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

Comparing AnyEvent/Changes (file contents):
Revision 1.248 by root, Thu Jun 25 14:27:18 2009 UTC vs.
Revision 1.605 by root, Thu Jul 16 12:48:38 2015 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: hosts always read? that's not expected
12TODO: more formats for parse_ipv6 (single ipv6 address without port, ...p80, ...#80)
13
147.11 Thu Jul 16 14:36:00 CEST 2015
15 - AnyEvent::Socket::parse_ipv6 could accept malformed ipv6
16 addresses (extra "::" at end and similar cases).
17 - add a more explicit warning to AnyEvent::Handle that it doesn't
18 work on files, people keep getting confused.
19 - new function AnyEvent::Socket::tcp_bind.
20 - new functions AnyEvent::fh_block and AnyEvent::fh_unblock.
21 - aligned ipv6 address formatting with RFC 5952 (by not shortening
22 a single :0: to ::).
23 - added stability canary support.
24
257.09 Sat May 2 16:38:53 CEST 2015
26 - AnyEvent::Debug called an internal function (AnyEvent::Log::ft)
27 that was renamed to AnyEvent::Log:format_time. uinder its old name
28 (analyzed by Michael Stovenour).
29 - update AnyEvent::DNS fallback resolver addresses:
30 seems google effectively killed most other free dns resolvers,
31 so remove them, but add cable and wireless (ecrc) since it was
32 stable for 20 years or so, official or not, and there should be
33 an alternative to google.
34 - perl5porters broke windows error codes in 5.20, and mapped
35 WSAEWOULDBLOCK on the (different) EWOULDBLOCK error code, and
36 WSAEINPROGRESS into the incompatible ERINPROGRESS code, probably
37 because they were so cool. They probably broke other error codes
38 for no reason, too, but I didn't care for checking, it's too
39 depressing. This version only works around the WSAEWOULDBLOCK
40 issue, because I don't have a nice way to work around the
41 WSAEINPROGRESS bug.
42
437.08 Wed Dec 10 05:27:17 CET 2014
44 - work around a newly introduced bug in Socket 2.011 (an
45 errornous sun_length check) (analyzed by Maxime Soulé).
46 - AnyEvent::TLS didn't load (but refer to) AnyEvent::Socket
47 (analyzed by Ben Magistro).
48 - AnyEvent::Strict will now confess, not croak. This is in line with
49 it being a development/debugging tool.
50 - work around a number of libglib bugs (debug builds of libglib enforce
51 certain undocumented behaviour patterns such as not being able to
52 remove a child watch source after it has fired, which we will try
53 to emulate to avoid "criticals". what where they thinking...).
54 - mention json security issues in AnyEvent::Handle, now that Douglas
55 Crockford has foolishly and incompatibly changed JSON.
56 - changed default dns resolver "max_outstanding" value from 1 to 10,
57 the latter beinfg the intended value all along
58 (reported by Ilya Chesnokov).
59 - added new "AnyEvent::Impl::UV" interface module to the UV event lib
60 (written by Mike Lowell).
61
627.07 Tue Dec 17 17:45:02 CET 2013
63 - the documentation for custom tls verify schemes was wrong. make it agree
64 with the code (reported by Maxime Soulé).
65 - added cbor read and write types to AnyEvent::Handle (using CBOR::XS).
66 - work around an API change in openssl that could cause wrong tls connection
67 aborts, likely on windows only (analyzed by sten).
68 - calling AnyEvent->now_update with AnyEvent::Impl::Perl caused an
69 endless loop (reported by Dietrich Rebmann).
70 - add tlsv1_1 and tlsv1_2 protocols to AnyEvent::TLS
71 (patch by Maxime Soulé).
72 - document AnyEvent::Impl::IOAsync::set_loop and
73 $AnyEvent::Impl::IOAsync::LOOP. Though only documented now, this
74 functionality has _always_ been available.
75 - force a toplevel domain name in t/81_hosts.t (analyzed by
76 David Jack Wange Olrik).
77 - document that AnyEvent::Log uses AnyEvent::IO.
78 - warn about AnyEvent::Filesys::Notify performance.
79 - praise the joys of AnyEvent::Fork::*.
80 - time for an =encoding directive.
81 - do no longer use JSON to create a default json coder, use
82 JSON::XS or JSON::PP directly.
83
847.05 Wed Aug 21 10:38:08 CEST 2013
85 - uts46data.pl couldn't be found due to wrong naming of the file
86 (reported by Fulko Hew).
87 - handle lone \015's properly in AE::Handle's default line read
88 (reported by various people).
89 - untaint ip addresses found in /etc/hosts (patch by José Micó).
90 - the memleak fix in 7.03 caused resolving via /etc/hosts to always
91 fail on first use (reported and testcase by Andrew Whatson).
92 - expose AnyEvent::Log::format_time, and allow users to redefine it.
93 - expose AnyEvent::Log::default_format, and allow redefinition.
94 - expose AnyEvent::Log::fatal_exit, to allow redefinition.
95 - AnyEvent::Debug shell can now run coro shell commands, if available.
