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

Comparing AnyEvent/Changes (file contents):
Revision 1.252 by root, Mon Jun 29 11:04:09 2009 UTC vs.
Revision 1.618 by root, Sat Nov 26 03:35:23 2016 UTC

1Revision history for Perl extension AnyEvent. 1Revision history for Perl extension AnyEvent.
2 2
3TODO: document TCP_*** constants
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: tcp_bind error message
17TODO: tcp_bind unix domain socket gets deleted instantly without guard
18 - fix a crash bug in AnyEvent::Handle with openssl 1.1.0 (patched
19 by Paul Howarth and apparently tracked down by Mike McCauley).
20 - AnyEvent::Handle->keepalive was documented (and defined) twice
21 (found by Matt Merhar).
22
237.13 Sat Sep 17 04:31:49 CEST 2016
24 - Only call tlsext_host_name for non-empty common names (reported
25 by Maxime Soulé).
26 - log a (single) notice message if SNI is not supported.
27 - upgrade to UTS-46:9.0.0 draft and switch to non-transitional
28 behaviour, beating thunderbird, ie, edge, chrome and safari to it :)
29 (see also https://bugzilla.mozilla.org/show_bug.cgi?id=1218179)
30 - turns out the UTS-46 IDNA testcase failures were indeed bugs in the
31 testcases and the specification and not in the code - the post-9.0.0
32 unicode files have all known problems fixed, so finally the AnyEvent
33 IDNA implementation can pass the full IDNA testsuite - without needing
34 a single fix :)
35 - guarantee (and document) that condvar callbacks will be removed
36 on invocation - important to avoid circular references.
37
387.12 Wed Jan 27 19:12:26 CET 2016
39 - use common name as hostname for tls connects, if Net::SSLeay
40 supports SNI.
41 - fix documentation of tls_autostart read type in AnyEvent::Handle,
42 analyzed by Felix Ostmann.
43
447.11 Thu Jul 16 14:36:00 CEST 2015
45 - AnyEvent::Socket::parse_ipv6 could accept malformed ipv6
46 addresses (extra "::" at end and similar cases).
47 - add a more explicit warning to AnyEvent::Handle that it doesn't
48 work on files, people keep getting confused.
49 - new function AnyEvent::Socket::tcp_bind.
50 - new functions AnyEvent::fh_block and AnyEvent::fh_unblock.
51 - aligned ipv6 address formatting with RFC 5952 (by not shortening
52 a single :0: to ::).
53 - added stability canary support.
54
557.09 Sat May 2 16:38:53 CEST 2015
56 - AnyEvent::Debug called an internal function (AnyEvent::Log::ft)
57 that was renamed to AnyEvent::Log:format_time. uinder its old name
58 (analyzed by Michael Stovenour).
59 - update AnyEvent::DNS fallback resolver addresses:
60 seems google effectively killed most other free dns resolvers,
61 so remove them, but add cable and wireless (ecrc) since it was
62 stable for 20 years or so, official or not, and there should be
63 an alternative to google.
64 - perl5porters broke windows error codes in 5.20, and mapped
65 WSAEWOULDBLOCK on the (different) EWOULDBLOCK error code, and
66 WSAEINPROGRESS into the incompatible ERINPROGRESS code, probably
67 because they were so cool. They probably broke other error codes
68 for no reason, too, but I didn't care for checking, it's too
69 depressing. This version only works around the WSAEWOULDBLOCK
70 issue, because I don't have a nice way to work around the
71 WSAEINPROGRESS bug.
72
737.08 Wed Dec 10 05:27:17 CET 2014
74 - work around a newly introduced bug in Socket 2.011 (an
75 errornous sun_length check) (analyzed by Maxime Soulé).
76 - AnyEvent::TLS didn't load (but refer to) AnyEvent::Socket
77 (analyzed by Ben Magistro).
78 - AnyEvent::Strict will now confess, not croak. This is in line with
79 it being a development/debugging tool.
80 - work around a number of libglib bugs (debug builds of libglib enforce
81 certain undocumented behaviour patterns such as not being able to
82 remove a child watch source after it has fired, which we will try
83 to emulate to avoid "criticals". what where they thinking...).
84 - mention json security issues in AnyEvent::Handle, now that Douglas
85 Crockford has foolishly and incompatibly changed JSON.
86 - changed default dns resolver "max_outstanding" value from 1 to 10,
87 the latter beinfg the intended value all along
88 (reported by Ilya Chesnokov).
89 - added new "AnyEvent::Impl::UV" interface module to the UV event lib
90 (written by Mike Lowell).
91
927.07 Tue Dec 17 17:45:02 CET 2013
93 - the documentation for custom tls verify schemes was wrong. make it agree
94 with the code (reported by Maxime Soulé).
95 - added cbor read and write types to AnyEvent::Handle (using CBOR::XS).
96 - work around an API change in openssl that could cause wrong tls connection
97 aborts, likely on windows only (analyzed by sten).
98 - calling AnyEvent->now_update with AnyEvent::Impl::Perl caused an
99 endless loop (reported by Dietrich Rebmann).
100 - add tlsv1_1 and tlsv1_2 protocols to AnyEvent::TLS
101 (patch by Maxime Soulé).
