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

Comparing AnyEvent/Changes (file contents):
Revision 1.249 by root, Fri Jun 26 06:33:17 2009 UTC vs.
Revision 1.624 by root, Sat Jul 1 15:48:51 2017 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
15 - AnyEvent::Socket::parse_ipv6 would NOT, as advertised, accept ipv4
16 addresses. It now does and converts them to ipv4 mapped addresses.
17
187.14 Sat Jun 24 01:44:19 CEST 2017
19 - fix a crash bug in AnyEvent::Handle with openssl 1.1.0 (patched
20 by Paul Howarth and apparently tracked down by Mike McCauley).
21 - AnyEvent::Handle->keepalive was documented (and defined) twice
22 (found by Matt Merhar).
23 - AnyEvent::Socket::tcp_bind/tcp_server would immediately unlink a unix
24 listening socket unless a guard is used. Change this so that no
25 cleanup will be performed unless a guard is used and document this more
26 clearly.
27 - make tcp_bind/tcp_server error messages more regular.
28 - work around latest perl upstream breakage - this time, perl5porters were
29 apparently bullied by a bogus CVE (CVE-2016-1238), and since distros
30 stupidly apply anything that has "security fix" stamped on it, it's likely
31 going to be a problem in practise (and working around it is trivial on
32 windows and unix).
33 - add TCP_FASTOPEN/MSG_FASTOPEN and MSG_NOSIGNAL/MSG_MORE/MSG_DONTWAIT
34 to constants.
35 - update warnings to common::sense 3.74 standards.
36
377.13 Sat Sep 17 04:31:49 CEST 2016
38 - Only call tlsext_host_name for non-empty common names (reported
39 by Maxime Soulé).
40 - log a (single) notice message if SNI is not supported.
41 - upgrade to UTS-46:9.0.0 draft and switch to non-transitional
42 behaviour, beating thunderbird, ie, edge, chrome and safari to it :)
43 (see also https://bugzilla.mozilla.org/show_bug.cgi?id=1218179)
44 - turns out the UTS-46 IDNA testcase failures were indeed bugs in the
45 testcases and the specification and not in the code - the post-9.0.0
46 unicode files have all known problems fixed, so finally the AnyEvent
47 IDNA implementation can pass the full IDNA testsuite - without needing
48 a single fix :)
49 - guarantee (and document) that condvar callbacks will be removed
50 on invocation - important to avoid circular references.
51
527.12 Wed Jan 27 19:12:26 CET 2016
53 - use common name as hostname for tls connects, if Net::SSLeay
54 supports SNI.
55 - fix documentation of tls_autostart read type in AnyEvent::Handle,
56 analyzed by Felix Ostmann.
57
587.11 Thu Jul 16 14:36:00 CEST 2015
59 - AnyEvent::Socket::parse_ipv6 could accept malformed ipv6
60 addresses (extra "::" at end and similar cases).
61 - add a more explicit warning to AnyEvent::Handle that it doesn't
62 work on files, people keep getting confused.
63 - new function AnyEvent::Socket::tcp_bind.
64 - new functions AnyEvent::fh_block and AnyEvent::fh_unblock.
65 - aligned ipv6 address formatting with RFC 5952 (by not shortening
66 a single :0: to ::).
67 - added stability canary support.
68
697.09 Sat May 2 16:38:53 CEST 2015
70 - AnyEvent::Debug called an internal function (AnyEvent::Log::ft)
71 that was renamed to AnyEvent::Log:format_time. uinder its old name
72 (analyzed by Michael Stovenour).
73 - update AnyEvent::DNS fallback resolver addresses:
74 seems google effectively killed most other free dns resolvers,
75 so remove them, but add cable and wireless (ecrc) since it was
76 stable for 20 years or so, official or not, and there should be
77 an alternative to google.
78 - perl5porters broke windows error codes in 5.20, and mapped
79 WSAEWOULDBLOCK on the (different) EWOULDBLOCK error code, and
80 WSAEINPROGRESS into the incompatible ERINPROGRESS code, probably
81 because they were so cool. They probably broke other error codes
82 for no reason, too, but I didn't care for checking, it's too
83 depressing. This version only works around the WSAEWOULDBLOCK
84 issue, because I don't have a nice way to work around the
85 WSAEINPROGRESS bug.
86
877.08 Wed Dec 10 05:27:17 CET 2014
88 - work around a newly introduced bug in Socket 2.011 (an
89 errornous sun_length check) (analyzed by Maxime Soulé).
90 - AnyEvent::TLS didn't load (but refer to) AnyEvent::Socket
91 (analyzed by Ben Magistro).
92 - AnyEvent::Strict will now confess, not croak. This is in line with
93 it being a development/debugging tool.
94 - work around a number of libglib bugs (debug builds of libglib enforce
95 certain undocumented behaviour patterns such as not being able to
96 remove a child watch source after it has fired, which we will try
97 to emulate to avoid "criticals". what where they thinking...).
98 - mention json security issues in AnyEvent::Handle, now that Douglas
99 Crockford has foolishly and incompatibly changed JSON.
100 - changed default dns resolver "max_outstanding" value from 1 to 10,
101 the latter beinfg the intended value all along
102 (reported by Ilya Chesnokov).
103 - added new "AnyEvent::Impl::UV" interface module to the UV event lib
104 (written by Mike Lowell).
