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

Comparing AnyEvent/Changes (file contents):
Revision 1.308 by root, Wed Jul 29 13:10:58 2009 UTC vs.
Revision 1.612 by root, Thu Jun 16 13:03:21 2016 UTC

1Revision history for Perl extension AnyEvent. 1Revision history for Perl extension AnyEvent.
2 2
3TODO: release a:i 1.0 3TODO: document TCP_*** constants
4TODO: ae::strict should chekc for signal names (enforced) 4TODO: add debug/trace logging to some modules
5TODO: use the eval-with-cb-call trick also to speed up JSON?
6TODO: FAQ: common pitfalls? invoke-before-return
7TODO: FAQ: scope-ids? hosts file order?
8TODO: AnyEvent::Handle with AnyEvent::IO?
9TODO: invalid. and localhost. specialcasing inside AEDNS and not AESocket (rfc6761)
10TODO: maybe implement env variable to give hosts precedence
11TODO: more formats for parse_ipv6 (single ipv6 address without port, ...p80, ...#80)
12
13TODO: inet_aton maybe pack C4? no, add a comment why not
14TODO: hosts read multiple times for initial concurrent tcp_connect's?
15
16TODO: check that uts46/unicode 9.0.0 is released
17 - Only call tlsext_host_name for non-empty common names (reported
18 by Maxime Soulé).
19 - log a (single) notice message if SNI is not supported.
20 - upgrade to UTS-46:9.0.0 draft and switch to non-transitional
21 behaviour, beating thunderbird, ie, edge, chrome and safari to it :)
22 (see also https://bugzilla.mozilla.org/show_bug.cgi?id=1218179)
23 - turns out the UTS-46 IDNA testcase failures were indeed bugs in the
24 testcases and the specification and not in the code - the post-9.0.0
25 unicode files have all known problems fixed, so finally the AnyEvent
26 IDNA implementation can pass the full IDNA testsuite - without needing
27 a single fix :)
28
297.12 Wed Jan 27 19:12:26 CET 2016
30 - use common name as hostname for tls connects, if Net::SSLeay
31 supports SNI.
32 - fix documentation of tls_autostart read type in AnyEvent::Handle,
33 analyzed by Felix Ostmann.
34
357.11 Thu Jul 16 14:36:00 CEST 2015
36 - AnyEvent::Socket::parse_ipv6 could accept malformed ipv6
37 addresses (extra "::" at end and similar cases).
38 - add a more explicit warning to AnyEvent::Handle that it doesn't
39 work on files, people keep getting confused.
40 - new function AnyEvent::Socket::tcp_bind.
41 - new functions AnyEvent::fh_block and AnyEvent::fh_unblock.
42 - aligned ipv6 address formatting with RFC 5952 (by not shortening
43 a single :0: to ::).
44 - added stability canary support.
45
467.09 Sat May 2 16:38:53 CEST 2015
47 - AnyEvent::Debug called an internal function (AnyEvent::Log::ft)
48 that was renamed to AnyEvent::Log:format_time. uinder its old name
49 (analyzed by Michael Stovenour).
50 - update AnyEvent::DNS fallback resolver addresses:
51 seems google effectively killed most other free dns resolvers,
52 so remove them, but add cable and wireless (ecrc) since it was
53 stable for 20 years or so, official or not, and there should be
54 an alternative to google.
55 - perl5porters broke windows error codes in 5.20, and mapped
56 WSAEWOULDBLOCK on the (different) EWOULDBLOCK error code, and
57 WSAEINPROGRESS into the incompatible ERINPROGRESS code, probably
58 because they were so cool. They probably broke other error codes
59 for no reason, too, but I didn't care for checking, it's too
60 depressing. This version only works around the WSAEWOULDBLOCK
61 issue, because I don't have a nice way to work around the
62 WSAEINPROGRESS bug.
63
647.08 Wed Dec 10 05:27:17 CET 2014
65 - work around a newly introduced bug in Socket 2.011 (an
66 errornous sun_length check) (analyzed by Maxime Soulé).
67 - AnyEvent::TLS didn't load (but refer to) AnyEvent::Socket
68 (analyzed by Ben Magistro).
69 - AnyEvent::Strict will now confess, not croak. This is in line with
70 it being a development/debugging tool.
71 - work around a number of libglib bugs (debug builds of libglib enforce
72 certain undocumented behaviour patterns such as not being able to
73 remove a child watch source after it has fired, which we will try
74 to emulate to avoid "criticals". what where they thinking...).
75 - mention json security issues in AnyEvent::Handle, now that Douglas
76 Crockford has foolishly and incompatibly changed JSON.
77 - changed default dns resolver "max_outstanding" value from 1 to 10,
78 the latter beinfg the intended value all along
79 (reported by Ilya Chesnokov).