96 - t/63* tests were wrongly in MANIFEST.
97 - kernel.org's finger server went MIA, switch to freebsd.org and
98 icculus.org.
99 - clarify that IO::AIO and AnyEvent::AIO are needed for AnyEvent::IO
100 to function asynchronously (reported by Daniel Carosone).
101 - hard-disable $^W in most tests, it generates too much garbage output.
102 - use a (hopefully) more future-proof method to emulate common::sense.
103 - upgrade to UTS-46:6.2.0.
104 - switch to INSTLIB from INSTLIBDIR, as INSTLIBDIR was wrongly
105 documented. should not affect anything.
106
1077.04 Thu Nov 15 02:23:58 CET 2012
108 - debugging code left in.
109
1107.03 Thu Nov 15 02:16:11 CET 2012
111 - AnyEvent::Socket::inet_aton did not work when dns resolution was
112 used to find the addresses (analyzed and patch by Maxime Soulé).
113 - fix a memory leak in the /etc/hosts lookup code when hosts don't
114 resolve and are not in hosts (analyzed by sten).
115
1167.02 Tue Aug 14 04:27:58 CEST 2012
117 - AnyEvent::Util::run_cmd could block indefinitely (analyzed and test
118 program by Yorhel).
119 - verified that AnyEvent::Socket follows RFC5952.
120 - try to parse "ADDR#PORT" in addition to "ADDR PORT".
121
1227.01 Sun May 13 01:03:17 CEST 2012
123 - fail with EPROTO in AnyEvent::Handle wqhen TLS is requested but
124 not available, instead of throwing an exception.
125 - use File::Spec to get the tmpdir in t/*, to avoid needless
126 failures on (most, not mine :) windows boxes.
127 - new handle read types: tls_detect and tls_autostart.
128
1297.0 Fri Apr 13 06:33:30 CEST 2012
130 - child watchers are broken in POE 1.352 (also many earlier
131 versions) and there seems to be no way to work around it, as POE
132 itself is inherently racy. Document this regression and add a delay
133 in t/68_poe_03_child.t for the time being.
134 - new module AnyEvent::IO, that is a frontend to either a pure-perl
135 synchronous I/O implementation (AnyEvent::IO::Perl), or to the
136 asynchronous I/O extension IO::AIO (AnyEvent::IO::IOAIO).
137 - load /etc/hosts only when DNS has no answer.
138 - stat /etc/hosts on every access and reload it if it changed.
139 - load /etc/hosts via AnyEvent::IO - potentially asynchronous.
140 - fix a buggy croak in the dh parameter reading in AnyEvent::TLS.
141 - AnyEvent::Log log_to_file and log_to_path now use AnyEvent::IO. As
142 a side effect, they now use true appending as opposed to libc
143 appending, and the file might not have been opened when the function
144 returns.
145 - the default logging level was not properly documented in a variety of
146 places, this has hopefully been rectified.
147 - updated uts46data.pl for unicode 6.1.0.
148 - made log messages generated by AnyEvent submodules not
149 include the package name anymore, as it will be logged
150 by default already.
151 - upgrade to the trick used by common-sense 3.5 to work
152 around extra warning torture/breakage under perl 5.15.x.
153 - log messages by anyevent are now ucfirst, are usually full
154 sentences and do no longer include the package name.
155 - the storable read type would sometimes throw an exception instead
156 of causing EBADMSG (analyzed by Maxime Soulé).
157 - work around a bug in openssl 1.0.1 which enforces a minimum rsa
158 keysize (reported by Paul Howarth).
159 - documented Rocco fabricating statements about AnyEvent and me.
160
1616.14 Tue Jan 31 20:00:24 CET 2012
162 - AnyEvent::Impl::Tk was broken due to a mysterious "die" inside,
163 probably an editing mistake (reported by Darin McBride).
164
1656.13 Thu Jan 12 07:27:01 CET 2012
166 - AnyEvent::Util::fork_call checked for POSIX availability in the wrong way,
167 causing it to fail if POSIX wasn't loaded already (analysed by Rock Power).
168 - AnyEvent::Handle::push_read (line => did pass $1 directly, so
169 regex-matching inside the callback would change the parameter, despite
170 $1 being dynamically scoped per-block. this perl bug is now being
171 worked around (testcase by Cindy Wang).
172
1736.12 Mon Dec 12 13:21:10 CET 2011
174 - $! was clobbered by subsequent calls in tcp_connect, due to
175 the postpone not saving/restoring it (analyzed by Richard Garnish).
176
1776.11 Tue Nov 22 10:36:05 CET 2011
178 - Tk cannot create windows when tainted unless you set a title, so
179 set a dummy title for AnyEvent's dummy mainwindow. How dummy.
180 - escape any nonprintable/nonascii characters when stringifying
181 backtraces.
182 - log the reason loading an interface module fails to load at level debug.
183
1846.1 Tue Oct 4 19:44:30 CEST 2011
185 - INCOMPATIBLE CHANGE: the default log level is now "4" (error
186 and above), and some messages inside AnyEvent have been elevated
187 to higher log levels to print by default.
