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

Comparing AnyEvent/Changes (file contents):
Revision 1.255 by root, Mon Jun 29 21:00:32 2009 UTC vs.
Revision 1.581 by root, Thu Oct 3 17:54:05 2013 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
11
12TODO: mention AnyEvent::fork etc.
13 - calling AnyEvent->now_update with AnyEvent::Impl::Perl caused an
14 endless loop (reported by Dietrich Rebmann).
15 - add tlsv1_1 and tlsv1_2 protocols to AnyEvent::TLS
16 (patch by Maxime Soulé).
17 - force a toplevel domain name in t/81_hosts.t (analyzed by
18 David Jack Wange Olrik).
19 - document that AnyEvent::Log uses AnyEvent::IO.
20
217.05 Wed Aug 21 10:38:08 CEST 2013
22 - uts46data.pl couldn't be found due to wrong naming of the file
23 (reported by Fulko Hew).
24 - handle lone \015's properly in AE::Handle's default line read
25 (reported by various people).
26 - untaint ip addresses found in /etc/hosts (patch by José Micó).
27 - the memleak fix in 7.03 caused resolving via /etc/hosts to always
28 fail on first use (reported and testcase by Andrew Whatson).
29 - expose AnyEvent::Log::format_time, and allow users to redefine it.
30 - expose AnyEvent::Log::default_format, and allow redefinition.
31 - expose AnyEvent::Log::fatal_exit, to allow redefinition.
32 - AnyEvent::Debug shell can now run coro shell commands, if available.
33 - t/63* tests were wrongly in MANIFEST.
34 - kernel.org's finger server went MIA, switch to freebsd.org and
35 icculus.org.
36 - clarify that IO::AIO and AnyEvent::AIO are needed for AnyEvent::IO
37 to function asynchronously (reported by Daniel Carosone).
38 - hard-disable $^W in most tests, it generates too much garbage output.
39 - use a (hopefully) more future-proof method to emulate common::sense.
40 - upgrade to UTS-46:6.2.0.
41 - switch to INSTLIB from INSTLIBDIR, as INSTLIBDIR was wrongly
42 documented. should not affect anything.
43
447.04 Thu Nov 15 02:23:58 CET 2012
45 - debugging code left in.
46
477.03 Thu Nov 15 02:16:11 CET 2012
48 - AnyEvent::Socket::inet_aton did not work when dns resolution was
49 used to find the addresses (analyzed and patch by Maxime Soulé).
50 - fix a memory leak in the /etc/hosts lookup code when hosts don't
51 resolve and are not in hosts (analyzed by sten).
52
537.02 Tue Aug 14 04:27:58 CEST 2012
54 - AnyEvent::Util::run_cmd could block indefinitely (analyzed and test
55 program by Yorhel).
56 - verified that AnyEvent::Socket follows RFC5952.
57 - try to parse "ADDR#PORT" in addition to "ADDR PORT".
58
597.01 Sun May 13 01:03:17 CEST 2012
60 - fail with EPROTO in AnyEvent::Handle wqhen TLS is requested but
61 not available, instead of throwing an exception.
62 - use File::Spec to get the tmpdir in t/*, to avoid needless
63 failures on (most, not mine :) windows boxes.
64 - new handle read types: tls_detect and tls_autostart.
65
667.0 Fri Apr 13 06:33:30 CEST 2012
67 - child watchers are broken in POE 1.352 (also many earlier
68 versions) and there seems to be no way to work around it, as POE
69 itself is inherently racy. Document this regression and add a delay
70 in t/68_poe_03_child.t for the time being.
71 - new module AnyEvent::IO, that is a frontend to either a pure-perl
72 synchronous I/O implementation (AnyEvent::IO::Perl), or to the
73 asynchronous I/O extension IO::AIO (AnyEvent::IO::IOAIO).
74 - load /etc/hosts only when DNS has no answer.
75 - stat /etc/hosts on every access and reload it if it changed.
76 - load /etc/hosts via AnyEvent::IO - potentially asynchronous.
77 - fix a buggy croak in the dh parameter reading in AnyEvent::TLS.
78 - AnyEvent::Log log_to_file and log_to_path now use AnyEvent::IO. As
79 a side effect, they now use true appending as opposed to libc
80 appending, and the file might not have been opened when the function
81 returns.
82 - the default logging level was not properly documented in a variety of
83 places, this has hopefully been rectified.
84 - updated uts46data.pl for unicode 6.1.0.
85 - made log messages generated by AnyEvent submodules not
86 include the package name anymore, as it will be logged
87 by default already.
