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

Comparing AnyEvent/Changes (file contents):
Revision 1.74 by elmex, Thu May 15 09:03:43 2008 UTC vs.
Revision 1.503 by root, Thu Sep 1 22:38:11 2011 UTC

1Revision history for Perl extension AnyEvent. 1Revision history for Perl extension AnyEvent.
2 2
3TODO: document TCP_*** constants
4
5TODO: ae::stritc, regulalry check for invalid fds?
6TOOD: stop internals abusing
7TODO: add google ipv6 public dns
8 - INCOMPATIBLE CHANGE: the default log level is now "3" (critical
9 and above), and some emssages inside AnyEvent have bene elevated
10 to critical, so.
11 - give guidelines on how to chose a logging level.
12 - increase signal latency test timer considerably for poor little osx.
13 - mention some debugging aids in the tutorial.
14
156.02 Fri Aug 26 20:08:31 CEST 2011
16 - log_to_syslog now accepts facility strings.
17 - log_to_syslog would not allow facilities - "$facility|$level"
18 *literally* means that string. sheesh.
19 - the CPAN parser got confused and thought we provide Sys::Syslog, try
20 to work around.
21 - renamed AnyEvent::Impl::FLTK2 to FLTK.
22
236.01 Fri Aug 26 07:04:11 CEST 2011
24 - INCOMPATIBLE CHANGE: AnyEvent::DNS resource records now include
25 the TTL as fourth element - this affects ->resolve, ->request
26 and ->dns_unpack indirectly. Convenience functions such as
27 AnyEvent::DNS::{a,txt,mx,srv} are not affected, neither is
28 AnyEvent::Socket.
29 - INCOMPATIBLE CHANGE: backend authors now should not implement
30 one_event or loop, but instead the AnyEvent::CondVar::_wait and _poll
31 methods.
32 - INCOMPATIBLE CHANGE: verbosity levels have been redefined, see
33 AnyEvent (PERL_ANYEVENT_VERBOSE) and AnyEvent::Log for details.
34 - DEPRECIATION: please change
35 "use AnyEvent::Impl::Perl" to "use AnyEvent::Loop" and
36 "AnyEvent::Impl::Perl::loop" to "AnyEvent::Loop::run"
37 in your programs, for future compatibility.
38 - *blush* idle watcher emulation was borked.
39 - AnyEvent::Strict now checks the AE::xxx functions as well.
40 - querying too long domain names would barf deep inside AnyEvent::DNS - since
41 this is hard to avoid for applications, AnyEvent:.DNS::request now checks for
42 for too long domain names explicitly and passes an undef to the callback
43 instead (reported by James Bromberger).
44 - make AnyEvent::Util more robust against EINTR (analyzed by Andrew Suffield).
45 - new functions: AnyEvent::postpone, AnyEvent::log.
46 - new module: AnyEvent::Log, for simple logging needs.
47 - new env variable, PERL_ANYEVENT_LOG, for fine-grained logging config.
48 - new env variable, PERL_ANYEVENT_DEBUG_SHELL, to automatically start
49 a debug shell.
50 - new env variable, PERL_ANYEVENT_DEBUG_WRAP, to automatically enable
51 watcher instrumentation/wrapping (see AnyEvent::Debug::wrap).
52 - new env variable, PERL_ANYEVENT_HOSTS, to override the place where
53 F</etc/hosts> can be found.
54 - an empty PERL_ANYEVENT_RESOLV_CONF now selects the default
55 configuration.
56 - env variables can now be specified using AE_xyz instead of
57 PERL_ANYEVENT_xyz, subject to some rules.
58 - lots of new AnyEvent::Debug functionality.
59 - AnyEvent::Strict now checks for modifications $_ by the callback
60 (by setting it to readonly while executing the callback, if possible).
61 - IO::Async sometimes cancels the wrong timer when you call cancel_timer.
62 Invest extra overhead to work around this peculiar behaviour.
63 - split the pure perl backend into a loop module and an impl
64 module (AnyEvent::Loop and AnyEvent::Impl::Perl).
65 - some read types (regex, netstring, json, storable) could
66 errornously keep an AE::Handle object alive on their own
67 (reported by Mohammad Toossi).
68 - AnyEvent::Socket::resolve_sockaddr and all functions using it now
69 supports /etc/hosts overrides.
70 - add more workarounds around child watcher bugs in IO::Async
71 and (newly introduced) in POE.
72 - use glib child watchers and try to work around its limitations,
73 if possible (based on a patch by Kevin Ryde).
74 - add support for FLTK (fltk2), via AnyEvent::Impl::FLTK2.
75 - many read types in AnyEvent::Handle would malfunction during
76 transitions to TLS because of a "delete $self->{rbuf}".
77 (reported by Gerald Galster).
78 - improve condvar blocking wait performance (probably only noticable
79 with faster backends), also make it easier for event loops that
80 do not support blocking waits.
81 - PERL_ANYEVENT_MODEL now also allows full module names, although
82 the usefulness of this feature is not yet clear.
83 - add tests for all supported event loops, conditional on PERL_ANYEVENT_LOOP_TESTS.
84 - add t/80_ssltext.t.
85 - no longer use AUTOLOAD for the initial loading (for no very good reason).
86 - tcp_server now tries to unlink unix domain sockets when it gets destroyed.
87 - do not run event loop detection multiple times when programs call their cached
88 copies of AnyEvent::detect.
89 - parse_hostport now accepts absolute paths als unix domain sockets.
90 - suppress the idiotic warning inside POE by patching the POE::Kernel
91 run flags instead of asking the user to jump through contortions.
92 - add Task::Weaken as dependency (as recommends only at this time).
93 - replace PF_UNSPEC by 0, for good old minix.
94
956.0 (not officially released, but was on CPAN for about 15 minutes :/)
96
975.34 Thu May 12 09:59:41 CEST 2011
98 - AE::TLS didn't support both cert_file and key at the same time,
99 due to some copy&paste bug (reported and analyzed by vti).