188 - AnyEvent::log, unlike AnyEvent::Log::log, did not always exit on fatal.
189 - work around yet another signal race bug, newly introduced by POE. My
190 how that module sucks.
191 - add google public ipv6 dns as fallback servers, so AE::DNS can finally
192 support ipv6-only operation even with defaults (also use ::1 instead
193 of 127.0.0.1).
194 - give better guidelines on how to chose a logging level.
195 - AnyEvent::Log can now cap (limit) log message priority by context.
196 - AE::log will now save and restore $! and $@, so callers don't have to.
197 - AE::Strict now checks that registered fds are still valid at regular
198 intervals.
199 - skip frequently called 'require' calls if the module appears loaded.
200 - increase signal latency test timer considerably for poor little osx,
201 which can't be bothered with basic posix functionality such as signal
202 delivery.
203 - mention some debugging aids in the tutorial.
204 - new env variable PERL_ANYEVENT_MAX_SIGNAL_LATENCY.
205 - private (%namespace) log contexts in PERL_ANYEVENT_LOG now have the name
206 of the context as title.
207 - AnyEvent::Strict would not restore read-write access to $_ when a callback
208 died - this is a user bug, but since AE::Strict exists to help dbeugging...
209 - detect the extremely evil IO::Async::Loop::AnyEvent module and refuse
210 to work with it - the author is well aware of the issues but refuses
211 to fix them. This extreme step has been taken because the module actively
212 attacks the goals of AnyEvent and makes it harder for module authors to use
213 AnyEvent.
214
2156.02 Fri Aug 26 20:08:31 CEST 2011
216 - log_to_syslog now accepts facility strings.
217 - log_to_syslog would not allow facilities - "$facility|$level"
218 *literally* means that string. sheesh.
219 - the CPAN parser got confused and thought we provide Sys::Syslog, try
220 to work around.
221 - renamed AnyEvent::Impl::FLTK2 to FLTK.
222
2236.01 Fri Aug 26 07:04:11 CEST 2011
224 - INCOMPATIBLE CHANGE: AnyEvent::DNS resource records now include
225 the TTL as fourth element - this affects ->resolve, ->request
226 and ->dns_unpack indirectly. Convenience functions such as
227 AnyEvent::DNS::{a,txt,mx,srv} are not affected, neither is
228 AnyEvent::Socket.
229 - INCOMPATIBLE CHANGE: backend authors now should not implement
230 one_event or loop, but instead the AnyEvent::CondVar::_wait and _poll
231 methods.
232 - INCOMPATIBLE CHANGE: verbosity levels have been redefined, see
233 AnyEvent (PERL_ANYEVENT_VERBOSE) and AnyEvent::Log for details.
234 - DEPRECIATION: please change
235 "use AnyEvent::Impl::Perl" to "use AnyEvent::Loop" and
236 "AnyEvent::Impl::Perl::loop" to "AnyEvent::Loop::run"
237 in your programs, for future compatibility.
238 - *blush* idle watcher emulation was borked.
239 - AnyEvent::Strict now checks the AE::xxx functions as well.
240 - querying too long domain names would barf deep inside AnyEvent::DNS - since
241 this is hard to avoid for applications, AnyEvent:.DNS::request now checks for
242 for too long domain names explicitly and passes an undef to the callback
243 instead (reported by James Bromberger).
244 - make AnyEvent::Util more robust against EINTR (analyzed by Andrew Suffield).
245 - new functions: AnyEvent::postpone, AnyEvent::log.
246 - new module: AnyEvent::Log, for simple logging needs.
247 - new env variable, PERL_ANYEVENT_LOG, for fine-grained logging config.
248 - new env variable, PERL_ANYEVENT_DEBUG_SHELL, to automatically start
249 a debug shell.
250 - new env variable, PERL_ANYEVENT_DEBUG_WRAP, to automatically enable
251 watcher instrumentation/wrapping (see AnyEvent::Debug::wrap).
252 - new env variable, PERL_ANYEVENT_HOSTS, to override the place where
253 /etc/hosts can be found.
254 - an empty PERL_ANYEVENT_RESOLV_CONF now selects the default
255 configuration.
256 - env variables can now be specified using AE_xyz instead of
257 PERL_ANYEVENT_xyz, subject to some rules.
258 - lots of new AnyEvent::Debug functionality.
259 - AnyEvent::Strict now checks for modifications of $_ by the callback
260 (by setting it to readonly while executing the callback, if possible).
261 - IO::Async sometimes cancels the wrong timer when you call cancel_timer.
262 Invest extra overhead to work around this peculiar behaviour.
263 - split the pure perl backend into a loop module and an impl
264 module (AnyEvent::Loop and AnyEvent::Impl::Perl).
265 - some read types (regex, netstring, json, storable) could
266 errornously keep an AE::Handle object alive on their own
267 (reported by Mohammad Toossi).
268 - AnyEvent::Socket::resolve_sockaddr and all functions using it now
269 supports /etc/hosts overrides.
