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

Comparing AnyEvent/Changes (file contents):
Revision 1.276 by root, Thu Jul 9 22:37:53 2009 UTC vs.
Revision 1.576 by root, Wed Aug 21 08:40:28 2013 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
11
127.05 Wed Aug 21 10:38:08 CEST 2013
13 - uts46data.pl couldn't be found due to wrong naming of the file
14 (reported by Fulko Hew).
15 - handle lone \015's properly in AE::Handle's default line read
16 (reported by various people).
17 - untaint ip addresses found in /etc/hosts (patch by José Micó).
18 - the memleak fix in 7.03 caused resolving via /etc/hosts to always
19 fail on first use (reported and testcase by Andrew Whatson).
20 - expose AnyEvent::Log::format_time, and allow users to redefine it.
21 - expose AnyEvent::Log::default_format, and allow redefinition.
22 - expose AnyEvent::Log::fatal_exit, to allow redefinition.
23 - AnyEvent::Debug shell can now run coro shell commands, if available.
24 - t/63* tests were wrongly in MANIFEST.
25 - kernel.org's finger server went MIA, switch to freebsd.org and
26 icculus.org.
27 - clarify that IO::AIO and AnyEvent::AIO are needed for AnyEvent::IO
28 to function asynchronously (reported by Daniel Carosone).
29 - hard-disable $^W in most tests, it generates too much garbage output.
30 - use a (hopefully) more future-proof method to emulate common::sense.
31 - upgrade to UTS-46:6.2.0.
32 - switch to INSTLIB from INSTLIBDIR, as INSTLIBDIR was wrongly
33 documented. should not affect anything.
34
357.04 Thu Nov 15 02:23:58 CET 2012
36 - debugging code left in.
37
387.03 Thu Nov 15 02:16:11 CET 2012
39 - AnyEvent::Socket::inet_aton did not work when dns resolution was
40 used to find the addresses (analyzed and patch by Maxime Soulé).
41 - fix a memory leak in the /etc/hosts lookup code when hosts don't
42 resolve and are not in hosts (analyzed by sten).
43
447.02 Tue Aug 14 04:27:58 CEST 2012
45 - AnyEvent::Util::run_cmd could block indefinitely (analyzed and test
46 program by Yorhel).
47 - verified that AnyEvent::Socket follows RFC5952.
48 - try to parse "ADDR#PORT" in addition to "ADDR PORT".
49
507.01 Sun May 13 01:03:17 CEST 2012
51 - fail with EPROTO in AnyEvent::Handle wqhen TLS is requested but
52 not available, instead of throwing an exception.
53 - use File::Spec to get the tmpdir in t/*, to avoid needless
54 failures on (most, not mine :) windows boxes.
55 - new handle read types: tls_detect and tls_autostart.
56
577.0 Fri Apr 13 06:33:30 CEST 2012
58 - child watchers are broken in POE 1.352 (also many earlier
59 versions) and there seems to be no way to work around it, as POE
60 itself is inherently racy. Document this regression and add a delay
61 in t/68_poe_03_child.t for the time being.
62 - new module AnyEvent::IO, that is a frontend to either a pure-perl
63 synchronous I/O implementation (AnyEvent::IO::Perl), or to the
64 asynchronous I/O extension IO::AIO (AnyEvent::IO::IOAIO).
65 - load /etc/hosts only when DNS has no answer.
66 - stat /etc/hosts on every access and reload it if it changed.
67 - load /etc/hosts via AnyEvent::IO - potentially asynchronous.
68 - fix a buggy croak in the dh parameter reading in AnyEvent::TLS.
69 - AnyEvent::Log log_to_file and log_to_path now use AnyEvent::IO. As
70 a side effect, they now use true appending as opposed to libc
71 appending, and the file might not have been opened when the function
72 returns.
73 - the default logging level was not properly documented in a variety of
74 places, this has hopefully been rectified.
75 - updated uts46data.pl for unicode 6.1.0.
76 - made log messages generated by AnyEvent submodules not
77 include the package name anymore, as it will be logged
78 by default already.
