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

Comparing AnyEvent/Changes (file contents):
Revision 1.237 by root, Wed May 13 15:19:43 2009 UTC vs.
Revision 1.616 by root, Fri Nov 11 01:24:20 2016 UTC

1Revision history for Perl extension AnyEvent. 1Revision history for Perl extension AnyEvent.
2 2
34.41 3TODO: document TCP_*** constants
4 - AnyEvent::DNS: add the "dname" resource record name 4TODO: add debug/trace logging to some modules
5 for cosmetic reasons. 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 - AnyEvent::Handle->keepalive was documented (and defined) twice
19 (found by Matt Merhar).
20
217.13 Sat Sep 17 04:31:49 CEST 2016
22 - Only call tlsext_host_name for non-empty common names (reported
23 by Maxime Soulé).
24 - log a (single) notice message if SNI is not supported.
25 - upgrade to UTS-46:9.0.0 draft and switch to non-transitional
26 behaviour, beating thunderbird, ie, edge, chrome and safari to it :)
27 (see also https://bugzilla.mozilla.org/show_bug.cgi?id=1218179)
28 - turns out the UTS-46 IDNA testcase failures were indeed bugs in the
29 testcases and the specification and not in the code - the post-9.0.0
30 unicode files have all known problems fixed, so finally the AnyEvent
31 IDNA implementation can pass the full IDNA testsuite - without needing
32 a single fix :)
33 - guarantee (and document) that condvar callbacks will be removed
34 on invocation - important to avoid circular references.
35
367.12 Wed Jan 27 19:12:26 CET 2016
37 - use common name as hostname for tls connects, if Net::SSLeay
38 supports SNI.
39 - fix documentation of tls_autostart read type in AnyEvent::Handle,
40 analyzed by Felix Ostmann.
41
427.11 Thu Jul 16 14:36:00 CEST 2015
43 - AnyEvent::Socket::parse_ipv6 could accept malformed ipv6
44 addresses (extra "::" at end and similar cases).
45 - add a more explicit warning to AnyEvent::Handle that it doesn't
46 work on files, people keep getting confused.
47 - new function AnyEvent::Socket::tcp_bind.
48 - new functions AnyEvent::fh_block and AnyEvent::fh_unblock.
49 - aligned ipv6 address formatting with RFC 5952 (by not shortening
50 a single :0: to ::).
51 - added stability canary support.
52
537.09 Sat May 2 16:38:53 CEST 2015
54 - AnyEvent::Debug called an internal function (AnyEvent::Log::ft)
55 that was renamed to AnyEvent::Log:format_time. uinder its old name
56 (analyzed by Michael Stovenour).
57 - update AnyEvent::DNS fallback resolver addresses:
58 seems google effectively killed most other free dns resolvers,
59 so remove them, but add cable and wireless (ecrc) since it was
60 stable for 20 years or so, official or not, and there should be
61 an alternative to google.
62 - perl5porters broke windows error codes in 5.20, and mapped
63 WSAEWOULDBLOCK on the (different) EWOULDBLOCK error code, and
64 WSAEINPROGRESS into the incompatible ERINPROGRESS code, probably
65 because they were so cool. They probably broke other error codes
66 for no reason, too, but I didn't care for checking, it's too
67 depressing. This version only works around the WSAEWOULDBLOCK
68 issue, because I don't have a nice way to work around the
69 WSAEINPROGRESS bug.
70
717.08 Wed Dec 10 05:27:17 CET 2014
72 - work around a newly introduced bug in Socket 2.011 (an
73 errornous sun_length check) (analyzed by Maxime Soulé).
74 - AnyEvent::TLS didn't load (but refer to) AnyEvent::Socket
75 (analyzed by Ben Magistro).
76 - AnyEvent::Strict will now confess, not croak. This is in line with
77 it being a development/debugging tool.
78 - work around a number of libglib bugs (debug builds of libglib enforce
79 certain undocumented behaviour patterns such as not being able to
80 remove a child watch source after it has fired, which we will try
81 to emulate to avoid "criticals". what where they thinking...).
82 - mention json security issues in AnyEvent::Handle, now that Douglas
83 Crockford has foolishly and incompatibly changed JSON.
84 - changed default dns resolver "max_outstanding" value from 1 to 10,
85 the latter beinfg the intended value all along
86 (reported by Ilya Chesnokov).
87 - added new "AnyEvent::Impl::UV" interface module to the UV event lib
88 (written by Mike Lowell).
89
907.07 Tue Dec 17 17:45:02 CET 2013
91 - the documentation for custom tls verify schemes was wrong. make it agree
92 with the code (reported by Maxime Soulé).
93 - added cbor read and write types to AnyEvent::Handle (using CBOR::XS).
94 - work around an API change in openssl that could cause wrong tls connection
95 aborts, likely on windows only (analyzed by sten).
96 - calling AnyEvent->now_update with AnyEvent::Impl::Perl caused an
97 endless loop (reported by Dietrich Rebmann).
98 - add tlsv1_1 and tlsv1_2 protocols to AnyEvent::TLS
99 (patch by Maxime Soulé).
100 - document AnyEvent::Impl::IOAsync::set_loop and
101 $AnyEvent::Impl::IOAsync::LOOP. Though only documented now, this
102 functionality has _always_ been available.
103 - force a toplevel domain name in t/81_hosts.t (analyzed by
104 David Jack Wange Olrik).
105 - document that AnyEvent::Log uses AnyEvent::IO.
106 - warn about AnyEvent::Filesys::Notify performance.
