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

Comparing AnyEvent/Changes (file contents):
Revision 1.284 by root, Thu Jul 16 23:27:46 2009 UTC vs.
Revision 1.625 by root, Thu Aug 17 02:12:36 2017 UTC

1Revision history for Perl extension AnyEvent. 1Revision history for Perl extension AnyEvent.
2 2
3TODO: document TCP_*** constants
4TODO: add debug/trace logging to some modules
5TODO: use the eval-with-cb-call trick also to speed up JSON?
6TODO: FAQ: common pitfalls? invoke-before-return
7TODO: FAQ: scope-ids? hosts file order?
8TODO: AnyEvent::Handle with AnyEvent::IO?
9TODO: invalid. and localhost. specialcasing inside AEDNS and not AESocket (rfc6761)
10TODO: maybe implement env variable to give hosts precedence
11TODO: more formats for parse_ipv6 (single ipv6 address without port, ...p80, ...#80)
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 - support CAA records, based on patch by Steve Atkins.
18
197.14 Sat Jun 24 01:44:19 CEST 2017
20 - fix a crash bug in AnyEvent::Handle with openssl 1.1.0 (patched
21 by Paul Howarth and apparently tracked down by Mike McCauley).
22 - AnyEvent::Handle->keepalive was documented (and defined) twice
23 (found by Matt Merhar).
24 - AnyEvent::Socket::tcp_bind/tcp_server would immediately unlink a unix
25 listening socket unless a guard is used. Change this so that no
26 cleanup will be performed unless a guard is used and document this more
27 clearly.
28 - make tcp_bind/tcp_server error messages more regular.
29 - work around latest perl upstream breakage - this time, perl5porters were
30 apparently bullied by a bogus CVE (CVE-2016-1238), and since distros
31 stupidly apply anything that has "security fix" stamped on it, it's likely
32 going to be a problem in practise (and working around it is trivial on
33 windows and unix).
34 - add TCP_FASTOPEN/MSG_FASTOPEN and MSG_NOSIGNAL/MSG_MORE/MSG_DONTWAIT
35 to constants.
36 - update warnings to common::sense 3.74 standards.
37
387.13 Sat Sep 17 04:31:49 CEST 2016
39 - Only call tlsext_host_name for non-empty common names (reported
40 by Maxime Soulé).
41 - log a (single) notice message if SNI is not supported.
42 - upgrade to UTS-46:9.0.0 draft and switch to non-transitional
43 behaviour, beating thunderbird, ie, edge, chrome and safari to it :)
44 (see also https://bugzilla.mozilla.org/show_bug.cgi?id=1218179)
45 - turns out the UTS-46 IDNA testcase failures were indeed bugs in the
46 testcases and the specification and not in the code - the post-9.0.0
47 unicode files have all known problems fixed, so finally the AnyEvent
48 IDNA implementation can pass the full IDNA testsuite - without needing
49 a single fix :)
50 - guarantee (and document) that condvar callbacks will be removed
51 on invocation - important to avoid circular references.
52
537.12 Wed Jan 27 19:12:26 CET 2016
54 - use common name as hostname for tls connects, if Net::SSLeay
55 supports SNI.
56 - fix documentation of tls_autostart read type in AnyEvent::Handle,
57 analyzed by Felix Ostmann.
58
597.11 Thu Jul 16 14:36:00 CEST 2015
60 - AnyEvent::Socket::parse_ipv6 could accept malformed ipv6
61 addresses (extra "::" at end and similar cases).
62 - add a more explicit warning to AnyEvent::Handle that it doesn't
63 work on files, people keep getting confused.
64 - new function AnyEvent::Socket::tcp_bind.
65 - new functions AnyEvent::fh_block and AnyEvent::fh_unblock.
66 - aligned ipv6 address formatting with RFC 5952 (by not shortening
67 a single :0: to ::).
68 - added stability canary support.
69
707.09 Sat May 2 16:38:53 CEST 2015
71 - AnyEvent::Debug called an internal function (AnyEvent::Log::ft)
72 that was renamed to AnyEvent::Log:format_time. uinder its old name
73 (analyzed by Michael Stovenour).
74 - update AnyEvent::DNS fallback resolver addresses:
75 seems google effectively killed most other free dns resolvers,
76 so remove them, but add cable and wireless (ecrc) since it was
77 stable for 20 years or so, official or not, and there should be
78 an alternative to google.
79 - perl5porters broke windows error codes in 5.20, and mapped
80 WSAEWOULDBLOCK on the (different) EWOULDBLOCK error code, and
81 WSAEINPROGRESS into the incompatible ERINPROGRESS code, probably
82 because they were so cool. They probably broke other error codes
83 for no reason, too, but I didn't care for checking, it's too
84 depressing. This version only works around the WSAEWOULDBLOCK
85 issue, because I don't have a nice way to work around the
86 WSAEINPROGRESS bug.
87
887.08 Wed Dec 10 05:27:17 CET 2014
89 - work around a newly introduced bug in Socket 2.011 (an
90 errornous sun_length check) (analyzed by Maxime Soulé).