105
1067.07 Tue Dec 17 17:45:02 CET 2013
107 - the documentation for custom tls verify schemes was wrong. make it agree
108 with the code (reported by Maxime Soulé).
109 - added cbor read and write types to AnyEvent::Handle (using CBOR::XS).
110 - work around an API change in openssl that could cause wrong tls connection
111 aborts, likely on windows only (analyzed by sten).
112 - calling AnyEvent->now_update with AnyEvent::Impl::Perl caused an
113 endless loop (reported by Dietrich Rebmann).
114 - add tlsv1_1 and tlsv1_2 protocols to AnyEvent::TLS
115 (patch by Maxime Soulé).
116 - document AnyEvent::Impl::IOAsync::set_loop and
117 $AnyEvent::Impl::IOAsync::LOOP. Though only documented now, this
118 functionality has _always_ been available.
119 - force a toplevel domain name in t/81_hosts.t (analyzed by
120 David Jack Wange Olrik).
121 - document that AnyEvent::Log uses AnyEvent::IO.
122 - warn about AnyEvent::Filesys::Notify performance.
123 - praise the joys of AnyEvent::Fork::*.
124 - time for an =encoding directive.
125 - do no longer use JSON to create a default json coder, use
126 JSON::XS or JSON::PP directly.
127
1287.05 Wed Aug 21 10:38:08 CEST 2013
129 - uts46data.pl couldn't be found due to wrong naming of the file
130 (reported by Fulko Hew).
131 - handle lone \015's properly in AE::Handle's default line read
132 (reported by various people).
133 - untaint ip addresses found in /etc/hosts (patch by José Micó).
134 - the memleak fix in 7.03 caused resolving via /etc/hosts to always
135 fail on first use (reported and testcase by Andrew Whatson).
136 - expose AnyEvent::Log::format_time, and allow users to redefine it.
137 - expose AnyEvent::Log::default_format, and allow redefinition.
138 - expose AnyEvent::Log::fatal_exit, to allow redefinition.
139 - AnyEvent::Debug shell can now run coro shell commands, if available.
140 - t/63* tests were wrongly in MANIFEST.
141 - kernel.org's finger server went MIA, switch to freebsd.org and
142 icculus.org.
143 - clarify that IO::AIO and AnyEvent::AIO are needed for AnyEvent::IO
144 to function asynchronously (reported by Daniel Carosone).
145 - hard-disable $^W in most tests, it generates too much garbage output.
146 - use a (hopefully) more future-proof method to emulate common::sense.
147 - upgrade to UTS-46:6.2.0.
148 - switch to INSTLIB from INSTLIBDIR, as INSTLIBDIR was wrongly
149 documented. should not affect anything.
150
1517.04 Thu Nov 15 02:23:58 CET 2012
152 - debugging code left in.
153
1547.03 Thu Nov 15 02:16:11 CET 2012
155 - AnyEvent::Socket::inet_aton did not work when dns resolution was
156 used to find the addresses (analyzed and patch by Maxime Soulé).
157 - fix a memory leak in the /etc/hosts lookup code when hosts don't
158 resolve and are not in hosts (analyzed by sten).
159
1607.02 Tue Aug 14 04:27:58 CEST 2012
161 - AnyEvent::Util::run_cmd could block indefinitely (analyzed and test
162 program by Yorhel).
163 - verified that AnyEvent::Socket follows RFC5952.
164 - try to parse "ADDR#PORT" in addition to "ADDR PORT".
165
1667.01 Sun May 13 01:03:17 CEST 2012
167 - fail with EPROTO in AnyEvent::Handle wqhen TLS is requested but
168 not available, instead of throwing an exception.
169 - use File::Spec to get the tmpdir in t/*, to avoid needless
170 failures on (most, not mine :) windows boxes.
171 - new handle read types: tls_detect and tls_autostart.
172
1737.0 Fri Apr 13 06:33:30 CEST 2012
174 - child watchers are broken in POE 1.352 (also many earlier
175 versions) and there seems to be no way to work around it, as POE
176 itself is inherently racy. Document this regression and add a delay
177 in t/68_poe_03_child.t for the time being.
178 - new module AnyEvent::IO, that is a frontend to either a pure-perl
179 synchronous I/O implementation (AnyEvent::IO::Perl), or to the
180 asynchronous I/O extension IO::AIO (AnyEvent::IO::IOAIO).
181 - load /etc/hosts only when DNS has no answer.
182 - stat /etc/hosts on every access and reload it if it changed.
183 - load /etc/hosts via AnyEvent::IO - potentially asynchronous.
184 - fix a buggy croak in the dh parameter reading in AnyEvent::TLS.
185 - AnyEvent::Log log_to_file and log_to_path now use AnyEvent::IO. As
186 a side effect, they now use true appending as opposed to libc
187 appending, and the file might not have been opened when the function
188 returns.
189 - the default logging level was not properly documented in a variety of
190 places, this has hopefully been rectified.
191 - updated uts46data.pl for unicode 6.1.0.
192 - made log messages generated by AnyEvent submodules not
193 include the package name anymore, as it will be logged
194 by default already.
195 - upgrade to the trick used by common-sense 3.5 to work
196 around extra warning torture/breakage under perl 5.15.x.
197 - log messages by anyevent are now ucfirst, are usually full
198 sentences and do no longer include the package name.