107 - praise the joys of AnyEvent::Fork::*.
108 - time for an =encoding directive.
109 - do no longer use JSON to create a default json coder, use
110 JSON::XS or JSON::PP directly.
111
1127.05 Wed Aug 21 10:38:08 CEST 2013
113 - uts46data.pl couldn't be found due to wrong naming of the file
114 (reported by Fulko Hew).
115 - handle lone \015's properly in AE::Handle's default line read
116 (reported by various people).
117 - untaint ip addresses found in /etc/hosts (patch by José Micó).
118 - the memleak fix in 7.03 caused resolving via /etc/hosts to always
119 fail on first use (reported and testcase by Andrew Whatson).
120 - expose AnyEvent::Log::format_time, and allow users to redefine it.
121 - expose AnyEvent::Log::default_format, and allow redefinition.
122 - expose AnyEvent::Log::fatal_exit, to allow redefinition.
123 - AnyEvent::Debug shell can now run coro shell commands, if available.
124 - t/63* tests were wrongly in MANIFEST.
125 - kernel.org's finger server went MIA, switch to freebsd.org and
126 icculus.org.
127 - clarify that IO::AIO and AnyEvent::AIO are needed for AnyEvent::IO
128 to function asynchronously (reported by Daniel Carosone).
129 - hard-disable $^W in most tests, it generates too much garbage output.
130 - use a (hopefully) more future-proof method to emulate common::sense.
131 - upgrade to UTS-46:6.2.0.
132 - switch to INSTLIB from INSTLIBDIR, as INSTLIBDIR was wrongly
133 documented. should not affect anything.
134
1357.04 Thu Nov 15 02:23:58 CET 2012
136 - debugging code left in.
137
1387.03 Thu Nov 15 02:16:11 CET 2012
139 - AnyEvent::Socket::inet_aton did not work when dns resolution was
140 used to find the addresses (analyzed and patch by Maxime Soulé).
141 - fix a memory leak in the /etc/hosts lookup code when hosts don't
142 resolve and are not in hosts (analyzed by sten).
143
1447.02 Tue Aug 14 04:27:58 CEST 2012
145 - AnyEvent::Util::run_cmd could block indefinitely (analyzed and test
146 program by Yorhel).
147 - verified that AnyEvent::Socket follows RFC5952.
148 - try to parse "ADDR#PORT" in addition to "ADDR PORT".
149
1507.01 Sun May 13 01:03:17 CEST 2012
151 - fail with EPROTO in AnyEvent::Handle wqhen TLS is requested but
152 not available, instead of throwing an exception.
153 - use File::Spec to get the tmpdir in t/*, to avoid needless
154 failures on (most, not mine :) windows boxes.
155 - new handle read types: tls_detect and tls_autostart.
156
1577.0 Fri Apr 13 06:33:30 CEST 2012
158 - child watchers are broken in POE 1.352 (also many earlier
159 versions) and there seems to be no way to work around it, as POE
160 itself is inherently racy. Document this regression and add a delay
161 in t/68_poe_03_child.t for the time being.
162 - new module AnyEvent::IO, that is a frontend to either a pure-perl
163 synchronous I/O implementation (AnyEvent::IO::Perl), or to the
164 asynchronous I/O extension IO::AIO (AnyEvent::IO::IOAIO).
165 - load /etc/hosts only when DNS has no answer.
166 - stat /etc/hosts on every access and reload it if it changed.
167 - load /etc/hosts via AnyEvent::IO - potentially asynchronous.
168 - fix a buggy croak in the dh parameter reading in AnyEvent::TLS.
169 - AnyEvent::Log log_to_file and log_to_path now use AnyEvent::IO. As
170 a side effect, they now use true appending as opposed to libc
171 appending, and the file might not have been opened when the function
172 returns.
173 - the default logging level was not properly documented in a variety of
174 places, this has hopefully been rectified.
175 - updated uts46data.pl for unicode 6.1.0.
176 - made log messages generated by AnyEvent submodules not
177 include the package name anymore, as it will be logged
178 by default already.
179 - upgrade to the trick used by common-sense 3.5 to work
180 around extra warning torture/breakage under perl 5.15.x.
181 - log messages by anyevent are now ucfirst, are usually full
182 sentences and do no longer include the package name.
183 - the storable read type would sometimes throw an exception instead
184 of causing EBADMSG (analyzed by Maxime Soulé).
185 - work around a bug in openssl 1.0.1 which enforces a minimum rsa
186 keysize (reported by Paul Howarth).
187 - documented Rocco fabricating statements about AnyEvent and me.
188
1896.14 Tue Jan 31 20:00:24 CET 2012
190 - AnyEvent::Impl::Tk was broken due to a mysterious "die" inside,
191 probably an editing mistake (reported by Darin McBride).
192
1936.13 Thu Jan 12 07:27:01 CET 2012
194 - AnyEvent::Util::fork_call checked for POSIX availability in the wrong way,
195 causing it to fail if POSIX wasn't loaded already (analysed by Rock Power).
196 - AnyEvent::Handle::push_read (line => did pass $1 directly, so
197 regex-matching inside the callback would change the parameter, despite
198 $1 being dynamically scoped per-block. this perl bug is now being
199 worked around (testcase by Cindy Wang).
200
2016.12 Mon Dec 12 13:21:10 CET 2011
202 - $! was clobbered by subsequent calls in tcp_connect, due to
203 the postpone not saving/restoring it (analyzed by Richard Garnish).