91 - AnyEvent::TLS didn't load (but refer to) AnyEvent::Socket
92 (analyzed by Ben Magistro).
93 - AnyEvent::Strict will now confess, not croak. This is in line with
94 it being a development/debugging tool.
95 - work around a number of libglib bugs (debug builds of libglib enforce
96 certain undocumented behaviour patterns such as not being able to
97 remove a child watch source after it has fired, which we will try
98 to emulate to avoid "criticals". what where they thinking...).
99 - mention json security issues in AnyEvent::Handle, now that Douglas
100 Crockford has foolishly and incompatibly changed JSON.
101 - changed default dns resolver "max_outstanding" value from 1 to 10,
102 the latter beinfg the intended value all along
103 (reported by Ilya Chesnokov).
104 - added new "AnyEvent::Impl::UV" interface module to the UV event lib
105 (written by Mike Lowell).
106
1077.07 Tue Dec 17 17:45:02 CET 2013
108 - the documentation for custom tls verify schemes was wrong. make it agree
109 with the code (reported by Maxime Soulé).
110 - added cbor read and write types to AnyEvent::Handle (using CBOR::XS).
111 - work around an API change in openssl that could cause wrong tls connection
112 aborts, likely on windows only (analyzed by sten).
113 - calling AnyEvent->now_update with AnyEvent::Impl::Perl caused an
114 endless loop (reported by Dietrich Rebmann).
115 - add tlsv1_1 and tlsv1_2 protocols to AnyEvent::TLS
116 (patch by Maxime Soulé).
117 - document AnyEvent::Impl::IOAsync::set_loop and
118 $AnyEvent::Impl::IOAsync::LOOP. Though only documented now, this
119 functionality has _always_ been available.
120 - force a toplevel domain name in t/81_hosts.t (analyzed by
121 David Jack Wange Olrik).
122 - document that AnyEvent::Log uses AnyEvent::IO.
123 - warn about AnyEvent::Filesys::Notify performance.
124 - praise the joys of AnyEvent::Fork::*.
125 - time for an =encoding directive.
126 - do no longer use JSON to create a default json coder, use
127 JSON::XS or JSON::PP directly.
128
1297.05 Wed Aug 21 10:38:08 CEST 2013
130 - uts46data.pl couldn't be found due to wrong naming of the file
131 (reported by Fulko Hew).
132 - handle lone \015's properly in AE::Handle's default line read
133 (reported by various people).
134 - untaint ip addresses found in /etc/hosts (patch by José Micó).
135 - the memleak fix in 7.03 caused resolving via /etc/hosts to always
136 fail on first use (reported and testcase by Andrew Whatson).
137 - expose AnyEvent::Log::format_time, and allow users to redefine it.
138 - expose AnyEvent::Log::default_format, and allow redefinition.
139 - expose AnyEvent::Log::fatal_exit, to allow redefinition.
140 - AnyEvent::Debug shell can now run coro shell commands, if available.
141 - t/63* tests were wrongly in MANIFEST.
142 - kernel.org's finger server went MIA, switch to freebsd.org and
143 icculus.org.
144 - clarify that IO::AIO and AnyEvent::AIO are needed for AnyEvent::IO
145 to function asynchronously (reported by Daniel Carosone).
146 - hard-disable $^W in most tests, it generates too much garbage output.
147 - use a (hopefully) more future-proof method to emulate common::sense.
148 - upgrade to UTS-46:6.2.0.
149 - switch to INSTLIB from INSTLIBDIR, as INSTLIBDIR was wrongly
150 documented. should not affect anything.
151
1527.04 Thu Nov 15 02:23:58 CET 2012
153 - debugging code left in.
154
1557.03 Thu Nov 15 02:16:11 CET 2012
156 - AnyEvent::Socket::inet_aton did not work when dns resolution was
157 used to find the addresses (analyzed and patch by Maxime Soulé).
158 - fix a memory leak in the /etc/hosts lookup code when hosts don't
159 resolve and are not in hosts (analyzed by sten).
160
1617.02 Tue Aug 14 04:27:58 CEST 2012
162 - AnyEvent::Util::run_cmd could block indefinitely (analyzed and test
163 program by Yorhel).
164 - verified that AnyEvent::Socket follows RFC5952.
165 - try to parse "ADDR#PORT" in addition to "ADDR PORT".
166
1677.01 Sun May 13 01:03:17 CEST 2012
168 - fail with EPROTO in AnyEvent::Handle wqhen TLS is requested but
169 not available, instead of throwing an exception.
170 - use File::Spec to get the tmpdir in t/*, to avoid needless
171 failures on (most, not mine :) windows boxes.
172 - new handle read types: tls_detect and tls_autostart.
173
1747.0 Fri Apr 13 06:33:30 CEST 2012
175 - child watchers are broken in POE 1.352 (also many earlier
176 versions) and there seems to be no way to work around it, as POE
177 itself is inherently racy. Document this regression and add a delay
178 in t/68_poe_03_child.t for the time being.
