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

Comparing AnyEvent/Changes (file contents):
Revision 1.152 by root, Thu Jun 5 18:30:08 2008 UTC vs.
Revision 1.582 by root, Tue Dec 10 15:54:51 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: newer openssl versions don't like 0-byte writes, work around
13TODO: mention AnyEvent::fork etc.
14TODO: mention AnyEvent::Impl::IOAsync::set_loop (also implement get_loop).
15TODO: From: Maxime Soulé <btik-cpan@scoubidou.com> verify_peername => [ 2, 0, 0 ]
16 - added cbor read and write types to AnyEvent::Handle (using CBOR::XS).
17 - calling AnyEvent->now_update with AnyEvent::Impl::Perl caused an
18 endless loop (reported by Dietrich Rebmann).
19 - add tlsv1_1 and tlsv1_2 protocols to AnyEvent::TLS
20 (patch by Maxime Soulé).
21 - force a toplevel domain name in t/81_hosts.t (analyzed by
22 David Jack Wange Olrik).
23 - document that AnyEvent::Log uses AnyEvent::IO.
24 - time for an =encoding directive.
25
267.05 Wed Aug 21 10:38:08 CEST 2013
27 - uts46data.pl couldn't be found due to wrong naming of the file
28 (reported by Fulko Hew).
29 - handle lone \015's properly in AE::Handle's default line read
30 (reported by various people).
31 - untaint ip addresses found in /etc/hosts (patch by José Micó).
32 - the memleak fix in 7.03 caused resolving via /etc/hosts to always
33 fail on first use (reported and testcase by Andrew Whatson).
34 - expose AnyEvent::Log::format_time, and allow users to redefine it.
35 - expose AnyEvent::Log::default_format, and allow redefinition.
36 - expose AnyEvent::Log::fatal_exit, to allow redefinition.
37 - AnyEvent::Debug shell can now run coro shell commands, if available.
38 - t/63* tests were wrongly in MANIFEST.
39 - kernel.org's finger server went MIA, switch to freebsd.org and
40 icculus.org.
41 - clarify that IO::AIO and AnyEvent::AIO are needed for AnyEvent::IO
42 to function asynchronously (reported by Daniel Carosone).
43 - hard-disable $^W in most tests, it generates too much garbage output.
44 - use a (hopefully) more future-proof method to emulate common::sense.
45 - upgrade to UTS-46:6.2.0.
46 - switch to INSTLIB from INSTLIBDIR, as INSTLIBDIR was wrongly
47 documented. should not affect anything.
48
497.04 Thu Nov 15 02:23:58 CET 2012
50 - debugging code left in.
51
527.03 Thu Nov 15 02:16:11 CET 2012
53 - AnyEvent::Socket::inet_aton did not work when dns resolution was
54 used to find the addresses (analyzed and patch by Maxime Soulé).
55 - fix a memory leak in the /etc/hosts lookup code when hosts don't
56 resolve and are not in hosts (analyzed by sten).
57
587.02 Tue Aug 14 04:27:58 CEST 2012
59 - AnyEvent::Util::run_cmd could block indefinitely (analyzed and test
60 program by Yorhel).
61 - verified that AnyEvent::Socket follows RFC5952.
62 - try to parse "ADDR#PORT" in addition to "ADDR PORT".
63
647.01 Sun May 13 01:03:17 CEST 2012
65 - fail with EPROTO in AnyEvent::Handle wqhen TLS is requested but
66 not available, instead of throwing an exception.
67 - use File::Spec to get the tmpdir in t/*, to avoid needless
68 failures on (most, not mine :) windows boxes.
69 - new handle read types: tls_detect and tls_autostart.
70
717.0 Fri Apr 13 06:33:30 CEST 2012
72 - child watchers are broken in POE 1.352 (also many earlier
73 versions) and there seems to be no way to work around it, as POE
74 itself is inherently racy. Document this regression and add a delay
75 in t/68_poe_03_child.t for the time being.
76 - new module AnyEvent::IO, that is a frontend to either a pure-perl
77 synchronous I/O implementation (AnyEvent::IO::Perl), or to the
78 asynchronous I/O extension IO::AIO (AnyEvent::IO::IOAIO).
79 - load /etc/hosts only when DNS has no answer.
80 - stat /etc/hosts on every access and reload it if it changed.
81 - load /etc/hosts via AnyEvent::IO - potentially asynchronous.