204
2056.11 Tue Nov 22 10:36:05 CET 2011
206 - Tk cannot create windows when tainted unless you set a title, so
207 set a dummy title for AnyEvent's dummy mainwindow. How dummy.
208 - escape any nonprintable/nonascii characters when stringifying
209 backtraces.
210 - log the reason loading an interface module fails to load at level debug.
211
2126.1 Tue Oct 4 19:44:30 CEST 2011
213 - INCOMPATIBLE CHANGE: the default log level is now "4" (error
214 and above), and some messages inside AnyEvent have been elevated
215 to higher log levels to print by default.
216 - AnyEvent::log, unlike AnyEvent::Log::log, did not always exit on fatal.
217 - work around yet another signal race bug, newly introduced by POE. My
218 how that module sucks.
219 - add google public ipv6 dns as fallback servers, so AE::DNS can finally
220 support ipv6-only operation even with defaults (also use ::1 instead
221 of 127.0.0.1).
222 - give better guidelines on how to chose a logging level.
223 - AnyEvent::Log can now cap (limit) log message priority by context.
224 - AE::log will now save and restore $! and $@, so callers don't have to.
225 - AE::Strict now checks that registered fds are still valid at regular
226 intervals.
227 - skip frequently called 'require' calls if the module appears loaded.
228 - increase signal latency test timer considerably for poor little osx,
229 which can't be bothered with basic posix functionality such as signal
230 delivery.
231 - mention some debugging aids in the tutorial.
232 - new env variable PERL_ANYEVENT_MAX_SIGNAL_LATENCY.
233 - private (%namespace) log contexts in PERL_ANYEVENT_LOG now have the name
234 of the context as title.
235 - AnyEvent::Strict would not restore read-write access to $_ when a callback
236 died - this is a user bug, but since AE::Strict exists to help dbeugging...
237 - detect the extremely evil IO::Async::Loop::AnyEvent module and refuse
238 to work with it - the author is well aware of the issues but refuses
239 to fix them. This extreme step has been taken because the module actively
240 attacks the goals of AnyEvent and makes it harder for module authors to use
241 AnyEvent.
242
2436.02 Fri Aug 26 20:08:31 CEST 2011
244 - log_to_syslog now accepts facility strings.
245 - log_to_syslog would not allow facilities - "$facility|$level"
246 *literally* means that string. sheesh.
247 - the CPAN parser got confused and thought we provide Sys::Syslog, try
248 to work around.
249 - renamed AnyEvent::Impl::FLTK2 to FLTK.
250
2516.01 Fri Aug 26 07:04:11 CEST 2011
252 - INCOMPATIBLE CHANGE: AnyEvent::DNS resource records now include
253 the TTL as fourth element - this affects ->resolve, ->request
254 and ->dns_unpack indirectly. Convenience functions such as
255 AnyEvent::DNS::{a,txt,mx,srv} are not affected, neither is
256 AnyEvent::Socket.
257 - INCOMPATIBLE CHANGE: backend authors now should not implement
258 one_event or loop, but instead the AnyEvent::CondVar::_wait and _poll
259 methods.
260 - INCOMPATIBLE CHANGE: verbosity levels have been redefined, see
261 AnyEvent (PERL_ANYEVENT_VERBOSE) and AnyEvent::Log for details.
262 - DEPRECIATION: please change
263 "use AnyEvent::Impl::Perl" to "use AnyEvent::Loop" and
264 "AnyEvent::Impl::Perl::loop" to "AnyEvent::Loop::run"
265 in your programs, for future compatibility.
266 - *blush* idle watcher emulation was borked.
267 - AnyEvent::Strict now checks the AE::xxx functions as well.
268 - querying too long domain names would barf deep inside AnyEvent::DNS - since
269 this is hard to avoid for applications, AnyEvent:.DNS::request now checks for
270 for too long domain names explicitly and passes an undef to the callback
271 instead (reported by James Bromberger).
272 - make AnyEvent::Util more robust against EINTR (analyzed by Andrew Suffield).
273 - new functions: AnyEvent::postpone, AnyEvent::log.
274 - new module: AnyEvent::Log, for simple logging needs.
275 - new env variable, PERL_ANYEVENT_LOG, for fine-grained logging config.
276 - new env variable, PERL_ANYEVENT_DEBUG_SHELL, to automatically start
277 a debug shell.
278 - new env variable, PERL_ANYEVENT_DEBUG_WRAP, to automatically enable
279 watcher instrumentation/wrapping (see AnyEvent::Debug::wrap).
280 - new env variable, PERL_ANYEVENT_HOSTS, to override the place where
281 /etc/hosts can be found.
282 - an empty PERL_ANYEVENT_RESOLV_CONF now selects the default
283 configuration.
284 - env variables can now be specified using AE_xyz instead of
285 PERL_ANYEVENT_xyz, subject to some rules.
286 - lots of new AnyEvent::Debug functionality.
287 - AnyEvent::Strict now checks for modifications of $_ by the callback
288 (by setting it to readonly while executing the callback, if possible).
289 - IO::Async sometimes cancels the wrong timer when you call cancel_timer.
290 Invest extra overhead to work around this peculiar behaviour.
291 - split the pure perl backend into a loop module and an impl
292 module (AnyEvent::Loop and AnyEvent::Impl::Perl).
293 - some read types (regex, netstring, json, storable) could
294 errornously keep an AE::Handle object alive on their own
295 (reported by Mohammad Toossi).
296 - AnyEvent::Socket::resolve_sockaddr and all functions using it now
297 supports /etc/hosts overrides.