270 - add more workarounds around child watcher bugs in IO::Async
271 and (newly introduced) in POE.
272 - use glib child watchers and try to work around its limitations,
273 if possible (based on a patch by Kevin Ryde).
274 - add support for FLTK (fltk2), via AnyEvent::Impl::FLTK2.
275 - many read types in AnyEvent::Handle would malfunction during
276 transitions to TLS because of a "delete $self->{rbuf}".
277 (reported by Gerald Galster).
278 - improve condvar blocking wait performance (probably only noticable
279 with faster backends), also make it easier for event loops that
280 do not support blocking waits.
281 - PERL_ANYEVENT_MODEL now also allows full module names, although
282 the usefulness of this feature is not yet clear.
283 - add tests for all supported event loops, conditional on PERL_ANYEVENT_LOOP_TESTS.
284 - add t/80_ssltext.t.
285 - no longer use AUTOLOAD for the initial loading (for no very good reason).
286 - tcp_server now tries to unlink unix domain sockets when it gets destroyed.
287 - do not run event loop detection multiple times when programs call their cached
288 copies of AnyEvent::detect.
289 - parse_hostport now accepts absolute paths als unix domain sockets.
290 - suppress the idiotic warning inside POE by patching the POE::Kernel
291 run flags instead of asking the user to jump through contortions.
292 - add Task::Weaken as dependency (as recommends only at this time).
293 - replace PF_UNSPEC by 0, for good old minix.
294
2956.0 (not officially released, but was on CPAN for about 15 minutes :/)
296
2975.34 Thu May 12 09:59:41 CEST 2011
298 - AE::TLS didn't support both cert_file and key at the same time,
299 due to some copy&paste bug (reported and analyzed by vti).
300 - improve AE::Handle timeout documentation slightly.
301
3025.33 Mon Feb 7 21:51:41 CET 2011
303 - never appeared on CPAN, for unknown reasons.
304 - AnyEvent::Handle on_connect_error - do not try to destroy a handle
305 object when it's already gone (analyzed by Ryan Bullock).
306 - update idna algorithm(s) to UTS#46 version 6.0.0.
307 - preserve trailing dot in idn_to_ascii - testsuite says to preserve,
308 uts46 says it's an error, and testsuite makes more sense.
309 - remove leading dots in idn_to_ascii - testsuite says to remove,
310 uts46 says it's an error, and testsuite makes more sense.
311 - fixed a bug in idn ignored character processing.
312 - passes the uts46 6.0.0 testsuite except for three cases which
313 apparently are bugs in the testsuite itself.
314 - passes the uts46 6.0.1 testsuite except for eleven cases which
315 apparently are bugs in the testsuite itself.
316
3175.31 Sun Jan 23 11:49:19 CET 2011
318 - work around a POE bug where POE would first poll for new events and
319 invoke handlers BEFORE handling the message queue.
320 - fix parsing of "domain" statement in resolv.conf (Johannes Plunien).
321 - stop_read now does just that in TLS mode. see its
322 documentation for caveats.
323 - update to common::sense 3.4 warning set (some warnings in perl impose
324 an arbitrary 32kb limit on regex matches, which affects
325 AnyEvent::Handle).
326 - AE::Handle had a broken rbuf_max method and missing wbuf_max method.
327 - new experimental AE::Handle::resettls method.
328
3295.3 Fri Dec 31 05:48:13 CET 2010
330 - major IOAsync patch by Paul Evans - 0.33 is required, as the major
331 issues with IO::Async have been solved in that version, including
332 the availabilty of a default loop!
333 - implement new wbuf_max parameter for AnyEvent::Handle.
334 - added Cocoa::EventLoop backend by Daisuke Murase.
335 - the default testsuite now honors PERL_ANYEVENT_MODEL (Paul Evans).
336 - clarify handle on_prepare documentation.
337 - set CLOEXEC flag on parent-side fd's in AnyEvent::Util::run_cmd,
338 to avoid leaking them to any child processes.
339 - try to force files into the arch-dependend lib directory, for the
340 half-"turly obsessive compulsive person"(s) that depend on it.
341
3425.29 Sun Dec 5 10:49:21 CET 2010
343 - convert EV backend to EV 4.00 API (so better upgrade EV too).
344 - AE::Handle: implement dynamic read_size adjustment, add
345 max_read_size parameter, reduce default read_size to 2048.
346 - add some FAQ entries for servers and/or on_eof vs. on_error.
347 - work around OS bugs (cygwin again) when lingering in AE::Handle.
348 - reduce memory usage slightly when connect was used in AE::Handle.
349 - wrok around more segfaults in Tk.
350 - document tls_ctx => undef as valid.
351 - detect WNOHANG value at compiletime.
352 - include some probably linux-only support for building AnyEvent
353 as part of the perl core.
354 - improve compatibility of t/08_idna.t to perl 5.8.
355 - make AnyEvent::Strict truly optional (in case somebody builds a
356 minimal perl...).