79 - upgrade to the trick used by common-sense 3.5 to work
80 around extra warning torture/breakage under perl 5.15.x.
81 - log messages by anyevent are now ucfirst, are usually full
82 sentences and do no longer include the package name.
83 - the storable read type would sometimes throw an exception instead
84 of causing EBADMSG (analyzed by Maxime Soulé).
85 - work around a bug in openssl 1.0.1 which enforces a minimum rsa
86 keysize (reported by Paul Howarth).
87 - documented Rocco fabricating statements about AnyEvent and me.
88
896.14 Tue Jan 31 20:00:24 CET 2012
90 - AnyEvent::Impl::Tk was broken due to a mysterious "die" inside,
91 probably an editing mistake (reported by Darin McBride).
92
936.13 Thu Jan 12 07:27:01 CET 2012
94 - AnyEvent::Util::fork_call checked for POSIX availability in the wrong way,
95 causing it to fail if POSIX wasn't loaded already (analysed by Rock Power).
96 - AnyEvent::Handle::push_read (line => did pass $1 directly, so
97 regex-matching inside the callback would change the parameter, despite
98 $1 being dynamically scoped per-block. this perl bug is now being
99 worked around (testcase by Cindy Wang).
100
1016.12 Mon Dec 12 13:21:10 CET 2011
102 - $! was clobbered by subsequent calls in tcp_connect, due to
103 the postpone not saving/restoring it (analyzed by Richard Garnish).
104
1056.11 Tue Nov 22 10:36:05 CET 2011
106 - Tk cannot create windows when tainted unless you set a title, so
107 set a dummy title for AnyEvent's dummy mainwindow. How dummy.
108 - escape any nonprintable/nonascii characters when stringifying
109 backtraces.
110 - log the reason loading an interface module fails to load at level debug.
111
1126.1 Tue Oct 4 19:44:30 CEST 2011
113 - INCOMPATIBLE CHANGE: the default log level is now "4" (error
114 and above), and some messages inside AnyEvent have been elevated
115 to higher log levels to print by default.
116 - AnyEvent::log, unlike AnyEvent::Log::log, did not always exit on fatal.
117 - work around yet another signal race bug, newly introduced by POE. My
118 how that module sucks.
119 - add google public ipv6 dns as fallback servers, so AE::DNS can finally
120 support ipv6-only operation even with defaults (also use ::1 instead
121 of 127.0.0.1).
122 - give better guidelines on how to chose a logging level.
123 - AnyEvent::Log can now cap (limit) log message priority by context.
124 - AE::log will now save and restore $! and $@, so callers don't have to.
125 - AE::Strict now checks that registered fds are still valid at regular
126 intervals.
127 - skip frequently called 'require' calls if the module appears loaded.
128 - increase signal latency test timer considerably for poor little osx,
129 which can't be bothered with basic posix functionality such as signal
130 delivery.
131 - mention some debugging aids in the tutorial.
132 - new env variable PERL_ANYEVENT_MAX_SIGNAL_LATENCY.
133 - private (%namespace) log contexts in PERL_ANYEVENT_LOG now have the name
134 of the context as title.
135 - AnyEvent::Strict would not restore read-write access to $_ when a callback
136 died - this is a user bug, but since AE::Strict exists to help dbeugging...
137 - detect the extremely evil IO::Async::Loop::AnyEvent module and refuse
138 to work with it - the author is well aware of the issues but refuses
139 to fix them. This extreme step has been taken because the module actively
140 attacks the goals of AnyEvent and makes it harder for module authors to use
141 AnyEvent.
142
1436.02 Fri Aug 26 20:08:31 CEST 2011
144 - log_to_syslog now accepts facility strings.
145 - log_to_syslog would not allow facilities - "$facility|$level"
146 *literally* means that string. sheesh.
147 - the CPAN parser got confused and thought we provide Sys::Syslog, try
148 to work around.
149 - renamed AnyEvent::Impl::FLTK2 to FLTK.