298 - add more workarounds around child watcher bugs in IO::Async
299 and (newly introduced) in POE.
300 - use glib child watchers and try to work around its limitations,
301 if possible (based on a patch by Kevin Ryde).
302 - add support for FLTK (fltk2), via AnyEvent::Impl::FLTK2.
303 - many read types in AnyEvent::Handle would malfunction during
304 transitions to TLS because of a "delete $self->{rbuf}".
305 (reported by Gerald Galster).
306 - improve condvar blocking wait performance (probably only noticable
307 with faster backends), also make it easier for event loops that
308 do not support blocking waits.
309 - PERL_ANYEVENT_MODEL now also allows full module names, although
310 the usefulness of this feature is not yet clear.
311 - add tests for all supported event loops, conditional on PERL_ANYEVENT_LOOP_TESTS.
312 - add t/80_ssltext.t.
313 - no longer use AUTOLOAD for the initial loading (for no very good reason).
314 - tcp_server now tries to unlink unix domain sockets when it gets destroyed.
315 - do not run event loop detection multiple times when programs call their cached
316 copies of AnyEvent::detect.
317 - parse_hostport now accepts absolute paths als unix domain sockets.
318 - suppress the idiotic warning inside POE by patching the POE::Kernel
319 run flags instead of asking the user to jump through contortions.
320 - add Task::Weaken as dependency (as recommends only at this time).
321 - replace PF_UNSPEC by 0, for good old minix.
322
3236.0 (not officially released, but was on CPAN for about 15 minutes :/)
324
3255.34 Thu May 12 09:59:41 CEST 2011
326 - AE::TLS didn't support both cert_file and key at the same time,
327 due to some copy&paste bug (reported and analyzed by vti).
328 - improve AE::Handle timeout documentation slightly.
329
3305.33 Mon Feb 7 21:51:41 CET 2011
331 - never appeared on CPAN, for unknown reasons.
332 - AnyEvent::Handle on_connect_error - do not try to destroy a handle
333 object when it's already gone (analyzed by Ryan Bullock).
334 - update idna algorithm(s) to UTS#46 version 6.0.0.
335 - preserve trailing dot in idn_to_ascii - testsuite says to preserve,
336 uts46 says it's an error, and testsuite makes more sense.
337 - remove leading dots in idn_to_ascii - testsuite says to remove,
338 uts46 says it's an error, and testsuite makes more sense.
339 - fixed a bug in idn ignored character processing.
340 - passes the uts46 6.0.0 testsuite except for three cases which
341 apparently are bugs in the testsuite itself.
342 - passes the uts46 6.0.1 testsuite except for eleven cases which
343 apparently are bugs in the testsuite itself.
344
3455.31 Sun Jan 23 11:49:19 CET 2011
346 - work around a POE bug where POE would first poll for new events and
347 invoke handlers BEFORE handling the message queue.
348 - fix parsing of "domain" statement in resolv.conf (Johannes Plunien).
349 - stop_read now does just that in TLS mode. see its
350 documentation for caveats.
351 - update to common::sense 3.4 warning set (some warnings in perl impose
352 an arbitrary 32kb limit on regex matches, which affects
353 AnyEvent::Handle).
354 - AE::Handle had a broken rbuf_max method and missing wbuf_max method.
355 - new experimental AE::Handle::resettls method.
356
3575.3 Fri Dec 31 05:48:13 CET 2010
358 - major IOAsync patch by Paul Evans - 0.33 is required, as the major
359 issues with IO::Async have been solved in that version, including
360 the availabilty of a default loop!
361 - implement new wbuf_max parameter for AnyEvent::Handle.
362 - added Cocoa::EventLoop backend by Daisuke Murase.
363 - the default testsuite now honors PERL_ANYEVENT_MODEL (Paul Evans).
364 - clarify handle on_prepare documentation.
365 - set CLOEXEC flag on parent-side fd's in AnyEvent::Util::run_cmd,
366 to avoid leaking them to any child processes.
367 - try to force files into the arch-dependend lib directory, for the
368 half-"turly obsessive compulsive person"(s) that depend on it.
369
3705.29 Sun Dec 5 10:49:21 CET 2010
371 - convert EV backend to EV 4.00 API (so better upgrade EV too).
372 - AE::Handle: implement dynamic read_size adjustment, add
373 max_read_size parameter, reduce default read_size to 2048.
374 - add some FAQ entries for servers and/or on_eof vs. on_error.
375 - work around OS bugs (cygwin again) when lingering in AE::Handle.
376 - reduce memory usage slightly when connect was used in AE::Handle.
377 - wrok around more segfaults in Tk.
378 - document tls_ctx => undef as valid.
379 - detect WNOHANG value at compiletime.
380 - include some probably linux-only support for building AnyEvent
381 as part of the perl core.
382 - improve compatibility of t/08_idna.t to perl 5.8.
383 - make AnyEvent::Strict truly optional (in case somebody builds a
384 minimal perl...).
385
3865.28 Wed Oct 13 04:14:23 CEST 2010
387 - due to a glitch, AnyEvent's internal getprotobyname wasn't
388 actually used in tcp_server.
389 - implement AnyEvent::CondVar->new method, which was documented but
390 missing.
391 - support autoloading of read types in unshift_read.
392 - AnyEvent::Strict: do the test for invalid fh arguments differently,
393 resulting in more informative error messages.
394 - new AnyEvent::FAQ manpage.
395 - clarify that you can actually call push_read etc. in on_read.