82 - fix a buggy croak in the dh parameter reading in AnyEvent::TLS.
83 - AnyEvent::Log log_to_file and log_to_path now use AnyEvent::IO. As
84 a side effect, they now use true appending as opposed to libc
85 appending, and the file might not have been opened when the function
86 returns.
87 - the default logging level was not properly documented in a variety of
88 places, this has hopefully been rectified.
89 - updated uts46data.pl for unicode 6.1.0.
90 - made log messages generated by AnyEvent submodules not
91 include the package name anymore, as it will be logged
92 by default already.
93 - upgrade to the trick used by common-sense 3.5 to work
94 around extra warning torture/breakage under perl 5.15.x.
95 - log messages by anyevent are now ucfirst, are usually full
96 sentences and do no longer include the package name.
97 - the storable read type would sometimes throw an exception instead
98 of causing EBADMSG (analyzed by Maxime Soulé).
99 - work around a bug in openssl 1.0.1 which enforces a minimum rsa
100 keysize (reported by Paul Howarth).
101 - documented Rocco fabricating statements about AnyEvent and me.
102
1036.14 Tue Jan 31 20:00:24 CET 2012
104 - AnyEvent::Impl::Tk was broken due to a mysterious "die" inside,
105 probably an editing mistake (reported by Darin McBride).
106
1076.13 Thu Jan 12 07:27:01 CET 2012
108 - AnyEvent::Util::fork_call checked for POSIX availability in the wrong way,
109 causing it to fail if POSIX wasn't loaded already (analysed by Rock Power).
110 - AnyEvent::Handle::push_read (line => did pass $1 directly, so
111 regex-matching inside the callback would change the parameter, despite
112 $1 being dynamically scoped per-block. this perl bug is now being
113 worked around (testcase by Cindy Wang).
114
1156.12 Mon Dec 12 13:21:10 CET 2011
116 - $! was clobbered by subsequent calls in tcp_connect, due to
117 the postpone not saving/restoring it (analyzed by Richard Garnish).
118
1196.11 Tue Nov 22 10:36:05 CET 2011
120 - Tk cannot create windows when tainted unless you set a title, so
121 set a dummy title for AnyEvent's dummy mainwindow. How dummy.
122 - escape any nonprintable/nonascii characters when stringifying
123 backtraces.
124 - log the reason loading an interface module fails to load at level debug.
125
1266.1 Tue Oct 4 19:44:30 CEST 2011
127 - INCOMPATIBLE CHANGE: the default log level is now "4" (error
128 and above), and some messages inside AnyEvent have been elevated
129 to higher log levels to print by default.
130 - AnyEvent::log, unlike AnyEvent::Log::log, did not always exit on fatal.
131 - work around yet another signal race bug, newly introduced by POE. My
132 how that module sucks.
133 - add google public ipv6 dns as fallback servers, so AE::DNS can finally
134 support ipv6-only operation even with defaults (also use ::1 instead
135 of 127.0.0.1).
136 - give better guidelines on how to chose a logging level.
137 - AnyEvent::Log can now cap (limit) log message priority by context.
138 - AE::log will now save and restore $! and $@, so callers don't have to.
139 - AE::Strict now checks that registered fds are still valid at regular
140 intervals.
141 - skip frequently called 'require' calls if the module appears loaded.
142 - increase signal latency test timer considerably for poor little osx,
143 which can't be bothered with basic posix functionality such as signal
144 delivery.
145 - mention some debugging aids in the tutorial.
146 - new env variable PERL_ANYEVENT_MAX_SIGNAL_LATENCY.
147 - private (%namespace) log contexts in PERL_ANYEVENT_LOG now have the name
148 of the context as title.
149 - AnyEvent::Strict would not restore read-write access to $_ when a callback
150 died - this is a user bug, but since AE::Strict exists to help dbeugging...
151 - detect the extremely evil IO::Async::Loop::AnyEvent module and refuse
152 to work with it - the author is well aware of the issues but refuses
153 to fix them. This extreme step has been taken because the module actively
154 attacks the goals of AnyEvent and makes it harder for module authors to use
155 AnyEvent.
156
1576.02 Fri Aug 26 20:08:31 CEST 2011
158 - log_to_syslog now accepts facility strings.
159 - log_to_syslog would not allow facilities - "$facility|$level"
160 *literally* means that string. sheesh.
161 - the CPAN parser got confused and thought we provide Sys::Syslog, try
162 to work around.
