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

Comparing AnyEvent/Changes (file contents):
Revision 1.304 by root, Tue Jul 28 11:59:35 2009 UTC vs.
Revision 1.627 by root, Tue Apr 17 19:04:50 2018 UTC

1Revision history for Perl extension AnyEvent. 1Revision history for Perl extension AnyEvent.
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)
12TODO: inet_aton maybe pack C4? no, add a comment why not
13TODO: socket prepare callbacks, can they easily signal failure as well?
14
15TODO: stuff dns packet into decoded structure
16TODO: allow client programs to extend the dns encoder/decoder list
17TODO: Олег Г <verdrehung@gmail.com> test nameserver
18
19TODO: tls_detetc callback version passed wrongly != documented
20 - AnyEvent::Socket::parse_ipv6 would NOT, as advertised, accept ipv4
21 addresses. It now does and converts them to ipv4 mapped addresses.
22 - support CAA records, based on patch by Steve Atkins.
23 - add freenom and cloudflare nameservers as dns fallback.
24
257.14 Sat Jun 24 01:44:19 CEST 2017
26 - fix a crash bug in AnyEvent::Handle with openssl 1.1.0 (patched
27 by Paul Howarth and apparently tracked down by Mike McCauley).
28 - AnyEvent::Handle->keepalive was documented (and defined) twice
29 (found by Matt Merhar).
30 - AnyEvent::Socket::tcp_bind/tcp_server would immediately unlink a unix
31 listening socket unless a guard is used. Change this so that no
32 cleanup will be performed unless a guard is used and document this more
33 clearly.
34 - make tcp_bind/tcp_server error messages more regular.
35 - work around latest perl upstream breakage - this time, perl5porters were
36 apparently bullied by a bogus CVE (CVE-2016-1238), and since distros
37 stupidly apply anything that has "security fix" stamped on it, it's likely
38 going to be a problem in practise (and working around it is trivial on
39 windows and unix).
40 - add TCP_FASTOPEN/MSG_FASTOPEN and MSG_NOSIGNAL/MSG_MORE/MSG_DONTWAIT
41 to constants.
42 - update warnings to common::sense 3.74 standards.
43
447.13 Sat Sep 17 04:31:49 CEST 2016
45 - Only call tlsext_host_name for non-empty common names (reported
46 by Maxime Soulé).
47 - log a (single) notice message if SNI is not supported.
48 - upgrade to UTS-46:9.0.0 draft and switch to non-transitional
49 behaviour, beating thunderbird, ie, edge, chrome and safari to it :)
50 (see also https://bugzilla.mozilla.org/show_bug.cgi?id=1218179)
51 - turns out the UTS-46 IDNA testcase failures were indeed bugs in the
52 testcases and the specification and not in the code - the post-9.0.0
53 unicode files have all known problems fixed, so finally the AnyEvent
54 IDNA implementation can pass the full IDNA testsuite - without needing
55 a single fix :)
56 - guarantee (and document) that condvar callbacks will be removed
57 on invocation - important to avoid circular references.
58
597.12 Wed Jan 27 19:12:26 CET 2016
60 - use common name as hostname for tls connects, if Net::SSLeay
61 supports SNI.
62 - fix documentation of tls_autostart read type in AnyEvent::Handle,
63 analyzed by Felix Ostmann.
64
657.11 Thu Jul 16 14:36:00 CEST 2015
66 - AnyEvent::Socket::parse_ipv6 could accept malformed ipv6
67 addresses (extra "::" at end and similar cases).
68 - add a more explicit warning to AnyEvent::Handle that it doesn't
69 work on files, people keep getting confused.
70 - new function AnyEvent::Socket::tcp_bind.
71 - new functions AnyEvent::fh_block and AnyEvent::fh_unblock.
72 - aligned ipv6 address formatting with RFC 5952 (by not shortening
73 a single :0: to ::).
74 - added stability canary support.
75
767.09 Sat May 2 16:38:53 CEST 2015
77 - AnyEvent::Debug called an internal function (AnyEvent::Log::ft)
78 that was renamed to AnyEvent::Log:format_time. uinder its old name
79 (analyzed by Michael Stovenour).
80 - update AnyEvent::DNS fallback resolver addresses:
81 seems google effectively killed most other free dns resolvers,
82 so remove them, but add cable and wireless (ecrc) since it was
83 stable for 20 years or so, official or not, and there should be
84 an alternative to google.
85 - perl5porters broke windows error codes in 5.20, and mapped
86 WSAEWOULDBLOCK on the (different) EWOULDBLOCK error code, and
87 WSAEINPROGRESS into the incompatible ERINPROGRESS code, probably
88 because they were so cool. They probably broke other error codes
89 for no reason, too, but I didn't care for checking, it's too
90 depressing. This version only works around the WSAEWOULDBLOCK
91 issue, because I don't have a nice way to work around the
92 WSAEINPROGRESS bug.
93
947.08 Wed Dec 10 05:27:17 CET 2014
95 - work around a newly introduced bug in Socket 2.011 (an
96 errornous sun_length check) (analyzed by Maxime Soulé).