88 - upgrade to the trick used by common-sense 3.5 to work
89 around extra warning torture/breakage under perl 5.15.x.
90 - log messages by anyevent are now ucfirst, are usually full
91 sentences and do no longer include the package name.
92 - the storable read type would sometimes throw an exception instead
93 of causing EBADMSG (analyzed by Maxime Soulé).
94 - work around a bug in openssl 1.0.1 which enforces a minimum rsa
95 keysize (reported by Paul Howarth).
96 - documented Rocco fabricating statements about AnyEvent and me.
97
986.14 Tue Jan 31 20:00:24 CET 2012
99 - AnyEvent::Impl::Tk was broken due to a mysterious "die" inside,
100 probably an editing mistake (reported by Darin McBride).
101
1026.13 Thu Jan 12 07:27:01 CET 2012
103 - AnyEvent::Util::fork_call checked for POSIX availability in the wrong way,
104 causing it to fail if POSIX wasn't loaded already (analysed by Rock Power).
105 - AnyEvent::Handle::push_read (line => did pass $1 directly, so
106 regex-matching inside the callback would change the parameter, despite
107 $1 being dynamically scoped per-block. this perl bug is now being
108 worked around (testcase by Cindy Wang).
109
1106.12 Mon Dec 12 13:21:10 CET 2011
111 - $! was clobbered by subsequent calls in tcp_connect, due to
112 the postpone not saving/restoring it (analyzed by Richard Garnish).
113
1146.11 Tue Nov 22 10:36:05 CET 2011
115 - Tk cannot create windows when tainted unless you set a title, so
116 set a dummy title for AnyEvent's dummy mainwindow. How dummy.
117 - escape any nonprintable/nonascii characters when stringifying
118 backtraces.
119 - log the reason loading an interface module fails to load at level debug.
120
1216.1 Tue Oct 4 19:44:30 CEST 2011
122 - INCOMPATIBLE CHANGE: the default log level is now "4" (error
123 and above), and some messages inside AnyEvent have been elevated
124 to higher log levels to print by default.
125 - AnyEvent::log, unlike AnyEvent::Log::log, did not always exit on fatal.
126 - work around yet another signal race bug, newly introduced by POE. My
127 how that module sucks.
128 - add google public ipv6 dns as fallback servers, so AE::DNS can finally
129 support ipv6-only operation even with defaults (also use ::1 instead
130 of 127.0.0.1).
131 - give better guidelines on how to chose a logging level.
132 - AnyEvent::Log can now cap (limit) log message priority by context.
133 - AE::log will now save and restore $! and $@, so callers don't have to.
134 - AE::Strict now checks that registered fds are still valid at regular
135 intervals.
136 - skip frequently called 'require' calls if the module appears loaded.
137 - increase signal latency test timer considerably for poor little osx,
138 which can't be bothered with basic posix functionality such as signal
139 delivery.
140 - mention some debugging aids in the tutorial.
141 - new env variable PERL_ANYEVENT_MAX_SIGNAL_LATENCY.
142 - private (%namespace) log contexts in PERL_ANYEVENT_LOG now have the name
143 of the context as title.
144 - AnyEvent::Strict would not restore read-write access to $_ when a callback
145 died - this is a user bug, but since AE::Strict exists to help dbeugging...
146 - detect the extremely evil IO::Async::Loop::AnyEvent module and refuse
147 to work with it - the author is well aware of the issues but refuses
148 to fix them. This extreme step has been taken because the module actively
149 attacks the goals of AnyEvent and makes it harder for module authors to use
150 AnyEvent.
151
1526.02 Fri Aug 26 20:08:31 CEST 2011
153 - log_to_syslog now accepts facility strings.
154 - log_to_syslog would not allow facilities - "$facility|$level"
155 *literally* means that string. sheesh.
156 - the CPAN parser got confused and thought we provide Sys::Syslog, try
157 to work around.
158 - renamed AnyEvent::Impl::FLTK2 to FLTK.
159
1606.01 Fri Aug 26 07:04:11 CEST 2011
161 - INCOMPATIBLE CHANGE: AnyEvent::DNS resource records now include
162 the TTL as fourth element - this affects ->resolve, ->request
163 and ->dns_unpack indirectly. Convenience functions such as
164 AnyEvent::DNS::{a,txt,mx,srv} are not affected, neither is
165 AnyEvent::Socket.
166 - INCOMPATIBLE CHANGE: backend authors now should not implement
167 one_event or loop, but instead the AnyEvent::CondVar::_wait and _poll
168 methods.