163 - renamed AnyEvent::Impl::FLTK2 to FLTK.
164
1656.01 Fri Aug 26 07:04:11 CEST 2011
166 - INCOMPATIBLE CHANGE: AnyEvent::DNS resource records now include
167 the TTL as fourth element - this affects ->resolve, ->request
168 and ->dns_unpack indirectly. Convenience functions such as
169 AnyEvent::DNS::{a,txt,mx,srv} are not affected, neither is
170 AnyEvent::Socket.
171 - INCOMPATIBLE CHANGE: backend authors now should not implement
172 one_event or loop, but instead the AnyEvent::CondVar::_wait and _poll
173 methods.
174 - INCOMPATIBLE CHANGE: verbosity levels have been redefined, see
175 AnyEvent (PERL_ANYEVENT_VERBOSE) and AnyEvent::Log for details.
176 - DEPRECIATION: please change
177 "use AnyEvent::Impl::Perl" to "use AnyEvent::Loop" and
178 "AnyEvent::Impl::Perl::loop" to "AnyEvent::Loop::run"
179 in your programs, for future compatibility.
180 - *blush* idle watcher emulation was borked.
181 - AnyEvent::Strict now checks the AE::xxx functions as well.
182 - querying too long domain names would barf deep inside AnyEvent::DNS - since
183 this is hard to avoid for applications, AnyEvent:.DNS::request now checks for
184 for too long domain names explicitly and passes an undef to the callback
185 instead (reported by James Bromberger).
186 - make AnyEvent::Util more robust against EINTR (analyzed by Andrew Suffield).
187 - new functions: AnyEvent::postpone, AnyEvent::log.
188 - new module: AnyEvent::Log, for simple logging needs.
189 - new env variable, PERL_ANYEVENT_LOG, for fine-grained logging config.
190 - new env variable, PERL_ANYEVENT_DEBUG_SHELL, to automatically start
191 a debug shell.
192 - new env variable, PERL_ANYEVENT_DEBUG_WRAP, to automatically enable
193 watcher instrumentation/wrapping (see AnyEvent::Debug::wrap).
194 - new env variable, PERL_ANYEVENT_HOSTS, to override the place where
195 /etc/hosts can be found.
196 - an empty PERL_ANYEVENT_RESOLV_CONF now selects the default
197 configuration.
198 - env variables can now be specified using AE_xyz instead of
199 PERL_ANYEVENT_xyz, subject to some rules.
200 - lots of new AnyEvent::Debug functionality.
201 - AnyEvent::Strict now checks for modifications of $_ by the callback
202 (by setting it to readonly while executing the callback, if possible).
203 - IO::Async sometimes cancels the wrong timer when you call cancel_timer.
204 Invest extra overhead to work around this peculiar behaviour.
205 - split the pure perl backend into a loop module and an impl
206 module (AnyEvent::Loop and AnyEvent::Impl::Perl).
207 - some read types (regex, netstring, json, storable) could
208 errornously keep an AE::Handle object alive on their own
209 (reported by Mohammad Toossi).
210 - AnyEvent::Socket::resolve_sockaddr and all functions using it now
211 supports /etc/hosts overrides.
212 - add more workarounds around child watcher bugs in IO::Async
213 and (newly introduced) in POE.
214 - use glib child watchers and try to work around its limitations,
215 if possible (based on a patch by Kevin Ryde).
216 - add support for FLTK (fltk2), via AnyEvent::Impl::FLTK2.
217 - many read types in AnyEvent::Handle would malfunction during
218 transitions to TLS because of a "delete $self->{rbuf}".
219 (reported by Gerald Galster).
220 - improve condvar blocking wait performance (probably only noticable
221 with faster backends), also make it easier for event loops that
222 do not support blocking waits.
223 - PERL_ANYEVENT_MODEL now also allows full module names, although
224 the usefulness of this feature is not yet clear.
225 - add tests for all supported event loops, conditional on PERL_ANYEVENT_LOOP_TESTS.
226 - add t/80_ssltext.t.
227 - no longer use AUTOLOAD for the initial loading (for no very good reason).
228 - tcp_server now tries to unlink unix domain sockets when it gets destroyed.
229 - do not run event loop detection multiple times when programs call their cached
230 copies of AnyEvent::detect.
231 - parse_hostport now accepts absolute paths als unix domain sockets.
232 - suppress the idiotic warning inside POE by patching the POE::Kernel
233 run flags instead of asking the user to jump through contortions.