80 - added new "AnyEvent::Impl::UV" interface module to the UV event lib
81 (written by Mike Lowell).
82
837.07 Tue Dec 17 17:45:02 CET 2013
84 - the documentation for custom tls verify schemes was wrong. make it agree
85 with the code (reported by Maxime Soulé).
86 - added cbor read and write types to AnyEvent::Handle (using CBOR::XS).
87 - work around an API change in openssl that could cause wrong tls connection
88 aborts, likely on windows only (analyzed by sten).
89 - calling AnyEvent->now_update with AnyEvent::Impl::Perl caused an
90 endless loop (reported by Dietrich Rebmann).
91 - add tlsv1_1 and tlsv1_2 protocols to AnyEvent::TLS
92 (patch by Maxime Soulé).
93 - document AnyEvent::Impl::IOAsync::set_loop and
94 $AnyEvent::Impl::IOAsync::LOOP. Though only documented now, this
95 functionality has _always_ been available.
96 - force a toplevel domain name in t/81_hosts.t (analyzed by
97 David Jack Wange Olrik).
98 - document that AnyEvent::Log uses AnyEvent::IO.
99 - warn about AnyEvent::Filesys::Notify performance.
100 - praise the joys of AnyEvent::Fork::*.
101 - time for an =encoding directive.
102 - do no longer use JSON to create a default json coder, use
103 JSON::XS or JSON::PP directly.
104
1057.05 Wed Aug 21 10:38:08 CEST 2013
106 - uts46data.pl couldn't be found due to wrong naming of the file
107 (reported by Fulko Hew).
108 - handle lone \015's properly in AE::Handle's default line read
109 (reported by various people).
110 - untaint ip addresses found in /etc/hosts (patch by José Micó).
111 - the memleak fix in 7.03 caused resolving via /etc/hosts to always
112 fail on first use (reported and testcase by Andrew Whatson).
113 - expose AnyEvent::Log::format_time, and allow users to redefine it.
114 - expose AnyEvent::Log::default_format, and allow redefinition.
115 - expose AnyEvent::Log::fatal_exit, to allow redefinition.
116 - AnyEvent::Debug shell can now run coro shell commands, if available.
117 - t/63* tests were wrongly in MANIFEST.
118 - kernel.org's finger server went MIA, switch to freebsd.org and
119 icculus.org.
120 - clarify that IO::AIO and AnyEvent::AIO are needed for AnyEvent::IO
121 to function asynchronously (reported by Daniel Carosone).
122 - hard-disable $^W in most tests, it generates too much garbage output.
123 - use a (hopefully) more future-proof method to emulate common::sense.
124 - upgrade to UTS-46:6.2.0.
125 - switch to INSTLIB from INSTLIBDIR, as INSTLIBDIR was wrongly
126 documented. should not affect anything.
127
1287.04 Thu Nov 15 02:23:58 CET 2012
129 - debugging code left in.
130
1317.03 Thu Nov 15 02:16:11 CET 2012
132 - AnyEvent::Socket::inet_aton did not work when dns resolution was
133 used to find the addresses (analyzed and patch by Maxime Soulé).
134 - fix a memory leak in the /etc/hosts lookup code when hosts don't
135 resolve and are not in hosts (analyzed by sten).
136
1377.02 Tue Aug 14 04:27:58 CEST 2012
138 - AnyEvent::Util::run_cmd could block indefinitely (analyzed and test
139 program by Yorhel).
140 - verified that AnyEvent::Socket follows RFC5952.
141 - try to parse "ADDR#PORT" in addition to "ADDR PORT".
142
1437.01 Sun May 13 01:03:17 CEST 2012
144 - fail with EPROTO in AnyEvent::Handle wqhen TLS is requested but
145 not available, instead of throwing an exception.
146 - use File::Spec to get the tmpdir in t/*, to avoid needless
147 failures on (most, not mine :) windows boxes.
148 - new handle read types: tls_detect and tls_autostart.
149
1507.0 Fri Apr 13 06:33:30 CEST 2012
151 - child watchers are broken in POE 1.352 (also many earlier
152 versions) and there seems to be no way to work around it, as POE
153 itself is inherently racy. Document this regression and add a delay
154 in t/68_poe_03_child.t for the time being.
155 - new module AnyEvent::IO, that is a frontend to either a pure-perl
156 synchronous I/O implementation (AnyEvent::IO::Perl), or to the
157 asynchronous I/O extension IO::AIO (AnyEvent::IO::IOAIO).
158 - load /etc/hosts only when DNS has no answer.
159 - stat /etc/hosts on every access and reload it if it changed.
160 - load /etc/hosts via AnyEvent::IO - potentially asynchronous.