357
3585.28 Wed Oct 13 04:14:23 CEST 2010
359 - due to a glitch, AnyEvent's internal getprotobyname wasn't
360 actually used in tcp_server.
361 - implement AnyEvent::CondVar->new method, which was documented but
362 missing.
363 - support autoloading of read types in unshift_read.
364 - AnyEvent::Strict: do the test for invalid fh arguments differently,
365 resulting in more informative error messages.
366 - new AnyEvent::FAQ manpage.
367 - clarify that you can actually call push_read etc. in on_read.
368 - put keys and certs in /tmp, not into the current directory,
369 and document this in the SECURITY CONSIDERATIONS sections
370 of AnyEvent::TLS.
371 - support ";" as resolv.conf comment character in AnyEvent::DNS, as
372 requested by Ron Isaacson.
373 - document $AnyEvent::DNS::RESOLVER variable.
374 - incorporated major typo patches by Abhijit Menon-Sen.
375 - AnyEvent::Handle now croaks when negative timeouts are passed.
376 - add a shitload of TCP_xxx sockopt name constants to AnyEvent::Util,
377 undocumented, but who knows when they might come in handy.
378
3795.271 Tue Jun 8 12:05:46 CEST 2010
380 - backport to perl 5.8.x.
381
3825.27 Sun Jun 6 12:12:05 CEST 2010
383 - postpone differently in AnyEvent::Socket now, as
384 when not, canceling the connection attempt might fail
385 (found by Felix Antonius Wilhelm Ostmann).
386 - explicitly check for non-stream sockets in AE::Handle, too many
387 clueless people fell into the trap of this somehow working.
388 - simplified and reworked the "OTHER MODULES" section.
389 - better/more condvar examples.
390
3915.261 Wed Apr 28 16:13:36 CEST 2010
392 - AF_INET6 was not properly used from Socket6 during configuration
393 time (found by Andy Grundman).
394
3955.26 Mon Apr 12 04:49:35 CEST 2010
396 - don't generate spurious readyness notifications when select
397 returns EINTR in the pure perl backend (analysed and testcase
398 by Bas Denissen).
399 - give same examples in SYNOPSIS sections of AnyEvent and AE.
400 - provider faster implementations for AE::time/now/now_update for
401 EV and Perl backends.
402 - provide faster AE::cv and AE::time implementations for all
403 backends.
404 - no longer support lower/mixed-case signal names (this was
405 never documented nor universally supported).
406 - some more memory µ-optimisations, and somewhat less messy
407 code to cope with both AnyEvent and AE APIs.
408
4095.251 Sat Mar 13 00:58:21 CET 2010
410 - make test failed if it was run before make install on perl <=
411 5.8.8, and older versions of perl make our live very difficult,
412 so write constants.pl during Makefile.PL time.
413
4145.25 Sat Mar 13 00:23:14 CET 2010
415 - fix a race condition in AnyEvent::Handle that would cause
416 a "bio_read: unsupported method" error (in _another_
417 ssl connection) after user code threw an exception.
418 - added AnyEvent::Handle->destroyed method.
419 - speed up AnyEvent::Socket::format_address and ::format_ipv6.
420 - the AnyEvent::Util::fh_nonblocking function worked only
421 by lucky accident on win32.
422 - smaller and faster AnyEvent::Util::fh_nonblocking.
423 - when the (required!) Time::HiRes module is missing, AnyEvent
424 did not fall back to built-in time properly.
425 - do not load Fcntl at runtime, saving memory and loading time.
426 - precompile a number of constants and use them instead of runtime
427 detection and eval.
428 - free detection code after detection and similar memory optimisations.
429 - Perl backend timer interval best effort drift has been improved
430 (same algorithm as EV).
431 - update unicode idna mapping table.
432
4335.24 Tue Jan 5 11:39:43 CET 2010
434 - cygwin never reports errors from failed connects. 1.5 just gives
435 you continous readyness and EAGAIN, 1.7 is even more broken and
436 just hangs. work around both issues in a cygwin-specific hack.
437 - improve idle watcher documentation slightly.
438
4395.23 Sun Dec 20 23:48:00 CET 2009
440 - support IDNs in resolve_sockaddr, and therefore in tcp_connect.
441 - implement punycode_encode/decode, idn_nameprep,
442 idn_to_ascii and idn_to_unicode operations in AnyEvent::Util.
443 - provide $AE::VERSION.
444 - removed traces of "no strict 'refs'".
445
4465.22 Sat Dec 5 03:51:13 CET 2009
447 - downgrade-or-fail in AnyEvent::Handle::push_write, to
448 diagnose encoding failures earlier and more succinctly.
449 (this works around bugs in perl, throwing away encoding info
450 when passing scalar data to extensions).
451 - add more examples to AnyEvent::Socket manpage.
452 - upgrade internal warning set to the same as common::sense 2.03.
453 - use pack "n/a*" for pre-5.8.9 perl compatibility in AnyEvent::DNS
454 (John Beppu).
455 - AnyEvent::Socket::inet_aton now properly supports ipv6, as documented.