234 - add Task::Weaken as dependency (as recommends only at this time).
235 - replace PF_UNSPEC by 0, for good old minix.
236
2376.0 (not officially released, but was on CPAN for about 15 minutes :/)
238
2395.34 Thu May 12 09:59:41 CEST 2011
240 - AE::TLS didn't support both cert_file and key at the same time,
241 due to some copy&paste bug (reported and analyzed by vti).
242 - improve AE::Handle timeout documentation slightly.
243
2445.33 Mon Feb 7 21:51:41 CET 2011
245 - never appeared on CPAN, for unknown reasons.
246 - AnyEvent::Handle on_connect_error - do not try to destroy a handle
247 object when it's already gone (analyzed by Ryan Bullock).
248 - update idna algorithm(s) to UTS#46 version 6.0.0.
249 - preserve trailing dot in idn_to_ascii - testsuite says to preserve,
250 uts46 says it's an error, and testsuite makes more sense.
251 - remove leading dots in idn_to_ascii - testsuite says to remove,
252 uts46 says it's an error, and testsuite makes more sense.
253 - fixed a bug in idn ignored character processing.
254 - passes the uts46 6.0.0 testsuite except for three cases which
255 apparently are bugs in the testsuite itself.
256 - passes the uts46 6.0.1 testsuite except for eleven cases which
257 apparently are bugs in the testsuite itself.
258
2595.31 Sun Jan 23 11:49:19 CET 2011
260 - work around a POE bug where POE would first poll for new events and
261 invoke handlers BEFORE handling the message queue.
262 - fix parsing of "domain" statement in resolv.conf (Johannes Plunien).
263 - stop_read now does just that in TLS mode. see its
264 documentation for caveats.
265 - update to common::sense 3.4 warning set (some warnings in perl impose
266 an arbitrary 32kb limit on regex matches, which affects
267 AnyEvent::Handle).
268 - AE::Handle had a broken rbuf_max method and missing wbuf_max method.
269 - new experimental AE::Handle::resettls method.
270
2715.3 Fri Dec 31 05:48:13 CET 2010
272 - major IOAsync patch by Paul Evans - 0.33 is required, as the major
273 issues with IO::Async have been solved in that version, including
274 the availabilty of a default loop!
275 - implement new wbuf_max parameter for AnyEvent::Handle.
276 - added Cocoa::EventLoop backend by Daisuke Murase.
277 - the default testsuite now honors PERL_ANYEVENT_MODEL (Paul Evans).
278 - clarify handle on_prepare documentation.
279 - set CLOEXEC flag on parent-side fd's in AnyEvent::Util::run_cmd,
280 to avoid leaking them to any child processes.
281 - try to force files into the arch-dependend lib directory, for the
282 half-"turly obsessive compulsive person"(s) that depend on it.
283
2845.29 Sun Dec 5 10:49:21 CET 2010
285 - convert EV backend to EV 4.00 API (so better upgrade EV too).
286 - AE::Handle: implement dynamic read_size adjustment, add
287 max_read_size parameter, reduce default read_size to 2048.
288 - add some FAQ entries for servers and/or on_eof vs. on_error.
289 - work around OS bugs (cygwin again) when lingering in AE::Handle.
290 - reduce memory usage slightly when connect was used in AE::Handle.
291 - wrok around more segfaults in Tk.
292 - document tls_ctx => undef as valid.
293 - detect WNOHANG value at compiletime.
294 - include some probably linux-only support for building AnyEvent
295 as part of the perl core.
296 - improve compatibility of t/08_idna.t to perl 5.8.
297 - make AnyEvent::Strict truly optional (in case somebody builds a
298 minimal perl...).
299
3005.28 Wed Oct 13 04:14:23 CEST 2010
301 - due to a glitch, AnyEvent's internal getprotobyname wasn't
302 actually used in tcp_server.
303 - implement AnyEvent::CondVar->new method, which was documented but
304 missing.
305 - support autoloading of read types in unshift_read.
306 - AnyEvent::Strict: do the test for invalid fh arguments differently,
307 resulting in more informative error messages.
308 - new AnyEvent::FAQ manpage.
309 - clarify that you can actually call push_read etc. in on_read.
310 - put keys and certs in /tmp, not into the current directory,
311 and document this in the SECURITY CONSIDERATIONS sections
312 of AnyEvent::TLS.