102 - document AnyEvent::Impl::IOAsync::set_loop and
103 $AnyEvent::Impl::IOAsync::LOOP. Though only documented now, this
104 functionality has _always_ been available.
105 - force a toplevel domain name in t/81_hosts.t (analyzed by
106 David Jack Wange Olrik).
107 - document that AnyEvent::Log uses AnyEvent::IO.
108 - warn about AnyEvent::Filesys::Notify performance.
109 - praise the joys of AnyEvent::Fork::*.
110 - time for an =encoding directive.
111 - do no longer use JSON to create a default json coder, use
112 JSON::XS or JSON::PP directly.
113
1147.05 Wed Aug 21 10:38:08 CEST 2013
115 - uts46data.pl couldn't be found due to wrong naming of the file
116 (reported by Fulko Hew).
117 - handle lone \015's properly in AE::Handle's default line read
118 (reported by various people).
119 - untaint ip addresses found in /etc/hosts (patch by José Micó).
120 - the memleak fix in 7.03 caused resolving via /etc/hosts to always
121 fail on first use (reported and testcase by Andrew Whatson).
122 - expose AnyEvent::Log::format_time, and allow users to redefine it.
123 - expose AnyEvent::Log::default_format, and allow redefinition.
124 - expose AnyEvent::Log::fatal_exit, to allow redefinition.
125 - AnyEvent::Debug shell can now run coro shell commands, if available.
126 - t/63* tests were wrongly in MANIFEST.
127 - kernel.org's finger server went MIA, switch to freebsd.org and
128 icculus.org.
129 - clarify that IO::AIO and AnyEvent::AIO are needed for AnyEvent::IO
130 to function asynchronously (reported by Daniel Carosone).
131 - hard-disable $^W in most tests, it generates too much garbage output.
132 - use a (hopefully) more future-proof method to emulate common::sense.
133 - upgrade to UTS-46:6.2.0.
134 - switch to INSTLIB from INSTLIBDIR, as INSTLIBDIR was wrongly
135 documented. should not affect anything.
136
1377.04 Thu Nov 15 02:23:58 CET 2012
138 - debugging code left in.
139
1407.03 Thu Nov 15 02:16:11 CET 2012
141 - AnyEvent::Socket::inet_aton did not work when dns resolution was
142 used to find the addresses (analyzed and patch by Maxime Soulé).
143 - fix a memory leak in the /etc/hosts lookup code when hosts don't
144 resolve and are not in hosts (analyzed by sten).
145
1467.02 Tue Aug 14 04:27:58 CEST 2012
147 - AnyEvent::Util::run_cmd could block indefinitely (analyzed and test
148 program by Yorhel).
149 - verified that AnyEvent::Socket follows RFC5952.
150 - try to parse "ADDR#PORT" in addition to "ADDR PORT".
151
1527.01 Sun May 13 01:03:17 CEST 2012
153 - fail with EPROTO in AnyEvent::Handle wqhen TLS is requested but
154 not available, instead of throwing an exception.
155 - use File::Spec to get the tmpdir in t/*, to avoid needless
156 failures on (most, not mine :) windows boxes.
157 - new handle read types: tls_detect and tls_autostart.
158
1597.0 Fri Apr 13 06:33:30 CEST 2012
160 - child watchers are broken in POE 1.352 (also many earlier
161 versions) and there seems to be no way to work around it, as POE
162 itself is inherently racy. Document this regression and add a delay
163 in t/68_poe_03_child.t for the time being.
164 - new module AnyEvent::IO, that is a frontend to either a pure-perl
165 synchronous I/O implementation (AnyEvent::IO::Perl), or to the
166 asynchronous I/O extension IO::AIO (AnyEvent::IO::IOAIO).
167 - load /etc/hosts only when DNS has no answer.
168 - stat /etc/hosts on every access and reload it if it changed.
169 - load /etc/hosts via AnyEvent::IO - potentially asynchronous.
170 - fix a buggy croak in the dh parameter reading in AnyEvent::TLS.
171 - AnyEvent::Log log_to_file and log_to_path now use AnyEvent::IO. As
172 a side effect, they now use true appending as opposed to libc
173 appending, and the file might not have been opened when the function
174 returns.
175 - the default logging level was not properly documented in a variety of
176 places, this has hopefully been rectified.
177 - updated uts46data.pl for unicode 6.1.0.
178 - made log messages generated by AnyEvent submodules not
179 include the package name anymore, as it will be logged
180 by default already.
181 - upgrade to the trick used by common-sense 3.5 to work
182 around extra warning torture/breakage under perl 5.15.x.
183 - log messages by anyevent are now ucfirst, are usually full
184 sentences and do no longer include the package name.
185 - the storable read type would sometimes throw an exception instead
186 of causing EBADMSG (analyzed by Maxime Soulé).
187 - work around a bug in openssl 1.0.1 which enforces a minimum rsa
188 keysize (reported by Paul Howarth).
189 - documented Rocco fabricating statements about AnyEvent and me.
190
1916.14 Tue Jan 31 20:00:24 CET 2012
192 - AnyEvent::Impl::Tk was broken due to a mysterious "die" inside,
193 probably an editing mistake (reported by Darin McBride).