179 - new module AnyEvent::IO, that is a frontend to either a pure-perl
180 synchronous I/O implementation (AnyEvent::IO::Perl), or to the
181 asynchronous I/O extension IO::AIO (AnyEvent::IO::IOAIO).
182 - load /etc/hosts only when DNS has no answer.
183 - stat /etc/hosts on every access and reload it if it changed.
184 - load /etc/hosts via AnyEvent::IO - potentially asynchronous.
185 - fix a buggy croak in the dh parameter reading in AnyEvent::TLS.
186 - AnyEvent::Log log_to_file and log_to_path now use AnyEvent::IO. As
187 a side effect, they now use true appending as opposed to libc
188 appending, and the file might not have been opened when the function
189 returns.
190 - the default logging level was not properly documented in a variety of
191 places, this has hopefully been rectified.
192 - updated uts46data.pl for unicode 6.1.0.
193 - made log messages generated by AnyEvent submodules not
194 include the package name anymore, as it will be logged
195 by default already.
196 - upgrade to the trick used by common-sense 3.5 to work
197 around extra warning torture/breakage under perl 5.15.x.
198 - log messages by anyevent are now ucfirst, are usually full
199 sentences and do no longer include the package name.
200 - the storable read type would sometimes throw an exception instead
201 of causing EBADMSG (analyzed by Maxime Soulé).
202 - work around a bug in openssl 1.0.1 which enforces a minimum rsa
203 keysize (reported by Paul Howarth).
204 - documented Rocco fabricating statements about AnyEvent and me.
205
2066.14 Tue Jan 31 20:00:24 CET 2012
207 - AnyEvent::Impl::Tk was broken due to a mysterious "die" inside,
208 probably an editing mistake (reported by Darin McBride).
209
2106.13 Thu Jan 12 07:27:01 CET 2012
211 - AnyEvent::Util::fork_call checked for POSIX availability in the wrong way,
212 causing it to fail if POSIX wasn't loaded already (analysed by Rock Power).
213 - AnyEvent::Handle::push_read (line => did pass $1 directly, so
214 regex-matching inside the callback would change the parameter, despite
215 $1 being dynamically scoped per-block. this perl bug is now being
216 worked around (testcase by Cindy Wang).
217
2186.12 Mon Dec 12 13:21:10 CET 2011
219 - $! was clobbered by subsequent calls in tcp_connect, due to
220 the postpone not saving/restoring it (analyzed by Richard Garnish).
221
2226.11 Tue Nov 22 10:36:05 CET 2011
223 - Tk cannot create windows when tainted unless you set a title, so
224 set a dummy title for AnyEvent's dummy mainwindow. How dummy.
225 - escape any nonprintable/nonascii characters when stringifying
226 backtraces.
227 - log the reason loading an interface module fails to load at level debug.
228
2296.1 Tue Oct 4 19:44:30 CEST 2011
230 - INCOMPATIBLE CHANGE: the default log level is now "4" (error
231 and above), and some messages inside AnyEvent have been elevated
232 to higher log levels to print by default.
233 - AnyEvent::log, unlike AnyEvent::Log::log, did not always exit on fatal.
234 - work around yet another signal race bug, newly introduced by POE. My
235 how that module sucks.
236 - add google public ipv6 dns as fallback servers, so AE::DNS can finally
237 support ipv6-only operation even with defaults (also use ::1 instead
238 of 127.0.0.1).
239 - give better guidelines on how to chose a logging level.
240 - AnyEvent::Log can now cap (limit) log message priority by context.
241 - AE::log will now save and restore $! and $@, so callers don't have to.
242 - AE::Strict now checks that registered fds are still valid at regular
243 intervals.
244 - skip frequently called 'require' calls if the module appears loaded.
245 - increase signal latency test timer considerably for poor little osx,
246 which can't be bothered with basic posix functionality such as signal
247 delivery.
248 - mention some debugging aids in the tutorial.
249 - new env variable PERL_ANYEVENT_MAX_SIGNAL_LATENCY.
250 - private (%namespace) log contexts in PERL_ANYEVENT_LOG now have the name
251 of the context as title.
252 - AnyEvent::Strict would not restore read-write access to $_ when a callback
253 died - this is a user bug, but since AE::Strict exists to help dbeugging...
254 - detect the extremely evil IO::Async::Loop::AnyEvent module and refuse
255 to work with it - the author is well aware of the issues but refuses
256 to fix them. This extreme step has been taken because the module actively
257 attacks the goals of AnyEvent and makes it harder for module authors to use
258 AnyEvent.
259
2606.02 Fri Aug 26 20:08:31 CEST 2011
261 - log_to_syslog now accepts facility strings.
262 - log_to_syslog would not allow facilities - "$facility|$level"
263 *literally* means that string. sheesh.
264 - the CPAN parser got confused and thought we provide Sys::Syslog, try
265 to work around.
266 - renamed AnyEvent::Impl::FLTK2 to FLTK.