150
1516.01 Fri Aug 26 07:04:11 CEST 2011
152 - INCOMPATIBLE CHANGE: AnyEvent::DNS resource records now include
153 the TTL as fourth element - this affects ->resolve, ->request
154 and ->dns_unpack indirectly. Convenience functions such as
155 AnyEvent::DNS::{a,txt,mx,srv} are not affected, neither is
156 AnyEvent::Socket.
157 - INCOMPATIBLE CHANGE: backend authors now should not implement
158 one_event or loop, but instead the AnyEvent::CondVar::_wait and _poll
159 methods.
160 - INCOMPATIBLE CHANGE: verbosity levels have been redefined, see
161 AnyEvent (PERL_ANYEVENT_VERBOSE) and AnyEvent::Log for details.
162 - DEPRECIATION: please change
163 "use AnyEvent::Impl::Perl" to "use AnyEvent::Loop" and
164 "AnyEvent::Impl::Perl::loop" to "AnyEvent::Loop::run"
165 in your programs, for future compatibility.
166 - *blush* idle watcher emulation was borked.
167 - AnyEvent::Strict now checks the AE::xxx functions as well.
168 - querying too long domain names would barf deep inside AnyEvent::DNS - since
169 this is hard to avoid for applications, AnyEvent:.DNS::request now checks for
170 for too long domain names explicitly and passes an undef to the callback
171 instead (reported by James Bromberger).
172 - make AnyEvent::Util more robust against EINTR (analyzed by Andrew Suffield).
173 - new functions: AnyEvent::postpone, AnyEvent::log.
174 - new module: AnyEvent::Log, for simple logging needs.
175 - new env variable, PERL_ANYEVENT_LOG, for fine-grained logging config.
176 - new env variable, PERL_ANYEVENT_DEBUG_SHELL, to automatically start
177 a debug shell.
178 - new env variable, PERL_ANYEVENT_DEBUG_WRAP, to automatically enable
179 watcher instrumentation/wrapping (see AnyEvent::Debug::wrap).
180 - new env variable, PERL_ANYEVENT_HOSTS, to override the place where
181 /etc/hosts can be found.
182 - an empty PERL_ANYEVENT_RESOLV_CONF now selects the default
183 configuration.
184 - env variables can now be specified using AE_xyz instead of
185 PERL_ANYEVENT_xyz, subject to some rules.
186 - lots of new AnyEvent::Debug functionality.
187 - AnyEvent::Strict now checks for modifications of $_ by the callback
188 (by setting it to readonly while executing the callback, if possible).
189 - IO::Async sometimes cancels the wrong timer when you call cancel_timer.
190 Invest extra overhead to work around this peculiar behaviour.
191 - split the pure perl backend into a loop module and an impl
192 module (AnyEvent::Loop and AnyEvent::Impl::Perl).
193 - some read types (regex, netstring, json, storable) could
194 errornously keep an AE::Handle object alive on their own
195 (reported by Mohammad Toossi).
196 - AnyEvent::Socket::resolve_sockaddr and all functions using it now
197 supports /etc/hosts overrides.
198 - add more workarounds around child watcher bugs in IO::Async
199 and (newly introduced) in POE.
200 - use glib child watchers and try to work around its limitations,
201 if possible (based on a patch by Kevin Ryde).
202 - add support for FLTK (fltk2), via AnyEvent::Impl::FLTK2.
203 - many read types in AnyEvent::Handle would malfunction during
204 transitions to TLS because of a "delete $self->{rbuf}".
205 (reported by Gerald Galster).
206 - improve condvar blocking wait performance (probably only noticable
207 with faster backends), also make it easier for event loops that
208 do not support blocking waits.
209 - PERL_ANYEVENT_MODEL now also allows full module names, although
210 the usefulness of this feature is not yet clear.
211 - add tests for all supported event loops, conditional on PERL_ANYEVENT_LOOP_TESTS.
212 - add t/80_ssltext.t.
213 - no longer use AUTOLOAD for the initial loading (for no very good reason).
214 - tcp_server now tries to unlink unix domain sockets when it gets destroyed.
215 - do not run event loop detection multiple times when programs call their cached
216 copies of AnyEvent::detect.
217 - parse_hostport now accepts absolute paths als unix domain sockets.