199 - the storable read type would sometimes throw an exception instead
200 of causing EBADMSG (analyzed by Maxime Soulé).
201 - work around a bug in openssl 1.0.1 which enforces a minimum rsa
202 keysize (reported by Paul Howarth).
203 - documented Rocco fabricating statements about AnyEvent and me.
204
2056.14 Tue Jan 31 20:00:24 CET 2012
206 - AnyEvent::Impl::Tk was broken due to a mysterious "die" inside,
207 probably an editing mistake (reported by Darin McBride).
208
2096.13 Thu Jan 12 07:27:01 CET 2012
210 - AnyEvent::Util::fork_call checked for POSIX availability in the wrong way,
211 causing it to fail if POSIX wasn't loaded already (analysed by Rock Power).
212 - AnyEvent::Handle::push_read (line => did pass $1 directly, so
213 regex-matching inside the callback would change the parameter, despite
214 $1 being dynamically scoped per-block. this perl bug is now being
215 worked around (testcase by Cindy Wang).
216
2176.12 Mon Dec 12 13:21:10 CET 2011
218 - $! was clobbered by subsequent calls in tcp_connect, due to
219 the postpone not saving/restoring it (analyzed by Richard Garnish).
220
2216.11 Tue Nov 22 10:36:05 CET 2011
222 - Tk cannot create windows when tainted unless you set a title, so
223 set a dummy title for AnyEvent's dummy mainwindow. How dummy.
224 - escape any nonprintable/nonascii characters when stringifying
225 backtraces.
226 - log the reason loading an interface module fails to load at level debug.
227
2286.1 Tue Oct 4 19:44:30 CEST 2011
229 - INCOMPATIBLE CHANGE: the default log level is now "4" (error
230 and above), and some messages inside AnyEvent have been elevated
231 to higher log levels to print by default.
232 - AnyEvent::log, unlike AnyEvent::Log::log, did not always exit on fatal.
233 - work around yet another signal race bug, newly introduced by POE. My
234 how that module sucks.
235 - add google public ipv6 dns as fallback servers, so AE::DNS can finally
236 support ipv6-only operation even with defaults (also use ::1 instead
237 of 127.0.0.1).
238 - give better guidelines on how to chose a logging level.
239 - AnyEvent::Log can now cap (limit) log message priority by context.
240 - AE::log will now save and restore $! and $@, so callers don't have to.
241 - AE::Strict now checks that registered fds are still valid at regular
242 intervals.
243 - skip frequently called 'require' calls if the module appears loaded.
244 - increase signal latency test timer considerably for poor little osx,
245 which can't be bothered with basic posix functionality such as signal
246 delivery.
247 - mention some debugging aids in the tutorial.
248 - new env variable PERL_ANYEVENT_MAX_SIGNAL_LATENCY.
249 - private (%namespace) log contexts in PERL_ANYEVENT_LOG now have the name
250 of the context as title.
251 - AnyEvent::Strict would not restore read-write access to $_ when a callback
252 died - this is a user bug, but since AE::Strict exists to help dbeugging...
253 - detect the extremely evil IO::Async::Loop::AnyEvent module and refuse
254 to work with it - the author is well aware of the issues but refuses
255 to fix them. This extreme step has been taken because the module actively
256 attacks the goals of AnyEvent and makes it harder for module authors to use
257 AnyEvent.
258
2596.02 Fri Aug 26 20:08:31 CEST 2011
260 - log_to_syslog now accepts facility strings.
261 - log_to_syslog would not allow facilities - "$facility|$level"
262 *literally* means that string. sheesh.
263 - the CPAN parser got confused and thought we provide Sys::Syslog, try
264 to work around.
265 - renamed AnyEvent::Impl::FLTK2 to FLTK.
266
2676.01 Fri Aug 26 07:04:11 CEST 2011
268 - INCOMPATIBLE CHANGE: AnyEvent::DNS resource records now include
269 the TTL as fourth element - this affects ->resolve, ->request
270 and ->dns_unpack indirectly. Convenience functions such as
271 AnyEvent::DNS::{a,txt,mx,srv} are not affected, neither is
272 AnyEvent::Socket.
273 - INCOMPATIBLE CHANGE: backend authors now should not implement
274 one_event or loop, but instead the AnyEvent::CondVar::_wait and _poll
275 methods.
276 - INCOMPATIBLE CHANGE: verbosity levels have been redefined, see
277 AnyEvent (PERL_ANYEVENT_VERBOSE) and AnyEvent::Log for details.
278 - DEPRECIATION: please change
279 "use AnyEvent::Impl::Perl" to "use AnyEvent::Loop" and
280 "AnyEvent::Impl::Perl::loop" to "AnyEvent::Loop::run"
281 in your programs, for future compatibility.
282 - *blush* idle watcher emulation was borked.
283 - AnyEvent::Strict now checks the AE::xxx functions as well.
284 - querying too long domain names would barf deep inside AnyEvent::DNS - since
285 this is hard to avoid for applications, AnyEvent:.DNS::request now checks for
286 for too long domain names explicitly and passes an undef to the callback
287 instead (reported by James Bromberger).
288 - make AnyEvent::Util more robust against EINTR (analyzed by Andrew Suffield).
289 - new functions: AnyEvent::postpone, AnyEvent::log.