267
2686.01 Fri Aug 26 07:04:11 CEST 2011
269 - INCOMPATIBLE CHANGE: AnyEvent::DNS resource records now include
270 the TTL as fourth element - this affects ->resolve, ->request
271 and ->dns_unpack indirectly. Convenience functions such as
272 AnyEvent::DNS::{a,txt,mx,srv} are not affected, neither is
273 AnyEvent::Socket.
274 - INCOMPATIBLE CHANGE: backend authors now should not implement
275 one_event or loop, but instead the AnyEvent::CondVar::_wait and _poll
276 methods.
277 - INCOMPATIBLE CHANGE: verbosity levels have been redefined, see
278 AnyEvent (PERL_ANYEVENT_VERBOSE) and AnyEvent::Log for details.
279 - DEPRECIATION: please change
280 "use AnyEvent::Impl::Perl" to "use AnyEvent::Loop" and
281 "AnyEvent::Impl::Perl::loop" to "AnyEvent::Loop::run"
282 in your programs, for future compatibility.
283 - *blush* idle watcher emulation was borked.
284 - AnyEvent::Strict now checks the AE::xxx functions as well.
285 - querying too long domain names would barf deep inside AnyEvent::DNS - since
286 this is hard to avoid for applications, AnyEvent:.DNS::request now checks for
287 for too long domain names explicitly and passes an undef to the callback
288 instead (reported by James Bromberger).
289 - make AnyEvent::Util more robust against EINTR (analyzed by Andrew Suffield).
290 - new functions: AnyEvent::postpone, AnyEvent::log.
291 - new module: AnyEvent::Log, for simple logging needs.
292 - new env variable, PERL_ANYEVENT_LOG, for fine-grained logging config.
293 - new env variable, PERL_ANYEVENT_DEBUG_SHELL, to automatically start
294 a debug shell.
295 - new env variable, PERL_ANYEVENT_DEBUG_WRAP, to automatically enable
296 watcher instrumentation/wrapping (see AnyEvent::Debug::wrap).
297 - new env variable, PERL_ANYEVENT_HOSTS, to override the place where
298 /etc/hosts can be found.
299 - an empty PERL_ANYEVENT_RESOLV_CONF now selects the default
300 configuration.
301 - env variables can now be specified using AE_xyz instead of
302 PERL_ANYEVENT_xyz, subject to some rules.
303 - lots of new AnyEvent::Debug functionality.
304 - AnyEvent::Strict now checks for modifications of $_ by the callback
305 (by setting it to readonly while executing the callback, if possible).
306 - IO::Async sometimes cancels the wrong timer when you call cancel_timer.
307 Invest extra overhead to work around this peculiar behaviour.
308 - split the pure perl backend into a loop module and an impl
309 module (AnyEvent::Loop and AnyEvent::Impl::Perl).
310 - some read types (regex, netstring, json, storable) could
311 errornously keep an AE::Handle object alive on their own
312 (reported by Mohammad Toossi).
313 - AnyEvent::Socket::resolve_sockaddr and all functions using it now
314 supports /etc/hosts overrides.
315 - add more workarounds around child watcher bugs in IO::Async
316 and (newly introduced) in POE.
317 - use glib child watchers and try to work around its limitations,
318 if possible (based on a patch by Kevin Ryde).
319 - add support for FLTK (fltk2), via AnyEvent::Impl::FLTK2.
320 - many read types in AnyEvent::Handle would malfunction during
321 transitions to TLS because of a "delete $self->{rbuf}".
322 (reported by Gerald Galster).
323 - improve condvar blocking wait performance (probably only noticable
324 with faster backends), also make it easier for event loops that
325 do not support blocking waits.
326 - PERL_ANYEVENT_MODEL now also allows full module names, although
327 the usefulness of this feature is not yet clear.
328 - add tests for all supported event loops, conditional on PERL_ANYEVENT_LOOP_TESTS.
329 - add t/80_ssltext.t.
330 - no longer use AUTOLOAD for the initial loading (for no very good reason).
331 - tcp_server now tries to unlink unix domain sockets when it gets destroyed.
332 - do not run event loop detection multiple times when programs call their cached
333 copies of AnyEvent::detect.
334 - parse_hostport now accepts absolute paths als unix domain sockets.
335 - suppress the idiotic warning inside POE by patching the POE::Kernel
336 run flags instead of asking the user to jump through contortions.
337 - add Task::Weaken as dependency (as recommends only at this time).
338 - replace PF_UNSPEC by 0, for good old minix.
339
3406.0 (not officially released, but was on CPAN for about 15 minutes :/)
341
3425.34 Thu May 12 09:59:41 CEST 2011
343 - AE::TLS didn't support both cert_file and key at the same time,
344 due to some copy&paste bug (reported and analyzed by vti).
345 - improve AE::Handle timeout documentation slightly.
346
3475.33 Mon Feb 7 21:51:41 CET 2011
348 - never appeared on CPAN, for unknown reasons.
349 - AnyEvent::Handle on_connect_error - do not try to destroy a handle
350 object when it's already gone (analyzed by Ryan Bullock).