97 - AnyEvent::TLS didn't load (but refer to) AnyEvent::Socket
98 (analyzed by Ben Magistro).
99 - AnyEvent::Strict will now confess, not croak. This is in line with
100 it being a development/debugging tool.
101 - work around a number of libglib bugs (debug builds of libglib enforce
102 certain undocumented behaviour patterns such as not being able to
103 remove a child watch source after it has fired, which we will try
104 to emulate to avoid "criticals". what where they thinking...).
105 - mention json security issues in AnyEvent::Handle, now that Douglas
106 Crockford has foolishly and incompatibly changed JSON.
107 - changed default dns resolver "max_outstanding" value from 1 to 10,
108 the latter beinfg the intended value all along
109 (reported by Ilya Chesnokov).
110 - added new "AnyEvent::Impl::UV" interface module to the UV event lib
111 (written by Mike Lowell).
112
1137.07 Tue Dec 17 17:45:02 CET 2013
114 - the documentation for custom tls verify schemes was wrong. make it agree
115 with the code (reported by Maxime Soulé).
116 - added cbor read and write types to AnyEvent::Handle (using CBOR::XS).
117 - work around an API change in openssl that could cause wrong tls connection
118 aborts, likely on windows only (analyzed by sten).
119 - calling AnyEvent->now_update with AnyEvent::Impl::Perl caused an
120 endless loop (reported by Dietrich Rebmann).
121 - add tlsv1_1 and tlsv1_2 protocols to AnyEvent::TLS
122 (patch by Maxime Soulé).
123 - document AnyEvent::Impl::IOAsync::set_loop and
124 $AnyEvent::Impl::IOAsync::LOOP. Though only documented now, this
125 functionality has _always_ been available.
126 - force a toplevel domain name in t/81_hosts.t (analyzed by
127 David Jack Wange Olrik).
128 - document that AnyEvent::Log uses AnyEvent::IO.
129 - warn about AnyEvent::Filesys::Notify performance.
130 - praise the joys of AnyEvent::Fork::*.
131 - time for an =encoding directive.
132 - do no longer use JSON to create a default json coder, use
133 JSON::XS or JSON::PP directly.
134
1357.05 Wed Aug 21 10:38:08 CEST 2013
136 - uts46data.pl couldn't be found due to wrong naming of the file
137 (reported by Fulko Hew).
138 - handle lone \015's properly in AE::Handle's default line read
139 (reported by various people).
140 - untaint ip addresses found in /etc/hosts (patch by José Micó).
141 - the memleak fix in 7.03 caused resolving via /etc/hosts to always
142 fail on first use (reported and testcase by Andrew Whatson).
143 - expose AnyEvent::Log::format_time, and allow users to redefine it.
144 - expose AnyEvent::Log::default_format, and allow redefinition.
145 - expose AnyEvent::Log::fatal_exit, to allow redefinition.
146 - AnyEvent::Debug shell can now run coro shell commands, if available.
147 - t/63* tests were wrongly in MANIFEST.
148 - kernel.org's finger server went MIA, switch to freebsd.org and
149 icculus.org.
150 - clarify that IO::AIO and AnyEvent::AIO are needed for AnyEvent::IO
151 to function asynchronously (reported by Daniel Carosone).
152 - hard-disable $^W in most tests, it generates too much garbage output.
153 - use a (hopefully) more future-proof method to emulate common::sense.
154 - upgrade to UTS-46:6.2.0.
155 - switch to INSTLIB from INSTLIBDIR, as INSTLIBDIR was wrongly
156 documented. should not affect anything.
157
1587.04 Thu Nov 15 02:23:58 CET 2012
159 - debugging code left in.
160
1617.03 Thu Nov 15 02:16:11 CET 2012
162 - AnyEvent::Socket::inet_aton did not work when dns resolution was
163 used to find the addresses (analyzed and patch by Maxime Soulé).
164 - fix a memory leak in the /etc/hosts lookup code when hosts don't
165 resolve and are not in hosts (analyzed by sten).
166
1677.02 Tue Aug 14 04:27:58 CEST 2012
168 - AnyEvent::Util::run_cmd could block indefinitely (analyzed and test
169 program by Yorhel).
170 - verified that AnyEvent::Socket follows RFC5952.
171 - try to parse "ADDR#PORT" in addition to "ADDR PORT".
172
1737.01 Sun May 13 01:03:17 CEST 2012
174 - fail with EPROTO in AnyEvent::Handle wqhen TLS is requested but
175 not available, instead of throwing an exception.