456 - add google public dns servers to fallback server set and make sure
457 we load-balance properly between the three sets. also add all
458 fallback dns servers, not just a random one, to each dns config.
459
4605.21 Thu Nov 19 02:48:47 CET 2009
461 - fix a problem where socket constants were called with parameters
462 (spotted by David Friedland).
463 - fork_call never use'd POSIX (reported by Daisuke Maki).
464 - improve perl 5.6 compatibility further (but it still won't work
465 unless you rip out everything but the core).
466 - prefer Net::DNS::Resolver over ipconfig, if installed, on win32.
467 uses 10MB of RAM, but doesn't open a console window. *sigh*.
468
4695.202 Wed Oct 14 22:35:44 CEST 2009
470 - AnyEvent::DNS would unexpectedly clobber $_ under windows
471 (analysed by Matthias Waldorf).
472 - AnyEvent::Handle::run_cmd can now pass the PID of the
473 newly-created process, which is much less useful than it might
474 sound (based on patch by Yann Kerherve).
475
4765.201 Tue Sep 29 12:09:25 CEST 2009
477 - AnyEvent:Handle::on_starttls/on_stoptls methods were broken
478 (reported by Torsten Foertsch).
479 - common::sense 2.0 could cause tcp_server to throw an exception
480 (analysed by elmex).
481
4825.2 Mon Sep 14 07:04:49 CEST 2009
483 - INCOMPATIBLE CHANGE: do no longer support register_read_type
484 and register_write_type in AnyEvent::Handle, instead support
485 package names (the facility was mostly abused).
486 - implement "packagename-as-read/write type" support in
487 AnyEvent::Handle.
488 - AnyEvent::Handle: new options "keepalive" and "oobinline".
489 - oobinline set by default to avoid security issues.
490 - the pure-perl event loop backend wrongly detected times() fallback
491 support (spotted by Pavel Boldin).
492
4935.12 Tue Sep 1 20:26:50 CEST 2009
494 - be more lenient when parsing resolv.conf files, as some people
495 use hashmarks as comment indicator inside directives (reported
496 by Michael S. Fischer).
497 - use same set of warnings as common::sense 2.0.
498 - fix a potential 32 bit overflow issue due to perl having problems
499 with large hex constants in the Perl backend.
500 - do not use the slower 5.6 version fo the enc_name code in AnyEvent::DNS
501 by default.
502 - fix some prototypes.
503 - (optionally) require Async::Interrupt 1.02.
504 - replace opendns fallbacks by verizon/level3 public dns servers.
505
5065.112 Fri Aug 21 13:59:04 CEST 2009
507 - AnyEvent::Handle could cause hard crashes in Net::SSLeay when
508 using starttls with an illegal/uncreatable tls context.
509 - port to broken win32 perls.
510
5115.111 Sun Aug 16 18:44:36 CEST 2009
512 - perl errornously requires sockaddr_un structures returned by the
513 kernel to have a specific length. work around this bug by padding
514 sockaddr_un structures (found to fail on OS X, as perl only uses
515 the correct code for GNU/Linux, but likely to fail on other
516 OSes as well).
517
5185.11 Wed Aug 12 17:49:37 CEST 2009
519 - fixed a bug in run_cmd with input from filename.
520 - create a json coder object when none is given in write direction
521 to reduce dependency on JSON.
522
5235.1 Tue Aug 11 03:17:32 CEST 2009
524 - tcp_connect now reports the correct error in $!, not an unrelated
525 one (was broken since 4.91).
526 - AE::Handle did not properly clear rtimeout/wtimeout watchers
527 when retrying a tcp_connect.
528 - new functions: AnyEvent::Util::run_cmd and close_all_fds_except.
529 - fixed wrong documentation in AE::Socket.
530
5315.01 Mon Aug 10 03:16:32 CEST 2009
532 - last release broke EV child watchers :/.
533
5345.0 Sun Aug 9 17:32:09 CEST 2009
535 - officially introduce and document the AE API.
536 - lowering the timeout at runtime did not have immediate effect
537 in AnyEvent::Handle.
538 - AE::Handle now has separate and independent read and write timeout
539 settings.
540 - probe for Guard module when AnyEvent::Util is loaded, not at
541 runtime (Event doesn't like eval "use" from callbacks, and it avoids
542 exporting the probe function).
543 - Impl::IOAsync's io watchers did not use the filehandle emulation
544 code to work around it's watcher limits.
545 - work around IO::Async's problems with overloaded objects.
546 - try to disable wearnings inside IO::Async, it's just too buggy.
547 - do no longer use Event by default, only EV or Perl, Event is too buggy.
548 - hackishly provide fast AE:: interface when EV or Perl is the backend.
549 - inherit VERSION from AnyEvent::VERSION in most modules.
550
5514.91 Thu Aug 6 15:42:45 CEST 2009
552 - AE::Handle::starttls could get out of sync when the read buffer
553 already contains some TLS handshake.
554 - AE::Handle did not properly free the TLS session, causing bigger
555 memory leaks in Net::SSLeay than expected (reported by toaster).