161 - fix a buggy croak in the dh parameter reading in AnyEvent::TLS.
162 - AnyEvent::Log log_to_file and log_to_path now use AnyEvent::IO. As
163 a side effect, they now use true appending as opposed to libc
164 appending, and the file might not have been opened when the function
165 returns.
166 - the default logging level was not properly documented in a variety of
167 places, this has hopefully been rectified.
168 - updated uts46data.pl for unicode 6.1.0.
169 - made log messages generated by AnyEvent submodules not
170 include the package name anymore, as it will be logged
171 by default already.
172 - upgrade to the trick used by common-sense 3.5 to work
173 around extra warning torture/breakage under perl 5.15.x.
174 - log messages by anyevent are now ucfirst, are usually full
175 sentences and do no longer include the package name.
176 - the storable read type would sometimes throw an exception instead
177 of causing EBADMSG (analyzed by Maxime Soulé).
178 - work around a bug in openssl 1.0.1 which enforces a minimum rsa
179 keysize (reported by Paul Howarth).
180 - documented Rocco fabricating statements about AnyEvent and me.
181
1826.14 Tue Jan 31 20:00:24 CET 2012
183 - AnyEvent::Impl::Tk was broken due to a mysterious "die" inside,
184 probably an editing mistake (reported by Darin McBride).
185
1866.13 Thu Jan 12 07:27:01 CET 2012
187 - AnyEvent::Util::fork_call checked for POSIX availability in the wrong way,
188 causing it to fail if POSIX wasn't loaded already (analysed by Rock Power).
189 - AnyEvent::Handle::push_read (line => did pass $1 directly, so
190 regex-matching inside the callback would change the parameter, despite
191 $1 being dynamically scoped per-block. this perl bug is now being
192 worked around (testcase by Cindy Wang).
193
1946.12 Mon Dec 12 13:21:10 CET 2011
195 - $! was clobbered by subsequent calls in tcp_connect, due to
196 the postpone not saving/restoring it (analyzed by Richard Garnish).
197
1986.11 Tue Nov 22 10:36:05 CET 2011
199 - Tk cannot create windows when tainted unless you set a title, so
200 set a dummy title for AnyEvent's dummy mainwindow. How dummy.
201 - escape any nonprintable/nonascii characters when stringifying
202 backtraces.
203 - log the reason loading an interface module fails to load at level debug.
204
2056.1 Tue Oct 4 19:44:30 CEST 2011
206 - INCOMPATIBLE CHANGE: the default log level is now "4" (error
207 and above), and some messages inside AnyEvent have been elevated
208 to higher log levels to print by default.
209 - AnyEvent::log, unlike AnyEvent::Log::log, did not always exit on fatal.
210 - work around yet another signal race bug, newly introduced by POE. My
211 how that module sucks.
212 - add google public ipv6 dns as fallback servers, so AE::DNS can finally
213 support ipv6-only operation even with defaults (also use ::1 instead
214 of 127.0.0.1).
215 - give better guidelines on how to chose a logging level.
216 - AnyEvent::Log can now cap (limit) log message priority by context.
217 - AE::log will now save and restore $! and $@, so callers don't have to.
218 - AE::Strict now checks that registered fds are still valid at regular
219 intervals.
220 - skip frequently called 'require' calls if the module appears loaded.
221 - increase signal latency test timer considerably for poor little osx,
222 which can't be bothered with basic posix functionality such as signal
223 delivery.
224 - mention some debugging aids in the tutorial.
225 - new env variable PERL_ANYEVENT_MAX_SIGNAL_LATENCY.
226 - private (%namespace) log contexts in PERL_ANYEVENT_LOG now have the name
227 of the context as title.
228 - AnyEvent::Strict would not restore read-write access to $_ when a callback
229 died - this is a user bug, but since AE::Strict exists to help dbeugging...
230 - detect the extremely evil IO::Async::Loop::AnyEvent module and refuse
231 to work with it - the author is well aware of the issues but refuses
232 to fix them. This extreme step has been taken because the module actively
233 attacks the goals of AnyEvent and makes it harder for module authors to use
234 AnyEvent.
235
2366.02 Fri Aug 26 20:08:31 CEST 2011
237 - log_to_syslog now accepts facility strings.
238 - log_to_syslog would not allow facilities - "$facility|$level"
239 *literally* means that string. sheesh.
240 - the CPAN parser got confused and thought we provide Sys::Syslog, try
241 to work around.
242 - renamed AnyEvent::Impl::FLTK2 to FLTK.