218 - suppress the idiotic warning inside POE by patching the POE::Kernel
219 run flags instead of asking the user to jump through contortions.
220 - add Task::Weaken as dependency (as recommends only at this time).
221 - replace PF_UNSPEC by 0, for good old minix.
222
2236.0 (not officially released, but was on CPAN for about 15 minutes :/)
224
2255.34 Thu May 12 09:59:41 CEST 2011
226 - AE::TLS didn't support both cert_file and key at the same time,
227 due to some copy&paste bug (reported and analyzed by vti).
228 - improve AE::Handle timeout documentation slightly.
229
2305.33 Mon Feb 7 21:51:41 CET 2011
231 - never appeared on CPAN, for unknown reasons.
232 - AnyEvent::Handle on_connect_error - do not try to destroy a handle
233 object when it's already gone (analyzed by Ryan Bullock).
234 - update idna algorithm(s) to UTS#46 version 6.0.0.
235 - preserve trailing dot in idn_to_ascii - testsuite says to preserve,
236 uts46 says it's an error, and testsuite makes more sense.
237 - remove leading dots in idn_to_ascii - testsuite says to remove,
238 uts46 says it's an error, and testsuite makes more sense.
239 - fixed a bug in idn ignored character processing.
240 - passes the uts46 6.0.0 testsuite except for three cases which
241 apparently are bugs in the testsuite itself.
242 - passes the uts46 6.0.1 testsuite except for eleven cases which
243 apparently are bugs in the testsuite itself.
244
2455.31 Sun Jan 23 11:49:19 CET 2011
246 - work around a POE bug where POE would first poll for new events and
247 invoke handlers BEFORE handling the message queue.
248 - fix parsing of "domain" statement in resolv.conf (Johannes Plunien).
249 - stop_read now does just that in TLS mode. see its
250 documentation for caveats.
251 - update to common::sense 3.4 warning set (some warnings in perl impose
252 an arbitrary 32kb limit on regex matches, which affects
253 AnyEvent::Handle).
254 - AE::Handle had a broken rbuf_max method and missing wbuf_max method.
255 - new experimental AE::Handle::resettls method.
256
2575.3 Fri Dec 31 05:48:13 CET 2010
258 - major IOAsync patch by Paul Evans - 0.33 is required, as the major
259 issues with IO::Async have been solved in that version, including
260 the availabilty of a default loop!
261 - implement new wbuf_max parameter for AnyEvent::Handle.
262 - added Cocoa::EventLoop backend by Daisuke Murase.
263 - the default testsuite now honors PERL_ANYEVENT_MODEL (Paul Evans).
264 - clarify handle on_prepare documentation.
265 - set CLOEXEC flag on parent-side fd's in AnyEvent::Util::run_cmd,
266 to avoid leaking them to any child processes.
267 - try to force files into the arch-dependend lib directory, for the
268 half-"turly obsessive compulsive person"(s) that depend on it.
269
2705.29 Sun Dec 5 10:49:21 CET 2010
271 - convert EV backend to EV 4.00 API (so better upgrade EV too).
272 - AE::Handle: implement dynamic read_size adjustment, add
273 max_read_size parameter, reduce default read_size to 2048.
274 - add some FAQ entries for servers and/or on_eof vs. on_error.
275 - work around OS bugs (cygwin again) when lingering in AE::Handle.
276 - reduce memory usage slightly when connect was used in AE::Handle.
277 - wrok around more segfaults in Tk.
278 - document tls_ctx => undef as valid.
279 - detect WNOHANG value at compiletime.
280 - include some probably linux-only support for building AnyEvent
281 as part of the perl core.
282 - improve compatibility of t/08_idna.t to perl 5.8.
283 - make AnyEvent::Strict truly optional (in case somebody builds a
284 minimal perl...).
285
2865.28 Wed Oct 13 04:14:23 CEST 2010
287 - due to a glitch, AnyEvent's internal getprotobyname wasn't
288 actually used in tcp_server.
289 - implement AnyEvent::CondVar->new method, which was documented but
290 missing.
291 - support autoloading of read types in unshift_read.
292 - AnyEvent::Strict: do the test for invalid fh arguments differently,
293 resulting in more informative error messages.