169 - INCOMPATIBLE CHANGE: verbosity levels have been redefined, see
170 AnyEvent (PERL_ANYEVENT_VERBOSE) and AnyEvent::Log for details.
171 - DEPRECIATION: please change
172 "use AnyEvent::Impl::Perl" to "use AnyEvent::Loop" and
173 "AnyEvent::Impl::Perl::loop" to "AnyEvent::Loop::run"
174 in your programs, for future compatibility.
175 - *blush* idle watcher emulation was borked.
176 - AnyEvent::Strict now checks the AE::xxx functions as well.
177 - querying too long domain names would barf deep inside AnyEvent::DNS - since
178 this is hard to avoid for applications, AnyEvent:.DNS::request now checks for
179 for too long domain names explicitly and passes an undef to the callback
180 instead (reported by James Bromberger).
181 - make AnyEvent::Util more robust against EINTR (analyzed by Andrew Suffield).
182 - new functions: AnyEvent::postpone, AnyEvent::log.
183 - new module: AnyEvent::Log, for simple logging needs.
184 - new env variable, PERL_ANYEVENT_LOG, for fine-grained logging config.
185 - new env variable, PERL_ANYEVENT_DEBUG_SHELL, to automatically start
186 a debug shell.
187 - new env variable, PERL_ANYEVENT_DEBUG_WRAP, to automatically enable
188 watcher instrumentation/wrapping (see AnyEvent::Debug::wrap).
189 - new env variable, PERL_ANYEVENT_HOSTS, to override the place where
190 /etc/hosts can be found.
191 - an empty PERL_ANYEVENT_RESOLV_CONF now selects the default
192 configuration.
193 - env variables can now be specified using AE_xyz instead of
194 PERL_ANYEVENT_xyz, subject to some rules.
195 - lots of new AnyEvent::Debug functionality.
196 - AnyEvent::Strict now checks for modifications of $_ by the callback
197 (by setting it to readonly while executing the callback, if possible).
198 - IO::Async sometimes cancels the wrong timer when you call cancel_timer.
199 Invest extra overhead to work around this peculiar behaviour.
200 - split the pure perl backend into a loop module and an impl
201 module (AnyEvent::Loop and AnyEvent::Impl::Perl).
202 - some read types (regex, netstring, json, storable) could
203 errornously keep an AE::Handle object alive on their own
204 (reported by Mohammad Toossi).
205 - AnyEvent::Socket::resolve_sockaddr and all functions using it now
206 supports /etc/hosts overrides.
207 - add more workarounds around child watcher bugs in IO::Async
208 and (newly introduced) in POE.
209 - use glib child watchers and try to work around its limitations,
210 if possible (based on a patch by Kevin Ryde).
211 - add support for FLTK (fltk2), via AnyEvent::Impl::FLTK2.
212 - many read types in AnyEvent::Handle would malfunction during
213 transitions to TLS because of a "delete $self->{rbuf}".
214 (reported by Gerald Galster).
215 - improve condvar blocking wait performance (probably only noticable
216 with faster backends), also make it easier for event loops that
217 do not support blocking waits.
218 - PERL_ANYEVENT_MODEL now also allows full module names, although
219 the usefulness of this feature is not yet clear.
220 - add tests for all supported event loops, conditional on PERL_ANYEVENT_LOOP_TESTS.
221 - add t/80_ssltext.t.
222 - no longer use AUTOLOAD for the initial loading (for no very good reason).
223 - tcp_server now tries to unlink unix domain sockets when it gets destroyed.
224 - do not run event loop detection multiple times when programs call their cached
225 copies of AnyEvent::detect.
226 - parse_hostport now accepts absolute paths als unix domain sockets.
227 - suppress the idiotic warning inside POE by patching the POE::Kernel
228 run flags instead of asking the user to jump through contortions.
229 - add Task::Weaken as dependency (as recommends only at this time).
230 - replace PF_UNSPEC by 0, for good old minix.
231
2326.0 (not officially released, but was on CPAN for about 15 minutes :/)
233
2345.34 Thu May 12 09:59:41 CEST 2011
235 - AE::TLS didn't support both cert_file and key at the same time,
236 due to some copy&paste bug (reported and analyzed by vti).
237 - improve AE::Handle timeout documentation slightly.
238
2395.33 Mon Feb 7 21:51:41 CET 2011
240 - never appeared on CPAN, for unknown reasons.