290 - new module: AnyEvent::Log, for simple logging needs.
291 - new env variable, PERL_ANYEVENT_LOG, for fine-grained logging config.
292 - new env variable, PERL_ANYEVENT_DEBUG_SHELL, to automatically start
293 a debug shell.
294 - new env variable, PERL_ANYEVENT_DEBUG_WRAP, to automatically enable
295 watcher instrumentation/wrapping (see AnyEvent::Debug::wrap).
296 - new env variable, PERL_ANYEVENT_HOSTS, to override the place where
297 /etc/hosts can be found.
298 - an empty PERL_ANYEVENT_RESOLV_CONF now selects the default
299 configuration.
300 - env variables can now be specified using AE_xyz instead of
301 PERL_ANYEVENT_xyz, subject to some rules.
302 - lots of new AnyEvent::Debug functionality.
303 - AnyEvent::Strict now checks for modifications of $_ by the callback
304 (by setting it to readonly while executing the callback, if possible).
305 - IO::Async sometimes cancels the wrong timer when you call cancel_timer.
306 Invest extra overhead to work around this peculiar behaviour.
307 - split the pure perl backend into a loop module and an impl
308 module (AnyEvent::Loop and AnyEvent::Impl::Perl).
309 - some read types (regex, netstring, json, storable) could
310 errornously keep an AE::Handle object alive on their own
311 (reported by Mohammad Toossi).
312 - AnyEvent::Socket::resolve_sockaddr and all functions using it now
313 supports /etc/hosts overrides.
314 - add more workarounds around child watcher bugs in IO::Async
315 and (newly introduced) in POE.
316 - use glib child watchers and try to work around its limitations,
317 if possible (based on a patch by Kevin Ryde).
318 - add support for FLTK (fltk2), via AnyEvent::Impl::FLTK2.
319 - many read types in AnyEvent::Handle would malfunction during
320 transitions to TLS because of a "delete $self->{rbuf}".
321 (reported by Gerald Galster).
322 - improve condvar blocking wait performance (probably only noticable
323 with faster backends), also make it easier for event loops that
324 do not support blocking waits.
325 - PERL_ANYEVENT_MODEL now also allows full module names, although
326 the usefulness of this feature is not yet clear.
327 - add tests for all supported event loops, conditional on PERL_ANYEVENT_LOOP_TESTS.
328 - add t/80_ssltext.t.
329 - no longer use AUTOLOAD for the initial loading (for no very good reason).
330 - tcp_server now tries to unlink unix domain sockets when it gets destroyed.
331 - do not run event loop detection multiple times when programs call their cached
332 copies of AnyEvent::detect.
333 - parse_hostport now accepts absolute paths als unix domain sockets.
334 - suppress the idiotic warning inside POE by patching the POE::Kernel
335 run flags instead of asking the user to jump through contortions.
336 - add Task::Weaken as dependency (as recommends only at this time).
337 - replace PF_UNSPEC by 0, for good old minix.
338
3396.0 (not officially released, but was on CPAN for about 15 minutes :/)
340
3415.34 Thu May 12 09:59:41 CEST 2011
342 - AE::TLS didn't support both cert_file and key at the same time,
343 due to some copy&paste bug (reported and analyzed by vti).
344 - improve AE::Handle timeout documentation slightly.
345
3465.33 Mon Feb 7 21:51:41 CET 2011
347 - never appeared on CPAN, for unknown reasons.
348 - AnyEvent::Handle on_connect_error - do not try to destroy a handle
349 object when it's already gone (analyzed by Ryan Bullock).
350 - update idna algorithm(s) to UTS#46 version 6.0.0.
351 - preserve trailing dot in idn_to_ascii - testsuite says to preserve,
352 uts46 says it's an error, and testsuite makes more sense.
353 - remove leading dots in idn_to_ascii - testsuite says to remove,
354 uts46 says it's an error, and testsuite makes more sense.
355 - fixed a bug in idn ignored character processing.
356 - passes the uts46 6.0.0 testsuite except for three cases which
357 apparently are bugs in the testsuite itself.
358 - passes the uts46 6.0.1 testsuite except for eleven cases which
359 apparently are bugs in the testsuite itself.
360
3615.31 Sun Jan 23 11:49:19 CET 2011
362 - work around a POE bug where POE would first poll for new events and
363 invoke handlers BEFORE handling the message queue.
364 - fix parsing of "domain" statement in resolv.conf (Johannes Plunien).
365 - stop_read now does just that in TLS mode. see its
366 documentation for caveats.
367 - update to common::sense 3.4 warning set (some warnings in perl impose
368 an arbitrary 32kb limit on regex matches, which affects
369 AnyEvent::Handle).
370 - AE::Handle had a broken rbuf_max method and missing wbuf_max method.
371 - new experimental AE::Handle::resettls method.
372
3735.3 Fri Dec 31 05:48:13 CET 2010
374 - major IOAsync patch by Paul Evans - 0.33 is required, as the major
375 issues with IO::Async have been solved in that version, including
376 the availabilty of a default loop!
377 - implement new wbuf_max parameter for AnyEvent::Handle.
378 - added Cocoa::EventLoop backend by Daisuke Murase.
379 - the default testsuite now honors PERL_ANYEVENT_MODEL (Paul Evans).
380 - clarify handle on_prepare documentation.