294 - new AnyEvent::FAQ manpage.
295 - clarify that you can actually call push_read etc. in on_read.
296 - put keys and certs in /tmp, not into the current directory,
297 and document this in the SECURITY CONSIDERATIONS sections
298 of AnyEvent::TLS.
299 - support ";" as resolv.conf comment character in AnyEvent::DNS, as
300 requested by Ron Isaacson.
301 - document $AnyEvent::DNS::RESOLVER variable.
302 - incorporated major typo patches by Abhijit Menon-Sen.
303 - AnyEvent::Handle now croaks when negative timeouts are passed.
304 - add a shitload of TCP_xxx sockopt name constants to AnyEvent::Util,
305 undocumented, but who knows when they might come in handy.
306
3075.271 Tue Jun 8 12:05:46 CEST 2010
308 - backport to perl 5.8.x.
309
3105.27 Sun Jun 6 12:12:05 CEST 2010
311 - postpone differently in AnyEvent::Socket now, as
312 when not, canceling the connection attempt might fail
313 (found by Felix Antonius Wilhelm Ostmann).
314 - explicitly check for non-stream sockets in AE::Handle, too many
315 clueless people fell into the trap of this somehow working.
316 - simplified and reworked the "OTHER MODULES" section.
317 - better/more condvar examples.
318
3195.261 Wed Apr 28 16:13:36 CEST 2010
320 - AF_INET6 was not properly used from Socket6 during configuration
321 time (found by Andy Grundman).
322
3235.26 Mon Apr 12 04:49:35 CEST 2010
324 - don't generate spurious readyness notifications when select
325 returns EINTR in the pure perl backend (analysed and testcase
326 by Bas Denissen).
327 - give same examples in SYNOPSIS sections of AnyEvent and AE.
328 - provider faster implementations for AE::time/now/now_update for
329 EV and Perl backends.
330 - provide faster AE::cv and AE::time implementations for all
331 backends.
332 - no longer support lower/mixed-case signal names (this was
333 never documented nor universally supported).
334 - some more memory µ-optimisations, and somewhat less messy
335 code to cope with both AnyEvent and AE APIs.
336
3375.251 Sat Mar 13 00:58:21 CET 2010
338 - make test failed if it was run before make install on perl <=
339 5.8.8, and older versions of perl make our live very difficult,
340 so write constants.pl during Makefile.PL time.
341
3425.25 Sat Mar 13 00:23:14 CET 2010
343 - fix a race condition in AnyEvent::Handle that would cause
344 a "bio_read: unsupported method" error (in _another_
345 ssl connection) after user code threw an exception.
346 - added AnyEvent::Handle->destroyed method.
347 - speed up AnyEvent::Socket::format_address and ::format_ipv6.
348 - the AnyEvent::Util::fh_nonblocking function worked only
349 by lucky accident on win32.
350 - smaller and faster AnyEvent::Util::fh_nonblocking.
351 - when the (required!) Time::HiRes module is missing, AnyEvent
352 did not fall back to built-in time properly.
353 - do not load Fcntl at runtime, saving memory and loading time.
354 - precompile a number of constants and use them instead of runtime
355 detection and eval.
356 - free detection code after detection and similar memory optimisations.
357 - Perl backend timer interval best effort drift has been improved
358 (same algorithm as EV).
359 - update unicode idna mapping table.
360
3615.24 Tue Jan 5 11:39:43 CET 2010
362 - cygwin never reports errors from failed connects. 1.5 just gives
363 you continous readyness and EAGAIN, 1.7 is even more broken and
364 just hangs. work around both issues in a cygwin-specific hack.
365 - improve idle watcher documentation slightly.
366
3675.23 Sun Dec 20 23:48:00 CET 2009
368 - support IDNs in resolve_sockaddr, and therefore in tcp_connect.
369 - implement punycode_encode/decode, idn_nameprep,
370 idn_to_ascii and idn_to_unicode operations in AnyEvent::Util.
371 - provide $AE::VERSION.
372 - removed traces of "no strict 'refs'".