241 - AnyEvent::Handle on_connect_error - do not try to destroy a handle
242 object when it's already gone (analyzed by Ryan Bullock).
243 - update idna algorithm(s) to UTS#46 version 6.0.0.
244 - preserve trailing dot in idn_to_ascii - testsuite says to preserve,
245 uts46 says it's an error, and testsuite makes more sense.
246 - remove leading dots in idn_to_ascii - testsuite says to remove,
247 uts46 says it's an error, and testsuite makes more sense.
248 - fixed a bug in idn ignored character processing.
249 - passes the uts46 6.0.0 testsuite except for three cases which
250 apparently are bugs in the testsuite itself.
251 - passes the uts46 6.0.1 testsuite except for eleven cases which
252 apparently are bugs in the testsuite itself.
253
2545.31 Sun Jan 23 11:49:19 CET 2011
255 - work around a POE bug where POE would first poll for new events and
256 invoke handlers BEFORE handling the message queue.
257 - fix parsing of "domain" statement in resolv.conf (Johannes Plunien).
258 - stop_read now does just that in TLS mode. see its
259 documentation for caveats.
260 - update to common::sense 3.4 warning set (some warnings in perl impose
261 an arbitrary 32kb limit on regex matches, which affects
262 AnyEvent::Handle).
263 - AE::Handle had a broken rbuf_max method and missing wbuf_max method.
264 - new experimental AE::Handle::resettls method.
265
2665.3 Fri Dec 31 05:48:13 CET 2010
267 - major IOAsync patch by Paul Evans - 0.33 is required, as the major
268 issues with IO::Async have been solved in that version, including
269 the availabilty of a default loop!
270 - implement new wbuf_max parameter for AnyEvent::Handle.
271 - added Cocoa::EventLoop backend by Daisuke Murase.
272 - the default testsuite now honors PERL_ANYEVENT_MODEL (Paul Evans).
273 - clarify handle on_prepare documentation.
274 - set CLOEXEC flag on parent-side fd's in AnyEvent::Util::run_cmd,
275 to avoid leaking them to any child processes.
276 - try to force files into the arch-dependend lib directory, for the
277 half-"turly obsessive compulsive person"(s) that depend on it.
278
2795.29 Sun Dec 5 10:49:21 CET 2010
280 - convert EV backend to EV 4.00 API (so better upgrade EV too).
281 - AE::Handle: implement dynamic read_size adjustment, add
282 max_read_size parameter, reduce default read_size to 2048.
283 - add some FAQ entries for servers and/or on_eof vs. on_error.
284 - work around OS bugs (cygwin again) when lingering in AE::Handle.
285 - reduce memory usage slightly when connect was used in AE::Handle.
286 - wrok around more segfaults in Tk.
287 - document tls_ctx => undef as valid.
288 - detect WNOHANG value at compiletime.
289 - include some probably linux-only support for building AnyEvent
290 as part of the perl core.
291 - improve compatibility of t/08_idna.t to perl 5.8.
292 - make AnyEvent::Strict truly optional (in case somebody builds a
293 minimal perl...).
294
2955.28 Wed Oct 13 04:14:23 CEST 2010
296 - due to a glitch, AnyEvent's internal getprotobyname wasn't
297 actually used in tcp_server.
298 - implement AnyEvent::CondVar->new method, which was documented but
299 missing.
300 - support autoloading of read types in unshift_read.
301 - AnyEvent::Strict: do the test for invalid fh arguments differently,
302 resulting in more informative error messages.
303 - new AnyEvent::FAQ manpage.
304 - clarify that you can actually call push_read etc. in on_read.
305 - put keys and certs in /tmp, not into the current directory,
306 and document this in the SECURITY CONSIDERATIONS sections
307 of AnyEvent::TLS.
308 - support ";" as resolv.conf comment character in AnyEvent::DNS, as
309 requested by Ron Isaacson.
310 - document $AnyEvent::DNS::RESOLVER variable.
311 - incorporated major typo patches by Abhijit Menon-Sen.
312 - AnyEvent::Handle now croaks when negative timeouts are passed.
313 - add a shitload of TCP_xxx sockopt name constants to AnyEvent::Util,
314 undocumented, but who knows when they might come in handy.
315
3165.271 Tue Jun 8 12:05:46 CEST 2010
317 - backport to perl 5.8.x.
318
3195.27 Sun Jun 6 12:12:05 CEST 2010
320 - postpone differently in AnyEvent::Socket now, as
321 when not, canceling the connection attempt might fail
322 (found by Felix Antonius Wilhelm Ostmann).