396 - put keys and certs in /tmp, not into the current directory,
397 and document this in the SECURITY CONSIDERATIONS sections
398 of AnyEvent::TLS.
399 - support ";" as resolv.conf comment character in AnyEvent::DNS, as
400 requested by Ron Isaacson.
401 - document $AnyEvent::DNS::RESOLVER variable.
402 - incorporated major typo patches by Abhijit Menon-Sen.
403 - AnyEvent::Handle now croaks when negative timeouts are passed.
404 - add a shitload of TCP_xxx sockopt name constants to AnyEvent::Util,
405 undocumented, but who knows when they might come in handy.
406
4075.271 Tue Jun 8 12:05:46 CEST 2010
408 - backport to perl 5.8.x.
409
4105.27 Sun Jun 6 12:12:05 CEST 2010
411 - postpone differently in AnyEvent::Socket now, as
412 when not, canceling the connection attempt might fail
413 (found by Felix Antonius Wilhelm Ostmann).
414 - explicitly check for non-stream sockets in AE::Handle, too many
415 clueless people fell into the trap of this somehow working.
416 - simplified and reworked the "OTHER MODULES" section.
417 - better/more condvar examples.
418
4195.261 Wed Apr 28 16:13:36 CEST 2010
420 - AF_INET6 was not properly used from Socket6 during configuration
421 time (found by Andy Grundman).
422
4235.26 Mon Apr 12 04:49:35 CEST 2010
424 - don't generate spurious readyness notifications when select
425 returns EINTR in the pure perl backend (analysed and testcase
426 by Bas Denissen).
427 - give same examples in SYNOPSIS sections of AnyEvent and AE.
428 - provider faster implementations for AE::time/now/now_update for
429 EV and Perl backends.
430 - provide faster AE::cv and AE::time implementations for all
431 backends.
432 - no longer support lower/mixed-case signal names (this was
433 never documented nor universally supported).
434 - some more memory µ-optimisations, and somewhat less messy
435 code to cope with both AnyEvent and AE APIs.
436
4375.251 Sat Mar 13 00:58:21 CET 2010
438 - make test failed if it was run before make install on perl <=
439 5.8.8, and older versions of perl make our live very difficult,
440 so write constants.pl during Makefile.PL time.
441
4425.25 Sat Mar 13 00:23:14 CET 2010
443 - fix a race condition in AnyEvent::Handle that would cause
444 a "bio_read: unsupported method" error (in _another_
445 ssl connection) after user code threw an exception.
446 - added AnyEvent::Handle->destroyed method.
447 - speed up AnyEvent::Socket::format_address and ::format_ipv6.
448 - the AnyEvent::Util::fh_nonblocking function worked only
449 by lucky accident on win32.
450 - smaller and faster AnyEvent::Util::fh_nonblocking.
451 - when the (required!) Time::HiRes module is missing, AnyEvent
452 did not fall back to built-in time properly.
453 - do not load Fcntl at runtime, saving memory and loading time.
454 - precompile a number of constants and use them instead of runtime
455 detection and eval.
456 - free detection code after detection and similar memory optimisations.
457 - Perl backend timer interval best effort drift has been improved
458 (same algorithm as EV).
459 - update unicode idna mapping table.
460
4615.24 Tue Jan 5 11:39:43 CET 2010
462 - cygwin never reports errors from failed connects. 1.5 just gives
463 you continous readyness and EAGAIN, 1.7 is even more broken and
464 just hangs. work around both issues in a cygwin-specific hack.
465 - improve idle watcher documentation slightly.
466
4675.23 Sun Dec 20 23:48:00 CET 2009
468 - support IDNs in resolve_sockaddr, and therefore in tcp_connect.
469 - implement punycode_encode/decode, idn_nameprep,
470 idn_to_ascii and idn_to_unicode operations in AnyEvent::Util.
471 - provide $AE::VERSION.
472 - removed traces of "no strict 'refs'".
473
4745.22 Sat Dec 5 03:51:13 CET 2009
475 - downgrade-or-fail in AnyEvent::Handle::push_write, to
476 diagnose encoding failures earlier and more succinctly.
477 (this works around bugs in perl, throwing away encoding info
478 when passing scalar data to extensions).
479 - add more examples to AnyEvent::Socket manpage.
480 - upgrade internal warning set to the same as common::sense 2.03.
481 - use pack "n/a*" for pre-5.8.9 perl compatibility in AnyEvent::DNS
482 (John Beppu).
483 - AnyEvent::Socket::inet_aton now properly supports ipv6, as documented.
484 - add google public dns servers to fallback server set and make sure
485 we load-balance properly between the three sets. also add all
486 fallback dns servers, not just a random one, to each dns config.
487
4885.21 Thu Nov 19 02:48:47 CET 2009
489 - fix a problem where socket constants were called with parameters
490 (spotted by David Friedland).
491 - fork_call never use'd POSIX (reported by Daisuke Maki).
492 - improve perl 5.6 compatibility further (but it still won't work
493 unless you rip out everything but the core).
494 - prefer Net::DNS::Resolver over ipconfig, if installed, on win32.
495 uses 10MB of RAM, but doesn't open a console window. *sigh*.
496
4975.202 Wed Oct 14 22:35:44 CEST 2009
498 - AnyEvent::DNS would unexpectedly clobber $_ under windows
499 (analysed by Matthias Waldorf).
500 - AnyEvent::Handle::run_cmd can now pass the PID of the
501 newly-created process, which is much less useful than it might
502 sound (based on patch by Yann Kerherve).