373
3745.22 Sat Dec 5 03:51:13 CET 2009
375 - downgrade-or-fail in AnyEvent::Handle::push_write, to
376 diagnose encoding failures earlier and more succinctly.
377 (this works around bugs in perl, throwing away encoding info
378 when passing scalar data to extensions).
379 - add more examples to AnyEvent::Socket manpage.
380 - upgrade internal warning set to the same as common::sense 2.03.
381 - use pack "n/a*" for pre-5.8.9 perl compatibility in AnyEvent::DNS
382 (John Beppu).
383 - AnyEvent::Socket::inet_aton now properly supports ipv6, as documented.
384 - add google public dns servers to fallback server set and make sure
385 we load-balance properly between the three sets. also add all
386 fallback dns servers, not just a random one, to each dns config.
387
3885.21 Thu Nov 19 02:48:47 CET 2009
389 - fix a problem where socket constants were called with parameters
390 (spotted by David Friedland).
391 - fork_call never use'd POSIX (reported by Daisuke Maki).
392 - improve perl 5.6 compatibility further (but it still won't work
393 unless you rip out everything but the core).
394 - prefer Net::DNS::Resolver over ipconfig, if installed, on win32.
395 uses 10MB of RAM, but doesn't open a console window. *sigh*.
396
3975.202 Wed Oct 14 22:35:44 CEST 2009
398 - AnyEvent::DNS would unexpectedly clobber $_ under windows
399 (analysed by Matthias Waldorf).
400 - AnyEvent::Handle::run_cmd can now pass the PID of the
401 newly-created process, which is much less useful than it might
402 sound (based on patch by Yann Kerherve).
403
4045.201 Tue Sep 29 12:09:25 CEST 2009
405 - AnyEvent:Handle::on_starttls/on_stoptls methods were broken
406 (reported by Torsten Foertsch).
407 - common::sense 2.0 could cause tcp_server to throw an exception
408 (analysed by elmex).
409
4105.2 Mon Sep 14 07:04:49 CEST 2009
411 - INCOMPATIBLE CHANGE: do no longer support register_read_type
412 and register_write_type in AnyEvent::Handle, instead support
413 package names (the facility was mostly abused).
414 - implement "packagename-as-read/write type" support in
415 AnyEvent::Handle.
416 - AnyEvent::Handle: new options "keepalive" and "oobinline".
417 - oobinline set by default to avoid security issues.
418 - the pure-perl event loop backend wrongly detected times() fallback
419 support (spotted by Pavel Boldin).
420
4215.12 Tue Sep 1 20:26:50 CEST 2009
422 - be more lenient when parsing resolv.conf files, as some people
423 use hashmarks as comment indicator inside directives (reported
424 by Michael S. Fischer).
425 - use same set of warnings as common::sense 2.0.
426 - fix a potential 32 bit overflow issue due to perl having problems
427 with large hex constants in the Perl backend.
428 - do not use the slower 5.6 version fo the enc_name code in AnyEvent::DNS
429 by default.
430 - fix some prototypes.
431 - (optionally) require Async::Interrupt 1.02.
432 - replace opendns fallbacks by verizon/level3 public dns servers.
433
4345.112 Fri Aug 21 13:59:04 CEST 2009
435 - AnyEvent::Handle could cause hard crashes in Net::SSLeay when
436 using starttls with an illegal/uncreatable tls context.
437 - port to broken win32 perls.
438
4395.111 Sun Aug 16 18:44:36 CEST 2009
440 - perl errornously requires sockaddr_un structures returned by the
441 kernel to have a specific length. work around this bug by padding
442 sockaddr_un structures (found to fail on OS X, as perl only uses
443 the correct code for GNU/Linux, but likely to fail on other
444 OSes as well).
445
4465.11 Wed Aug 12 17:49:37 CEST 2009
447 - fixed a bug in run_cmd with input from filename.
448 - create a json coder object when none is given in write direction
449 to reduce dependency on JSON.
450
4515.1 Tue Aug 11 03:17:32 CEST 2009
452 - tcp_connect now reports the correct error in $!, not an unrelated
453 one (was broken since 4.91).