323 - explicitly check for non-stream sockets in AE::Handle, too many
324 clueless people fell into the trap of this somehow working.
325 - simplified and reworked the "OTHER MODULES" section.
326 - better/more condvar examples.
327
3285.261 Wed Apr 28 16:13:36 CEST 2010
329 - AF_INET6 was not properly used from Socket6 during configuration
330 time (found by Andy Grundman).
331
3325.26 Mon Apr 12 04:49:35 CEST 2010
333 - don't generate spurious readyness notifications when select
334 returns EINTR in the pure perl backend (analysed and testcase
335 by Bas Denissen).
336 - give same examples in SYNOPSIS sections of AnyEvent and AE.
337 - provider faster implementations for AE::time/now/now_update for
338 EV and Perl backends.
339 - provide faster AE::cv and AE::time implementations for all
340 backends.
341 - no longer support lower/mixed-case signal names (this was
342 never documented nor universally supported).
343 - some more memory µ-optimisations, and somewhat less messy
344 code to cope with both AnyEvent and AE APIs.
345
3465.251 Sat Mar 13 00:58:21 CET 2010
347 - make test failed if it was run before make install on perl <=
348 5.8.8, and older versions of perl make our live very difficult,
349 so write constants.pl during Makefile.PL time.
350
3515.25 Sat Mar 13 00:23:14 CET 2010
352 - fix a race condition in AnyEvent::Handle that would cause
353 a "bio_read: unsupported method" error (in _another_
354 ssl connection) after user code threw an exception.
355 - added AnyEvent::Handle->destroyed method.
356 - speed up AnyEvent::Socket::format_address and ::format_ipv6.
357 - the AnyEvent::Util::fh_nonblocking function worked only
358 by lucky accident on win32.
359 - smaller and faster AnyEvent::Util::fh_nonblocking.
360 - when the (required!) Time::HiRes module is missing, AnyEvent
361 did not fall back to built-in time properly.
362 - do not load Fcntl at runtime, saving memory and loading time.
363 - precompile a number of constants and use them instead of runtime
364 detection and eval.
365 - free detection code after detection and similar memory optimisations.
366 - Perl backend timer interval best effort drift has been improved
367 (same algorithm as EV).
368 - update unicode idna mapping table.
369
3705.24 Tue Jan 5 11:39:43 CET 2010
371 - cygwin never reports errors from failed connects. 1.5 just gives
372 you continous readyness and EAGAIN, 1.7 is even more broken and
373 just hangs. work around both issues in a cygwin-specific hack.
374 - improve idle watcher documentation slightly.
375
3765.23 Sun Dec 20 23:48:00 CET 2009
377 - support IDNs in resolve_sockaddr, and therefore in tcp_connect.
378 - implement punycode_encode/decode, idn_nameprep,
379 idn_to_ascii and idn_to_unicode operations in AnyEvent::Util.
380 - provide $AE::VERSION.
381 - removed traces of "no strict 'refs'".
382
3835.22 Sat Dec 5 03:51:13 CET 2009
384 - downgrade-or-fail in AnyEvent::Handle::push_write, to
385 diagnose encoding failures earlier and more succinctly.
386 (this works around bugs in perl, throwing away encoding info
387 when passing scalar data to extensions).
388 - add more examples to AnyEvent::Socket manpage.
389 - upgrade internal warning set to the same as common::sense 2.03.
390 - use pack "n/a*" for pre-5.8.9 perl compatibility in AnyEvent::DNS
391 (John Beppu).
392 - AnyEvent::Socket::inet_aton now properly supports ipv6, as documented.
393 - add google public dns servers to fallback server set and make sure
394 we load-balance properly between the three sets. also add all
395 fallback dns servers, not just a random one, to each dns config.
396
3975.21 Thu Nov 19 02:48:47 CET 2009
398 - fix a problem where socket constants were called with parameters
399 (spotted by David Friedland).
400 - fork_call never use'd POSIX (reported by Daisuke Maki).
401 - improve perl 5.6 compatibility further (but it still won't work
402 unless you rip out everything but the core).
403 - prefer Net::DNS::Resolver over ipconfig, if installed, on win32.
404 uses 10MB of RAM, but doesn't open a console window. *sigh*.
405
4065.202 Wed Oct 14 22:35:44 CEST 2009
407 - AnyEvent::DNS would unexpectedly clobber $_ under windows
408 (analysed by Matthias Waldorf).