381 - set CLOEXEC flag on parent-side fd's in AnyEvent::Util::run_cmd,
382 to avoid leaking them to any child processes.
383 - try to force files into the arch-dependend lib directory, for the
384 half-"turly obsessive compulsive person"(s) that depend on it.
385
3865.29 Sun Dec 5 10:49:21 CET 2010
387 - convert EV backend to EV 4.00 API (so better upgrade EV too).
388 - AE::Handle: implement dynamic read_size adjustment, add
389 max_read_size parameter, reduce default read_size to 2048.
390 - add some FAQ entries for servers and/or on_eof vs. on_error.
391 - work around OS bugs (cygwin again) when lingering in AE::Handle.
392 - reduce memory usage slightly when connect was used in AE::Handle.
393 - wrok around more segfaults in Tk.
394 - document tls_ctx => undef as valid.
395 - detect WNOHANG value at compiletime.
396 - include some probably linux-only support for building AnyEvent
397 as part of the perl core.
398 - improve compatibility of t/08_idna.t to perl 5.8.
399 - make AnyEvent::Strict truly optional (in case somebody builds a
400 minimal perl...).
401
4025.28 Wed Oct 13 04:14:23 CEST 2010
403 - due to a glitch, AnyEvent's internal getprotobyname wasn't
404 actually used in tcp_server.
405 - implement AnyEvent::CondVar->new method, which was documented but
406 missing.
407 - support autoloading of read types in unshift_read.
408 - AnyEvent::Strict: do the test for invalid fh arguments differently,
409 resulting in more informative error messages.
410 - new AnyEvent::FAQ manpage.
411 - clarify that you can actually call push_read etc. in on_read.
412 - put keys and certs in /tmp, not into the current directory,
413 and document this in the SECURITY CONSIDERATIONS sections
414 of AnyEvent::TLS.
415 - support ";" as resolv.conf comment character in AnyEvent::DNS, as
416 requested by Ron Isaacson.
417 - document $AnyEvent::DNS::RESOLVER variable.
418 - incorporated major typo patches by Abhijit Menon-Sen.
419 - AnyEvent::Handle now croaks when negative timeouts are passed.
420 - add a shitload of TCP_xxx sockopt name constants to AnyEvent::Util,
421 undocumented, but who knows when they might come in handy.
422
4235.271 Tue Jun 8 12:05:46 CEST 2010
424 - backport to perl 5.8.x.
425
4265.27 Sun Jun 6 12:12:05 CEST 2010
427 - postpone differently in AnyEvent::Socket now, as
428 when not, canceling the connection attempt might fail
429 (found by Felix Antonius Wilhelm Ostmann).
430 - explicitly check for non-stream sockets in AE::Handle, too many
431 clueless people fell into the trap of this somehow working.
432 - simplified and reworked the "OTHER MODULES" section.
433 - better/more condvar examples.
434
4355.261 Wed Apr 28 16:13:36 CEST 2010
436 - AF_INET6 was not properly used from Socket6 during configuration
437 time (found by Andy Grundman).
438
4395.26 Mon Apr 12 04:49:35 CEST 2010
440 - don't generate spurious readyness notifications when select
441 returns EINTR in the pure perl backend (analysed and testcase
442 by Bas Denissen).
443 - give same examples in SYNOPSIS sections of AnyEvent and AE.
444 - provider faster implementations for AE::time/now/now_update for
445 EV and Perl backends.
446 - provide faster AE::cv and AE::time implementations for all
447 backends.
448 - no longer support lower/mixed-case signal names (this was
449 never documented nor universally supported).
450 - some more memory µ-optimisations, and somewhat less messy
451 code to cope with both AnyEvent and AE APIs.
452
4535.251 Sat Mar 13 00:58:21 CET 2010
454 - make test failed if it was run before make install on perl <=
455 5.8.8, and older versions of perl make our live very difficult,
456 so write constants.pl during Makefile.PL time.
457
4585.25 Sat Mar 13 00:23:14 CET 2010
459 - fix a race condition in AnyEvent::Handle that would cause
460 a "bio_read: unsupported method" error (in _another_
461 ssl connection) after user code threw an exception.
462 - added AnyEvent::Handle->destroyed method.
463 - speed up AnyEvent::Socket::format_address and ::format_ipv6.
464 - the AnyEvent::Util::fh_nonblocking function worked only
465 by lucky accident on win32.
466 - smaller and faster AnyEvent::Util::fh_nonblocking.
467 - when the (required!) Time::HiRes module is missing, AnyEvent
468 did not fall back to built-in time properly.
469 - do not load Fcntl at runtime, saving memory and loading time.
470 - precompile a number of constants and use them instead of runtime
471 detection and eval.
472 - free detection code after detection and similar memory optimisations.
473 - Perl backend timer interval best effort drift has been improved
474 (same algorithm as EV).
475 - update unicode idna mapping table.
476
4775.24 Tue Jan 5 11:39:43 CET 2010
478 - cygwin never reports errors from failed connects. 1.5 just gives
479 you continous readyness and EAGAIN, 1.7 is even more broken and
480 just hangs. work around both issues in a cygwin-specific hack.
481 - improve idle watcher documentation slightly.
482
4835.23 Sun Dec 20 23:48:00 CET 2009
484 - support IDNs in resolve_sockaddr, and therefore in tcp_connect.