194
1956.13 Thu Jan 12 07:27:01 CET 2012
196 - AnyEvent::Util::fork_call checked for POSIX availability in the wrong way,
197 causing it to fail if POSIX wasn't loaded already (analysed by Rock Power).
198 - AnyEvent::Handle::push_read (line => did pass $1 directly, so
199 regex-matching inside the callback would change the parameter, despite
200 $1 being dynamically scoped per-block. this perl bug is now being
201 worked around (testcase by Cindy Wang).
202
2036.12 Mon Dec 12 13:21:10 CET 2011
204 - $! was clobbered by subsequent calls in tcp_connect, due to
205 the postpone not saving/restoring it (analyzed by Richard Garnish).
206
2076.11 Tue Nov 22 10:36:05 CET 2011
208 - Tk cannot create windows when tainted unless you set a title, so
209 set a dummy title for AnyEvent's dummy mainwindow. How dummy.
210 - escape any nonprintable/nonascii characters when stringifying
211 backtraces.
212 - log the reason loading an interface module fails to load at level debug.
213
2146.1 Tue Oct 4 19:44:30 CEST 2011
215 - INCOMPATIBLE CHANGE: the default log level is now "4" (error
216 and above), and some messages inside AnyEvent have been elevated
217 to higher log levels to print by default.
218 - AnyEvent::log, unlike AnyEvent::Log::log, did not always exit on fatal.
219 - work around yet another signal race bug, newly introduced by POE. My
220 how that module sucks.
221 - add google public ipv6 dns as fallback servers, so AE::DNS can finally
222 support ipv6-only operation even with defaults (also use ::1 instead
223 of 127.0.0.1).
224 - give better guidelines on how to chose a logging level.
225 - AnyEvent::Log can now cap (limit) log message priority by context.
226 - AE::log will now save and restore $! and $@, so callers don't have to.
227 - AE::Strict now checks that registered fds are still valid at regular
228 intervals.
229 - skip frequently called 'require' calls if the module appears loaded.
230 - increase signal latency test timer considerably for poor little osx,
231 which can't be bothered with basic posix functionality such as signal
232 delivery.
233 - mention some debugging aids in the tutorial.
234 - new env variable PERL_ANYEVENT_MAX_SIGNAL_LATENCY.
235 - private (%namespace) log contexts in PERL_ANYEVENT_LOG now have the name
236 of the context as title.
237 - AnyEvent::Strict would not restore read-write access to $_ when a callback
238 died - this is a user bug, but since AE::Strict exists to help dbeugging...
239 - detect the extremely evil IO::Async::Loop::AnyEvent module and refuse
240 to work with it - the author is well aware of the issues but refuses
241 to fix them. This extreme step has been taken because the module actively
242 attacks the goals of AnyEvent and makes it harder for module authors to use
243 AnyEvent.
244
2456.02 Fri Aug 26 20:08:31 CEST 2011
246 - log_to_syslog now accepts facility strings.
247 - log_to_syslog would not allow facilities - "$facility|$level"
248 *literally* means that string. sheesh.
249 - the CPAN parser got confused and thought we provide Sys::Syslog, try
250 to work around.
251 - renamed AnyEvent::Impl::FLTK2 to FLTK.
252
2536.01 Fri Aug 26 07:04:11 CEST 2011
254 - INCOMPATIBLE CHANGE: AnyEvent::DNS resource records now include
255 the TTL as fourth element - this affects ->resolve, ->request
256 and ->dns_unpack indirectly. Convenience functions such as
257 AnyEvent::DNS::{a,txt,mx,srv} are not affected, neither is
258 AnyEvent::Socket.
259 - INCOMPATIBLE CHANGE: backend authors now should not implement
260 one_event or loop, but instead the AnyEvent::CondVar::_wait and _poll
261 methods.
262 - INCOMPATIBLE CHANGE: verbosity levels have been redefined, see
263 AnyEvent (PERL_ANYEVENT_VERBOSE) and AnyEvent::Log for details.
264 - DEPRECIATION: please change
265 "use AnyEvent::Impl::Perl" to "use AnyEvent::Loop" and
266 "AnyEvent::Impl::Perl::loop" to "AnyEvent::Loop::run"
267 in your programs, for future compatibility.
268 - *blush* idle watcher emulation was borked.
269 - AnyEvent::Strict now checks the AE::xxx functions as well.
270 - querying too long domain names would barf deep inside AnyEvent::DNS - since
271 this is hard to avoid for applications, AnyEvent:.DNS::request now checks for
272 for too long domain names explicitly and passes an undef to the callback
273 instead (reported by James Bromberger).
274 - make AnyEvent::Util more robust against EINTR (analyzed by Andrew Suffield).
275 - new functions: AnyEvent::postpone, AnyEvent::log.
276 - new module: AnyEvent::Log, for simple logging needs.
277 - new env variable, PERL_ANYEVENT_LOG, for fine-grained logging config.
278 - new env variable, PERL_ANYEVENT_DEBUG_SHELL, to automatically start
279 a debug shell.
280 - new env variable, PERL_ANYEVENT_DEBUG_WRAP, to automatically enable
281 watcher instrumentation/wrapping (see AnyEvent::Debug::wrap).
282 - new env variable, PERL_ANYEVENT_HOSTS, to override the place where
283 /etc/hosts can be found.