176 - use File::Spec to get the tmpdir in t/*, to avoid needless
177 failures on (most, not mine :) windows boxes.
178 - new handle read types: tls_detect and tls_autostart.
179
1807.0 Fri Apr 13 06:33:30 CEST 2012
181 - child watchers are broken in POE 1.352 (also many earlier
182 versions) and there seems to be no way to work around it, as POE
183 itself is inherently racy. Document this regression and add a delay
184 in t/68_poe_03_child.t for the time being.
185 - new module AnyEvent::IO, that is a frontend to either a pure-perl
186 synchronous I/O implementation (AnyEvent::IO::Perl), or to the
187 asynchronous I/O extension IO::AIO (AnyEvent::IO::IOAIO).
188 - load /etc/hosts only when DNS has no answer.
189 - stat /etc/hosts on every access and reload it if it changed.
190 - load /etc/hosts via AnyEvent::IO - potentially asynchronous.
191 - fix a buggy croak in the dh parameter reading in AnyEvent::TLS.
192 - AnyEvent::Log log_to_file and log_to_path now use AnyEvent::IO. As
193 a side effect, they now use true appending as opposed to libc
194 appending, and the file might not have been opened when the function
195 returns.
196 - the default logging level was not properly documented in a variety of
197 places, this has hopefully been rectified.
198 - updated uts46data.pl for unicode 6.1.0.
199 - made log messages generated by AnyEvent submodules not
200 include the package name anymore, as it will be logged
201 by default already.
202 - upgrade to the trick used by common-sense 3.5 to work
203 around extra warning torture/breakage under perl 5.15.x.
204 - log messages by anyevent are now ucfirst, are usually full
205 sentences and do no longer include the package name.
206 - the storable read type would sometimes throw an exception instead
207 of causing EBADMSG (analyzed by Maxime Soulé).
208 - work around a bug in openssl 1.0.1 which enforces a minimum rsa
209 keysize (reported by Paul Howarth).
210 - documented Rocco fabricating statements about AnyEvent and me.
211
2126.14 Tue Jan 31 20:00:24 CET 2012
213 - AnyEvent::Impl::Tk was broken due to a mysterious "die" inside,
214 probably an editing mistake (reported by Darin McBride).
215
2166.13 Thu Jan 12 07:27:01 CET 2012
217 - AnyEvent::Util::fork_call checked for POSIX availability in the wrong way,
218 causing it to fail if POSIX wasn't loaded already (analysed by Rock Power).
219 - AnyEvent::Handle::push_read (line => did pass $1 directly, so
220 regex-matching inside the callback would change the parameter, despite
221 $1 being dynamically scoped per-block. this perl bug is now being
222 worked around (testcase by Cindy Wang).
223
2246.12 Mon Dec 12 13:21:10 CET 2011
225 - $! was clobbered by subsequent calls in tcp_connect, due to
226 the postpone not saving/restoring it (analyzed by Richard Garnish).
227
2286.11 Tue Nov 22 10:36:05 CET 2011
229 - Tk cannot create windows when tainted unless you set a title, so
230 set a dummy title for AnyEvent's dummy mainwindow. How dummy.
231 - escape any nonprintable/nonascii characters when stringifying
232 backtraces.
233 - log the reason loading an interface module fails to load at level debug.
234
2356.1 Tue Oct 4 19:44:30 CEST 2011
236 - INCOMPATIBLE CHANGE: the default log level is now "4" (error
237 and above), and some messages inside AnyEvent have been elevated
238 to higher log levels to print by default.
239 - AnyEvent::log, unlike AnyEvent::Log::log, did not always exit on fatal.
240 - work around yet another signal race bug, newly introduced by POE. My
241 how that module sucks.
242 - add google public ipv6 dns as fallback servers, so AE::DNS can finally
243 support ipv6-only operation even with defaults (also use ::1 instead
244 of 127.0.0.1).
245 - give better guidelines on how to chose a logging level.
246 - AnyEvent::Log can now cap (limit) log message priority by context.
247 - AE::log will now save and restore $! and $@, so callers don't have to.
248 - AE::Strict now checks that registered fds are still valid at regular
249 intervals.
250 - skip frequently called 'require' calls if the module appears loaded.
251 - increase signal latency test timer considerably for poor little osx,
252 which can't be bothered with basic posix functionality such as signal
253 delivery.
254 - mention some debugging aids in the tutorial.
255 - new env variable PERL_ANYEVENT_MAX_SIGNAL_LATENCY.
256 - private (%namespace) log contexts in PERL_ANYEVENT_LOG now have the name
257 of the context as title.
258 - AnyEvent::Strict would not restore read-write access to $_ when a callback
259 died - this is a user bug, but since AE::Strict exists to help dbeugging...