351 - update idna algorithm(s) to UTS#46 version 6.0.0.
352 - preserve trailing dot in idn_to_ascii - testsuite says to preserve,
353 uts46 says it's an error, and testsuite makes more sense.
354 - remove leading dots in idn_to_ascii - testsuite says to remove,
355 uts46 says it's an error, and testsuite makes more sense.
356 - fixed a bug in idn ignored character processing.
357 - passes the uts46 6.0.0 testsuite except for three cases which
358 apparently are bugs in the testsuite itself.
359 - passes the uts46 6.0.1 testsuite except for eleven cases which
360 apparently are bugs in the testsuite itself.
361
3625.31 Sun Jan 23 11:49:19 CET 2011
363 - work around a POE bug where POE would first poll for new events and
364 invoke handlers BEFORE handling the message queue.
365 - fix parsing of "domain" statement in resolv.conf (Johannes Plunien).
366 - stop_read now does just that in TLS mode. see its
367 documentation for caveats.
368 - update to common::sense 3.4 warning set (some warnings in perl impose
369 an arbitrary 32kb limit on regex matches, which affects
370 AnyEvent::Handle).
371 - AE::Handle had a broken rbuf_max method and missing wbuf_max method.
372 - new experimental AE::Handle::resettls method.
373
3745.3 Fri Dec 31 05:48:13 CET 2010
375 - major IOAsync patch by Paul Evans - 0.33 is required, as the major
376 issues with IO::Async have been solved in that version, including
377 the availabilty of a default loop!
378 - implement new wbuf_max parameter for AnyEvent::Handle.
379 - added Cocoa::EventLoop backend by Daisuke Murase.
380 - the default testsuite now honors PERL_ANYEVENT_MODEL (Paul Evans).
381 - clarify handle on_prepare documentation.
382 - set CLOEXEC flag on parent-side fd's in AnyEvent::Util::run_cmd,
383 to avoid leaking them to any child processes.
384 - try to force files into the arch-dependend lib directory, for the
385 half-"turly obsessive compulsive person"(s) that depend on it.
386
3875.29 Sun Dec 5 10:49:21 CET 2010
388 - convert EV backend to EV 4.00 API (so better upgrade EV too).
389 - AE::Handle: implement dynamic read_size adjustment, add
390 max_read_size parameter, reduce default read_size to 2048.
391 - add some FAQ entries for servers and/or on_eof vs. on_error.
392 - work around OS bugs (cygwin again) when lingering in AE::Handle.
393 - reduce memory usage slightly when connect was used in AE::Handle.
394 - wrok around more segfaults in Tk.
395 - document tls_ctx => undef as valid.
396 - detect WNOHANG value at compiletime.
397 - include some probably linux-only support for building AnyEvent
398 as part of the perl core.
399 - improve compatibility of t/08_idna.t to perl 5.8.
400 - make AnyEvent::Strict truly optional (in case somebody builds a
401 minimal perl...).
402
4035.28 Wed Oct 13 04:14:23 CEST 2010
404 - due to a glitch, AnyEvent's internal getprotobyname wasn't
405 actually used in tcp_server.
406 - implement AnyEvent::CondVar->new method, which was documented but
407 missing.
408 - support autoloading of read types in unshift_read.
409 - AnyEvent::Strict: do the test for invalid fh arguments differently,
410 resulting in more informative error messages.
411 - new AnyEvent::FAQ manpage.
412 - clarify that you can actually call push_read etc. in on_read.
413 - put keys and certs in /tmp, not into the current directory,
414 and document this in the SECURITY CONSIDERATIONS sections
415 of AnyEvent::TLS.
416 - support ";" as resolv.conf comment character in AnyEvent::DNS, as
417 requested by Ron Isaacson.
418 - document $AnyEvent::DNS::RESOLVER variable.
419 - incorporated major typo patches by Abhijit Menon-Sen.
420 - AnyEvent::Handle now croaks when negative timeouts are passed.
421 - add a shitload of TCP_xxx sockopt name constants to AnyEvent::Util,
422 undocumented, but who knows when they might come in handy.
423
4245.271 Tue Jun 8 12:05:46 CEST 2010
425 - backport to perl 5.8.x.
426
4275.27 Sun Jun 6 12:12:05 CEST 2010
428 - postpone differently in AnyEvent::Socket now, as
429 when not, canceling the connection attempt might fail
430 (found by Felix Antonius Wilhelm Ostmann).
431 - explicitly check for non-stream sockets in AE::Handle, too many
432 clueless people fell into the trap of this somehow working.
433 - simplified and reworked the "OTHER MODULES" section.
434 - better/more condvar examples.
435
4365.261 Wed Apr 28 16:13:36 CEST 2010
437 - AF_INET6 was not properly used from Socket6 during configuration
438 time (found by Andy Grundman).
439
4405.26 Mon Apr 12 04:49:35 CEST 2010
441 - don't generate spurious readyness notifications when select
442 returns EINTR in the pure perl backend (analysed and testcase
443 by Bas Denissen).