100 - improve AE::Handle timeout documentation slightly.
101
1025.33 Mon Feb 7 21:51:41 CET 2011
103 - never appeared on CPAN, for unknown reasons.
104 - AnyEvent::Handle on_connect_error - do not try to destroy a handle
105 object when it's already gone (analyzed by Ryan Bullock).
106 - update idna algorithm(s) to UTS#46 version 6.0.0.
107 - preserve trailing dot in idn_to_ascii - testsuite says to preserve,
108 uts46 says it's an error, and testsuite makes more sense.
109 - remove leading dots in idn_to_ascii - testsuite says to remove,
110 uts46 says it's an error, and testsuite makes more sense.
111 - fixed a bug in idn ignored character processing.
112 - passes the uts46 6.0.0 testsuite except for three cases which
113 apparently are bugs in the testsuite itself.
114 - passes the uts46 6.0.1 testsuite except for eleven cases which
115 apparently are bugs in the testsuite itself.
116
1175.31 Sun Jan 23 11:49:19 CET 2011
118 - work around a POE bug where POE would first poll for new events and
119 invoke handlers BEFORE handling the message queue.
120 - fix parsing of "domain" statement in resolv.conf (Johannes Plunien).
121 - stop_read now does just that in TLS mode. see its
122 documentation for caveats.
123 - update to common::sense 3.4 warning set (some warnings in perl impose
124 an arbitrary 32kb limit on regex matches, which affects
125 AnyEvent::Handle).
126 - AE::Handle had a broken rbuf_max method and missing wbuf_max method.
127 - new experimental AE::Handle::resettls method.
128
1295.3 Fri Dec 31 05:48:13 CET 2010
130 - major IOAsync patch by Paul Evans - 0.33 is required, as the major
131 issues with IO::Async have been solved in that version, including
132 the availabilty of a default loop!
133 - implement new wbuf_max parameter for AnyEvent::Handle.
134 - added Cocoa::EventLoop backend by Daisuke Murase.
135 - the default testsuite now honors PERL_ANYEVENT_MODEL (Paul Evans).
136 - clarify handle on_prepare documentation.
137 - set CLOEXEC flag on parent-side fd's in AnyEvent::Util::run_cmd,
138 to avoid leaking them to any child processes.
139 - try to force files into the arch-dependend lib directory, for the
140 half-"turly obsessive compulsive person"(s) that depend on it.
141
1425.29 Sun Dec 5 10:49:21 CET 2010
143 - convert EV backend to EV 4.00 API (so better upgrade EV too).
144 - AE::Handle: implement dynamic read_size adjustment, add
145 max_read_size parameter, reduce default read_size to 2048.
146 - add some FAQ entries for servers and/or on_eof vs. on_error.
147 - work around OS bugs (cygwin again) when lingering in AE::Handle.
148 - reduce memory usage slightly when connect was used in AE::Handle.
149 - wrok around more segfaults in Tk.
150 - document tls_ctx => undef as valid.
151 - detect WNOHANG value at compiletime.
152 - include some probably linux-only support for building AnyEvent
153 as part of the perl core.
154 - improve compatibility of t/08_idna.t to perl 5.8.
155 - make AnyEvent::Strict truly optional (in case somebody builds a
156 minimal perl...).
157
1585.28 Wed Oct 13 04:14:23 CEST 2010
159 - due to a glitch, AnyEvent's internal getprotobyname wasn't
160 actually used in tcp_server.
161 - implement AnyEvent::CondVar->new method, which was documented but
162 missing.
163 - support autoloading of read types in unshift_read.
164 - AnyEvent::Strict: do the test for invalid fh arguments differently,
165 resulting in more informative error messages.
166 - new AnyEvent::FAQ manpage.
167 - clarify that you can actually call push_read etc. in on_read.
168 - put keys and certs in /tmp, not into the current directory,
169 and document this in the SECURITY CONSIDERATIONS sections
170 of AnyEvent::TLS.
171 - support ";" as resolv.conf comment character in AnyEvent::DNS, as
172 requested by Ron Isaacson.
173 - document $AnyEvent::DNS::RESOLVER variable.
174 - incorporated major typo patches by Abhijit Menon-Sen.
175 - AnyEvent::Handle now croaks when negative timeouts are passed.
176 - add a shitload of TCP_xxx sockopt name constants to AnyEvent::Util,
177 undocumented, but who knows when they might come in handy.
178
1795.271 Tue Jun 8 12:05:46 CEST 2010
180 - backport to perl 5.8.x.
181
1825.27 Sun Jun 6 12:12:05 CEST 2010
183 - postpone differently in AnyEvent::Socket now, as
184 when not, canceling the connection attempt might fail
185 (found by Felix Antonius Wilhelm Ostmann).
186 - explicitly check for non-stream sockets in AE::Handle, too many
187 clueless people fell into the trap of this somehow working.
188 - simplified and reworked the "OTHER MODULES" section.
189 - better/more condvar examples.
190
1915.261 Wed Apr 28 16:13:36 CEST 2010
192 - AF_INET6 was not properly used from Socket6 during configuration
193 time (found by Andy Grundman).
194
1955.26 Mon Apr 12 04:49:35 CEST 2010
196 - don't generate spurious readyness notifications when select
197 returns EINTR in the pure perl backend (analysed and testcase
198 by Bas Denissen).
199 - give same examples in SYNOPSIS sections of AnyEvent and AE.
200 - provider faster implementations for AE::time/now/now_update for
201 EV and Perl backends.
202 - provide faster AE::cv and AE::time implementations for all
203 backends.
204 - no longer support lower/mixed-case signal names (this was
205 never documented nor universally supported).
206 - some more memory µ-optimisations, and somewhat less messy
207 code to cope with both AnyEvent and AE APIs.