556 - AE::Socket::tcp_connect will now artificially delay invoking
557 the callback to avoid returning after invoking the callback.
558 - convert many internal watcher uses to AE API.
559
5604.9 Sat Aug 1 11:07:01 CEST 2009
561 - Glib silently fails when registering a timeout with
562 a negative value, so avoid that.
563 - call condvar callback immediately when it is set
564 after the condition is already signalled.
565 - check rbuf_max condition only after trying to consume
566 data.
567 - normalise signal numbers to names when using the ae's
568 signal handling, but do not document this (yet).
569 - pure perl signal emulation did not properly set nonblocking
570 mode on the signal pipe on win32 (this is unlikely to have caused
571 any issues).
572 - new module: AnyEvent::Debug.
573 - AnyEvent::Strict now checks that a signal was specified
574 by name (not name or number).
575 - reduce memory footprint in typical cases by ~50kb by
576 only compiling signal code when necessary.
577 - add AnyEvent::Handle->rbuf_max.
578 - grab the AE namespace for future new API, implement stubs for
579 said future/alternative API.
580 - new function: AnyEvent::Socket::format_hostport.
581
5824.881 Tue Jul 28 12:51:53 CEST 2009
583 - work around a bug in local in pre-5.10 perls, causing
584 AnyEvent::Handle to recurse when it should not
585 (analysed by elmex).
586
5874.88 Tue Jul 28 04:04:37 CEST 2009
588 - re-bless the handle into a dummy package after calling
589 AnyEvent::Handle::destroy, so the user does not need to check
590 for errors after every push_write etc.
591 - do not attempt to run t/02_signals.t on obviously broken
592 platforms.
593
5944.87 Sun Jul 26 02:06:16 CEST 2009
595 - do not attempt to linger when there is no longer a valid fh.
596 - enforce tls mode to be either accept or connect, do not simply
597 segfault in Net::SSLeay.
598 - AnyEvent::Handle can now call tcp_connect itself (new parameters
599 connect, on_prepare, on_connect and on_connect_error). Updated
600 tutorial accordingly.
601 - add AnyEvent::Impl::Irssi backend.
602
6034.86 Mon Jul 20 23:52:29 CEST 2009
604 - since the verbose warning is not enough, explicitly document
605 that versions before 1.33 of Net::SSLeay are not secure.
606 - work around signal handling races in Event and (...) Event::Lib.
607 - try to align signal-race timer to full-second boundaries.
608 - work around Tk not liking negative timeouts.
609 - don't complain of different grades of environmental unfriendlyness
610 in IO::Async.
611
6124.85 Sat Jul 18 06:16:14 CEST 2009
613 - nail the signal race problem in perl once and for all
614 (see $AnyEvent::MAX_SIGNAL_LATENCY).
615 - take advantage of Async::Interrupt if it is available.
616 - load Time::HiRes and Guard modules on demand only.
617 - add optional/recommended modules section to AnyEvent
618 documentation.
619 - reduce memory usage considerably (and reduce startup penalty)
620 by not using "strict", "warnings" and "overload" modules.
621 - work around buggy windows/openbsd perls and provide EBADMSG
622 and EPROTO ourselves when missing.
623 - improve perl 5.6 compatibility of the core event loop.
624 - made Net::SSLeay version 1.33 a soft requirement.
625
6264.83 Fri Jul 17 16:56:26 CEST 2009
627 - implement AnyEvent::Socket::getprotobyname.
628 - AnyEvent::CondVar's will now detect recursive blocking
629 waits and will croak, as too many people fall into
630 this trap.
631 - AnyEvent::Handle will now call ->destroy on itself after
632 executing the on_error callback, instead of doing some
633 half-baked internal shutdown, for fatal errors.
634 - clarify on_eof behaviour w.r.t. the read queue and
635 on_read callbacks.
636 - ignore some possible spurious wake-ups in tcp_connect.
637
6384.82 Sat Jul 11 00:34:55 CEST 2009
639 - POE and Event backends didn't accept some callable objects as
640 callbacks.
641 - use Config module instead of POSIX module to detect signal names
642 in AnyEvent::Strict and AnyEvent::Impl::EventLib, as the POSIX
643 module doesn't even have all POSIX signals :/.
644 - use more workarounds around the many refcnt/corruption bugs in
645 Event::Lib.
646 - work around a race condition in perl's select, causing t/03_child.t
647 to rarely fail.
648
6494.81 Thu Jul 9 10:30:30 CEST 2009
650 - AnyEvent::Handle didn't properly diagnose write errors
651 (it expected -1 from syswrite, how lame... :).
652 - support file descriptors in addition to file handles
653 in AnyEvent->io.
654 - new env variables: PERL_ANYEVENT_RESOLV_CONF,
655 PERL_ANYEVENT_MAX_OUTSTANDING_DNS, PERL_ANYEVENT_CA_FILE
656 and PERL_ANYEVENT_CA_PATH.
657 - provide a sensible synopsis section for AnyEvent::TLS.
658 - add a "supported backends" section to the manpage.
659 - added simple io watcher test to testsuite, using a
660 portable_socketpair.