260 - detect the extremely evil IO::Async::Loop::AnyEvent module and refuse
261 to work with it - the author is well aware of the issues but refuses
262 to fix them. This extreme step has been taken because the module actively
263 attacks the goals of AnyEvent and makes it harder for module authors to use
264 AnyEvent.
265
2666.02 Fri Aug 26 20:08:31 CEST 2011
267 - log_to_syslog now accepts facility strings.
268 - log_to_syslog would not allow facilities - "$facility|$level"
269 *literally* means that string. sheesh.
270 - the CPAN parser got confused and thought we provide Sys::Syslog, try
271 to work around.
272 - renamed AnyEvent::Impl::FLTK2 to FLTK.
273
2746.01 Fri Aug 26 07:04:11 CEST 2011
275 - INCOMPATIBLE CHANGE: AnyEvent::DNS resource records now include
276 the TTL as fourth element - this affects ->resolve, ->request
277 and ->dns_unpack indirectly. Convenience functions such as
278 AnyEvent::DNS::{a,txt,mx,srv} are not affected, neither is
279 AnyEvent::Socket.
280 - INCOMPATIBLE CHANGE: backend authors now should not implement
281 one_event or loop, but instead the AnyEvent::CondVar::_wait and _poll
282 methods.
283 - INCOMPATIBLE CHANGE: verbosity levels have been redefined, see
284 AnyEvent (PERL_ANYEVENT_VERBOSE) and AnyEvent::Log for details.
285 - DEPRECIATION: please change
286 "use AnyEvent::Impl::Perl" to "use AnyEvent::Loop" and
287 "AnyEvent::Impl::Perl::loop" to "AnyEvent::Loop::run"
288 in your programs, for future compatibility.
289 - *blush* idle watcher emulation was borked.
290 - AnyEvent::Strict now checks the AE::xxx functions as well.
291 - querying too long domain names would barf deep inside AnyEvent::DNS - since
292 this is hard to avoid for applications, AnyEvent:.DNS::request now checks for
293 for too long domain names explicitly and passes an undef to the callback
294 instead (reported by James Bromberger).
295 - make AnyEvent::Util more robust against EINTR (analyzed by Andrew Suffield).
296 - new functions: AnyEvent::postpone, AnyEvent::log.
297 - new module: AnyEvent::Log, for simple logging needs.
298 - new env variable, PERL_ANYEVENT_LOG, for fine-grained logging config.
299 - new env variable, PERL_ANYEVENT_DEBUG_SHELL, to automatically start
300 a debug shell.
301 - new env variable, PERL_ANYEVENT_DEBUG_WRAP, to automatically enable
302 watcher instrumentation/wrapping (see AnyEvent::Debug::wrap).
303 - new env variable, PERL_ANYEVENT_HOSTS, to override the place where
304 /etc/hosts can be found.
305 - an empty PERL_ANYEVENT_RESOLV_CONF now selects the default
306 configuration.
307 - env variables can now be specified using AE_xyz instead of
308 PERL_ANYEVENT_xyz, subject to some rules.
309 - lots of new AnyEvent::Debug functionality.
310 - AnyEvent::Strict now checks for modifications of $_ by the callback
311 (by setting it to readonly while executing the callback, if possible).
312 - IO::Async sometimes cancels the wrong timer when you call cancel_timer.
313 Invest extra overhead to work around this peculiar behaviour.
314 - split the pure perl backend into a loop module and an impl
315 module (AnyEvent::Loop and AnyEvent::Impl::Perl).
316 - some read types (regex, netstring, json, storable) could
317 errornously keep an AE::Handle object alive on their own
318 (reported by Mohammad Toossi).
319 - AnyEvent::Socket::resolve_sockaddr and all functions using it now
320 supports /etc/hosts overrides.
321 - add more workarounds around child watcher bugs in IO::Async
322 and (newly introduced) in POE.
323 - use glib child watchers and try to work around its limitations,
324 if possible (based on a patch by Kevin Ryde).
325 - add support for FLTK (fltk2), via AnyEvent::Impl::FLTK2.
326 - many read types in AnyEvent::Handle would malfunction during
327 transitions to TLS because of a "delete $self->{rbuf}".
328 (reported by Gerald Galster).
329 - improve condvar blocking wait performance (probably only noticable
330 with faster backends), also make it easier for event loops that
331 do not support blocking waits.
332 - PERL_ANYEVENT_MODEL now also allows full module names, although
333 the usefulness of this feature is not yet clear.
334 - add tests for all supported event loops, conditional on PERL_ANYEVENT_LOOP_TESTS.
335 - add t/80_ssltext.t.
336 - no longer use AUTOLOAD for the initial loading (for no very good reason).
337 - tcp_server now tries to unlink unix domain sockets when it gets destroyed.
338 - do not run event loop detection multiple times when programs call their cached
339 copies of AnyEvent::detect.