444 - give same examples in SYNOPSIS sections of AnyEvent and AE.
445 - provider faster implementations for AE::time/now/now_update for
446 EV and Perl backends.
447 - provide faster AE::cv and AE::time implementations for all
448 backends.
449 - no longer support lower/mixed-case signal names (this was
450 never documented nor universally supported).
451 - some more memory µ-optimisations, and somewhat less messy
452 code to cope with both AnyEvent and AE APIs.
453
4545.251 Sat Mar 13 00:58:21 CET 2010
455 - make test failed if it was run before make install on perl <=
456 5.8.8, and older versions of perl make our live very difficult,
457 so write constants.pl during Makefile.PL time.
458
4595.25 Sat Mar 13 00:23:14 CET 2010
460 - fix a race condition in AnyEvent::Handle that would cause
461 a "bio_read: unsupported method" error (in _another_
462 ssl connection) after user code threw an exception.
463 - added AnyEvent::Handle->destroyed method.
464 - speed up AnyEvent::Socket::format_address and ::format_ipv6.
465 - the AnyEvent::Util::fh_nonblocking function worked only
466 by lucky accident on win32.
467 - smaller and faster AnyEvent::Util::fh_nonblocking.
468 - when the (required!) Time::HiRes module is missing, AnyEvent
469 did not fall back to built-in time properly.
470 - do not load Fcntl at runtime, saving memory and loading time.
471 - precompile a number of constants and use them instead of runtime
472 detection and eval.
473 - free detection code after detection and similar memory optimisations.
474 - Perl backend timer interval best effort drift has been improved
475 (same algorithm as EV).
476 - update unicode idna mapping table.
477
4785.24 Tue Jan 5 11:39:43 CET 2010
479 - cygwin never reports errors from failed connects. 1.5 just gives
480 you continous readyness and EAGAIN, 1.7 is even more broken and
481 just hangs. work around both issues in a cygwin-specific hack.
482 - improve idle watcher documentation slightly.
483
4845.23 Sun Dec 20 23:48:00 CET 2009
485 - support IDNs in resolve_sockaddr, and therefore in tcp_connect.
486 - implement punycode_encode/decode, idn_nameprep,
487 idn_to_ascii and idn_to_unicode operations in AnyEvent::Util.
488 - provide $AE::VERSION.
489 - removed traces of "no strict 'refs'".
490
4915.22 Sat Dec 5 03:51:13 CET 2009
492 - downgrade-or-fail in AnyEvent::Handle::push_write, to
493 diagnose encoding failures earlier and more succinctly.
494 (this works around bugs in perl, throwing away encoding info
495 when passing scalar data to extensions).
496 - add more examples to AnyEvent::Socket manpage.
497 - upgrade internal warning set to the same as common::sense 2.03.
498 - use pack "n/a*" for pre-5.8.9 perl compatibility in AnyEvent::DNS
499 (John Beppu).
500 - AnyEvent::Socket::inet_aton now properly supports ipv6, as documented.
501 - add google public dns servers to fallback server set and make sure
502 we load-balance properly between the three sets. also add all
503 fallback dns servers, not just a random one, to each dns config.
504
5055.21 Thu Nov 19 02:48:47 CET 2009
506 - fix a problem where socket constants were called with parameters
507 (spotted by David Friedland).
508 - fork_call never use'd POSIX (reported by Daisuke Maki).
509 - improve perl 5.6 compatibility further (but it still won't work
510 unless you rip out everything but the core).
511 - prefer Net::DNS::Resolver over ipconfig, if installed, on win32.
512 uses 10MB of RAM, but doesn't open a console window. *sigh*.
513
5145.202 Wed Oct 14 22:35:44 CEST 2009
515 - AnyEvent::DNS would unexpectedly clobber $_ under windows
516 (analysed by Matthias Waldorf).
517 - AnyEvent::Handle::run_cmd can now pass the PID of the
518 newly-created process, which is much less useful than it might
519 sound (based on patch by Yann Kerherve).
520
5215.201 Tue Sep 29 12:09:25 CEST 2009
522 - AnyEvent:Handle::on_starttls/on_stoptls methods were broken
523 (reported by Torsten Foertsch).
524 - common::sense 2.0 could cause tcp_server to throw an exception
525 (analysed by elmex).
526
5275.2 Mon Sep 14 07:04:49 CEST 2009
528 - INCOMPATIBLE CHANGE: do no longer support register_read_type
529 and register_write_type in AnyEvent::Handle, instead support
530 package names (the facility was mostly abused).
531 - implement "packagename-as-read/write type" support in
532 AnyEvent::Handle.
533 - AnyEvent::Handle: new options "keepalive" and "oobinline".
534 - oobinline set by default to avoid security issues.
535 - the pure-perl event loop backend wrongly detected times() fallback
536 support (spotted by Pavel Boldin).
537
5385.12 Tue Sep 1 20:26:50 CEST 2009
539 - be more lenient when parsing resolv.conf files, as some people
540 use hashmarks as comment indicator inside directives (reported
541 by Michael S. Fischer).