243
2446.01 Fri Aug 26 07:04:11 CEST 2011
245 - INCOMPATIBLE CHANGE: AnyEvent::DNS resource records now include
246 the TTL as fourth element - this affects ->resolve, ->request
247 and ->dns_unpack indirectly. Convenience functions such as
248 AnyEvent::DNS::{a,txt,mx,srv} are not affected, neither is
249 AnyEvent::Socket.
250 - INCOMPATIBLE CHANGE: backend authors now should not implement
251 one_event or loop, but instead the AnyEvent::CondVar::_wait and _poll
252 methods.
253 - INCOMPATIBLE CHANGE: verbosity levels have been redefined, see
254 AnyEvent (PERL_ANYEVENT_VERBOSE) and AnyEvent::Log for details.
255 - DEPRECIATION: please change
256 "use AnyEvent::Impl::Perl" to "use AnyEvent::Loop" and
257 "AnyEvent::Impl::Perl::loop" to "AnyEvent::Loop::run"
258 in your programs, for future compatibility.
259 - *blush* idle watcher emulation was borked.
260 - AnyEvent::Strict now checks the AE::xxx functions as well.
261 - querying too long domain names would barf deep inside AnyEvent::DNS - since
262 this is hard to avoid for applications, AnyEvent:.DNS::request now checks for
263 for too long domain names explicitly and passes an undef to the callback
264 instead (reported by James Bromberger).
265 - make AnyEvent::Util more robust against EINTR (analyzed by Andrew Suffield).
266 - new functions: AnyEvent::postpone, AnyEvent::log.
267 - new module: AnyEvent::Log, for simple logging needs.
268 - new env variable, PERL_ANYEVENT_LOG, for fine-grained logging config.
269 - new env variable, PERL_ANYEVENT_DEBUG_SHELL, to automatically start
270 a debug shell.
271 - new env variable, PERL_ANYEVENT_DEBUG_WRAP, to automatically enable
272 watcher instrumentation/wrapping (see AnyEvent::Debug::wrap).
273 - new env variable, PERL_ANYEVENT_HOSTS, to override the place where
274 /etc/hosts can be found.
275 - an empty PERL_ANYEVENT_RESOLV_CONF now selects the default
276 configuration.
277 - env variables can now be specified using AE_xyz instead of
278 PERL_ANYEVENT_xyz, subject to some rules.
279 - lots of new AnyEvent::Debug functionality.
280 - AnyEvent::Strict now checks for modifications of $_ by the callback
281 (by setting it to readonly while executing the callback, if possible).
282 - IO::Async sometimes cancels the wrong timer when you call cancel_timer.
283 Invest extra overhead to work around this peculiar behaviour.
284 - split the pure perl backend into a loop module and an impl
285 module (AnyEvent::Loop and AnyEvent::Impl::Perl).
286 - some read types (regex, netstring, json, storable) could
287 errornously keep an AE::Handle object alive on their own
288 (reported by Mohammad Toossi).
289 - AnyEvent::Socket::resolve_sockaddr and all functions using it now
290 supports /etc/hosts overrides.
291 - add more workarounds around child watcher bugs in IO::Async
292 and (newly introduced) in POE.
293 - use glib child watchers and try to work around its limitations,
294 if possible (based on a patch by Kevin Ryde).
295 - add support for FLTK (fltk2), via AnyEvent::Impl::FLTK2.
296 - many read types in AnyEvent::Handle would malfunction during
297 transitions to TLS because of a "delete $self->{rbuf}".
298 (reported by Gerald Galster).
299 - improve condvar blocking wait performance (probably only noticable
300 with faster backends), also make it easier for event loops that
301 do not support blocking waits.
302 - PERL_ANYEVENT_MODEL now also allows full module names, although
303 the usefulness of this feature is not yet clear.
304 - add tests for all supported event loops, conditional on PERL_ANYEVENT_LOOP_TESTS.
305 - add t/80_ssltext.t.
306 - no longer use AUTOLOAD for the initial loading (for no very good reason).
307 - tcp_server now tries to unlink unix domain sockets when it gets destroyed.
308 - do not run event loop detection multiple times when programs call their cached
309 copies of AnyEvent::detect.
310 - parse_hostport now accepts absolute paths als unix domain sockets.
311 - suppress the idiotic warning inside POE by patching the POE::Kernel
312 run flags instead of asking the user to jump through contortions.
313 - add Task::Weaken as dependency (as recommends only at this time).
314 - replace PF_UNSPEC by 0, for good old minix.
315
3166.0 (not officially released, but was on CPAN for about 15 minutes :/)
317
3185.34 Thu May 12 09:59:41 CEST 2011
319 - AE::TLS didn't support both cert_file and key at the same time,
320 due to some copy&paste bug (reported and analyzed by vti).
321 - improve AE::Handle timeout documentation slightly.
322
3235.33 Mon Feb 7 21:51:41 CET 2011
324 - never appeared on CPAN, for unknown reasons.