284 - an empty PERL_ANYEVENT_RESOLV_CONF now selects the default
285 configuration.
286 - env variables can now be specified using AE_xyz instead of
287 PERL_ANYEVENT_xyz, subject to some rules.
288 - lots of new AnyEvent::Debug functionality.
289 - AnyEvent::Strict now checks for modifications of $_ by the callback
290 (by setting it to readonly while executing the callback, if possible).
291 - IO::Async sometimes cancels the wrong timer when you call cancel_timer.
292 Invest extra overhead to work around this peculiar behaviour.
293 - split the pure perl backend into a loop module and an impl
294 module (AnyEvent::Loop and AnyEvent::Impl::Perl).
295 - some read types (regex, netstring, json, storable) could
296 errornously keep an AE::Handle object alive on their own
297 (reported by Mohammad Toossi).
298 - AnyEvent::Socket::resolve_sockaddr and all functions using it now
299 supports /etc/hosts overrides.
300 - add more workarounds around child watcher bugs in IO::Async
301 and (newly introduced) in POE.
302 - use glib child watchers and try to work around its limitations,
303 if possible (based on a patch by Kevin Ryde).
304 - add support for FLTK (fltk2), via AnyEvent::Impl::FLTK2.
305 - many read types in AnyEvent::Handle would malfunction during
306 transitions to TLS because of a "delete $self->{rbuf}".
307 (reported by Gerald Galster).
308 - improve condvar blocking wait performance (probably only noticable
309 with faster backends), also make it easier for event loops that
310 do not support blocking waits.
311 - PERL_ANYEVENT_MODEL now also allows full module names, although
312 the usefulness of this feature is not yet clear.
313 - add tests for all supported event loops, conditional on PERL_ANYEVENT_LOOP_TESTS.
314 - add t/80_ssltext.t.
315 - no longer use AUTOLOAD for the initial loading (for no very good reason).
316 - tcp_server now tries to unlink unix domain sockets when it gets destroyed.
317 - do not run event loop detection multiple times when programs call their cached
318 copies of AnyEvent::detect.
319 - parse_hostport now accepts absolute paths als unix domain sockets.
320 - suppress the idiotic warning inside POE by patching the POE::Kernel
321 run flags instead of asking the user to jump through contortions.
322 - add Task::Weaken as dependency (as recommends only at this time).
323 - replace PF_UNSPEC by 0, for good old minix.
324
3256.0 (not officially released, but was on CPAN for about 15 minutes :/)
326
3275.34 Thu May 12 09:59:41 CEST 2011
328 - AE::TLS didn't support both cert_file and key at the same time,
329 due to some copy&paste bug (reported and analyzed by vti).
330 - improve AE::Handle timeout documentation slightly.
331
3325.33 Mon Feb 7 21:51:41 CET 2011
333 - never appeared on CPAN, for unknown reasons.
334 - AnyEvent::Handle on_connect_error - do not try to destroy a handle
335 object when it's already gone (analyzed by Ryan Bullock).
336 - update idna algorithm(s) to UTS#46 version 6.0.0.
337 - preserve trailing dot in idn_to_ascii - testsuite says to preserve,
338 uts46 says it's an error, and testsuite makes more sense.
339 - remove leading dots in idn_to_ascii - testsuite says to remove,
340 uts46 says it's an error, and testsuite makes more sense.
341 - fixed a bug in idn ignored character processing.
342 - passes the uts46 6.0.0 testsuite except for three cases which
343 apparently are bugs in the testsuite itself.
344 - passes the uts46 6.0.1 testsuite except for eleven cases which
345 apparently are bugs in the testsuite itself.
346
3475.31 Sun Jan 23 11:49:19 CET 2011
348 - work around a POE bug where POE would first poll for new events and
349 invoke handlers BEFORE handling the message queue.
350 - fix parsing of "domain" statement in resolv.conf (Johannes Plunien).
351 - stop_read now does just that in TLS mode. see its
352 documentation for caveats.
353 - update to common::sense 3.4 warning set (some warnings in perl impose
354 an arbitrary 32kb limit on regex matches, which affects
355 AnyEvent::Handle).
356 - AE::Handle had a broken rbuf_max method and missing wbuf_max method.
357 - new experimental AE::Handle::resettls method.
358
3595.3 Fri Dec 31 05:48:13 CET 2010
360 - major IOAsync patch by Paul Evans - 0.33 is required, as the major
361 issues with IO::Async have been solved in that version, including
362 the availabilty of a default loop!
363 - implement new wbuf_max parameter for AnyEvent::Handle.
364 - added Cocoa::EventLoop backend by Daisuke Murase.
365 - the default testsuite now honors PERL_ANYEVENT_MODEL (Paul Evans).
366 - clarify handle on_prepare documentation.
367 - set CLOEXEC flag on parent-side fd's in AnyEvent::Util::run_cmd,
368 to avoid leaking them to any child processes.
369 - try to force files into the arch-dependend lib directory, for the
370 half-"turly obsessive compulsive person"(s) that depend on it.
371
3725.29 Sun Dec 5 10:49:21 CET 2010
373 - convert EV backend to EV 4.00 API (so better upgrade EV too).
374 - AE::Handle: implement dynamic read_size adjustment, add
375 max_read_size parameter, reduce default read_size to 2048.