542 - use same set of warnings as common::sense 2.0.
543 - fix a potential 32 bit overflow issue due to perl having problems
544 with large hex constants in the Perl backend.
545 - do not use the slower 5.6 version fo the enc_name code in AnyEvent::DNS
546 by default.
547 - fix some prototypes.
548 - (optionally) require Async::Interrupt 1.02.
549 - replace opendns fallbacks by verizon/level3 public dns servers.
550
5515.112 Fri Aug 21 13:59:04 CEST 2009
552 - AnyEvent::Handle could cause hard crashes in Net::SSLeay when
553 using starttls with an illegal/uncreatable tls context.
554 - port to broken win32 perls.
555
5565.111 Sun Aug 16 18:44:36 CEST 2009
557 - perl errornously requires sockaddr_un structures returned by the
558 kernel to have a specific length. work around this bug by padding
559 sockaddr_un structures (found to fail on OS X, as perl only uses
560 the correct code for GNU/Linux, but likely to fail on other
561 OSes as well).
562
5635.11 Wed Aug 12 17:49:37 CEST 2009
564 - fixed a bug in run_cmd with input from filename.
565 - create a json coder object when none is given in write direction
566 to reduce dependency on JSON.
567
5685.1 Tue Aug 11 03:17:32 CEST 2009
569 - tcp_connect now reports the correct error in $!, not an unrelated
570 one (was broken since 4.91).
571 - AE::Handle did not properly clear rtimeout/wtimeout watchers
572 when retrying a tcp_connect.
573 - new functions: AnyEvent::Util::run_cmd and close_all_fds_except.
574 - fixed wrong documentation in AE::Socket.
575
5765.01 Mon Aug 10 03:16:32 CEST 2009
577 - last release broke EV child watchers :/.
578
5795.0 Sun Aug 9 17:32:09 CEST 2009
580 - officially introduce and document the AE API.
581 - lowering the timeout at runtime did not have immediate effect
582 in AnyEvent::Handle.
583 - AE::Handle now has separate and independent read and write timeout
584 settings.
585 - probe for Guard module when AnyEvent::Util is loaded, not at
586 runtime (Event doesn't like eval "use" from callbacks, and it avoids
587 exporting the probe function).
588 - Impl::IOAsync's io watchers did not use the filehandle emulation
589 code to work around it's watcher limits.
590 - work around IO::Async's problems with overloaded objects.
591 - try to disable wearnings inside IO::Async, it's just too buggy.
592 - do no longer use Event by default, only EV or Perl, Event is too buggy.
593 - hackishly provide fast AE:: interface when EV or Perl is the backend.
594 - inherit VERSION from AnyEvent::VERSION in most modules.
595
5964.91 Thu Aug 6 15:42:45 CEST 2009
597 - AE::Handle::starttls could get out of sync when the read buffer
598 already contains some TLS handshake.
599 - AE::Handle did not properly free the TLS session, causing bigger
600 memory leaks in Net::SSLeay than expected (reported by toaster).
601 - AE::Socket::tcp_connect will now artificially delay invoking
602 the callback to avoid returning after invoking the callback.
603 - convert many internal watcher uses to AE API.
604
6054.9 Sat Aug 1 11:07:01 CEST 2009
606 - Glib silently fails when registering a timeout with
607 a negative value, so avoid that.
608 - call condvar callback immediately when it is set
609 after the condition is already signalled.
610 - check rbuf_max condition only after trying to consume
611 data.
612 - normalise signal numbers to names when using the ae's
613 signal handling, but do not document this (yet).
614 - pure perl signal emulation did not properly set nonblocking
615 mode on the signal pipe on win32 (this is unlikely to have caused
616 any issues).
617 - new module: AnyEvent::Debug.
618 - AnyEvent::Strict now checks that a signal was specified
619 by name (not name or number).
620 - reduce memory footprint in typical cases by ~50kb by
621 only compiling signal code when necessary.
622 - add AnyEvent::Handle->rbuf_max.
623 - grab the AE namespace for future new API, implement stubs for
624 said future/alternative API.
625 - new function: AnyEvent::Socket::format_hostport.
626
6274.881 Tue Jul 28 12:51:53 CEST 2009
628 - work around a bug in local in pre-5.10 perls, causing
629 AnyEvent::Handle to recurse when it should not
630 (analysed by elmex).
631
6324.88 Tue Jul 28 04:04:37 CEST 2009
633 - re-bless the handle into a dummy package after calling
634 AnyEvent::Handle::destroy, so the user does not need to check
635 for errors after every push_write etc.
636 - do not attempt to run t/02_signals.t on obviously broken
637 platforms.
638
6394.87 Sun Jul 26 02:06:16 CEST 2009
640 - do not attempt to linger when there is no longer a valid fh.
641 - enforce tls mode to be either accept or connect, do not simply
642 segfault in Net::SSLeay.
643 - AnyEvent::Handle can now call tcp_connect itself (new parameters
644 connect, on_prepare, on_connect and on_connect_error). Updated
645 tutorial accordingly.