208
2095.251 Sat Mar 13 00:58:21 CET 2010
210 - make test failed if it was run before make install on perl <=
211 5.8.8, and older versions of perl make our live very difficult,
212 so write constants.pl during Makefile.PL time.
213
2145.25 Sat Mar 13 00:23:14 CET 2010
215 - fix a race condition in AnyEvent::Handle that would cause
216 a "bio_read: unsupported method" error (in _another_
217 ssl connection) after user code threw an exception.
218 - added AnyEvent::Handle->destroyed method.
219 - speed up AnyEvent::Socket::format_address and ::format_ipv6.
220 - the AnyEvent::Util::fh_nonblocking function worked only
221 by lucky accident on win32.
222 - smaller and faster AnyEvent::Util::fh_nonblocking.
223 - when the (required!) Time::HiRes module is missing, AnyEvent
224 did not fall back to built-in time properly.
225 - do not load Fcntl at runtime, saving memory and loading time.
226 - precompile a number of constants and use them instead of runtime
227 detection and eval.
228 - free detection code after detection and similar memory optimisations.
229 - Perl backend timer interval best effort drift has been improved
230 (same algorithm as EV).
231 - update unicode idna mapping table.
232
2335.24 Tue Jan 5 11:39:43 CET 2010
234 - cygwin never reports errors from failed connects. 1.5 just gives
235 you continous readyness and EAGAIN, 1.7 is even more broken and
236 just hangs. work around both issues in a cygwin-specific hack.
237 - improve idle watcher documentation slightly.
238
2395.23 Sun Dec 20 23:48:00 CET 2009
240 - support IDNs in resolve_sockaddr, and therefore in tcp_connect.
241 - implement punycode_encode/decode, idn_nameprep,
242 idn_to_ascii and idn_to_unicode operations in AnyEvent::Util.
243 - provide $AE::VERSION.
244 - removed traces of "no strict 'refs'".
245
2465.22 Sat Dec 5 03:51:13 CET 2009
247 - downgrade-or-fail in AnyEvent::Handle::push_write, to
248 diagnose encoding failures earlier and more succinctly.
249 (this works around bugs in perl, throwing away encoding info
250 when passing scalar data to extensions).
251 - add more examples to AnyEvent::Socket manpage.
252 - upgrade internal warning set to the same as common::sense 2.03.
253 - use pack "n/a*" for pre-5.8.9 perl compatibility in AnyEvent::DNS
254 (John Beppu).
255 - AnyEvent::Socket::inet_aton now properly supports ipv6, as documented.
256 - add google public dns servers to fallback server set and make sure
257 we load-balance properly between the three sets. also add all
258 fallback dns servers, not just a random one, to each dns config.
259
2605.21 Thu Nov 19 02:48:47 CET 2009
261 - fix a problem where socket constants were called with parameters
262 (spotted by David Friedland).
263 - fork_call never use'd POSIX (reported by Daisuke Maki).
264 - improve perl 5.6 compatibility further (but it still won't work
265 unless you rip out everything but the core).
266 - prefer Net::DNS::Resolver over ipconfig, if installed, on win32.
267 uses 10MB of RAM, but doesn't open a console window. *sigh*.
268
2695.202 Wed Oct 14 22:35:44 CEST 2009
270 - AnyEvent::DNS would unexpectedly clobber $_ under windows
271 (analysed by Matthias Waldorf).
272 - AnyEvent::Handle::run_cmd can now pass the PID of the
273 newly-created process, which is much less useful than it might
274 sound (based on patch by Yann Kerherve).
275
2765.201 Tue Sep 29 12:09:25 CEST 2009
277 - AnyEvent:Handle::on_starttls/on_stoptls methods were broken
278 (reported by Torsten Foertsch).
279 - common::sense 2.0 could cause tcp_server to throw an exception
280 (analysed by elmex).
281
2825.2 Mon Sep 14 07:04:49 CEST 2009
283 - INCOMPATIBLE CHANGE: do no longer support register_read_type
284 and register_write_type in AnyEvent::Handle, instead support
285 package names (the facility was mostly abused).
286 - implement "packagename-as-read/write type" support in
287 AnyEvent::Handle.
288 - AnyEvent::Handle: new options "keepalive" and "oobinline".
289 - oobinline set by default to avoid security issues.
290 - the pure-perl event loop backend wrongly detected times() fallback
291 support (spotted by Pavel Boldin).
292
2935.12 Tue Sep 1 20:26:50 CEST 2009
294 - be more lenient when parsing resolv.conf files, as some people
295 use hashmarks as comment indicator inside directives (reported
296 by Michael S. Fischer).
297 - use same set of warnings as common::sense 2.0.
298 - fix a potential 32 bit overflow issue due to perl having problems
299 with large hex constants in the Perl backend.
300 - do not use the slower 5.6 version fo the enc_name code in AnyEvent::DNS
301 by default.
302 - fix some prototypes.
303 - (optionally) require Async::Interrupt 1.02.
304 - replace opendns fallbacks by verizon/level3 public dns servers.
305
3065.112 Fri Aug 21 13:59:04 CEST 2009
307 - AnyEvent::Handle could cause hard crashes in Net::SSLeay when
308 using starttls with an illegal/uncreatable tls context.
309 - port to broken win32 perls.
310
3115.111 Sun Aug 16 18:44:36 CEST 2009
312 - perl errornously requires sockaddr_un structures returned by the
313 kernel to have a specific length. work around this bug by padding
314 sockaddr_un structures (found to fail on OS X, as perl only uses
315 the correct code for GNU/Linux, but likely to fail on other
316 OSes as well).
317
3185.11 Wed Aug 12 17:49:37 CEST 2009
319 - fixed a bug in run_cmd with input from filename.
320 - create a json coder object when none is given in write direction
321 to reduce dependency on JSON.