409 - AnyEvent::Handle::run_cmd can now pass the PID of the
410 newly-created process, which is much less useful than it might
411 sound (based on patch by Yann Kerherve).
412
4135.201 Tue Sep 29 12:09:25 CEST 2009
414 - AnyEvent:Handle::on_starttls/on_stoptls methods were broken
415 (reported by Torsten Foertsch).
416 - common::sense 2.0 could cause tcp_server to throw an exception
417 (analysed by elmex).
418
4195.2 Mon Sep 14 07:04:49 CEST 2009
420 - INCOMPATIBLE CHANGE: do no longer support register_read_type
421 and register_write_type in AnyEvent::Handle, instead support
422 package names (the facility was mostly abused).
423 - implement "packagename-as-read/write type" support in
424 AnyEvent::Handle.
425 - AnyEvent::Handle: new options "keepalive" and "oobinline".
426 - oobinline set by default to avoid security issues.
427 - the pure-perl event loop backend wrongly detected times() fallback
428 support (spotted by Pavel Boldin).
429
4305.12 Tue Sep 1 20:26:50 CEST 2009
431 - be more lenient when parsing resolv.conf files, as some people
432 use hashmarks as comment indicator inside directives (reported
433 by Michael S. Fischer).
434 - use same set of warnings as common::sense 2.0.
435 - fix a potential 32 bit overflow issue due to perl having problems
436 with large hex constants in the Perl backend.
437 - do not use the slower 5.6 version fo the enc_name code in AnyEvent::DNS
438 by default.
439 - fix some prototypes.
440 - (optionally) require Async::Interrupt 1.02.
441 - replace opendns fallbacks by verizon/level3 public dns servers.
442
4435.112 Fri Aug 21 13:59:04 CEST 2009
444 - AnyEvent::Handle could cause hard crashes in Net::SSLeay when
445 using starttls with an illegal/uncreatable tls context.
446 - port to broken win32 perls.
447
4485.111 Sun Aug 16 18:44:36 CEST 2009
449 - perl errornously requires sockaddr_un structures returned by the
450 kernel to have a specific length. work around this bug by padding
451 sockaddr_un structures (found to fail on OS X, as perl only uses
452 the correct code for GNU/Linux, but likely to fail on other
453 OSes as well).
454
4555.11 Wed Aug 12 17:49:37 CEST 2009
456 - fixed a bug in run_cmd with input from filename.
457 - create a json coder object when none is given in write direction
458 to reduce dependency on JSON.
459
4605.1 Tue Aug 11 03:17:32 CEST 2009
461 - tcp_connect now reports the correct error in $!, not an unrelated
462 one (was broken since 4.91).
463 - AE::Handle did not properly clear rtimeout/wtimeout watchers
464 when retrying a tcp_connect.
465 - new functions: AnyEvent::Util::run_cmd and close_all_fds_except.
466 - fixed wrong documentation in AE::Socket.
467
4685.01 Mon Aug 10 03:16:32 CEST 2009
469 - last release broke EV child watchers :/.
470
4715.0 Sun Aug 9 17:32:09 CEST 2009
472 - officially introduce and document the AE API.
473 - lowering the timeout at runtime did not have immediate effect
474 in AnyEvent::Handle.
475 - AE::Handle now has separate and independent read and write timeout
476 settings.
477 - probe for Guard module when AnyEvent::Util is loaded, not at
478 runtime (Event doesn't like eval "use" from callbacks, and it avoids
479 exporting the probe function).
480 - Impl::IOAsync's io watchers did not use the filehandle emulation
481 code to work around it's watcher limits.
482 - work around IO::Async's problems with overloaded objects.
483 - try to disable wearnings inside IO::Async, it's just too buggy.
484 - do no longer use Event by default, only EV or Perl, Event is too buggy.
485 - hackishly provide fast AE:: interface when EV or Perl is the backend.
486 - inherit VERSION from AnyEvent::VERSION in most modules.
487
4884.91 Thu Aug 6 15:42:45 CEST 2009
489 - AE::Handle::starttls could get out of sync when the read buffer
490 already contains some TLS handshake.
491 - AE::Handle did not properly free the TLS session, causing bigger
492 memory leaks in Net::SSLeay than expected (reported by toaster).
493 - AE::Socket::tcp_connect will now artificially delay invoking
494 the callback to avoid returning after invoking the callback.
495 - convert many internal watcher uses to AE API.
496
4974.9 Sat Aug 1 11:07:01 CEST 2009
498 - Glib silently fails when registering a timeout with
499 a negative value, so avoid that.
