOF, putting the connection into a hung state. 552 as stream EOF, putting the connection into a hung state.
20 (reported and analyzed in an absolutely exemplary manner by Adam 553 (reported and analysed in an absolutely exemplary manner by Adam
21 Rosenstein). 554 Rosenstein).
22 - fix a potential segfault when the TLS context 555 - fix a potential segfault when the TLS context
23 would go missing in a read callback (e.g. due to stoptls) - 556 would go missing in a read callback (e.g. due to stoptls) -
24 Net::SSLeay of course makes no type checking whatsoever. 557 Net::SSLeay of course makes no type checking whatsoever.
25 - AnyEvent::Handle will not stop the read watcher in TLS mode. 558 - AnyEvent::Handle will not stop the read watcher in TLS mode.

Diff Legend

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