325 - AnyEvent::Handle on_connect_error - do not try to destroy a handle
326 object when it's already gone (analyzed by Ryan Bullock).
327 - update idna algorithm(s) to UTS#46 version 6.0.0.
328 - preserve trailing dot in idn_to_ascii - testsuite says to preserve,
329 uts46 says it's an error, and testsuite makes more sense.
330 - remove leading dots in idn_to_ascii - testsuite says to remove,
331 uts46 says it's an error, and testsuite makes more sense.
332 - fixed a bug in idn ignored character processing.
333 - passes the uts46 6.0.0 testsuite except for three cases which
334 apparently are bugs in the testsuite itself.
335 - passes the uts46 6.0.1 testsuite except for eleven cases which
336 apparently are bugs in the testsuite itself.
337
3385.31 Sun Jan 23 11:49:19 CET 2011
339 - work around a POE bug where POE would first poll for new events and
340 invoke handlers BEFORE handling the message queue.
341 - fix parsing of "domain" statement in resolv.conf (Johannes Plunien).
342 - stop_read now does just that in TLS mode. see its
343 documentation for caveats.
344 - update to common::sense 3.4 warning set (some warnings in perl impose
345 an arbitrary 32kb limit on regex matches, which affects
346 AnyEvent::Handle).
347 - AE::Handle had a broken rbuf_max method and missing wbuf_max method.
348 - new experimental AE::Handle::resettls method.
349
3505.3 Fri Dec 31 05:48:13 CET 2010
351 - major IOAsync patch by Paul Evans - 0.33 is required, as the major
352 issues with IO::Async have been solved in that version, including
353 the availabilty of a default loop!
354 - implement new wbuf_max parameter for AnyEvent::Handle.
355 - added Cocoa::EventLoop backend by Daisuke Murase.
356 - the default testsuite now honors PERL_ANYEVENT_MODEL (Paul Evans).
357 - clarify handle on_prepare documentation.
358 - set CLOEXEC flag on parent-side fd's in AnyEvent::Util::run_cmd,
359 to avoid leaking them to any child processes.
360 - try to force files into the arch-dependend lib directory, for the
361 half-"turly obsessive compulsive person"(s) that depend on it.
362
3635.29 Sun Dec 5 10:49:21 CET 2010
364 - convert EV backend to EV 4.00 API (so better upgrade EV too).
365 - AE::Handle: implement dynamic read_size adjustment, add
366 max_read_size parameter, reduce default read_size to 2048.
367 - add some FAQ entries for servers and/or on_eof vs. on_error.
368 - work around OS bugs (cygwin again) when lingering in AE::Handle.
369 - reduce memory usage slightly when connect was used in AE::Handle.
370 - wrok around more segfaults in Tk.
371 - document tls_ctx => undef as valid.
372 - detect WNOHANG value at compiletime.
373 - include some probably linux-only support for building AnyEvent
374 as part of the perl core.
375 - improve compatibility of t/08_idna.t to perl 5.8.
376 - make AnyEvent::Strict truly optional (in case somebody builds a
377 minimal perl...).
378
3795.28 Wed Oct 13 04:14:23 CEST 2010
380 - due to a glitch, AnyEvent's internal getprotobyname wasn't
381 actually used in tcp_server.
382 - implement AnyEvent::CondVar->new method, which was documented but
383 missing.
384 - support autoloading of read types in unshift_read.
385 - AnyEvent::Strict: do the test for invalid fh arguments differently,
386 resulting in more informative error messages.
387 - new AnyEvent::FAQ manpage.
388 - clarify that you can actually call push_read etc. in on_read.
389 - put keys and certs in /tmp, not into the current directory,
390 and document this in the SECURITY CONSIDERATIONS sections
391 of AnyEvent::TLS.
392 - support ";" as resolv.conf comment character in AnyEvent::DNS, as
393 requested by Ron Isaacson.
394 - document $AnyEvent::DNS::RESOLVER variable.
395 - incorporated major typo patches by Abhijit Menon-Sen.
396 - AnyEvent::Handle now croaks when negative timeouts are passed.
397 - add a shitload of TCP_xxx sockopt name constants to AnyEvent::Util,
398 undocumented, but who knows when they might come in handy.
399
4005.271 Tue Jun 8 12:05:46 CEST 2010
401 - backport to perl 5.8.x.
402
4035.27 Sun Jun 6 12:12:05 CEST 2010
404 - postpone differently in AnyEvent::Socket now, as
405 when not, canceling the connection attempt might fail
406 (found by Felix Antonius Wilhelm Ostmann).
407 - explicitly check for non-stream sockets in AE::Handle, too many
408 clueless people fell into the trap of this somehow working.