340 - parse_hostport now accepts absolute paths als unix domain sockets.
341 - suppress the idiotic warning inside POE by patching the POE::Kernel
342 run flags instead of asking the user to jump through contortions.
343 - add Task::Weaken as dependency (as recommends only at this time).
344 - replace PF_UNSPEC by 0, for good old minix.
345
3466.0 (not officially released, but was on CPAN for about 15 minutes :/)
347
3485.34 Thu May 12 09:59:41 CEST 2011
349 - AE::TLS didn't support both cert_file and key at the same time,
350 due to some copy&paste bug (reported and analyzed by vti).
351 - improve AE::Handle timeout documentation slightly.
352
3535.33 Mon Feb 7 21:51:41 CET 2011
354 - never appeared on CPAN, for unknown reasons.
355 - AnyEvent::Handle on_connect_error - do not try to destroy a handle
356 object when it's already gone (analyzed by Ryan Bullock).
357 - update idna algorithm(s) to UTS#46 version 6.0.0.
358 - preserve trailing dot in idn_to_ascii - testsuite says to preserve,
359 uts46 says it's an error, and testsuite makes more sense.
360 - remove leading dots in idn_to_ascii - testsuite says to remove,
361 uts46 says it's an error, and testsuite makes more sense.
362 - fixed a bug in idn ignored character processing.
363 - passes the uts46 6.0.0 testsuite except for three cases which
364 apparently are bugs in the testsuite itself.
365 - passes the uts46 6.0.1 testsuite except for eleven cases which
366 apparently are bugs in the testsuite itself.
367
3685.31 Sun Jan 23 11:49:19 CET 2011
369 - work around a POE bug where POE would first poll for new events and
370 invoke handlers BEFORE handling the message queue.
371 - fix parsing of "domain" statement in resolv.conf (Johannes Plunien).
372 - stop_read now does just that in TLS mode. see its
373 documentation for caveats.
374 - update to common::sense 3.4 warning set (some warnings in perl impose
375 an arbitrary 32kb limit on regex matches, which affects
376 AnyEvent::Handle).
377 - AE::Handle had a broken rbuf_max method and missing wbuf_max method.
378 - new experimental AE::Handle::resettls method.
379
3805.3 Fri Dec 31 05:48:13 CET 2010
381 - major IOAsync patch by Paul Evans - 0.33 is required, as the major
382 issues with IO::Async have been solved in that version, including
383 the availabilty of a default loop!
384 - implement new wbuf_max parameter for AnyEvent::Handle.
385 - added Cocoa::EventLoop backend by Daisuke Murase.
386 - the default testsuite now honors PERL_ANYEVENT_MODEL (Paul Evans).
387 - clarify handle on_prepare documentation.
388 - set CLOEXEC flag on parent-side fd's in AnyEvent::Util::run_cmd,
389 to avoid leaking them to any child processes.
390 - try to force files into the arch-dependend lib directory, for the
391 half-"turly obsessive compulsive person"(s) that depend on it.
392
3935.29 Sun Dec 5 10:49:21 CET 2010
394 - convert EV backend to EV 4.00 API (so better upgrade EV too).
395 - AE::Handle: implement dynamic read_size adjustment, add
396 max_read_size parameter, reduce default read_size to 2048.
397 - add some FAQ entries for servers and/or on_eof vs. on_error.
398 - work around OS bugs (cygwin again) when lingering in AE::Handle.
399 - reduce memory usage slightly when connect was used in AE::Handle.
400 - wrok around more segfaults in Tk.
401 - document tls_ctx => undef as valid.
402 - detect WNOHANG value at compiletime.
403 - include some probably linux-only support for building AnyEvent
404 as part of the perl core.
405 - improve compatibility of t/08_idna.t to perl 5.8.
406 - make AnyEvent::Strict truly optional (in case somebody builds a
407 minimal perl...).
408
4095.28 Wed Oct 13 04:14:23 CEST 2010
410 - due to a glitch, AnyEvent's internal getprotobyname wasn't
411 actually used in tcp_server.
412 - implement AnyEvent::CondVar->new method, which was documented but
413 missing.
414 - support autoloading of read types in unshift_read.
415 - AnyEvent::Strict: do the test for invalid fh arguments differently,
416 resulting in more informative error messages.
417 - new AnyEvent::FAQ manpage.
418 - clarify that you can actually call push_read etc. in on_read.
419 - put keys and certs in /tmp, not into the current directory,
420 and document this in the SECURITY CONSIDERATIONS sections
421 of AnyEvent::TLS.
422 - support ";" as resolv.conf comment character in AnyEvent::DNS, as
423 requested by Ron Isaacson.
424 - document $AnyEvent::DNS::RESOLVER variable.
425 - incorporated major typo patches by Abhijit Menon-Sen.