646 - add AnyEvent::Impl::Irssi backend.
647
6484.86 Mon Jul 20 23:52:29 CEST 2009
649 - since the verbose warning is not enough, explicitly document
650 that versions before 1.33 of Net::SSLeay are not secure.
651 - work around signal handling races in Event and (...) Event::Lib.
652 - try to align signal-race timer to full-second boundaries.
653 - work around Tk not liking negative timeouts.
654 - don't complain of different grades of environmental unfriendlyness
655 in IO::Async.
656
6574.85 Sat Jul 18 06:16:14 CEST 2009
658 - nail the signal race problem in perl once and for all
659 (see $AnyEvent::MAX_SIGNAL_LATENCY).
660 - take advantage of Async::Interrupt if it is available.
661 - load Time::HiRes and Guard modules on demand only.
662 - add optional/recommended modules section to AnyEvent
663 documentation.
664 - reduce memory usage considerably (and reduce startup penalty)
665 by not using "strict", "warnings" and "overload" modules.
666 - work around buggy windows/openbsd perls and provide EBADMSG
667 and EPROTO ourselves when missing.
668 - improve perl 5.6 compatibility of the core event loop.
669 - made Net::SSLeay version 1.33 a soft requirement.
670
6714.83 Fri Jul 17 16:56:26 CEST 2009
3 - implement AnyEvent::Socket::getprotobyname. 672 - implement AnyEvent::Socket::getprotobyname.
4 - AnyEvent::CondVar's will now detect recursive blocking 673 - AnyEvent::CondVar's will now detect recursive blocking
5 waits and will croak, as too many people fall into 674 waits and will croak, as too many people fall into
6 this trap. 675 this trap.
7 - AnyEvent::Handle will now call ->destroy on itself after 676 - AnyEvent::Handle will now call ->destroy on itself after
… …
54 in the common case. 723 in the common case.
55 - add AnyEvent::Handle->push_shutdown method. 724 - add AnyEvent::Handle->push_shutdown method.
56 - add an additional error message parameter to AnyEvent::Handle's 725 - add an additional error message parameter to AnyEvent::Handle's
57 on_error callback (for TLS, $! is still available). 726 on_error callback (for TLS, $! is still available).
58 - add AnyEvent::Handle on_starttls/on_stoptls callbacks. 727 - add AnyEvent::Handle on_starttls/on_stoptls callbacks.
59 - make AnyEvent::Handle more robust against early conenction 728 - make AnyEvent::Handle more robust against early connection
60 failures (during new), and return C<undef> in that case 729 failures (during new), and return C<undef> in that case
61 from the constructor. 730 from the constructor.
62 - AnyEvent::Handle will now try to load only JSON::XS first, 731 - AnyEvent::Handle will now try to load only JSON::XS first,
63 then fall back to JSON. 732 then fall back to JSON.
64 - format_ipv4/format_ipv6 are now exported by default, for symmetry, 733 - format_ipv4/format_ipv6 are now exported by default, for symmetry,
… …
1884.3 Fri Oct 3 09:18:43 CEST 2008 8574.3 Fri Oct 3 09:18:43 CEST 2008
189 - AnyEvent will now install a no-op signal handler for SIGPIPE, 858 - AnyEvent will now install a no-op signal handler for SIGPIPE,
190 unless one has been installed already. 859 unless one has been installed already.
191 - warn about not loaidng AnyEvent::Impl::POE early enough 860 - warn about not loaidng AnyEvent::Impl::POE early enough
192 (patch by Adam Rosenstein). 861 (patch by Adam Rosenstein).
193 - fixed a great number of bugs and corner cases in AnyEvent::Handle: 862 - fixed a great number of bugs and corner cases in AnyEvent::Handle.
194 - fix a bug in where in SSL connect mode, the 863 - fix a bug in where in SSL connect mode, the
195 client would first wait for some data by the server and 864 client would first wait for some data by the server and
196 otherwise hang. (reported and analyzed in an absolutely 865 otherwise hang. (reported and analysed in an absolutely
197 exemplary manner by Adam Rosenstein). 866 exemplary manner by Adam Rosenstein).
198 - fix a bug in where SSL EOF would not be treated 867 - fix a bug in where SSL EOF would not be treated
199 as stream EOF, putting the connection into a hung state. 868 as stream EOF, putting the connection into a hung state.
200 (reported and analyzed in an absolutely exemplary manner by Adam 869 (reported and analysed in an absolutely exemplary manner by Adam
201 Rosenstein). 870 Rosenstein).
202 - fix a potential segfault when the TLS context 871 - fix a potential segfault when the TLS context
203 would go missing in a read callback (e.g. due to stoptls) - 872 would go missing in a read callback (e.g. due to stoptls) -
204 Net::SSLeay of course makes no type checking whatsoever. 873 Net::SSLeay of course makes no type checking whatsoever.
205 - AnyEvent::Handle will not stop the read watcher in TLS mode. 874 - AnyEvent::Handle will not stop the read watcher in TLS mode.

Diff Legend

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