503
5045.201 Tue Sep 29 12:09:25 CEST 2009
505 - AnyEvent:Handle::on_starttls/on_stoptls methods were broken
506 (reported by Torsten Foertsch).
507 - common::sense 2.0 could cause tcp_server to throw an exception
508 (analysed by elmex).
509
5105.2 Mon Sep 14 07:04:49 CEST 2009
511 - INCOMPATIBLE CHANGE: do no longer support register_read_type
512 and register_write_type in AnyEvent::Handle, instead support
513 package names (the facility was mostly abused).
514 - implement "packagename-as-read/write type" support in
515 AnyEvent::Handle.
516 - AnyEvent::Handle: new options "keepalive" and "oobinline".
517 - oobinline set by default to avoid security issues.
518 - the pure-perl event loop backend wrongly detected times() fallback
519 support (spotted by Pavel Boldin).
520
5215.12 Tue Sep 1 20:26:50 CEST 2009
522 - be more lenient when parsing resolv.conf files, as some people
523 use hashmarks as comment indicator inside directives (reported
524 by Michael S. Fischer).
525 - use same set of warnings as common::sense 2.0.
526 - fix a potential 32 bit overflow issue due to perl having problems
527 with large hex constants in the Perl backend.
528 - do not use the slower 5.6 version fo the enc_name code in AnyEvent::DNS
529 by default.
530 - fix some prototypes.
531 - (optionally) require Async::Interrupt 1.02.
532 - replace opendns fallbacks by verizon/level3 public dns servers.
533
5345.112 Fri Aug 21 13:59:04 CEST 2009
535 - AnyEvent::Handle could cause hard crashes in Net::SSLeay when
536 using starttls with an illegal/uncreatable tls context.
537 - port to broken win32 perls.
538
5395.111 Sun Aug 16 18:44:36 CEST 2009
540 - perl errornously requires sockaddr_un structures returned by the
541 kernel to have a specific length. work around this bug by padding
542 sockaddr_un structures (found to fail on OS X, as perl only uses
543 the correct code for GNU/Linux, but likely to fail on other
544 OSes as well).
545
5465.11 Wed Aug 12 17:49:37 CEST 2009
547 - fixed a bug in run_cmd with input from filename.
548 - create a json coder object when none is given in write direction
549 to reduce dependency on JSON.
550
5515.1 Tue Aug 11 03:17:32 CEST 2009
552 - tcp_connect now reports the correct error in $!, not an unrelated
553 one (was broken since 4.91).
554 - AE::Handle did not properly clear rtimeout/wtimeout watchers
555 when retrying a tcp_connect.
556 - new functions: AnyEvent::Util::run_cmd and close_all_fds_except.
557 - fixed wrong documentation in AE::Socket.
558
5595.01 Mon Aug 10 03:16:32 CEST 2009
560 - last release broke EV child watchers :/.
561
5625.0 Sun Aug 9 17:32:09 CEST 2009
563 - officially introduce and document the AE API.
564 - lowering the timeout at runtime did not have immediate effect
565 in AnyEvent::Handle.
566 - AE::Handle now has separate and independent read and write timeout
567 settings.
568 - probe for Guard module when AnyEvent::Util is loaded, not at
569 runtime (Event doesn't like eval "use" from callbacks, and it avoids
570 exporting the probe function).
571 - Impl::IOAsync's io watchers did not use the filehandle emulation
572 code to work around it's watcher limits.
573 - work around IO::Async's problems with overloaded objects.
574 - try to disable wearnings inside IO::Async, it's just too buggy.
575 - do no longer use Event by default, only EV or Perl, Event is too buggy.
576 - hackishly provide fast AE:: interface when EV or Perl is the backend.
577 - inherit VERSION from AnyEvent::VERSION in most modules.
578
5794.91 Thu Aug 6 15:42:45 CEST 2009
580 - AE::Handle::starttls could get out of sync when the read buffer
581 already contains some TLS handshake.
582 - AE::Handle did not properly free the TLS session, causing bigger
583 memory leaks in Net::SSLeay than expected (reported by toaster).
584 - AE::Socket::tcp_connect will now artificially delay invoking
585 the callback to avoid returning after invoking the callback.
586 - convert many internal watcher uses to AE API.
587
5884.9 Sat Aug 1 11:07:01 CEST 2009
589 - Glib silently fails when registering a timeout with
590 a negative value, so avoid that.
591 - call condvar callback immediately when it is set
592 after the condition is already signalled.
593 - check rbuf_max condition only after trying to consume
594 data.
595 - normalise signal numbers to names when using the ae's
596 signal handling, but do not document this (yet).
597 - pure perl signal emulation did not properly set nonblocking
598 mode on the signal pipe on win32 (this is unlikely to have caused
599 any issues).
600 - new module: AnyEvent::Debug.
601 - AnyEvent::Strict now checks that a signal was specified
602 by name (not name or number).
603 - reduce memory footprint in typical cases by ~50kb by
604 only compiling signal code when necessary.
605 - add AnyEvent::Handle->rbuf_max.
606 - grab the AE namespace for future new API, implement stubs for
607 said future/alternative API.
608 - new function: AnyEvent::Socket::format_hostport.
609
6104.881 Tue Jul 28 12:51:53 CEST 2009
611 - work around a bug in local in pre-5.10 perls, causing
612 AnyEvent::Handle to recurse when it should not
613 (analysed by elmex).
614
6154.88 Tue Jul 28 04:04:37 CEST 2009
616 - re-bless the handle into a dummy package after calling
617 AnyEvent::Handle::destroy, so the user does not need to check
618 for errors after every push_write etc.