376 - add some FAQ entries for servers and/or on_eof vs. on_error.
377 - work around OS bugs (cygwin again) when lingering in AE::Handle.
378 - reduce memory usage slightly when connect was used in AE::Handle.
379 - wrok around more segfaults in Tk.
380 - document tls_ctx => undef as valid.
381 - detect WNOHANG value at compiletime.
382 - include some probably linux-only support for building AnyEvent
383 as part of the perl core.
384 - improve compatibility of t/08_idna.t to perl 5.8.
385 - make AnyEvent::Strict truly optional (in case somebody builds a
386 minimal perl...).
387
3885.28 Wed Oct 13 04:14:23 CEST 2010
389 - due to a glitch, AnyEvent's internal getprotobyname wasn't
390 actually used in tcp_server.
391 - implement AnyEvent::CondVar->new method, which was documented but
392 missing.
393 - support autoloading of read types in unshift_read.
394 - AnyEvent::Strict: do the test for invalid fh arguments differently,
395 resulting in more informative error messages.
396 - new AnyEvent::FAQ manpage.
397 - clarify that you can actually call push_read etc. in on_read.
398 - put keys and certs in /tmp, not into the current directory,
399 and document this in the SECURITY CONSIDERATIONS sections
400 of AnyEvent::TLS.
401 - support ";" as resolv.conf comment character in AnyEvent::DNS, as
402 requested by Ron Isaacson.
403 - document $AnyEvent::DNS::RESOLVER variable.
404 - incorporated major typo patches by Abhijit Menon-Sen.
405 - AnyEvent::Handle now croaks when negative timeouts are passed.
406 - add a shitload of TCP_xxx sockopt name constants to AnyEvent::Util,
407 undocumented, but who knows when they might come in handy.
408
4095.271 Tue Jun 8 12:05:46 CEST 2010
410 - backport to perl 5.8.x.
411
4125.27 Sun Jun 6 12:12:05 CEST 2010
413 - postpone differently in AnyEvent::Socket now, as
414 when not, canceling the connection attempt might fail
415 (found by Felix Antonius Wilhelm Ostmann).
416 - explicitly check for non-stream sockets in AE::Handle, too many
417 clueless people fell into the trap of this somehow working.
418 - simplified and reworked the "OTHER MODULES" section.
419 - better/more condvar examples.
420
4215.261 Wed Apr 28 16:13:36 CEST 2010
422 - AF_INET6 was not properly used from Socket6 during configuration
423 time (found by Andy Grundman).
424
4255.26 Mon Apr 12 04:49:35 CEST 2010
426 - don't generate spurious readyness notifications when select
427 returns EINTR in the pure perl backend (analysed and testcase
428 by Bas Denissen).
429 - give same examples in SYNOPSIS sections of AnyEvent and AE.
430 - provider faster implementations for AE::time/now/now_update for
431 EV and Perl backends.
432 - provide faster AE::cv and AE::time implementations for all
433 backends.
434 - no longer support lower/mixed-case signal names (this was
435 never documented nor universally supported).
436 - some more memory µ-optimisations, and somewhat less messy
437 code to cope with both AnyEvent and AE APIs.
438
4395.251 Sat Mar 13 00:58:21 CET 2010
440 - make test failed if it was run before make install on perl <=
441 5.8.8, and older versions of perl make our live very difficult,
442 so write constants.pl during Makefile.PL time.
443
4445.25 Sat Mar 13 00:23:14 CET 2010
445 - fix a race condition in AnyEvent::Handle that would cause
446 a "bio_read: unsupported method" error (in _another_
447 ssl connection) after user code threw an exception.
448 - added AnyEvent::Handle->destroyed method.
449 - speed up AnyEvent::Socket::format_address and ::format_ipv6.
450 - the AnyEvent::Util::fh_nonblocking function worked only
451 by lucky accident on win32.
452 - smaller and faster AnyEvent::Util::fh_nonblocking.
453 - when the (required!) Time::HiRes module is missing, AnyEvent
454 did not fall back to built-in time properly.
455 - do not load Fcntl at runtime, saving memory and loading time.
456 - precompile a number of constants and use them instead of runtime
457 detection and eval.
458 - free detection code after detection and similar memory optimisations.
459 - Perl backend timer interval best effort drift has been improved
460 (same algorithm as EV).
461 - update unicode idna mapping table.
462
4635.24 Tue Jan 5 11:39:43 CET 2010
464 - cygwin never reports errors from failed connects. 1.5 just gives
465 you continous readyness and EAGAIN, 1.7 is even more broken and
466 just hangs. work around both issues in a cygwin-specific hack.
467 - improve idle watcher documentation slightly.
468
4695.23 Sun Dec 20 23:48:00 CET 2009
470 - support IDNs in resolve_sockaddr, and therefore in tcp_connect.
471 - implement punycode_encode/decode, idn_nameprep,
472 idn_to_ascii and idn_to_unicode operations in AnyEvent::Util.
473 - provide $AE::VERSION.
474 - removed traces of "no strict 'refs'".
475
4765.22 Sat Dec 5 03:51:13 CET 2009
477 - downgrade-or-fail in AnyEvent::Handle::push_write, to
478 diagnose encoding failures earlier and more succinctly.