500 - call condvar callback immediately when it is set
501 after the condition is already signalled.
502 - check rbuf_max condition only after trying to consume
503 data.
504 - normalise signal numbers to names when using the ae's
505 signal handling, but do not document this (yet).
506 - pure perl signal emulation did not properly set nonblocking
507 mode on the signal pipe on win32 (this is unlikely to have caused
508 any issues).
509 - new module: AnyEvent::Debug.
510 - AnyEvent::Strict now checks that a signal was specified
511 by name (not name or number).
512 - reduce memory footprint in typical cases by ~50kb by
513 only compiling signal code when necessary.
514 - add AnyEvent::Handle->rbuf_max.
515 - grab the AE namespace for future new API, implement stubs for
516 said future/alternative API.
517 - new function: AnyEvent::Socket::format_hostport.
518
5194.881 Tue Jul 28 12:51:53 CEST 2009
520 - work around a bug in local in pre-5.10 perls, causing
521 AnyEvent::Handle to recurse when it should not
522 (analysed by elmex).
523
5244.88 Tue Jul 28 04:04:37 CEST 2009
525 - re-bless the handle into a dummy package after calling
526 AnyEvent::Handle::destroy, so the user does not need to check
527 for errors after every push_write etc.
528 - do not attempt to run t/02_signals.t on obviously broken
529 platforms.
530
5314.87 Sun Jul 26 02:06:16 CEST 2009
532 - do not attempt to linger when there is no longer a valid fh.
533 - enforce tls mode to be either accept or connect, do not simply
534 segfault in Net::SSLeay.
535 - AnyEvent::Handle can now call tcp_connect itself (new parameters
536 connect, on_prepare, on_connect and on_connect_error). Updated
537 tutorial accordingly.
538 - add AnyEvent::Impl::Irssi backend.
539
5404.86 Mon Jul 20 23:52:29 CEST 2009
541 - since the verbose warning is not enough, explicitly document
542 that versions before 1.33 of Net::SSLeay are not secure.
543 - work around signal handling races in Event and (...) Event::Lib.
544 - try to align signal-race timer to full-second boundaries.
545 - work around Tk not liking negative timeouts.
546 - don't complain of different grades of environmental unfriendlyness
547 in IO::Async.
548
5494.85 Sat Jul 18 06:16:14 CEST 2009
550 - nail the signal race problem in perl once and for all
551 (see $AnyEvent::MAX_SIGNAL_LATENCY).
552 - take advantage of Async::Interrupt if it is available.
553 - load Time::HiRes and Guard modules on demand only.
554 - add optional/recommended modules section to AnyEvent
555 documentation.
556 - reduce memory usage considerably (and reduce startup penalty)
557 by not using "strict", "warnings" and "overload" modules.
558 - work around buggy windows/openbsd perls and provide EBADMSG
559 and EPROTO ourselves when missing.
560 - improve perl 5.6 compatibility of the core event loop.
561 - made Net::SSLeay version 1.33 a soft requirement.
562
5634.83 Fri Jul 17 16:56:26 CEST 2009
564 - implement AnyEvent::Socket::getprotobyname.
565 - AnyEvent::CondVar's will now detect recursive blocking
566 waits and will croak, as too many people fall into
567 this trap.
568 - AnyEvent::Handle will now call ->destroy on itself after
569 executing the on_error callback, instead of doing some
570 half-baked internal shutdown, for fatal errors.
571 - clarify on_eof behaviour w.r.t. the read queue and
572 on_read callbacks.
573 - ignore some possible spurious wake-ups in tcp_connect.
574
5754.82 Sat Jul 11 00:34:55 CEST 2009
576 - POE and Event backends didn't accept some callable objects as
577 callbacks.
578 - use Config module instead of POSIX module to detect signal names
579 in AnyEvent::Strict and AnyEvent::Impl::EventLib, as the POSIX
580 module doesn't even have all POSIX signals :/.
581 - use more workarounds around the many refcnt/corruption bugs in
582 Event::Lib.
583 - work around a race condition in perl's select, causing t/03_child.t
584 to rarely fail.
585
5864.81 Thu Jul 9 10:30:30 CEST 2009
587 - AnyEvent::Handle didn't properly diagnose write errors
588 (it expected -1 from syswrite, how lame... :).
589 - support file descriptors in addition to file handles
590 in AnyEvent->io.
591 - new env variables: PERL_ANYEVENT_RESOLV_CONF,
592 PERL_ANYEVENT_MAX_OUTSTANDING_DNS, PERL_ANYEVENT_CA_FILE
593 and PERL_ANYEVENT_CA_PATH.