619 - do not attempt to run t/02_signals.t on obviously broken
620 platforms.
621
6224.87 Sun Jul 26 02:06:16 CEST 2009
623 - do not attempt to linger when there is no longer a valid fh.
624 - enforce tls mode to be either accept or connect, do not simply
625 segfault in Net::SSLeay.
626 - AnyEvent::Handle can now call tcp_connect itself (new parameters
627 connect, on_prepare, on_connect and on_connect_error). Updated
628 tutorial accordingly.
629 - add AnyEvent::Impl::Irssi backend.
630
6314.86 Mon Jul 20 23:52:29 CEST 2009
632 - since the verbose warning is not enough, explicitly document
633 that versions before 1.33 of Net::SSLeay are not secure.
634 - work around signal handling races in Event and (...) Event::Lib.
635 - try to align signal-race timer to full-second boundaries.
636 - work around Tk not liking negative timeouts.
637 - don't complain of different grades of environmental unfriendlyness
638 in IO::Async.
639
6404.85 Sat Jul 18 06:16:14 CEST 2009
641 - nail the signal race problem in perl once and for all
642 (see $AnyEvent::MAX_SIGNAL_LATENCY).
643 - take advantage of Async::Interrupt if it is available.
644 - load Time::HiRes and Guard modules on demand only.
645 - add optional/recommended modules section to AnyEvent
646 documentation.
647 - reduce memory usage considerably (and reduce startup penalty)
648 by not using "strict", "warnings" and "overload" modules.
649 - work around buggy windows/openbsd perls and provide EBADMSG
650 and EPROTO ourselves when missing.
651 - improve perl 5.6 compatibility of the core event loop.
652 - made Net::SSLeay version 1.33 a soft requirement.
653
6544.83 Fri Jul 17 16:56:26 CEST 2009
655 - implement AnyEvent::Socket::getprotobyname.
656 - AnyEvent::CondVar's will now detect recursive blocking
657 waits and will croak, as too many people fall into
658 this trap.
659 - AnyEvent::Handle will now call ->destroy on itself after
660 executing the on_error callback, instead of doing some
661 half-baked internal shutdown, for fatal errors.
662 - clarify on_eof behaviour w.r.t. the read queue and
663 on_read callbacks.
664 - ignore some possible spurious wake-ups in tcp_connect.
665
6664.82 Sat Jul 11 00:34:55 CEST 2009
667 - POE and Event backends didn't accept some callable objects as
668 callbacks.
669 - use Config module instead of POSIX module to detect signal names
670 in AnyEvent::Strict and AnyEvent::Impl::EventLib, as the POSIX
671 module doesn't even have all POSIX signals :/.
672 - use more workarounds around the many refcnt/corruption bugs in
673 Event::Lib.
674 - work around a race condition in perl's select, causing t/03_child.t
675 to rarely fail.
676
6774.81 Thu Jul 9 10:30:30 CEST 2009
678 - AnyEvent::Handle didn't properly diagnose write errors
679 (it expected -1 from syswrite, how lame... :).
680 - support file descriptors in addition to file handles
681 in AnyEvent->io.
682 - new env variables: PERL_ANYEVENT_RESOLV_CONF,
683 PERL_ANYEVENT_MAX_OUTSTANDING_DNS, PERL_ANYEVENT_CA_FILE
684 and PERL_ANYEVENT_CA_PATH.
685 - provide a sensible synopsis section for AnyEvent::TLS.
686 - add a "supported backends" section to the manpage.
687 - added simple io watcher test to testsuite, using a
688 portable_socketpair.
689 - tried to improve the stability of the Event::Lib backend,
690 YMMV.
691
6924.8 Mon Jul 6 23:45:16 CEST 2009
693 - AnyEvent::DNS did not properly follow CNAME records with
694 uppercase targets.
695 - AnyEvent::DNS would errornously return AAAA records
696 with v4 mapped addresses (a faulty record) as ipv4 addresses,
697 causing AnyEvent::Socket to throw an exception.
698 - added new module AnyEvent::TLS for easier SSL/TLS context
699 creation, with many options including hostname verification,
700 secure default configuration, lots of documentation and,
701 predefined diffie-hellman keys for perfect forward security
702 and much more. get it while it's still fresh!
703 - use AnyEvent::TLS in AnyEvent::Handle for context management.
704 - load AnyEvent::Handle only on demand in AnyEvent::DNS,
705 so AnyEvent::Socket users have smaller memory footprint
706 in the common case.
707 - add AnyEvent::Handle->push_shutdown method.
708 - add an additional error message parameter to AnyEvent::Handle's
709 on_error callback (for TLS, $! is still available).
710 - add AnyEvent::Handle on_starttls/on_stoptls callbacks.
711 - make AnyEvent::Handle more robust against early connection
712 failures (during new), and return C<undef> in that case
713 from the constructor.
714 - AnyEvent::Handle will now try to load only JSON::XS first,
715 then fall back to JSON.
716 - format_ipv4/format_ipv6 are now exported by default, for symmetry,
717 and because it was documented that way.
718
7194.451 Fri Jul 3 00:28:58 CEST 2009
720 - do not clear rbuf when shutting down an AnyEvent::Handle
721 object - doing so breaks AnyEvent::HTTP.
722
7234.45 Mon Jun 29 22:59:26 CEST 2009
724 - a write error could cause AnyEvent::Handle to create
725 an I/O watcher with an undefined $fh.