409 - simplified and reworked the "OTHER MODULES" section.
410 - better/more condvar examples.
411
4125.261 Wed Apr 28 16:13:36 CEST 2010
413 - AF_INET6 was not properly used from Socket6 during configuration
414 time (found by Andy Grundman).
415
4165.26 Mon Apr 12 04:49:35 CEST 2010
417 - don't generate spurious readyness notifications when select
418 returns EINTR in the pure perl backend (analysed and testcase
419 by Bas Denissen).
420 - give same examples in SYNOPSIS sections of AnyEvent and AE.
421 - provider faster implementations for AE::time/now/now_update for
422 EV and Perl backends.
423 - provide faster AE::cv and AE::time implementations for all
424 backends.
425 - no longer support lower/mixed-case signal names (this was
426 never documented nor universally supported).
427 - some more memory µ-optimisations, and somewhat less messy
428 code to cope with both AnyEvent and AE APIs.
429
4305.251 Sat Mar 13 00:58:21 CET 2010
431 - make test failed if it was run before make install on perl <=
432 5.8.8, and older versions of perl make our live very difficult,
433 so write constants.pl during Makefile.PL time.
434
4355.25 Sat Mar 13 00:23:14 CET 2010
436 - fix a race condition in AnyEvent::Handle that would cause
437 a "bio_read: unsupported method" error (in _another_
438 ssl connection) after user code threw an exception.
439 - added AnyEvent::Handle->destroyed method.
440 - speed up AnyEvent::Socket::format_address and ::format_ipv6.
441 - the AnyEvent::Util::fh_nonblocking function worked only
442 by lucky accident on win32.
443 - smaller and faster AnyEvent::Util::fh_nonblocking.
444 - when the (required!) Time::HiRes module is missing, AnyEvent
445 did not fall back to built-in time properly.
446 - do not load Fcntl at runtime, saving memory and loading time.
447 - precompile a number of constants and use them instead of runtime
448 detection and eval.
449 - free detection code after detection and similar memory optimisations.
450 - Perl backend timer interval best effort drift has been improved
451 (same algorithm as EV).
452 - update unicode idna mapping table.
453
4545.24 Tue Jan 5 11:39:43 CET 2010
455 - cygwin never reports errors from failed connects. 1.5 just gives
456 you continous readyness and EAGAIN, 1.7 is even more broken and
457 just hangs. work around both issues in a cygwin-specific hack.
458 - improve idle watcher documentation slightly.
459
4605.23 Sun Dec 20 23:48:00 CET 2009
461 - support IDNs in resolve_sockaddr, and therefore in tcp_connect.
462 - implement punycode_encode/decode, idn_nameprep,
463 idn_to_ascii and idn_to_unicode operations in AnyEvent::Util.
464 - provide $AE::VERSION.
465 - removed traces of "no strict 'refs'".
466
4675.22 Sat Dec 5 03:51:13 CET 2009
468 - downgrade-or-fail in AnyEvent::Handle::push_write, to
469 diagnose encoding failures earlier and more succinctly.
470 (this works around bugs in perl, throwing away encoding info
471 when passing scalar data to extensions).
472 - add more examples to AnyEvent::Socket manpage.
473 - upgrade internal warning set to the same as common::sense 2.03.
474 - use pack "n/a*" for pre-5.8.9 perl compatibility in AnyEvent::DNS
475 (John Beppu).
476 - AnyEvent::Socket::inet_aton now properly supports ipv6, as documented.
477 - add google public dns servers to fallback server set and make sure
478 we load-balance properly between the three sets. also add all
479 fallback dns servers, not just a random one, to each dns config.
480
4815.21 Thu Nov 19 02:48:47 CET 2009
482 - fix a problem where socket constants were called with parameters
483 (spotted by David Friedland).
484 - fork_call never use'd POSIX (reported by Daisuke Maki).
485 - improve perl 5.6 compatibility further (but it still won't work
486 unless you rip out everything but the core).
487 - prefer Net::DNS::Resolver over ipconfig, if installed, on win32.
488 uses 10MB of RAM, but doesn't open a console window. *sigh*.
489
4905.202 Wed Oct 14 22:35:44 CEST 2009
491 - AnyEvent::DNS would unexpectedly clobber $_ under windows
492 (analysed by Matthias Waldorf).
493 - AnyEvent::Handle::run_cmd can now pass the PID of the
494 newly-created process, which is much less useful than it might
495 sound (based on patch by Yann Kerherve).
496
4975.201 Tue Sep 29 12:09:25 CEST 2009
498 - AnyEvent:Handle::on_starttls/on_stoptls methods were broken
499 (reported by Torsten Foertsch).
500 - common::sense 2.0 could cause tcp_server to throw an exception
501 (analysed by elmex).