454 - AE::Handle did not properly clear rtimeout/wtimeout watchers
455 when retrying a tcp_connect.
456 - new functions: AnyEvent::Util::run_cmd and close_all_fds_except.
457 - fixed wrong documentation in AE::Socket.
458
4595.01 Mon Aug 10 03:16:32 CEST 2009
460 - last release broke EV child watchers :/.
461
4625.0 Sun Aug 9 17:32:09 CEST 2009
463 - officially introduce and document the AE API.
464 - lowering the timeout at runtime did not have immediate effect
465 in AnyEvent::Handle.
466 - AE::Handle now has separate and independent read and write timeout
467 settings.
468 - probe for Guard module when AnyEvent::Util is loaded, not at
469 runtime (Event doesn't like eval "use" from callbacks, and it avoids
470 exporting the probe function).
471 - Impl::IOAsync's io watchers did not use the filehandle emulation
472 code to work around it's watcher limits.
473 - work around IO::Async's problems with overloaded objects.
474 - try to disable wearnings inside IO::Async, it's just too buggy.
475 - do no longer use Event by default, only EV or Perl, Event is too buggy.
476 - hackishly provide fast AE:: interface when EV or Perl is the backend.
477 - inherit VERSION from AnyEvent::VERSION in most modules.
478
4794.91 Thu Aug 6 15:42:45 CEST 2009
480 - AE::Handle::starttls could get out of sync when the read buffer
481 already contains some TLS handshake.
482 - AE::Handle did not properly free the TLS session, causing bigger
483 memory leaks in Net::SSLeay than expected (reported by toaster).
484 - AE::Socket::tcp_connect will now artificially delay invoking
485 the callback to avoid returning after invoking the callback.
486 - convert many internal watcher uses to AE API.
487
4884.9 Sat Aug 1 11:07:01 CEST 2009
489 - Glib silently fails when registering a timeout with
490 a negative value, so avoid that.
491 - call condvar callback immediately when it is set
492 after the condition is already signalled.
493 - check rbuf_max condition only after trying to consume
494 data.
495 - normalise signal numbers to names when using the ae's
496 signal handling, but do not document this (yet).
497 - pure perl signal emulation did not properly set nonblocking
498 mode on the signal pipe on win32 (this is unlikely to have caused
499 any issues).
500 - new module: AnyEvent::Debug.
501 - AnyEvent::Strict now checks that a signal was specified
502 by name (not name or number).
503 - reduce memory footprint in typical cases by ~50kb by
504 only compiling signal code when necessary.
505 - add AnyEvent::Handle->rbuf_max.
506 - grab the AE namespace for future new API, implement stubs for
507 said future/alternative API.
508 - new function: AnyEvent::Socket::format_hostport.
509
5104.881 Tue Jul 28 12:51:53 CEST 2009
511 - work around a bug in local in pre-5.10 perls, causing
512 AnyEvent::Handle to recurse when it should not
513 (analysed by elmex).
514
5154.88 Tue Jul 28 04:04:37 CEST 2009
516 - re-bless the handle into a dummy package after calling
517 AnyEvent::Handle::destroy, so the user does not need to check
518 for errors after every push_write etc.
519 - do not attempt to run t/02_signals.t on obviously broken
520 platforms.
521
5224.87 Sun Jul 26 02:06:16 CEST 2009
523 - do not attempt to linger when there is no longer a valid fh.
524 - enforce tls mode to be either accept or connect, do not simply
525 segfault in Net::SSLeay.
526 - AnyEvent::Handle can now call tcp_connect itself (new parameters
527 connect, on_prepare, on_connect and on_connect_error). Updated
528 tutorial accordingly.
529 - add AnyEvent::Impl::Irssi backend.
530
5314.86 Mon Jul 20 23:52:29 CEST 2009
532 - since the verbose warning is not enough, explicitly document
533 that versions before 1.33 of Net::SSLeay are not secure.
534 - work around signal handling races in Event and (...) Event::Lib.
535 - try to align signal-race timer to full-second boundaries.
536 - work around Tk not liking negative timeouts.
537 - don't complain of different grades of environmental unfriendlyness
538 in IO::Async.