479 (this works around bugs in perl, throwing away encoding info
480 when passing scalar data to extensions).
481 - add more examples to AnyEvent::Socket manpage.
482 - upgrade internal warning set to the same as common::sense 2.03.
483 - use pack "n/a*" for pre-5.8.9 perl compatibility in AnyEvent::DNS
484 (John Beppu).
485 - AnyEvent::Socket::inet_aton now properly supports ipv6, as documented.
486 - add google public dns servers to fallback server set and make sure
487 we load-balance properly between the three sets. also add all
488 fallback dns servers, not just a random one, to each dns config.
489
4905.21 Thu Nov 19 02:48:47 CET 2009
491 - fix a problem where socket constants were called with parameters
492 (spotted by David Friedland).
493 - fork_call never use'd POSIX (reported by Daisuke Maki).
494 - improve perl 5.6 compatibility further (but it still won't work
495 unless you rip out everything but the core).
496 - prefer Net::DNS::Resolver over ipconfig, if installed, on win32.
497 uses 10MB of RAM, but doesn't open a console window. *sigh*.
498
4995.202 Wed Oct 14 22:35:44 CEST 2009
500 - AnyEvent::DNS would unexpectedly clobber $_ under windows
501 (analysed by Matthias Waldorf).
502 - AnyEvent::Handle::run_cmd can now pass the PID of the
503 newly-created process, which is much less useful than it might
504 sound (based on patch by Yann Kerherve).
505
5065.201 Tue Sep 29 12:09:25 CEST 2009
507 - AnyEvent:Handle::on_starttls/on_stoptls methods were broken
508 (reported by Torsten Foertsch).
509 - common::sense 2.0 could cause tcp_server to throw an exception
510 (analysed by elmex).
511
5125.2 Mon Sep 14 07:04:49 CEST 2009
513 - INCOMPATIBLE CHANGE: do no longer support register_read_type
514 and register_write_type in AnyEvent::Handle, instead support
515 package names (the facility was mostly abused).
516 - implement "packagename-as-read/write type" support in
517 AnyEvent::Handle.
518 - AnyEvent::Handle: new options "keepalive" and "oobinline".
519 - oobinline set by default to avoid security issues.
520 - the pure-perl event loop backend wrongly detected times() fallback
521 support (spotted by Pavel Boldin).
522
5235.12 Tue Sep 1 20:26:50 CEST 2009
524 - be more lenient when parsing resolv.conf files, as some people
525 use hashmarks as comment indicator inside directives (reported
526 by Michael S. Fischer).
527 - use same set of warnings as common::sense 2.0.
528 - fix a potential 32 bit overflow issue due to perl having problems
529 with large hex constants in the Perl backend.
530 - do not use the slower 5.6 version fo the enc_name code in AnyEvent::DNS
531 by default.
532 - fix some prototypes.
533 - (optionally) require Async::Interrupt 1.02.
534 - replace opendns fallbacks by verizon/level3 public dns servers.
535
5365.112 Fri Aug 21 13:59:04 CEST 2009
537 - AnyEvent::Handle could cause hard crashes in Net::SSLeay when
538 using starttls with an illegal/uncreatable tls context.
539 - port to broken win32 perls.
540
5415.111 Sun Aug 16 18:44:36 CEST 2009
542 - perl errornously requires sockaddr_un structures returned by the
543 kernel to have a specific length. work around this bug by padding
544 sockaddr_un structures (found to fail on OS X, as perl only uses
545 the correct code for GNU/Linux, but likely to fail on other
546 OSes as well).
547
5485.11 Wed Aug 12 17:49:37 CEST 2009
549 - fixed a bug in run_cmd with input from filename.
550 - create a json coder object when none is given in write direction
551 to reduce dependency on JSON.
552
5535.1 Tue Aug 11 03:17:32 CEST 2009
554 - tcp_connect now reports the correct error in $!, not an unrelated
555 one (was broken since 4.91).
556 - AE::Handle did not properly clear rtimeout/wtimeout watchers
557 when retrying a tcp_connect.
558 - new functions: AnyEvent::Util::run_cmd and close_all_fds_except.
559 - fixed wrong documentation in AE::Socket.
560
5615.01 Mon Aug 10 03:16:32 CEST 2009
562 - last release broke EV child watchers :/.
563
5645.0 Sun Aug 9 17:32:09 CEST 2009
565 - officially introduce and document the AE API.
566 - lowering the timeout at runtime did not have immediate effect
567 in AnyEvent::Handle.
568 - AE::Handle now has separate and independent read and write timeout
569 settings.
570 - probe for Guard module when AnyEvent::Util is loaded, not at
571 runtime (Event doesn't like eval "use" from callbacks, and it avoids
572 exporting the probe function).
573 - Impl::IOAsync's io watchers did not use the filehandle emulation
574 code to work around it's watcher limits.
575 - work around IO::Async's problems with overloaded objects.
576 - try to disable wearnings inside IO::Async, it's just too buggy.
577 - do no longer use Event by default, only EV or Perl, Event is too buggy.
578 - hackishly provide fast AE:: interface when EV or Perl is the backend.
579 - inherit VERSION from AnyEvent::VERSION in most modules.
580
5814.91 Thu Aug 6 15:42:45 CEST 2009
582 - AE::Handle::starttls could get out of sync when the read buffer
583 already contains some TLS handshake.