502
5035.2 Mon Sep 14 07:04:49 CEST 2009
504 - INCOMPATIBLE CHANGE: do no longer support register_read_type
505 and register_write_type in AnyEvent::Handle, instead support
506 package names (the facility was mostly abused).
507 - implement "packagename-as-read/write type" support in
508 AnyEvent::Handle.
509 - AnyEvent::Handle: new options "keepalive" and "oobinline".
510 - oobinline set by default to avoid security issues.
511 - the pure-perl event loop backend wrongly detected times() fallback
512 support (spotted by Pavel Boldin).
513
5145.12 Tue Sep 1 20:26:50 CEST 2009
515 - be more lenient when parsing resolv.conf files, as some people
516 use hashmarks as comment indicator inside directives (reported
517 by Michael S. Fischer).
518 - use same set of warnings as common::sense 2.0.
519 - fix a potential 32 bit overflow issue due to perl having problems
520 with large hex constants in the Perl backend.
521 - do not use the slower 5.6 version fo the enc_name code in AnyEvent::DNS
522 by default.
523 - fix some prototypes.
524 - (optionally) require Async::Interrupt 1.02.
525 - replace opendns fallbacks by verizon/level3 public dns servers.
526
5275.112 Fri Aug 21 13:59:04 CEST 2009
528 - AnyEvent::Handle could cause hard crashes in Net::SSLeay when
529 using starttls with an illegal/uncreatable tls context.
530 - port to broken win32 perls.
531
5325.111 Sun Aug 16 18:44:36 CEST 2009
533 - perl errornously requires sockaddr_un structures returned by the
534 kernel to have a specific length. work around this bug by padding
535 sockaddr_un structures (found to fail on OS X, as perl only uses
536 the correct code for GNU/Linux, but likely to fail on other
537 OSes as well).
538
5395.11 Wed Aug 12 17:49:37 CEST 2009
540 - fixed a bug in run_cmd with input from filename.
541 - create a json coder object when none is given in write direction
542 to reduce dependency on JSON.
543
5445.1 Tue Aug 11 03:17:32 CEST 2009
545 - tcp_connect now reports the correct error in $!, not an unrelated
546 one (was broken since 4.91).
547 - AE::Handle did not properly clear rtimeout/wtimeout watchers
548 when retrying a tcp_connect.
549 - new functions: AnyEvent::Util::run_cmd and close_all_fds_except.
550 - fixed wrong documentation in AE::Socket.
551
5525.01 Mon Aug 10 03:16:32 CEST 2009
553 - last release broke EV child watchers :/.
554
5555.0 Sun Aug 9 17:32:09 CEST 2009
556 - officially introduce and document the AE API.
557 - lowering the timeout at runtime did not have immediate effect
558 in AnyEvent::Handle.
559 - AE::Handle now has separate and independent read and write timeout
560 settings.
561 - probe for Guard module when AnyEvent::Util is loaded, not at
562 runtime (Event doesn't like eval "use" from callbacks, and it avoids
563 exporting the probe function).
564 - Impl::IOAsync's io watchers did not use the filehandle emulation
565 code to work around it's watcher limits.
566 - work around IO::Async's problems with overloaded objects.
567 - try to disable wearnings inside IO::Async, it's just too buggy.
568 - do no longer use Event by default, only EV or Perl, Event is too buggy.
569 - hackishly provide fast AE:: interface when EV or Perl is the backend.
570 - inherit VERSION from AnyEvent::VERSION in most modules.
571
5724.91 Thu Aug 6 15:42:45 CEST 2009
573 - AE::Handle::starttls could get out of sync when the read buffer
574 already contains some TLS handshake.
575 - AE::Handle did not properly free the TLS session, causing bigger
576 memory leaks in Net::SSLeay than expected (reported by toaster).
577 - AE::Socket::tcp_connect will now artificially delay invoking
578 the callback to avoid returning after invoking the callback.
579 - convert many internal watcher uses to AE API.
580
5814.9 Sat Aug 1 11:07:01 CEST 2009
5 - Glib silently fails when registering a timeout with 582 - Glib silently fails when registering a timeout with
6 a negative value, so avoid that. 583 a negative value, so avoid that.
584 - call condvar callback immediately when it is set
585 after the condition is already signalled.
586 - check rbuf_max condition only after trying to consume
587 data.
7 - normalise signal numbers to names when using the ae's 588 - normalise signal numbers to names when using the ae's
8 signal handling, but do not document this (yet). 589 signal handling, but do not document this (yet).
9 - pure perl signal emulation did not properly set nonblocking 590 - pure perl signal emulation did not properly set nonblocking
10 mode on the signal pipe on win32 (this is unlikely to have caused 591 mode on the signal pipe on win32 (this is unlikely to have caused
11 any issues). 592 any issues).