594 - provide a sensible synopsis section for AnyEvent::TLS.
595 - add a "supported backends" section to the manpage.
596 - added simple io watcher test to testsuite, using a
597 portable_socketpair.
598 - tried to improve the stability of the Event::Lib backend,
599 YMMV.
600
6014.8 Mon Jul 6 23:45:16 CEST 2009
602 - AnyEvent::DNS did not properly follow CNAME records with
603 uppercase targets.
604 - AnyEvent::DNS would errornously return AAAA records
605 with v4 mapped addresses (a faulty record) as ipv4 addresses,
606 causing AnyEvent::Socket to throw an exception.
607 - added new module AnyEvent::TLS for easier SSL/TLS context
608 creation, with many options including hostname verification,
609 secure default configuration, lots of documentation and,
610 predefined diffie-hellman keys for perfect forward security
611 and much more. get it while it's still fresh!
612 - use AnyEvent::TLS in AnyEvent::Handle for context management.
613 - load AnyEvent::Handle only on demand in AnyEvent::DNS,
614 so AnyEvent::Socket users have smaller memory footprint
615 in the common case.
616 - add AnyEvent::Handle->push_shutdown method.
617 - add an additional error message parameter to AnyEvent::Handle's
618 on_error callback (for TLS, $! is still available).
619 - add AnyEvent::Handle on_starttls/on_stoptls callbacks.
620 - make AnyEvent::Handle more robust against early conenction
621 failures (during new), and return C<undef> in that case
622 from the constructor.
623 - AnyEvent::Handle will now try to load only JSON::XS first,
624 then fall back to JSON.
625 - format_ipv4/format_ipv6 are now exported by default, for symmetry,
626 and because it was documented that way.
627
6284.451 Fri Jul 3 00:28:58 CEST 2009
629 - do not clear rbuf when shutting down an AnyEvent::Handle
630 object - doing so breaks AnyEvent::HTTP.
2 631
34.45 Mon Jun 29 22:59:26 CEST 2009 6324.45 Mon Jun 29 22:59:26 CEST 2009
4 - a write error could cause AnyEvent::Handle to create 633 - a write error could cause AnyEvent::Handle to create
5 an I/O watcher with an undefined $fh. 634 an I/O watcher with an undefined $fh.
6 - special-case mapped ipv4 addresses in both 635 - special-case mapped ipv4 addresses in both
… …
1204.3 Fri Oct 3 09:18:43 CEST 2008 7494.3 Fri Oct 3 09:18:43 CEST 2008
121 - AnyEvent will now install a no-op signal handler for SIGPIPE, 750 - AnyEvent will now install a no-op signal handler for SIGPIPE,
122 unless one has been installed already. 751 unless one has been installed already.
123 - warn about not loaidng AnyEvent::Impl::POE early enough 752 - warn about not loaidng AnyEvent::Impl::POE early enough
124 (patch by Adam Rosenstein). 753 (patch by Adam Rosenstein).
125 - fixed a great number of bugs and corner cases in AnyEvent::Handle: 754 - fixed a great number of bugs and corner cases in AnyEvent::Handle.
126 - fix a bug in where in SSL connect mode, the 755 - fix a bug in where in SSL connect mode, the
127 client would first wait for some data by the server and 756 client would first wait for some data by the server and
128 otherwise hang. (reported and analyzed in an absolutely 757 otherwise hang. (reported and analysed in an absolutely
129 exemplary manner by Adam Rosenstein). 758 exemplary manner by Adam Rosenstein).
130 - fix a bug in where SSL EOF would not be treated 759 - fix a bug in where SSL EOF would not be treated
131 as stream EOF, putting the connection into a hung state. 760 as stream EOF, putting the connection into a hung state.
132 (reported and analyzed in an absolutely exemplary manner by Adam 761 (reported and analysed in an absolutely exemplary manner by Adam
133 Rosenstein). 762 Rosenstein).
134 - fix a potential segfault when the TLS context 763 - fix a potential segfault when the TLS context
135 would go missing in a read callback (e.g. due to stoptls) - 764 would go missing in a read callback (e.g. due to stoptls) -
136 Net::SSLeay of course makes no type checking whatsoever. 765 Net::SSLeay of course makes no type checking whatsoever.
137 - AnyEvent::Handle will not stop the read watcher in TLS mode. 766 - AnyEvent::Handle will not stop the read watcher in TLS mode.

Diff Legend

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