584 - AE::Handle did not properly free the TLS session, causing bigger
585 memory leaks in Net::SSLeay than expected (reported by toaster).
586 - AE::Socket::tcp_connect will now artificially delay invoking
587 the callback to avoid returning after invoking the callback.
588 - convert many internal watcher uses to AE API.
589
5904.9 Sat Aug 1 11:07:01 CEST 2009
591 - Glib silently fails when registering a timeout with
592 a negative value, so avoid that.
593 - call condvar callback immediately when it is set
594 after the condition is already signalled.
595 - check rbuf_max condition only after trying to consume
596 data.
597 - normalise signal numbers to names when using the ae's
598 signal handling, but do not document this (yet).
599 - pure perl signal emulation did not properly set nonblocking
600 mode on the signal pipe on win32 (this is unlikely to have caused
601 any issues).
602 - new module: AnyEvent::Debug.
603 - AnyEvent::Strict now checks that a signal was specified
604 by name (not name or number).
605 - reduce memory footprint in typical cases by ~50kb by
606 only compiling signal code when necessary.
607 - add AnyEvent::Handle->rbuf_max.
608 - grab the AE namespace for future new API, implement stubs for
609 said future/alternative API.
610 - new function: AnyEvent::Socket::format_hostport.
611
6124.881 Tue Jul 28 12:51:53 CEST 2009
613 - work around a bug in local in pre-5.10 perls, causing
614 AnyEvent::Handle to recurse when it should not
615 (analysed by elmex).
616
6174.88 Tue Jul 28 04:04:37 CEST 2009
618 - re-bless the handle into a dummy package after calling
619 AnyEvent::Handle::destroy, so the user does not need to check
620 for errors after every push_write etc.
621 - do not attempt to run t/02_signals.t on obviously broken
622 platforms.
623
6244.87 Sun Jul 26 02:06:16 CEST 2009
625 - do not attempt to linger when there is no longer a valid fh.
626 - enforce tls mode to be either accept or connect, do not simply
627 segfault in Net::SSLeay.
628 - AnyEvent::Handle can now call tcp_connect itself (new parameters
629 connect, on_prepare, on_connect and on_connect_error). Updated
630 tutorial accordingly.
631 - add AnyEvent::Impl::Irssi backend.
632
6334.86 Mon Jul 20 23:52:29 CEST 2009
634 - since the verbose warning is not enough, explicitly document
635 that versions before 1.33 of Net::SSLeay are not secure.
636 - work around signal handling races in Event and (...) Event::Lib.
637 - try to align signal-race timer to full-second boundaries.
638 - work around Tk not liking negative timeouts.
639 - don't complain of different grades of environmental unfriendlyness
640 in IO::Async.
641
6424.85 Sat Jul 18 06:16:14 CEST 2009
643 - nail the signal race problem in perl once and for all
644 (see $AnyEvent::MAX_SIGNAL_LATENCY).
645 - take advantage of Async::Interrupt if it is available.
646 - load Time::HiRes and Guard modules on demand only.
647 - add optional/recommended modules section to AnyEvent
648 documentation.
649 - reduce memory usage considerably (and reduce startup penalty)
650 by not using "strict", "warnings" and "overload" modules.
651 - work around buggy windows/openbsd perls and provide EBADMSG
652 and EPROTO ourselves when missing.
653 - improve perl 5.6 compatibility of the core event loop.
654 - made Net::SSLeay version 1.33 a soft requirement.
655
6564.83 Fri Jul 17 16:56:26 CEST 2009
657 - implement AnyEvent::Socket::getprotobyname.
658 - AnyEvent::CondVar's will now detect recursive blocking
659 waits and will croak, as too many people fall into
660 this trap.
661 - AnyEvent::Handle will now call ->destroy on itself after
662 executing the on_error callback, instead of doing some
663 half-baked internal shutdown, for fatal errors.
664 - clarify on_eof behaviour w.r.t. the read queue and
665 on_read callbacks.
666 - ignore some possible spurious wake-ups in tcp_connect.
667
6684.82 Sat Jul 11 00:34:55 CEST 2009
669 - POE and Event backends didn't accept some callable objects as
670 callbacks.
671 - use Config module instead of POSIX module to detect signal names
672 in AnyEvent::Strict and AnyEvent::Impl::EventLib, as the POSIX
673 module doesn't even have all POSIX signals :/.
674 - use more workarounds around the many refcnt/corruption bugs in
675 Event::Lib.
676 - work around a race condition in perl's select, causing t/03_child.t
677 to rarely fail.
678
6794.81 Thu Jul 9 10:30:30 CEST 2009
680 - AnyEvent::Handle didn't properly diagnose write errors
681 (it expected -1 from syswrite, how lame... :).
682 - support file descriptors in addition to file handles
683 in AnyEvent->io.
684 - new env variables: PERL_ANYEVENT_RESOLV_CONF,
685 PERL_ANYEVENT_MAX_OUTSTANDING_DNS, PERL_ANYEVENT_CA_FILE
686 and PERL_ANYEVENT_CA_PATH.
687 - provide a sensible synopsis section for AnyEvent::TLS.
688 - add a "supported backends" section to the manpage.