593 - new module: AnyEvent::Debug.
594 - AnyEvent::Strict now checks that a signal was specified
595 by name (not name or number).
12 - reduce memory footprint int ypical cases by ~50kb by 596 - reduce memory footprint in typical cases by ~50kb by
13 only compiling signal code when necessary. 597 only compiling signal code when necessary.
598 - add AnyEvent::Handle->rbuf_max.
599 - grab the AE namespace for future new API, implement stubs for
600 said future/alternative API.
601 - new function: AnyEvent::Socket::format_hostport.
14 602
154.881 Tue Jul 28 12:51:53 CEST 2009 6034.881 Tue Jul 28 12:51:53 CEST 2009
16 - work around a bug in local in pre-5.10 perls, causing 604 - work around a bug in local in pre-5.10 perls, causing
17 AnyEvent::Handle to recurse when it should not 605 AnyEvent::Handle to recurse when it should not
18 (analyzed by elmex). 606 (analysed by elmex).
19 607
204.88 Tue Jul 28 04:04:37 CEST 2009 6084.88 Tue Jul 28 04:04:37 CEST 2009
21 - re-bless the handle into a dummy package after calling 609 - re-bless the handle into a dummy package after calling
22 AnyEvent::Handle::destroy, so the user does not need to check 610 AnyEvent::Handle::destroy, so the user does not need to check
23 for errors after every push_write etc. 611 for errors after every push_write etc.
… …
111 in the common case. 699 in the common case.
112 - add AnyEvent::Handle->push_shutdown method. 700 - add AnyEvent::Handle->push_shutdown method.
113 - add an additional error message parameter to AnyEvent::Handle's 701 - add an additional error message parameter to AnyEvent::Handle's
114 on_error callback (for TLS, $! is still available). 702 on_error callback (for TLS, $! is still available).
115 - add AnyEvent::Handle on_starttls/on_stoptls callbacks. 703 - add AnyEvent::Handle on_starttls/on_stoptls callbacks.
116 - make AnyEvent::Handle more robust against early conenction 704 - make AnyEvent::Handle more robust against early connection
117 failures (during new), and return C<undef> in that case 705 failures (during new), and return C<undef> in that case
118 from the constructor. 706 from the constructor.
119 - AnyEvent::Handle will now try to load only JSON::XS first, 707 - AnyEvent::Handle will now try to load only JSON::XS first,
120 then fall back to JSON. 708 then fall back to JSON.
121 - format_ipv4/format_ipv6 are now exported by default, for symmetry, 709 - format_ipv4/format_ipv6 are now exported by default, for symmetry,
… …
2454.3 Fri Oct 3 09:18:43 CEST 2008 8334.3 Fri Oct 3 09:18:43 CEST 2008
246 - AnyEvent will now install a no-op signal handler for SIGPIPE, 834 - AnyEvent will now install a no-op signal handler for SIGPIPE,
247 unless one has been installed already. 835 unless one has been installed already.
248 - warn about not loaidng AnyEvent::Impl::POE early enough 836 - warn about not loaidng AnyEvent::Impl::POE early enough
249 (patch by Adam Rosenstein). 837 (patch by Adam Rosenstein).
250 - fixed a great number of bugs and corner cases in AnyEvent::Handle: 838 - fixed a great number of bugs and corner cases in AnyEvent::Handle.
251 - fix a bug in where in SSL connect mode, the 839 - fix a bug in where in SSL connect mode, the
252 client would first wait for some data by the server and 840 client would first wait for some data by the server and
253 otherwise hang. (reported and analyzed in an absolutely 841 otherwise hang. (reported and analysed in an absolutely
254 exemplary manner by Adam Rosenstein). 842 exemplary manner by Adam Rosenstein).
255 - fix a bug in where SSL EOF would not be treated 843 - fix a bug in where SSL EOF would not be treated
256 as stream EOF, putting the connection into a hung state. 844 as stream EOF, putting the connection into a hung state.
257 (reported and analyzed in an absolutely exemplary manner by Adam 845 (reported and analysed in an absolutely exemplary manner by Adam
258 Rosenstein). 846 Rosenstein).
259 - fix a potential segfault when the TLS context 847 - fix a potential segfault when the TLS context
260 would go missing in a read callback (e.g. due to stoptls) - 848 would go missing in a read callback (e.g. due to stoptls) -
261 Net::SSLeay of course makes no type checking whatsoever. 849 Net::SSLeay of course makes no type checking whatsoever.
262 - AnyEvent::Handle will not stop the read watcher in TLS mode. 850 - AnyEvent::Handle will not stop the read watcher in TLS mode.

Diff Legend

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