322
3235.1 Tue Aug 11 03:17:32 CEST 2009
324 - tcp_connect now reports the correct error in $!, not an unrelated
325 one (was broken since 4.91).
326 - AE::Handle did not properly clear rtimeout/wtimeout watchers
327 when retrying a tcp_connect.
328 - new functions: AnyEvent::Util::run_cmd and close_all_fds_except.
329 - fixed wrong documentation in AE::Socket.
330
3315.01 Mon Aug 10 03:16:32 CEST 2009
332 - last release broke EV child watchers :/.
333
3345.0 Sun Aug 9 17:32:09 CEST 2009
335 - officially introduce and document the AE API.
336 - lowering the timeout at runtime did not have immediate effect
337 in AnyEvent::Handle.
338 - AE::Handle now has separate and independent read and write timeout
339 settings.
340 - probe for Guard module when AnyEvent::Util is loaded, not at
341 runtime (Event doesn't like eval "use" from callbacks, and it avoids
342 exporting the probe function).
343 - Impl::IOAsync's io watchers did not use the filehandle emulation
344 code to work around it's watcher limits.
345 - work around IO::Async's problems with overloaded objects.
346 - try to disable wearnings inside IO::Async, it's just too buggy.
347 - do no longer use Event by default, only EV or Perl, Event is too buggy.
348 - hackishly provide fast AE:: interface when EV or Perl is the backend.
349 - inherit VERSION from AnyEvent::VERSION in most modules.
350
3514.91 Thu Aug 6 15:42:45 CEST 2009
352 - AE::Handle::starttls could get out of sync when the read buffer
353 already contains some TLS handshake.
354 - AE::Handle did not properly free the TLS session, causing bigger
355 memory leaks in Net::SSLeay than expected (reported by toaster).
356 - AE::Socket::tcp_connect will now artificially delay invoking
357 the callback to avoid returning after invoking the callback.
358 - convert many internal watcher uses to AE API.
359
3604.9 Sat Aug 1 11:07:01 CEST 2009
361 - Glib silently fails when registering a timeout with
362 a negative value, so avoid that.
363 - call condvar callback immediately when it is set
364 after the condition is already signalled.
365 - check rbuf_max condition only after trying to consume
366 data.
367 - normalise signal numbers to names when using the ae's
368 signal handling, but do not document this (yet).
369 - pure perl signal emulation did not properly set nonblocking
370 mode on the signal pipe on win32 (this is unlikely to have caused
371 any issues).
372 - new module: AnyEvent::Debug.
373 - AnyEvent::Strict now checks that a signal was specified
374 by name (not name or number).
375 - reduce memory footprint in typical cases by ~50kb by
376 only compiling signal code when necessary.
377 - add AnyEvent::Handle->rbuf_max.
378 - grab the AE namespace for future new API, implement stubs for
379 said future/alternative API.
380 - new function: AnyEvent::Socket::format_hostport.
381
3824.881 Tue Jul 28 12:51:53 CEST 2009
383 - work around a bug in local in pre-5.10 perls, causing
384 AnyEvent::Handle to recurse when it should not
385 (analysed by elmex).
386
3874.88 Tue Jul 28 04:04:37 CEST 2009
388 - re-bless the handle into a dummy package after calling
389 AnyEvent::Handle::destroy, so the user does not need to check
390 for errors after every push_write etc.
391 - do not attempt to run t/02_signals.t on obviously broken
392 platforms.
393
3944.87 Sun Jul 26 02:06:16 CEST 2009
395 - do not attempt to linger when there is no longer a valid fh.
396 - enforce tls mode to be either accept or connect, do not simply
397 segfault in Net::SSLeay.
398 - AnyEvent::Handle can now call tcp_connect itself (new parameters
399 connect, on_prepare, on_connect and on_connect_error). Updated
400 tutorial accordingly.
401 - add AnyEvent::Impl::Irssi backend.
402
4034.86 Mon Jul 20 23:52:29 CEST 2009
404 - since the verbose warning is not enough, explicitly document
405 that versions before 1.33 of Net::SSLeay are not secure.
406 - work around signal handling races in Event and (...) Event::Lib.
407 - try to align signal-race timer to full-second boundaries.
408 - work around Tk not liking negative timeouts.
409 - don't complain of different grades of environmental unfriendlyness
410 in IO::Async.
411
4124.85 Sat Jul 18 06:16:14 CEST 2009
413 - nail the signal race problem in perl once and for all
414 (see $AnyEvent::MAX_SIGNAL_LATENCY).
415 - take advantage of Async::Interrupt if it is available.
416 - load Time::HiRes and Guard modules on demand only.
417 - add optional/recommended modules section to AnyEvent
418 documentation.
419 - reduce memory usage considerably (and reduce startup penalty)
420 by not using "strict", "warnings" and "overload" modules.
421 - work around buggy windows/openbsd perls and provide EBADMSG
422 and EPROTO ourselves when missing.
423 - improve perl 5.6 compatibility of the core event loop.
424 - made Net::SSLeay version 1.33 a soft requirement.
425
4264.83 Fri Jul 17 16:56:26 CEST 2009
427 - implement AnyEvent::Socket::getprotobyname.
428 - AnyEvent::CondVar's will now detect recursive blocking
429 waits and will croak, as too many people fall into
430 this trap.
431 - AnyEvent::Handle will now call ->destroy on itself after
432 executing the on_error callback, instead of doing some
433 half-baked internal shutdown, for fatal errors.
434 - clarify on_eof behaviour w.r.t. the read queue and
435 on_read callbacks.
436 - ignore some possible spurious wake-ups in tcp_connect.
437
4384.82 Sat Jul 11 00:34:55 CEST 2009
439 - POE and Event backends didn't accept some callable objects as
440 callbacks.
441 - use Config module instead of POSIX module to detect signal names
442 in AnyEvent::Strict and AnyEvent::Impl::EventLib, as the POSIX
443 module doesn't even have all POSIX signals :/.