426 - AnyEvent::Handle now croaks when negative timeouts are passed.
427 - add a shitload of TCP_xxx sockopt name constants to AnyEvent::Util,
428 undocumented, but who knows when they might come in handy.
429
4305.271 Tue Jun 8 12:05:46 CEST 2010
431 - backport to perl 5.8.x.
432
4335.27 Sun Jun 6 12:12:05 CEST 2010
434 - postpone differently in AnyEvent::Socket now, as
435 when not, canceling the connection attempt might fail
436 (found by Felix Antonius Wilhelm Ostmann).
437 - explicitly check for non-stream sockets in AE::Handle, too many
438 clueless people fell into the trap of this somehow working.
439 - simplified and reworked the "OTHER MODULES" section.
440 - better/more condvar examples.
441
4425.261 Wed Apr 28 16:13:36 CEST 2010
443 - AF_INET6 was not properly used from Socket6 during configuration
444 time (found by Andy Grundman).
445
4465.26 Mon Apr 12 04:49:35 CEST 2010
447 - don't generate spurious readyness notifications when select
448 returns EINTR in the pure perl backend (analysed and testcase
449 by Bas Denissen).
450 - give same examples in SYNOPSIS sections of AnyEvent and AE.
451 - provider faster implementations for AE::time/now/now_update for
452 EV and Perl backends.
453 - provide faster AE::cv and AE::time implementations for all
454 backends.
455 - no longer support lower/mixed-case signal names (this was
456 never documented nor universally supported).
457 - some more memory µ-optimisations, and somewhat less messy
458 code to cope with both AnyEvent and AE APIs.
459
4605.251 Sat Mar 13 00:58:21 CET 2010
461 - make test failed if it was run before make install on perl <=
462 5.8.8, and older versions of perl make our live very difficult,
463 so write constants.pl during Makefile.PL time.
464
4655.25 Sat Mar 13 00:23:14 CET 2010
466 - fix a race condition in AnyEvent::Handle that would cause
467 a "bio_read: unsupported method" error (in _another_
468 ssl connection) after user code threw an exception.
469 - added AnyEvent::Handle->destroyed method.
470 - speed up AnyEvent::Socket::format_address and ::format_ipv6.
471 - the AnyEvent::Util::fh_nonblocking function worked only
472 by lucky accident on win32.
473 - smaller and faster AnyEvent::Util::fh_nonblocking.
474 - when the (required!) Time::HiRes module is missing, AnyEvent
475 did not fall back to built-in time properly.
476 - do not load Fcntl at runtime, saving memory and loading time.
477 - precompile a number of constants and use them instead of runtime
478 detection and eval.
479 - free detection code after detection and similar memory optimisations.
480 - Perl backend timer interval best effort drift has been improved
481 (same algorithm as EV).
482 - update unicode idna mapping table.
483
4845.24 Tue Jan 5 11:39:43 CET 2010
485 - cygwin never reports errors from failed connects. 1.5 just gives
486 you continous readyness and EAGAIN, 1.7 is even more broken and
487 just hangs. work around both issues in a cygwin-specific hack.
488 - improve idle watcher documentation slightly.
489
4905.23 Sun Dec 20 23:48:00 CET 2009
491 - support IDNs in resolve_sockaddr, and therefore in tcp_connect.
492 - implement punycode_encode/decode, idn_nameprep,
493 idn_to_ascii and idn_to_unicode operations in AnyEvent::Util.
494 - provide $AE::VERSION.
495 - removed traces of "no strict 'refs'".
496
4975.22 Sat Dec 5 03:51:13 CET 2009
498 - downgrade-or-fail in AnyEvent::Handle::push_write, to
499 diagnose encoding failures earlier and more succinctly.
500 (this works around bugs in perl, throwing away encoding info
501 when passing scalar data to extensions).
502 - add more examples to AnyEvent::Socket manpage.
503 - upgrade internal warning set to the same as common::sense 2.03.
504 - use pack "n/a*" for pre-5.8.9 perl compatibility in AnyEvent::DNS
505 (John Beppu).
506 - AnyEvent::Socket::inet_aton now properly supports ipv6, as documented.
507 - add google public dns servers to fallback server set and make sure
508 we load-balance properly between the three sets. also add all
509 fallback dns servers, not just a random one, to each dns config.
510
5115.21 Thu Nov 19 02:48:47 CET 2009
512 - fix a problem where socket constants were called with parameters
513 (spotted by David Friedland).
514 - fork_call never use'd POSIX (reported by Daisuke Maki).
515 - improve perl 5.6 compatibility further (but it still won't work
516 unless you rip out everything but the core).
517 - prefer Net::DNS::Resolver over ipconfig, if installed, on win32.
518 uses 10MB of RAM, but doesn't open a console window. *sigh*.