485 - implement punycode_encode/decode, idn_nameprep,
486 idn_to_ascii and idn_to_unicode operations in AnyEvent::Util.
487 - provide $AE::VERSION.
488 - removed traces of "no strict 'refs'".
489
4905.22 Sat Dec 5 03:51:13 CET 2009
491 - downgrade-or-fail in AnyEvent::Handle::push_write, to
492 diagnose encoding failures earlier and more succinctly.
493 (this works around bugs in perl, throwing away encoding info
494 when passing scalar data to extensions).
495 - add more examples to AnyEvent::Socket manpage.
496 - upgrade internal warning set to the same as common::sense 2.03.
497 - use pack "n/a*" for pre-5.8.9 perl compatibility in AnyEvent::DNS
498 (John Beppu).
499 - AnyEvent::Socket::inet_aton now properly supports ipv6, as documented.
500 - add google public dns servers to fallback server set and make sure
501 we load-balance properly between the three sets. also add all
502 fallback dns servers, not just a random one, to each dns config.
503
5045.21 Thu Nov 19 02:48:47 CET 2009
505 - fix a problem where socket constants were called with parameters
506 (spotted by David Friedland).
507 - fork_call never use'd POSIX (reported by Daisuke Maki).
508 - improve perl 5.6 compatibility further (but it still won't work
509 unless you rip out everything but the core).
510 - prefer Net::DNS::Resolver over ipconfig, if installed, on win32.
511 uses 10MB of RAM, but doesn't open a console window. *sigh*.
512
5135.202 Wed Oct 14 22:35:44 CEST 2009
514 - AnyEvent::DNS would unexpectedly clobber $_ under windows
515 (analysed by Matthias Waldorf).
516 - AnyEvent::Handle::run_cmd can now pass the PID of the
517 newly-created process, which is much less useful than it might
518 sound (based on patch by Yann Kerherve).
519
5205.201 Tue Sep 29 12:09:25 CEST 2009
521 - AnyEvent:Handle::on_starttls/on_stoptls methods were broken
522 (reported by Torsten Foertsch).
523 - common::sense 2.0 could cause tcp_server to throw an exception
524 (analysed by elmex).
525
5265.2 Mon Sep 14 07:04:49 CEST 2009
527 - INCOMPATIBLE CHANGE: do no longer support register_read_type
528 and register_write_type in AnyEvent::Handle, instead support
529 package names (the facility was mostly abused).
530 - implement "packagename-as-read/write type" support in
531 AnyEvent::Handle.
532 - AnyEvent::Handle: new options "keepalive" and "oobinline".
533 - oobinline set by default to avoid security issues.
534 - the pure-perl event loop backend wrongly detected times() fallback
535 support (spotted by Pavel Boldin).
536
5375.12 Tue Sep 1 20:26:50 CEST 2009
538 - be more lenient when parsing resolv.conf files, as some people
539 use hashmarks as comment indicator inside directives (reported
540 by Michael S. Fischer).
541 - use same set of warnings as common::sense 2.0.
542 - fix a potential 32 bit overflow issue due to perl having problems
543 with large hex constants in the Perl backend.
544 - do not use the slower 5.6 version fo the enc_name code in AnyEvent::DNS
545 by default.
546 - fix some prototypes.
547 - (optionally) require Async::Interrupt 1.02.
548 - replace opendns fallbacks by verizon/level3 public dns servers.
549
5505.112 Fri Aug 21 13:59:04 CEST 2009
551 - AnyEvent::Handle could cause hard crashes in Net::SSLeay when
552 using starttls with an illegal/uncreatable tls context.
553 - port to broken win32 perls.
554
5555.111 Sun Aug 16 18:44:36 CEST 2009
556 - perl errornously requires sockaddr_un structures returned by the
557 kernel to have a specific length. work around this bug by padding
558 sockaddr_un structures (found to fail on OS X, as perl only uses
559 the correct code for GNU/Linux, but likely to fail on other
560 OSes as well).
561
5625.11 Wed Aug 12 17:49:37 CEST 2009
563 - fixed a bug in run_cmd with input from filename.
564 - create a json coder object when none is given in write direction
565 to reduce dependency on JSON.
566
5675.1 Tue Aug 11 03:17:32 CEST 2009
568 - tcp_connect now reports the correct error in $!, not an unrelated
569 one (was broken since 4.91).
570 - AE::Handle did not properly clear rtimeout/wtimeout watchers
571 when retrying a tcp_connect.
572 - new functions: AnyEvent::Util::run_cmd and close_all_fds_except.
573 - fixed wrong documentation in AE::Socket.
574
5755.01 Mon Aug 10 03:16:32 CEST 2009
576 - last release broke EV child watchers :/.
577
5785.0 Sun Aug 9 17:32:09 CEST 2009
579 - officially introduce and document the AE API.
580 - lowering the timeout at runtime did not have immediate effect
581 in AnyEvent::Handle.
582 - AE::Handle now has separate and independent read and write timeout
583 settings.
584 - probe for Guard module when AnyEvent::Util is loaded, not at
585 runtime (Event doesn't like eval "use" from callbacks, and it avoids
586 exporting the probe function).
587 - Impl::IOAsync's io watchers did not use the filehandle emulation
588 code to work around it's watcher limits.