689 - added simple io watcher test to testsuite, using a
690 portable_socketpair.
691 - tried to improve the stability of the Event::Lib backend,
692 YMMV.
693
6944.8 Mon Jul 6 23:45:16 CEST 2009
695 - AnyEvent::DNS did not properly follow CNAME records with
696 uppercase targets.
697 - AnyEvent::DNS would errornously return AAAA records
698 with v4 mapped addresses (a faulty record) as ipv4 addresses,
699 causing AnyEvent::Socket to throw an exception.
700 - added new module AnyEvent::TLS for easier SSL/TLS context
701 creation, with many options including hostname verification,
702 secure default configuration, lots of documentation and,
703 predefined diffie-hellman keys for perfect forward security
704 and much more. get it while it's still fresh!
705 - use AnyEvent::TLS in AnyEvent::Handle for context management.
706 - load AnyEvent::Handle only on demand in AnyEvent::DNS,
707 so AnyEvent::Socket users have smaller memory footprint
708 in the common case.
709 - add AnyEvent::Handle->push_shutdown method.
710 - add an additional error message parameter to AnyEvent::Handle's
711 on_error callback (for TLS, $! is still available).
712 - add AnyEvent::Handle on_starttls/on_stoptls callbacks.
713 - make AnyEvent::Handle more robust against early connection
714 failures (during new), and return C<undef> in that case
715 from the constructor.
716 - AnyEvent::Handle will now try to load only JSON::XS first,
717 then fall back to JSON.
718 - format_ipv4/format_ipv6 are now exported by default, for symmetry,
719 and because it was documented that way.
720
7214.451 Fri Jul 3 00:28:58 CEST 2009
722 - do not clear rbuf when shutting down an AnyEvent::Handle
723 object - doing so breaks AnyEvent::HTTP.
724
7254.45 Mon Jun 29 22:59:26 CEST 2009
726 - a write error could cause AnyEvent::Handle to create
727 an I/O watcher with an undefined $fh.
3 - special-case mapped ipv4 addresses in both 728 - special-case mapped ipv4 addresses in both
4 AnyEvent::Socket::format_address and parse_address, 729 AnyEvent::Socket::format_address and parse_address,
5 to treat them just like ipv4 addresses. 730 to treat them just like ipv4 addresses.
731 - updated and overhauled the AnyEvent::Intro doc.
6 - implement AnyEvent::Socket::format_ipv4/ipv6. 732 - implement AnyEvent::Socket::format_ipv4/ipv6.
7 - slightly speed up portable_pipe/socketpair. 733 - slightly speed up portable_pipe/socketpair.
8 - expand condvar ebgin/end documentation. 734 - expand condvar begin/end documentation.
9 735
104.42 Fri Jun 26 08:32:18 CEST 2009 7364.42 Fri Jun 26 08:32:18 CEST 2009
11 - preliminary and neccesarily incomplete support for IO::Async. 737 - preliminary and neccesarily incomplete support for IO::Async.
12 - reset SIGCHLD to DEFAULT when AnyEvent is loaded, in 738 - reset SIGCHLD to DEFAULT when AnyEvent is loaded, in
13 case it was set to IGNORE, to ensure we can catch child statuses 739 case it was set to IGNORE, to ensure we can catch child statuses
… …
1164.3 Fri Oct 3 09:18:43 CEST 2008 8424.3 Fri Oct 3 09:18:43 CEST 2008
117 - AnyEvent will now install a no-op signal handler for SIGPIPE, 843 - AnyEvent will now install a no-op signal handler for SIGPIPE,
118 unless one has been installed already. 844 unless one has been installed already.
119 - warn about not loaidng AnyEvent::Impl::POE early enough 845 - warn about not loaidng AnyEvent::Impl::POE early enough
120 (patch by Adam Rosenstein). 846 (patch by Adam Rosenstein).
121 - fixed a great number of bugs and corner cases in AnyEvent::Handle: 847 - fixed a great number of bugs and corner cases in AnyEvent::Handle.
122 - fix a bug in where in SSL connect mode, the 848 - fix a bug in where in SSL connect mode, the
123 client would first wait for some data by the server and 849 client would first wait for some data by the server and
124 otherwise hang. (reported and analyzed in an absolutely 850 otherwise hang. (reported and analysed in an absolutely
125 exemplary manner by Adam Rosenstein). 851 exemplary manner by Adam Rosenstein).
126 - fix a bug in where SSL EOF would not be treated 852 - fix a bug in where SSL EOF would not be treated
127 as stream EOF, putting the connection into a hung state. 853 as stream EOF, putting the connection into a hung state.
128 (reported and analyzed in an absolutely exemplary manner by Adam 854 (reported and analysed in an absolutely exemplary manner by Adam
129 Rosenstein). 855 Rosenstein).
130 - fix a potential segfault when the TLS context 856 - fix a potential segfault when the TLS context
131 would go missing in a read callback (e.g. due to stoptls) - 857 would go missing in a read callback (e.g. due to stoptls) -
132 Net::SSLeay of course makes no type checking whatsoever. 858 Net::SSLeay of course makes no type checking whatsoever.
133 - AnyEvent::Handle will not stop the read watcher in TLS mode. 859 - AnyEvent::Handle will not stop the read watcher in TLS mode.

Diff Legend

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