519
5205.202 Wed Oct 14 22:35:44 CEST 2009
521 - AnyEvent::DNS would unexpectedly clobber $_ under windows
522 (analysed by Matthias Waldorf).
523 - AnyEvent::Handle::run_cmd can now pass the PID of the
524 newly-created process, which is much less useful than it might
525 sound (based on patch by Yann Kerherve).
526
5275.201 Tue Sep 29 12:09:25 CEST 2009
528 - AnyEvent:Handle::on_starttls/on_stoptls methods were broken
529 (reported by Torsten Foertsch).
530 - common::sense 2.0 could cause tcp_server to throw an exception
531 (analysed by elmex).
532
5335.2 Mon Sep 14 07:04:49 CEST 2009
534 - INCOMPATIBLE CHANGE: do no longer support register_read_type
535 and register_write_type in AnyEvent::Handle, instead support
536 package names (the facility was mostly abused).
537 - implement "packagename-as-read/write type" support in
538 AnyEvent::Handle.
539 - AnyEvent::Handle: new options "keepalive" and "oobinline".
540 - oobinline set by default to avoid security issues.
541 - the pure-perl event loop backend wrongly detected times() fallback
542 support (spotted by Pavel Boldin).
543
5445.12 Tue Sep 1 20:26:50 CEST 2009
545 - be more lenient when parsing resolv.conf files, as some people
546 use hashmarks as comment indicator inside directives (reported
547 by Michael S. Fischer).
548 - use same set of warnings as common::sense 2.0.
549 - fix a potential 32 bit overflow issue due to perl having problems
550 with large hex constants in the Perl backend.
551 - do not use the slower 5.6 version fo the enc_name code in AnyEvent::DNS
552 by default.
553 - fix some prototypes.
554 - (optionally) require Async::Interrupt 1.02.
555 - replace opendns fallbacks by verizon/level3 public dns servers.
556
5575.112 Fri Aug 21 13:59:04 CEST 2009
558 - AnyEvent::Handle could cause hard crashes in Net::SSLeay when
559 using starttls with an illegal/uncreatable tls context.
560 - port to broken win32 perls.
561
5625.111 Sun Aug 16 18:44:36 CEST 2009
563 - perl errornously requires sockaddr_un structures returned by the
564 kernel to have a specific length. work around this bug by padding
565 sockaddr_un structures (found to fail on OS X, as perl only uses
566 the correct code for GNU/Linux, but likely to fail on other
567 OSes as well).
568
5695.11 Wed Aug 12 17:49:37 CEST 2009
570 - fixed a bug in run_cmd with input from filename.
571 - create a json coder object when none is given in write direction
572 to reduce dependency on JSON.
573
5745.1 Tue Aug 11 03:17:32 CEST 2009
575 - tcp_connect now reports the correct error in $!, not an unrelated
576 one (was broken since 4.91).
577 - AE::Handle did not properly clear rtimeout/wtimeout watchers
578 when retrying a tcp_connect.
579 - new functions: AnyEvent::Util::run_cmd and close_all_fds_except.
580 - fixed wrong documentation in AE::Socket.
581
5825.01 Mon Aug 10 03:16:32 CEST 2009
583 - last release broke EV child watchers :/.
584
5855.0 Sun Aug 9 17:32:09 CEST 2009
586 - officially introduce and document the AE API.
587 - lowering the timeout at runtime did not have immediate effect
588 in AnyEvent::Handle.
589 - AE::Handle now has separate and independent read and write timeout
590 settings.
591 - probe for Guard module when AnyEvent::Util is loaded, not at
592 runtime (Event doesn't like eval "use" from callbacks, and it avoids
593 exporting the probe function).
594 - Impl::IOAsync's io watchers did not use the filehandle emulation
595 code to work around it's watcher limits.
596 - work around IO::Async's problems with overloaded objects.
597 - try to disable wearnings inside IO::Async, it's just too buggy.
598 - do no longer use Event by default, only EV or Perl, Event is too buggy.
599 - hackishly provide fast AE:: interface when EV or Perl is the backend.
600 - inherit VERSION from AnyEvent::VERSION in most modules.
601
6024.91 Thu Aug 6 15:42:45 CEST 2009
603 - AE::Handle::starttls could get out of sync when the read buffer
604 already contains some TLS handshake.
605 - AE::Handle did not properly free the TLS session, causing bigger
606 memory leaks in Net::SSLeay than expected (reported by toaster).
607 - AE::Socket::tcp_connect will now artificially delay invoking
608 the callback to avoid returning after invoking the callback.
609 - convert many internal watcher uses to AE API.
610
6114.9 Sat Aug 1 11:07:01 CEST 2009
612 - Glib silently fails when registering a timeout with
613 a negative value, so avoid that.
614 - call condvar callback immediately when it is set
615 after the condition is already signalled.