539
5404.85 Sat Jul 18 06:16:14 CEST 2009
541 - nail the signal race problem in perl once and for all
542 (see $AnyEvent::MAX_SIGNAL_LATENCY).
543 - take advantage of Async::Interrupt if it is available.
544 - load Time::HiRes and Guard modules on demand only.
545 - add optional/recommended modules section to AnyEvent
546 documentation.
547 - reduce memory usage considerably (and reduce startup penalty)
548 by not using "strict", "warnings" and "overload" modules.
549 - work around buggy windows/openbsd perls and provide EBADMSG
550 and EPROTO ourselves when missing.
551 - improve perl 5.6 compatibility of the core event loop.
552 - made Net::SSLeay version 1.33 a soft requirement.
553
5544.83 Fri Jul 17 16:56:26 CEST 2009
555 - implement AnyEvent::Socket::getprotobyname.
556 - AnyEvent::CondVar's will now detect recursive blocking
557 waits and will croak, as too many people fall into
558 this trap.
559 - AnyEvent::Handle will now call ->destroy on itself after
560 executing the on_error callback, instead of doing some
561 half-baked internal shutdown, for fatal errors.
562 - clarify on_eof behaviour w.r.t. the read queue and
563 on_read callbacks.
564 - ignore some possible spurious wake-ups in tcp_connect.
565
5664.82 Sat Jul 11 00:34:55 CEST 2009
567 - POE and Event backends didn't accept some callable objects as
568 callbacks.
3 - use Config module instead of POSIX module to detect signal names 569 - use Config module instead of POSIX module to detect signal names
4 in AnyEvent::Strict and AnyEvent::Impl::EventLib, as the POSIX 570 in AnyEvent::Strict and AnyEvent::Impl::EventLib, as the POSIX
5 module doesn't even have all POSIX signals :/. 571 module doesn't even have all POSIX signals :/.
6 - use more workarounds around the many refcnt/corruption bugs in 572 - use more workarounds around the many refcnt/corruption bugs in
7 Event::Lib. 573 Event::Lib.
… …
1744.3 Fri Oct 3 09:18:43 CEST 2008 7404.3 Fri Oct 3 09:18:43 CEST 2008
175 - AnyEvent will now install a no-op signal handler for SIGPIPE, 741 - AnyEvent will now install a no-op signal handler for SIGPIPE,
176 unless one has been installed already. 742 unless one has been installed already.
177 - warn about not loaidng AnyEvent::Impl::POE early enough 743 - warn about not loaidng AnyEvent::Impl::POE early enough
178 (patch by Adam Rosenstein). 744 (patch by Adam Rosenstein).
179 - fixed a great number of bugs and corner cases in AnyEvent::Handle: 745 - fixed a great number of bugs and corner cases in AnyEvent::Handle.
180 - fix a bug in where in SSL connect mode, the 746 - fix a bug in where in SSL connect mode, the
181 client would first wait for some data by the server and 747 client would first wait for some data by the server and
182 otherwise hang. (reported and analyzed in an absolutely 748 otherwise hang. (reported and analysed in an absolutely
183 exemplary manner by Adam Rosenstein). 749 exemplary manner by Adam Rosenstein).
184 - fix a bug in where SSL EOF would not be treated 750 - fix a bug in where SSL EOF would not be treated
185 as stream EOF, putting the connection into a hung state. 751 as stream EOF, putting the connection into a hung state.
186 (reported and analyzed in an absolutely exemplary manner by Adam 752 (reported and analysed in an absolutely exemplary manner by Adam
187 Rosenstein). 753 Rosenstein).
188 - fix a potential segfault when the TLS context 754 - fix a potential segfault when the TLS context
189 would go missing in a read callback (e.g. due to stoptls) - 755 would go missing in a read callback (e.g. due to stoptls) -
190 Net::SSLeay of course makes no type checking whatsoever. 756 Net::SSLeay of course makes no type checking whatsoever.
191 - AnyEvent::Handle will not stop the read watcher in TLS mode. 757 - AnyEvent::Handle will not stop the read watcher in TLS mode.

Diff Legend

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