589 - work around IO::Async's problems with overloaded objects.
590 - try to disable wearnings inside IO::Async, it's just too buggy.
591 - do no longer use Event by default, only EV or Perl, Event is too buggy.
592 - hackishly provide fast AE:: interface when EV or Perl is the backend.
593 - inherit VERSION from AnyEvent::VERSION in most modules.
594
5954.91 Thu Aug 6 15:42:45 CEST 2009
596 - AE::Handle::starttls could get out of sync when the read buffer
597 already contains some TLS handshake.
598 - AE::Handle did not properly free the TLS session, causing bigger
599 memory leaks in Net::SSLeay than expected (reported by toaster).
600 - AE::Socket::tcp_connect will now artificially delay invoking
601 the callback to avoid returning after invoking the callback.
602 - convert many internal watcher uses to AE API.
603
6044.9 Sat Aug 1 11:07:01 CEST 2009
605 - Glib silently fails when registering a timeout with
606 a negative value, so avoid that.
607 - call condvar callback immediately when it is set
608 after the condition is already signalled.
609 - check rbuf_max condition only after trying to consume
610 data.
611 - normalise signal numbers to names when using the ae's
612 signal handling, but do not document this (yet).
613 - pure perl signal emulation did not properly set nonblocking
614 mode on the signal pipe on win32 (this is unlikely to have caused
615 any issues).
616 - new module: AnyEvent::Debug.
617 - AnyEvent::Strict now checks that a signal was specified
618 by name (not name or number).
619 - reduce memory footprint in typical cases by ~50kb by
620 only compiling signal code when necessary.
621 - add AnyEvent::Handle->rbuf_max.
622 - grab the AE namespace for future new API, implement stubs for
623 said future/alternative API.
624 - new function: AnyEvent::Socket::format_hostport.
625
6264.881 Tue Jul 28 12:51:53 CEST 2009
627 - work around a bug in local in pre-5.10 perls, causing
628 AnyEvent::Handle to recurse when it should not
629 (analysed by elmex).
630
6314.88 Tue Jul 28 04:04:37 CEST 2009
632 - re-bless the handle into a dummy package after calling
633 AnyEvent::Handle::destroy, so the user does not need to check
634 for errors after every push_write etc.
635 - do not attempt to run t/02_signals.t on obviously broken
636 platforms.
637
6384.87 Sun Jul 26 02:06:16 CEST 2009
639 - do not attempt to linger when there is no longer a valid fh.
640 - enforce tls mode to be either accept or connect, do not simply
641 segfault in Net::SSLeay.
642 - AnyEvent::Handle can now call tcp_connect itself (new parameters
643 connect, on_prepare, on_connect and on_connect_error). Updated
644 tutorial accordingly.
645 - add AnyEvent::Impl::Irssi backend.
646
6474.86 Mon Jul 20 23:52:29 CEST 2009
648 - since the verbose warning is not enough, explicitly document
649 that versions before 1.33 of Net::SSLeay are not secure.
650 - work around signal handling races in Event and (...) Event::Lib.
651 - try to align signal-race timer to full-second boundaries.
652 - work around Tk not liking negative timeouts.
653 - don't complain of different grades of environmental unfriendlyness
654 in IO::Async.
655
6564.85 Sat Jul 18 06:16:14 CEST 2009
657 - nail the signal race problem in perl once and for all
658 (see $AnyEvent::MAX_SIGNAL_LATENCY).
659 - take advantage of Async::Interrupt if it is available.
660 - load Time::HiRes and Guard modules on demand only.
661 - add optional/recommended modules section to AnyEvent
662 documentation.
663 - reduce memory usage considerably (and reduce startup penalty)
664 by not using "strict", "warnings" and "overload" modules.
665 - work around buggy windows/openbsd perls and provide EBADMSG
666 and EPROTO ourselves when missing.
667 - improve perl 5.6 compatibility of the core event loop.
668 - made Net::SSLeay version 1.33 a soft requirement.
669
6704.83 Fri Jul 17 16:56:26 CEST 2009
671 - implement AnyEvent::Socket::getprotobyname.
672 - AnyEvent::CondVar's will now detect recursive blocking
673 waits and will croak, as too many people fall into
674 this trap.
675 - AnyEvent::Handle will now call ->destroy on itself after
676 executing the on_error callback, instead of doing some
677 half-baked internal shutdown, for fatal errors.
678 - clarify on_eof behaviour w.r.t. the read queue and
679 on_read callbacks.
680 - ignore some possible spurious wake-ups in tcp_connect.
681
6824.82 Sat Jul 11 00:34:55 CEST 2009
683 - POE and Event backends didn't accept some callable objects as
684 callbacks.
685 - use Config module instead of POSIX module to detect signal names
686 in AnyEvent::Strict and AnyEvent::Impl::EventLib, as the POSIX
687 module doesn't even have all POSIX signals :/.
688 - use more workarounds around the many refcnt/corruption bugs in
689 Event::Lib.
690 - work around a race condition in perl's select, causing t/03_child.t
691 to rarely fail.
692
6934.81 Thu Jul 9 10:30:30 CEST 2009
694 - AnyEvent::Handle didn't properly diagnose write errors
695 (it expected -1 from syswrite, how lame... :).
696 - support file descriptors in addition to file handles
697 in AnyEvent->io.