616 - check rbuf_max condition only after trying to consume
617 data.
618 - normalise signal numbers to names when using the ae's
619 signal handling, but do not document this (yet).
620 - pure perl signal emulation did not properly set nonblocking
621 mode on the signal pipe on win32 (this is unlikely to have caused
622 any issues).
623 - new module: AnyEvent::Debug.
624 - AnyEvent::Strict now checks that a signal was specified
625 by name (not name or number).
626 - reduce memory footprint in typical cases by ~50kb by
627 only compiling signal code when necessary.
628 - add AnyEvent::Handle->rbuf_max.
629 - grab the AE namespace for future new API, implement stubs for
630 said future/alternative API.
631 - new function: AnyEvent::Socket::format_hostport.
2 632
34.881 Tue Jul 28 12:51:53 CEST 2009 6334.881 Tue Jul 28 12:51:53 CEST 2009
4 - work around a bug in local in pre-5.10 perls, causing 634 - work around a bug in local in pre-5.10 perls, causing
5 AnyEvent::Handle to recurse when it should not 635 AnyEvent::Handle to recurse when it should not
6 (analyzed by elmex). 636 (analysed by elmex).
7 637
84.88 Tue Jul 28 04:04:37 CEST 2009 6384.88 Tue Jul 28 04:04:37 CEST 2009
9 - re-bless the handle into a dummy package after calling 639 - re-bless the handle into a dummy package after calling
10 AnyEvent::Handle::destroy, so the user does not need to check 640 AnyEvent::Handle::destroy, so the user does not need to check
11 for errors after every push_write etc. 641 for errors after every push_write etc.
… …
99 in the common case. 729 in the common case.
100 - add AnyEvent::Handle->push_shutdown method. 730 - add AnyEvent::Handle->push_shutdown method.
101 - add an additional error message parameter to AnyEvent::Handle's 731 - add an additional error message parameter to AnyEvent::Handle's
102 on_error callback (for TLS, $! is still available). 732 on_error callback (for TLS, $! is still available).
103 - add AnyEvent::Handle on_starttls/on_stoptls callbacks. 733 - add AnyEvent::Handle on_starttls/on_stoptls callbacks.
104 - make AnyEvent::Handle more robust against early conenction 734 - make AnyEvent::Handle more robust against early connection
105 failures (during new), and return C<undef> in that case 735 failures (during new), and return C<undef> in that case
106 from the constructor. 736 from the constructor.
107 - AnyEvent::Handle will now try to load only JSON::XS first, 737 - AnyEvent::Handle will now try to load only JSON::XS first,
108 then fall back to JSON. 738 then fall back to JSON.
109 - format_ipv4/format_ipv6 are now exported by default, for symmetry, 739 - format_ipv4/format_ipv6 are now exported by default, for symmetry,
… …
2334.3 Fri Oct 3 09:18:43 CEST 2008 8634.3 Fri Oct 3 09:18:43 CEST 2008
234 - AnyEvent will now install a no-op signal handler for SIGPIPE, 864 - AnyEvent will now install a no-op signal handler for SIGPIPE,
235 unless one has been installed already. 865 unless one has been installed already.
236 - warn about not loaidng AnyEvent::Impl::POE early enough 866 - warn about not loaidng AnyEvent::Impl::POE early enough
237 (patch by Adam Rosenstein). 867 (patch by Adam Rosenstein).
238 - fixed a great number of bugs and corner cases in AnyEvent::Handle: 868 - fixed a great number of bugs and corner cases in AnyEvent::Handle.
239 - fix a bug in where in SSL connect mode, the 869 - fix a bug in where in SSL connect mode, the
240 client would first wait for some data by the server and 870 client would first wait for some data by the server and
241 otherwise hang. (reported and analyzed in an absolutely 871 otherwise hang. (reported and analysed in an absolutely
242 exemplary manner by Adam Rosenstein). 872 exemplary manner by Adam Rosenstein).
243 - fix a bug in where SSL EOF would not be treated 873 - fix a bug in where SSL EOF would not be treated
244 as stream EOF, putting the connection into a hung state. 874 as stream EOF, putting the connection into a hung state.
245 (reported and analyzed in an absolutely exemplary manner by Adam 875 (reported and analysed in an absolutely exemplary manner by Adam
246 Rosenstein). 876 Rosenstein).
247 - fix a potential segfault when the TLS context 877 - fix a potential segfault when the TLS context
248 would go missing in a read callback (e.g. due to stoptls) - 878 would go missing in a read callback (e.g. due to stoptls) -
249 Net::SSLeay of course makes no type checking whatsoever. 879 Net::SSLeay of course makes no type checking whatsoever.
250 - AnyEvent::Handle will not stop the read watcher in TLS mode. 880 - AnyEvent::Handle will not stop the read watcher in TLS mode.

Diff Legend

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