444 - use more workarounds around the many refcnt/corruption bugs in
445 Event::Lib.
446 - work around a race condition in perl's select, causing t/03_child.t
447 to rarely fail.
448
4494.81 Thu Jul 9 10:30:30 CEST 2009
450 - AnyEvent::Handle didn't properly diagnose write errors
451 (it expected -1 from syswrite, how lame... :).
452 - support file descriptors in addition to file handles
453 in AnyEvent->io.
454 - new env variables: PERL_ANYEVENT_RESOLV_CONF,
455 PERL_ANYEVENT_MAX_OUTSTANDING_DNS, PERL_ANYEVENT_CA_FILE
456 and PERL_ANYEVENT_CA_PATH.
457 - provide a sensible synopsis section for AnyEvent::TLS.
458 - add a "supported backends" section to the manpage.
459 - added simple io watcher test to testsuite, using a
460 portable_socketpair.
461 - tried to improve the stability of the Event::Lib backend,
462 YMMV.
463
4644.8 Mon Jul 6 23:45:16 CEST 2009
465 - AnyEvent::DNS did not properly follow CNAME records with
466 uppercase targets.
467 - AnyEvent::DNS would errornously return AAAA records
468 with v4 mapped addresses (a faulty record) as ipv4 addresses,
469 causing AnyEvent::Socket to throw an exception.
470 - added new module AnyEvent::TLS for easier SSL/TLS context
471 creation, with many options including hostname verification,
472 secure default configuration, lots of documentation and,
473 predefined diffie-hellman keys for perfect forward security
474 and much more. get it while it's still fresh!
475 - use AnyEvent::TLS in AnyEvent::Handle for context management.
476 - load AnyEvent::Handle only on demand in AnyEvent::DNS,
477 so AnyEvent::Socket users have smaller memory footprint
478 in the common case.
479 - add AnyEvent::Handle->push_shutdown method.
480 - add an additional error message parameter to AnyEvent::Handle's
481 on_error callback (for TLS, $! is still available).
482 - add AnyEvent::Handle on_starttls/on_stoptls callbacks.
483 - make AnyEvent::Handle more robust against early conenction
484 failures (during new), and return C<undef> in that case
485 from the constructor.
486 - AnyEvent::Handle will now try to load only JSON::XS first,
487 then fall back to JSON.
488 - format_ipv4/format_ipv6 are now exported by default, for symmetry,
489 and because it was documented that way.
490
4914.451 Fri Jul 3 00:28:58 CEST 2009
492 - do not clear rbuf when shutting down an AnyEvent::Handle
493 object - doing so breaks AnyEvent::HTTP.
494
4954.45 Mon Jun 29 22:59:26 CEST 2009
496 - a write error could cause AnyEvent::Handle to create
497 an I/O watcher with an undefined $fh.
498 - special-case mapped ipv4 addresses in both
499 AnyEvent::Socket::format_address and parse_address,
500 to treat them just like ipv4 addresses.
501 - updated and overhauled the AnyEvent::Intro doc.
502 - implement AnyEvent::Socket::format_ipv4/ipv6.
503 - slightly speed up portable_pipe/socketpair.
504 - expand condvar begin/end documentation.
505
5064.42 Fri Jun 26 08:32:18 CEST 2009
507 - preliminary and neccesarily incomplete support for IO::Async.
508 - reset SIGCHLD to DEFAULT when AnyEvent is loaded, in
509 case it was set to IGNORE, to ensure we can catch child statuses
510 even when the calling env acts stupidly.
511 - updated benchmarks with IO::Async, which performs very well.
512
5134.412 Wed Jun 24 01:35:57 CEST 2009
514 - support an "untaint" attribute for AnyEvent::DNS
515 and set it on the default resolver.
516 - implement and document AnyEvent::Impl::Perl::loop.
517 - remove all anyevent-env variables from %ENV
518 when running in tainted mode.
519 - mention and extend the IO::Lambda benchmark.
520
5214.411 Sun Jun 7 18:48:02 CEST 2009
522 - do not try to use F_SETFD on windows.
523
5244.41 Thu May 14 06:40:11 CEST 2009
525 - work around issues in older perls (5.8.0?) when
526 a signal handler is deleted from the %SIG hash.
527 - use POSIX::_exit in child test, to avoid running
528 destructors.
529 - speed up CHLD handling by relying on SIGCHLD being
530 synchronously delivered, even when we roll our own
531 implementation.
532 - AnyEvent::DNS: add the "dname" resource record name
533 for cosmetic reasons.
534
5354.4 Sun Apr 26 20:12:33 CEST 2009
536 - implemented idle watchers, where applicable.
537 - AnyEvent->time died when Event backend was in use.
538 - fix a memleak in the Tk backend.
539 - sped up Tk timer handling.
540 - clip negative "after" values to 0 in AnyEvent::Impl::Event
541 to avoid spamming out warning messages.
542 - fix Qt timers without interval.
543 - avoid Qt zero-timeout special case, in old libqt's.
544
5454.352 Mon Apr 20 16:31:11 CEST 2009
546 - fix AnyEvent::Strict error messages for child watchers.
547 - fix/update Makefile.PL "recommends" versions.
548 - add AnyEvent->now_update.
549
5504.351 Sat Apr 11 07:56:14 CEST 2009
551 - actually make the signal pipe work under win32.
552 - localise $! in signal handler to avoid changing $!, although
553 perl itself does not seem to save/restore errno either.
554 - set the cloexec flag on the signal pipe (normally set by perl too).
555
5564.35 Fri Mar 27 11:48:20 CET 2009
557 - event models relying on AnyEvent's signal watcher emulation did
558 invoke the callback asynchronously, contradicting documentation
559 and causing signals to get lost (this includes AnyEvent's own
560 event loop). AnyEvent now uses the standard pipe trick to make
561 callback execution synchronous to the event loop.