313 - support ";" as resolv.conf comment character in AnyEvent::DNS, as
314 requested by Ron Isaacson.
315 - document $AnyEvent::DNS::RESOLVER variable.
316 - incorporated major typo patches by Abhijit Menon-Sen.
317 - AnyEvent::Handle now croaks when negative timeouts are passed.
318 - add a shitload of TCP_xxx sockopt name constants to AnyEvent::Util,
319 undocumented, but who knows when they might come in handy.
320
3215.271 Tue Jun 8 12:05:46 CEST 2010
322 - backport to perl 5.8.x.
323
3245.27 Sun Jun 6 12:12:05 CEST 2010
325 - postpone differently in AnyEvent::Socket now, as
326 when not, canceling the connection attempt might fail
327 (found by Felix Antonius Wilhelm Ostmann).
328 - explicitly check for non-stream sockets in AE::Handle, too many
329 clueless people fell into the trap of this somehow working.
330 - simplified and reworked the "OTHER MODULES" section.
331 - better/more condvar examples.
332
3335.261 Wed Apr 28 16:13:36 CEST 2010
334 - AF_INET6 was not properly used from Socket6 during configuration
335 time (found by Andy Grundman).
336
3375.26 Mon Apr 12 04:49:35 CEST 2010
338 - don't generate spurious readyness notifications when select
339 returns EINTR in the pure perl backend (analysed and testcase
340 by Bas Denissen).
341 - give same examples in SYNOPSIS sections of AnyEvent and AE.
342 - provider faster implementations for AE::time/now/now_update for
343 EV and Perl backends.
344 - provide faster AE::cv and AE::time implementations for all
345 backends.
346 - no longer support lower/mixed-case signal names (this was
347 never documented nor universally supported).
348 - some more memory µ-optimisations, and somewhat less messy
349 code to cope with both AnyEvent and AE APIs.
350
3515.251 Sat Mar 13 00:58:21 CET 2010
352 - make test failed if it was run before make install on perl <=
353 5.8.8, and older versions of perl make our live very difficult,
354 so write constants.pl during Makefile.PL time.
355
3565.25 Sat Mar 13 00:23:14 CET 2010
357 - fix a race condition in AnyEvent::Handle that would cause
358 a "bio_read: unsupported method" error (in _another_
359 ssl connection) after user code threw an exception.
360 - added AnyEvent::Handle->destroyed method.
361 - speed up AnyEvent::Socket::format_address and ::format_ipv6.
362 - the AnyEvent::Util::fh_nonblocking function worked only
363 by lucky accident on win32.
364 - smaller and faster AnyEvent::Util::fh_nonblocking.
365 - when the (required!) Time::HiRes module is missing, AnyEvent
366 did not fall back to built-in time properly.
367 - do not load Fcntl at runtime, saving memory and loading time.
368 - precompile a number of constants and use them instead of runtime
369 detection and eval.
370 - free detection code after detection and similar memory optimisations.
371 - Perl backend timer interval best effort drift has been improved
372 (same algorithm as EV).
373 - update unicode idna mapping table.
374
3755.24 Tue Jan 5 11:39:43 CET 2010
376 - cygwin never reports errors from failed connects. 1.5 just gives
377 you continous readyness and EAGAIN, 1.7 is even more broken and
378 just hangs. work around both issues in a cygwin-specific hack.
379 - improve idle watcher documentation slightly.
380
3815.23 Sun Dec 20 23:48:00 CET 2009
382 - support IDNs in resolve_sockaddr, and therefore in tcp_connect.
383 - implement punycode_encode/decode, idn_nameprep,
384 idn_to_ascii and idn_to_unicode operations in AnyEvent::Util.
385 - provide $AE::VERSION.
386 - removed traces of "no strict 'refs'".
387
3885.22 Sat Dec 5 03:51:13 CET 2009
389 - downgrade-or-fail in AnyEvent::Handle::push_write, to
390 diagnose encoding failures earlier and more succinctly.
391 (this works around bugs in perl, throwing away encoding info
392 when passing scalar data to extensions).
393 - add more examples to AnyEvent::Socket manpage.
394 - upgrade internal warning set to the same as common::sense 2.03.
395 - use pack "n/a*" for pre-5.8.9 perl compatibility in AnyEvent::DNS
396 (John Beppu).
397 - AnyEvent::Socket::inet_aton now properly supports ipv6, as documented.