698 - new env variables: PERL_ANYEVENT_RESOLV_CONF,
699 PERL_ANYEVENT_MAX_OUTSTANDING_DNS, PERL_ANYEVENT_CA_FILE
700 and PERL_ANYEVENT_CA_PATH.
701 - provide a sensible synopsis section for AnyEvent::TLS.
702 - add a "supported backends" section to the manpage.
703 - added simple io watcher test to testsuite, using a
704 portable_socketpair.
705 - tried to improve the stability of the Event::Lib backend,
706 YMMV.
707
7084.8 Mon Jul 6 23:45:16 CEST 2009
709 - AnyEvent::DNS did not properly follow CNAME records with
710 uppercase targets.
711 - AnyEvent::DNS would errornously return AAAA records
712 with v4 mapped addresses (a faulty record) as ipv4 addresses,
713 causing AnyEvent::Socket to throw an exception.
714 - added new module AnyEvent::TLS for easier SSL/TLS context
715 creation, with many options including hostname verification,
716 secure default configuration, lots of documentation and,
717 predefined diffie-hellman keys for perfect forward security
718 and much more. get it while it's still fresh!
719 - use AnyEvent::TLS in AnyEvent::Handle for context management.
720 - load AnyEvent::Handle only on demand in AnyEvent::DNS,
721 so AnyEvent::Socket users have smaller memory footprint
722 in the common case.
723 - add AnyEvent::Handle->push_shutdown method.
724 - add an additional error message parameter to AnyEvent::Handle's
725 on_error callback (for TLS, $! is still available).
726 - add AnyEvent::Handle on_starttls/on_stoptls callbacks.
727 - make AnyEvent::Handle more robust against early connection
728 failures (during new), and return C<undef> in that case
729 from the constructor.
730 - AnyEvent::Handle will now try to load only JSON::XS first,
731 then fall back to JSON.
732 - format_ipv4/format_ipv6 are now exported by default, for symmetry,
733 and because it was documented that way.
734
7354.451 Fri Jul 3 00:28:58 CEST 2009
736 - do not clear rbuf when shutting down an AnyEvent::Handle
737 object - doing so breaks AnyEvent::HTTP.
738
7394.45 Mon Jun 29 22:59:26 CEST 2009
740 - a write error could cause AnyEvent::Handle to create
741 an I/O watcher with an undefined $fh.
742 - special-case mapped ipv4 addresses in both
743 AnyEvent::Socket::format_address and parse_address,
744 to treat them just like ipv4 addresses.
745 - updated and overhauled the AnyEvent::Intro doc.
746 - implement AnyEvent::Socket::format_ipv4/ipv6.
747 - slightly speed up portable_pipe/socketpair.
748 - expand condvar begin/end documentation.
2 749
34.42 Fri Jun 26 08:32:18 CEST 2009 7504.42 Fri Jun 26 08:32:18 CEST 2009
4 - preliminary and neccesarily incomplete support for IO::Async. 751 - preliminary and neccesarily incomplete support for IO::Async.
5 - reset SIGCHLD to DEFAULT when AnyEvent is loaded, in 752 - reset SIGCHLD to DEFAULT when AnyEvent is loaded, in
6 case it was set to IGNORE, to ensure we can catch child statuses 753 case it was set to IGNORE, to ensure we can catch child statuses
… …
1094.3 Fri Oct 3 09:18:43 CEST 2008 8564.3 Fri Oct 3 09:18:43 CEST 2008
110 - AnyEvent will now install a no-op signal handler for SIGPIPE, 857 - AnyEvent will now install a no-op signal handler for SIGPIPE,
111 unless one has been installed already. 858 unless one has been installed already.
112 - warn about not loaidng AnyEvent::Impl::POE early enough 859 - warn about not loaidng AnyEvent::Impl::POE early enough
113 (patch by Adam Rosenstein). 860 (patch by Adam Rosenstein).
114 - fixed a great number of bugs and corner cases in AnyEvent::Handle: 861 - fixed a great number of bugs and corner cases in AnyEvent::Handle.
115 - fix a bug in where in SSL connect mode, the 862 - fix a bug in where in SSL connect mode, the
116 client would first wait for some data by the server and 863 client would first wait for some data by the server and
117 otherwise hang. (reported and analyzed in an absolutely 864 otherwise hang. (reported and analysed in an absolutely
118 exemplary manner by Adam Rosenstein). 865 exemplary manner by Adam Rosenstein).
119 - fix a bug in where SSL EOF would not be treated 866 - fix a bug in where SSL EOF would not be treated
120 as stream EOF, putting the connection into a hung state. 867 as stream EOF, putting the connection into a hung state.
121 (reported and analyzed in an absolutely exemplary manner by Adam 868 (reported and analysed in an absolutely exemplary manner by Adam
122 Rosenstein). 869 Rosenstein).
123 - fix a potential segfault when the TLS context 870 - fix a potential segfault when the TLS context
124 would go missing in a read callback (e.g. due to stoptls) - 871 would go missing in a read callback (e.g. due to stoptls) -
125 Net::SSLeay of course makes no type checking whatsoever. 872 Net::SSLeay of course makes no type checking whatsoever.
126 - AnyEvent::Handle will not stop the read watcher in TLS mode. 873 - AnyEvent::Handle will not stop the read watcher in TLS mode.

Diff Legend

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