562 - AnyEvent::Handle didn't free TLS context data on DESTROY
563 (patch by Pavel Shaydo).
564 - work around the ever-incompatibly-changing API of MakeMaker.
565 - document that changing global variables without restoring them
566 is a bad idea in Perl (noted by Adam Rosenstein).
567 - AnyEvent::Strict now barfs if ->io is passed a file.
568
5694.34 Thu Feb 12 18:32:45 CET 2009
570 - separately buffer TLS read data, as otherwise the read queue could
571 deadlock as receiving data is not expected while draining the
572 read queue (which cna only happen with TLS).
573 - raise EBADMSG error on JSON decoding errors.
574 - fix some minor manpage bugs (reported by Maximilian Gaß).
575 - speed up select bitmask parsing quite a bit in the pure perl
576 backend.
577 - use CORE::select instead of just select in AnyEvent::Impl::Perl.
578
5794.331 Tue Jan 6 21:07:25 CET 2009
580 - socketpair fails on many vista machines because vista has
581 completely broken accept/getpeername and getsockname functions,
582 so we provide our own socketpair emulation that kind of works
583 (AnyEvent::Util::portable_pipe).
584 - new function: AnyEvent::Util::portable_socketpair.
585 - take advantage of the Guard module if it exists.
586
5874.33 Fri Nov 21 02:35:40 CET 2008
588 - AnyEvent::Strict did errornously flag a fileno of 0 as
589 illegal.
590 - reduce memory usage and slightly speed up the pure perl backend
591 by only storing the file descriptor, not the file handle.
592 - add missing autocork method to AnyEvent::Handle (reported
593 by Adam Rosenstein).
594 - AnyEvent::DNS->resolve errornously documented an $rcode result
595 argument, but there isn't (spotted by Henrik Krohns).
596 - the naked truth about Tk - it's basically unmaintained.
597
5984.32 Mon Nov 3 22:46:32 CET 2008
599 - fix AnyEvent::Socket::resolve_sockaddr to properly support
600 unix sockets again and choose a proper default for
601 the socket type.
602 - fix call to resolve_sockaddr in tcp_connect.
603
6044.31 Thu Oct 30 04:41:48 CET 2008
605 - implemented AnyEvent::Handle->destroy method.
606 - hint about unexpected effects in TLS mode.
607 - speed up AnyEvent::Impl::Perl by using more arrays
608 instead of hashes without the slightest loss of readability :)
609 - work around a perl argument refcounting bug.
610 - with some perl patches applied, AnyEvent now seems leak-free.
611
6124.3 Fri Oct 3 09:18:43 CEST 2008
613 - AnyEvent will now install a no-op signal handler for SIGPIPE,
614 unless one has been installed already.
615 - warn about not loaidng AnyEvent::Impl::POE early enough
616 (patch by Adam Rosenstein).
617 - fixed a great number of bugs and corner cases in AnyEvent::Handle.
618 - fix a bug in where in SSL connect mode, the
619 client would first wait for some data by the server and
620 otherwise hang. (reported and analysed in an absolutely
621 exemplary manner by Adam Rosenstein).
622 - fix a bug in where SSL EOF would not be treated
623 as stream EOF, putting the connection into a hung state.
624 (reported and analysed in an absolutely exemplary manner by Adam
625 Rosenstein).
626 - fix a potential segfault when the TLS context
627 would go missing in a read callback (e.g. due to stoptls) -
628 Net::SSLeay of course makes no type checking whatsoever.
629 - AnyEvent::Handle will not stop the read watcher in TLS mode.
630 - AnyEvent::Handle->stoptls will send a TLS close notify
631 instead of simply destroying the stream now.
632 - fix a bug where large blocks of data written in TLS mode
633 would not be sent unless triggered by receives.
634 - on_drain will now take the tls write buffer into account.
635 - SSL operations have been streamlined a bit and should be faster
636 now, more to come, though.
637 - added a FAQ section, will document nontrivial issues in there.
638
6394.234 Mon Sep 29 04:08:13 CEST 2008
640 - fix child watcher documentation: only child exits will be reported,
641 no trace events.
642 - mention SIGPIPE in AnyEvent::Handle.
643 - perl backend will now use POSIX::times as monotonic clock
644 when available and the monotonic clock option is not.
645 - run even without Time::HiRes in some cases now, but do not
646 provide sub-second accuracy (all *supported* perl versions have
647 Time::HiRes).
648
6494.233 Fri Aug 22 01:48:05 CEST 2008
650 - fix a bug in the testsuite, causing a hang.
651 - clarified lots of AnyEvent::Handle constructor arguments.
652
6534.232 Thu Aug 21 20:44:25 CEST 2008
654 - fix a bug in AnyEvent::Handle that could cause two fatal error
655 callbacks on EOF, which caused AnyEvent::HTTP to signal success twice
656 (testcase by Leon Brocard).
657 - clarified on_eof/on_error documentation a bit.
658
6594.231 Tue Jul 29 13:12:15 CEST 2008
660 - remove some debugging code left in AnyEvent::Util::fork_call
661 (and no, it's impossible to implement with the broken windows
662 perls without resource leaks or worse).
663
6644.23 Tue Jul 29 12:19:59 CEST 2008
665 - document the first parameter passed to condvar callbacks to be
666 the callback.
667 - add AnyEvent::Socket::{ntoa,aton} aliases.
668 - optimize the AE::Handle->push_read (line) for the default
669 eol marker.
670 - optimize push_read (packstring|storable) for small packets.
671 - invoke on_error callback when no on_eof callback is set.
672 - fix a bug in push_read (storable) of unknown impact.
673
6744.22 Sun Jul 20 16:34:13 CEST 2008
675 - new function AnyEvent::Socket::parse_hostport.
676 - as the bulkheads at microsoft can't even get getprotobyname reliably
677 working on their shitty fucking broken os we need to hardcode
678 some common protocol numbers in AnyEvent::Socket. How can
679 people even bother with such a pile of shit as windows.