726 - special-case mapped ipv4 addresses in both
727 AnyEvent::Socket::format_address and parse_address,
728 to treat them just like ipv4 addresses.
729 - updated and overhauled the AnyEvent::Intro doc.
730 - implement AnyEvent::Socket::format_ipv4/ipv6.
731 - slightly speed up portable_pipe/socketpair.
732 - expand condvar begin/end documentation.
733
7344.42 Fri Jun 26 08:32:18 CEST 2009
735 - preliminary and neccesarily incomplete support for IO::Async.
736 - reset SIGCHLD to DEFAULT when AnyEvent is loaded, in
737 case it was set to IGNORE, to ensure we can catch child statuses
738 even when the calling env acts stupidly.
739 - updated benchmarks with IO::Async, which performs very well.
740
7414.412 Wed Jun 24 01:35:57 CEST 2009
742 - support an "untaint" attribute for AnyEvent::DNS
743 and set it on the default resolver.
744 - implement and document AnyEvent::Impl::Perl::loop.
745 - remove all anyevent-env variables from %ENV
746 when running in tainted mode.
747 - mention and extend the IO::Lambda benchmark.
748
7494.411 Sun Jun 7 18:48:02 CEST 2009
750 - do not try to use F_SETFD on windows.
751
7524.41 Thu May 14 06:40:11 CEST 2009
6 - work around issues in older perls (5.8.0?) when 753 - work around issues in older perls (5.8.0?) when
7 a signal handler is deleted from the %SIG hash. 754 a signal handler is deleted from the %SIG hash.
8 - use POSIX::_exit in child test, to avoid running 755 - use POSIX::_exit in child test, to avoid running
9 destructors. 756 destructors.
10 - speed up CHLD handling by relying on SIGCHLD being 757 - speed up CHLD handling by relying on SIGCHLD being
11 synchronously delivered, even when we roll our own 758 synchronously delivered, even when we roll our own
12 implementation. 759 implementation.
760 - AnyEvent::DNS: add the "dname" resource record name
761 for cosmetic reasons.
13 762
144.4 Sun Apr 26 20:12:33 CEST 2009 7634.4 Sun Apr 26 20:12:33 CEST 2009
15 - implemented idle watchers, where applicable. 764 - implemented idle watchers, where applicable.
16 - AnyEvent->time died when Event backend was in use. 765 - AnyEvent->time died when Event backend was in use.
17 - fix a memleak in the Tk backend. 766 - fix a memleak in the Tk backend.
18 - sped up Tk timer handling. 767 - sped up Tk timer handling.
19 - clip negative "after" values to 0 in AnyEvent::Impl::Event 768 - clip negative "after" values to 0 in AnyEvent::Impl::Event
20 to avoid spamming out warning messages. 769 to avoid spamming out warning messages.
21 - fix Qt timers without interval. 770 - fix Qt timers without interval.
22 - avoid Qt zero-timeout specialcase, in old libqt's. 771 - avoid Qt zero-timeout special case, in old libqt's.
23 772
244.352 Mon Apr 20 16:31:11 CEST 2009 7734.352 Mon Apr 20 16:31:11 CEST 2009
25 - fix AnyEvent::Strict error messages for child watchers. 774 - fix AnyEvent::Strict error messages for child watchers.
26 - fix/update Makefile.PL "recommends" versions. 775 - fix/update Makefile.PL "recommends" versions.
27 - add AnyEvent->now_update. 776 - add AnyEvent->now_update.
… …
914.3 Fri Oct 3 09:18:43 CEST 2008 8404.3 Fri Oct 3 09:18:43 CEST 2008
92 - AnyEvent will now install a no-op signal handler for SIGPIPE, 841 - AnyEvent will now install a no-op signal handler for SIGPIPE,
93 unless one has been installed already. 842 unless one has been installed already.
94 - warn about not loaidng AnyEvent::Impl::POE early enough 843 - warn about not loaidng AnyEvent::Impl::POE early enough
95 (patch by Adam Rosenstein). 844 (patch by Adam Rosenstein).
96 - fixed a great number of bugs and corner cases in AnyEvent::Handle: 845 - fixed a great number of bugs and corner cases in AnyEvent::Handle.
97 - fix a bug in where in SSL connect mode, the 846 - fix a bug in where in SSL connect mode, the
98 client would first wait for some data by the server and 847 client would first wait for some data by the server and
99 otherwise hang. (reported and analyzed in an absolutely 848 otherwise hang. (reported and analysed in an absolutely
100 exemplary manner by Adam Rosenstein). 849 exemplary manner by Adam Rosenstein).
101 - fix a bug in where SSL EOF would not be treated 850 - fix a bug in where SSL EOF would not be treated
102 as stream EOF, putting the connection into a hung state. 851 as stream EOF, putting the connection into a hung state.
103 (reported and analyzed in an absolutely exemplary manner by Adam 852 (reported and analysed in an absolutely exemplary manner by Adam
104 Rosenstein). 853 Rosenstein).
105 - fix a potential segfault when the TLS context 854 - fix a potential segfault when the TLS context
106 would go missing in a read callback (e.g. due to stoptls) - 855 would go missing in a read callback (e.g. due to stoptls) -
107 Net::SSLeay of course makes no type checking whatsoever. 856 Net::SSLeay of course makes no type checking whatsoever.
108 - AnyEvent::Handle will not stop the read watcher in TLS mode. 857 - AnyEvent::Handle will not stop the read watcher in TLS mode.

Diff Legend

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