661 - tried to improve the stability of the Event::Lib backend,
662 YMMV.
663
6644.8 Mon Jul 6 23:45:16 CEST 2009
665 - AnyEvent::DNS did not properly follow CNAME records with
666 uppercase targets.
667 - AnyEvent::DNS would errornously return AAAA records
668 with v4 mapped addresses (a faulty record) as ipv4 addresses,
669 causing AnyEvent::Socket to throw an exception.
670 - added new module AnyEvent::TLS for easier SSL/TLS context
671 creation, with many options including hostname verification,
672 secure default configuration, lots of documentation and,
673 predefined diffie-hellman keys for perfect forward security
674 and much more. get it while it's still fresh!
675 - use AnyEvent::TLS in AnyEvent::Handle for context management.
676 - load AnyEvent::Handle only on demand in AnyEvent::DNS,
677 so AnyEvent::Socket users have smaller memory footprint
678 in the common case.
679 - add AnyEvent::Handle->push_shutdown method.
680 - add an additional error message parameter to AnyEvent::Handle's
681 on_error callback (for TLS, $! is still available).
682 - add AnyEvent::Handle on_starttls/on_stoptls callbacks.
683 - make AnyEvent::Handle more robust against early connection
684 failures (during new), and return C<undef> in that case
685 from the constructor.
686 - AnyEvent::Handle will now try to load only JSON::XS first,
687 then fall back to JSON.
688 - format_ipv4/format_ipv6 are now exported by default, for symmetry,
689 and because it was documented that way.
690
6914.451 Fri Jul 3 00:28:58 CEST 2009
692 - do not clear rbuf when shutting down an AnyEvent::Handle
693 object - doing so breaks AnyEvent::HTTP.
694
6954.45 Mon Jun 29 22:59:26 CEST 2009
696 - a write error could cause AnyEvent::Handle to create
697 an I/O watcher with an undefined $fh.
698 - special-case mapped ipv4 addresses in both
699 AnyEvent::Socket::format_address and parse_address,
700 to treat them just like ipv4 addresses.
701 - updated and overhauled the AnyEvent::Intro doc.
702 - implement AnyEvent::Socket::format_ipv4/ipv6.
703 - slightly speed up portable_pipe/socketpair.
704 - expand condvar begin/end documentation.
705
7064.42 Fri Jun 26 08:32:18 CEST 2009
3 - preliminary and neccesarily incomplete support for IO::Async. 707 - preliminary and neccesarily incomplete support for IO::Async.
4 - reset SIGCHLD to DEFAULT when AnyEvent is loaded, in 708 - reset SIGCHLD to DEFAULT when AnyEvent is loaded, in
5 case it was set to IGNORE, to ensure we can catch child statuses 709 case it was set to IGNORE, to ensure we can catch child statuses
6 even when the calling env acts stupidly. 710 even when the calling env acts stupidly.
7 - updated benchmarks with IO::Async, which performs very well. 711 - updated benchmarks with IO::Async, which performs very well.
… …
1084.3 Fri Oct 3 09:18:43 CEST 2008 8124.3 Fri Oct 3 09:18:43 CEST 2008
109 - AnyEvent will now install a no-op signal handler for SIGPIPE, 813 - AnyEvent will now install a no-op signal handler for SIGPIPE,
110 unless one has been installed already. 814 unless one has been installed already.
111 - warn about not loaidng AnyEvent::Impl::POE early enough 815 - warn about not loaidng AnyEvent::Impl::POE early enough
112 (patch by Adam Rosenstein). 816 (patch by Adam Rosenstein).
113 - fixed a great number of bugs and corner cases in AnyEvent::Handle: 817 - fixed a great number of bugs and corner cases in AnyEvent::Handle.
114 - fix a bug in where in SSL connect mode, the 818 - fix a bug in where in SSL connect mode, the
115 client would first wait for some data by the server and 819 client would first wait for some data by the server and
116 otherwise hang. (reported and analyzed in an absolutely 820 otherwise hang. (reported and analysed in an absolutely
117 exemplary manner by Adam Rosenstein). 821 exemplary manner by Adam Rosenstein).
118 - fix a bug in where SSL EOF would not be treated 822 - fix a bug in where SSL EOF would not be treated
119 as stream EOF, putting the connection into a hung state. 823 as stream EOF, putting the connection into a hung state.
120 (reported and analyzed in an absolutely exemplary manner by Adam 824 (reported and analysed in an absolutely exemplary manner by Adam
121 Rosenstein). 825 Rosenstein).
122 - fix a potential segfault when the TLS context 826 - fix a potential segfault when the TLS context
123 would go missing in a read callback (e.g. due to stoptls) - 827 would go missing in a read callback (e.g. due to stoptls) -
124 Net::SSLeay of course makes no type checking whatsoever. 828 Net::SSLeay of course makes no type checking whatsoever.
125 - AnyEvent::Handle will not stop the read watcher in TLS mode. 829 - AnyEvent::Handle will not stop the read watcher in TLS mode.

Diff Legend

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