680
6814.21 Thu Jul 17 14:40:05 CEST 2008
682 - INCOMPATIBLE CHANGE: AnyEvent::Util::fork_call now has a prototype
683 which is incompatible to previous invocation syntax.
684 - work around more windows perl bugs in fork_call. I cannot imagine
685 why anybody would bother implementing fork in such an obviously
686 unusable way.
687 - avoid calling $do_retry if the request has already finished in
688 AnyEvent::DNS, thus avoiding an exception (reported by Anatoly
689 K. Sharifulin).
690
6914.2 Sat Jul 12 22:42:11 CEST 2008
692 - support an "interval" argument for repeating timers.
693 - fix a bug in the pure perl event loop that caused timers to
694 be delayed under some circumstances.
695 - implement optional argument checking via PERL_ANYEVENT_STRICT.
696 - not importing anything from Event causes it to malfunction
697 and not use Time::HiRes, work around this bug.
698 - more bug workarounds against the endlessly broken Tk module,
699 seems to work now on 32 bit machines, but less so on 64 bit.
700 - minor optimisations applied to most backends.
701
7024.161 Fri Jul 4 14:24:48 CEST 2008
703 - fixed the t/handle/02_write.t, it read after write, with the
704 assumption that all data could be written before read
705 (reported by Jim Razmus).
706
7074.160 Thu Jul 3 04:02:21 CEST 2008
708 - re-registering signal watchers after unregistering one did
709 not work if a backend falls back on AnyEvent's default
710 implementation (also affected child watchers).
711 - new AnyEvent::Handle options: autocork and no_delay.
712
7134.152 Sun Jun 22 14:15:44 CEST 2008
714 - allow for 32-bit perls that implement shifts differently
715 on different architectures in parse_ipv4 (reported and
716 analysed by Keiichi DAIBA).
717
7184.151 Fri Jun 6 17:34:24 CEST 2008
719 - make sure specifying _only_ on_read and never pushing reads
720 works.
721
7224.15 Fri Jun 6 13:00:46 CEST 2008
723 - the pure perl backend would keep some watchers alive when more than
724 one watcher was registered for the same fd.
725 - new "packstring" and "storable" read and write types
726 for AnyEvent::Handle.
727 - allow on_eof handler to be called after on_error with EPIPE returns.
728 - do not immediately call on_read callback in handle constructor.
729
7304.14 Thu Jun 5 20:29:31 CEST 2008
731 - Fixed a bug in DNS SRV priority sorting.
732 - AnyEvent::Util::guard now reports runtime errors while
733 executing the guard block as warnings.
734 - handle 0-byte-reads just before EOF correctly in AnyEvent::Handle.
735
7364.13 Thu Jun 5 00:47:59 CEST 2008
737 - AnyEvent::DNS only followed cname chains with length 2,
738 contrary to documentation. bumped it up to 10, thanks to
739 microsoft, the current broken-dns-config-king.
740 - AnyEvent::DNS didn't check for socket return status
741 but instead relied on perl not creating filehandles in that
742 case - too bad it gives you a bogus file handle (reported
743 and analysed by Vladimir Timofeev).
744 - fix queue management logic in AnyEvent::Handle:
745 when on_read was registered, the queue was empty
746 and no progress could be made AnyEvent::Handle would
747 enter an endless loop.
748 - correctly start reading again when the handle became
749 busy again after idling.
750 - correctly treat tls shut-downs as EOF condition
751 (for the time being, would be nice to have a callback for that).
752 - correctly call eof callback under all conditions
753 (hopefully).
754 - the timeout callback did not expect that $self can go
755 away any time.
756
7574.12 Tue Jun 3 10:58:04 CEST 2008
758 - include AnyEvent::Intro, a tutorial for anyevent,
759 anyevent::socket and anyevent::handle.
760 - allow more options in on_error.
761
7624.11 Fri May 30 23:42:25 CEST 2008
763 - INCOMPATIBLE CHANGE: replace ptr by real PTR lookup, provide
764 reverse_lookup and reverse_verify to replace it, support
765 v4mapped and v4compat addresses.
766 - provide more documentation for the resolver class.
767 - really replace longest run of :0: by :: in format_address,
768 also properly convert :: and ::1 again.
769 - support NAPTR record name and decode it.
770 - implement random weight sampling for SRV records, as per
771 rfc 2782.
772 - correctly abort on srv-record targets of ".".
773 - added AnyEvent::DNS::wait_for_slot.
774 - in the unlikely event of a virtual circuit connection
775 being invalidated by a delayed udp reply, AnyEvent::DNS
776 could die. this has been fixed.
777 - plug a probable memleak in the DNS vc code.
778 - use configured timeout also for VC request phase.
779 - implement timeout and max_outstanding methods
780 for resolver class.
781 - update version numbers in all modules.
782
7834.1 Thu May 29 05:45:40 CEST 2008
784 - INCOMPATIBLE CHANGE: renamed xxx_ip functions to xxx_address
785 which mirrors their purpose better (old names still available
786 for a while). Also moved AnyEvent::DNS::addr to
787 AnyEvent::Socket::resolve_sockaddr.
788 - implement AnyEvent->time and AnyEvent->now.
789 - fix IPv6 support in pack_sockaddr.
790 - officially un-experimentalise all newly introduced code.
791 - support unix domain sockets everywhere by specifying
792 a host of "unix/" and the pathname as service.
793 - implement an activity timeout in AnyEvent::Handle.
794 - added a regex read type to AnyEvent::Handle.
795 - added a json read/write type to AnyEvent::Handle.
796 - always croak in on_error in AnyEvent::Handle.
797 - document how to subclass AnyEvent::Handle.
798 - implement AnyEvent::Util::fork_call.
799 - add support for IPv6 nameservers and nameserver statements.
800 - work around _yet_ _another_ windows perl bug in where empty
801 select masks cause errors in the pure perl backend (this
802 workaround cannot be implemented for the other event loops,
803 but EV also works around this).
804 - supply AnyEvent::Socket::address_type and make good use of it.
805 - clarify and fix the Handle documentation w.r.t $self vs. $handle
806 vs. callback arguments.
807 - add some recommends to the META.yml.
808
8094.05 Mon May 26 19:44:06 CEST 2008
810 - some platforms ignore the rfc and prepend an extra sa_len member to
811 the sockaddr structure, cater for those. also use sockaddr_family,
812 didn't know it was there, but comes in handy for the workaround.
813 - undo the import WIN32 hack.
814
8154.04 Mon May 26 08:03:31 CEST 2008
816 - try to work around yet another windows bug: failed connects are reported
817 as if it were out-of-band data. windows users: you suck. Only
818 EV, Glib, Event and the pure perl backend can handle this condition.
819 - optimize write algorithm in AnyEvent::Handle.
820 - properly parse PERL_ANYEVENT_PROTOCOLS and use the info correctly.
821 - tcp_connect now properly iterates through all targets.
822 - check wether IPv6 sockets can actually be created, otherwise, disable ipv6
823 support entirely. also hardcode some AF_INET6 constants for the
824 sake of perl 5.8.
825 - vastly improved nameserver/suffix detection algorithm for windows
826 (still a hack...)
827 - try to use a fallback dns server if no dns servers could be found.
828 - splatter around some local $SIG{__DIE__}'s to work around bad
829 die handlers.
830
8314.03 Sun May 25 05:05:57 CEST 2008
832 - work around perl bugs on windows, where perl returns undocumented
833 error codes for sysread, syswrite etc.
834 - fix AnyEvent::Handle::unshift_read to really unshift and not push.
835 - AnyEvent::Handle could sometimes stumble over 0-byte writes.
836 - fix unpack_sockaddr, now works for AF_INET6 sockaddr's.
837 - updated documentation.
838 - lots of workarounds for perl 5.8 quirks.
839 - implement a more flexible type system for both reads and writes
840 in AnyEvent::Handle.
841 - support IPv6 in tcp_server.
842 - pass local hostname and port to the prepare callback in tcp_server.
843 - make types extendable in AnyEvent::Handle.
844 - croak, not die, in AnyEvent::Handle;
845
8464.0 Sat May 24 19:58:08 CEST 2008
847 - added (experimental) AnyEvent::DNS module.
848 - added (experimental) AnyEvent::Socket module.
849 - added (experiemntal) support for TLS/SSL in AnyEvent::Handle.
850 - make condvar->end call ->send as documented.
851 - make condvar constructor support the documented cb => argument.
852 - overload condvars so you can use them instead of code references.
853 - support more of the POSIX forms instead of a pure dotted quad for
854 verbatim ip addresses.
855 - AnyEvent::Socket::inet_aton will use AnyEvent::DNS now and not fork anymore.
856 - try very hard not to freeze on broken windows perls.
857 - make on_eof optional in AnyEvent::Handle.
858 - support service names in tcp_connect and tcp_server.
859 - work around netbsd/openbsd bug causing unix sockets not to
860 become writable until empty.
861 - only make a dummy read for error detection when getpeername
862 returns ENOTCONN.
863 - defend AnyEvent::Util::inet_aton against Coro::LWP's brutal
864 override.
865 - new AnyEvent::Guard::cancel.
866
8673.5 Sat May 17 23:17:14 CEST 2008
868 - gracefully deal with systems not defining CLOCK_MONOTONIC
869 (as opposed to not supporting the clock itself, which worked in 3.41).
870 - added AnyEvent::Util::tcp_connect and AnyEvent::Util::tcp_server
871 helpers, to replace the ill-designed AnyEvent::Socket class.
872 - added some examples (eg/connect, eg/handle).
873 - added AnyEvent::Util::guard.
874 - AnyEvent::Util now exports most of its functions.
3 - fixed a bug in *_read_chunk, where the callback was called with $self as 875 - fixed a bug in *_read_chunk, where the callback was called with $self as
4 first and second argument. 876 first and second argument.
5 - added AnyEvent::Util::connect and some examples (eg/connect, eg/handle). 877 - simplified and fixed regex handling in AnyEvent::Handle::*_read_line.
6 - fix manpage condvars examples to use new syntax. 878 - fix manpage condvars examples to use new syntax.
879 - considerably improve inet_aton with EV::ADNS backend to support
880 cname chains (common dns configuration bug, e.g. with www.google.de)
881 and aaaa records if no a record could be found.
7 882
83.41 Sun May 11 19:53:13 CEST 2008 8833.41 Sun May 11 19:53:13 CEST 2008
9 - work around a bug in perl 5.8.8's local. Reported by Yi Ma Mao. 884 - work around a bug in perl 5.8.8's local. Reported by Yi Ma Mao.
10 885
113.4 Sun May 11 00:29:25 CEST 2008 8863.4 Sun May 11 00:29:25 CEST 2008
12 - complete re-implementation of condvars with a hopefully much mroe useful 887 - complete re-implementation of condvars with a hopefully much more useful
13 API (backward compatible functions exist but are undocumented). 888 API (backward compatible functions exist but are undocumented).
14 - AnyEvent::Handle fully rewritten with a hopefully more powerful API 889 - AnyEvent::Handle fully rewritten with a hopefully more powerful API
15 (still subject to change). 890 (still subject to change).
16 - take advantage of CLOCK_MONOTONIC, if available, in AnyEvent::Impl::Perl. 891 - take advantage of CLOCK_MONOTONIC, if available, in AnyEvent::Impl::Perl.
17 - provide AnyEvent::post_detect and @AnyEvent::post_detect, which 892 - provide AnyEvent::post_detect and @AnyEvent::post_detect, which

Diff Legend

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