398 - add google public dns servers to fallback server set and make sure
399 we load-balance properly between the three sets. also add all
400 fallback dns servers, not just a random one, to each dns config.
401
4025.21 Thu Nov 19 02:48:47 CET 2009
403 - fix a problem where socket constants were called with parameters
404 (spotted by David Friedland).
405 - fork_call never use'd POSIX (reported by Daisuke Maki).
406 - improve perl 5.6 compatibility further (but it still won't work
407 unless you rip out everything but the core).
408 - prefer Net::DNS::Resolver over ipconfig, if installed, on win32.
409 uses 10MB of RAM, but doesn't open a console window. *sigh*.
410
4115.202 Wed Oct 14 22:35:44 CEST 2009
412 - AnyEvent::DNS would unexpectedly clobber $_ under windows
413 (analysed by Matthias Waldorf).
414 - AnyEvent::Handle::run_cmd can now pass the PID of the
415 newly-created process, which is much less useful than it might
416 sound (based on patch by Yann Kerherve).
417
4185.201 Tue Sep 29 12:09:25 CEST 2009
419 - AnyEvent:Handle::on_starttls/on_stoptls methods were broken
420 (reported by Torsten Foertsch).
421 - common::sense 2.0 could cause tcp_server to throw an exception
422 (analysed by elmex).
423
4245.2 Mon Sep 14 07:04:49 CEST 2009
425 - INCOMPATIBLE CHANGE: do no longer support register_read_type
426 and register_write_type in AnyEvent::Handle, instead support
427 package names (the facility was mostly abused).
428 - implement "packagename-as-read/write type" support in
429 AnyEvent::Handle.
430 - AnyEvent::Handle: new options "keepalive" and "oobinline".
431 - oobinline set by default to avoid security issues.
432 - the pure-perl event loop backend wrongly detected times() fallback
433 support (spotted by Pavel Boldin).
434
4355.12 Tue Sep 1 20:26:50 CEST 2009
436 - be more lenient when parsing resolv.conf files, as some people
437 use hashmarks as comment indicator inside directives (reported
438 by Michael S. Fischer).
439 - use same set of warnings as common::sense 2.0.
440 - fix a potential 32 bit overflow issue due to perl having problems
441 with large hex constants in the Perl backend.
442 - do not use the slower 5.6 version fo the enc_name code in AnyEvent::DNS
443 by default.
444 - fix some prototypes.
445 - (optionally) require Async::Interrupt 1.02.
446 - replace opendns fallbacks by verizon/level3 public dns servers.
447
4485.112 Fri Aug 21 13:59:04 CEST 2009
449 - AnyEvent::Handle could cause hard crashes in Net::SSLeay when
450 using starttls with an illegal/uncreatable tls context.
451 - port to broken win32 perls.
452
4535.111 Sun Aug 16 18:44:36 CEST 2009
454 - perl errornously requires sockaddr_un structures returned by the
455 kernel to have a specific length. work around this bug by padding
456 sockaddr_un structures (found to fail on OS X, as perl only uses
457 the correct code for GNU/Linux, but likely to fail on other
458 OSes as well).
459
4605.11 Wed Aug 12 17:49:37 CEST 2009
461 - fixed a bug in run_cmd with input from filename.
462 - create a json coder object when none is given in write direction
463 to reduce dependency on JSON.
464
4655.1 Tue Aug 11 03:17:32 CEST 2009
466 - tcp_connect now reports the correct error in $!, not an unrelated
467 one (was broken since 4.91).
468 - AE::Handle did not properly clear rtimeout/wtimeout watchers
469 when retrying a tcp_connect.
470 - new functions: AnyEvent::Util::run_cmd and close_all_fds_except.
471 - fixed wrong documentation in AE::Socket.
472
4735.01 Mon Aug 10 03:16:32 CEST 2009
474 - last release broke EV child watchers :/.
475
4765.0 Sun Aug 9 17:32:09 CEST 2009
477 - officially introduce and document the AE API.
478 - lowering the timeout at runtime did not have immediate effect
479 in AnyEvent::Handle.
480 - AE::Handle now has separate and independent read and write timeout
481 settings.
482 - probe for Guard module when AnyEvent::Util is loaded, not at
483 runtime (Event doesn't like eval "use" from callbacks, and it avoids
484 exporting the probe function).
485 - Impl::IOAsync's io watchers did not use the filehandle emulation
486 code to work around it's watcher limits.
487 - work around IO::Async's problems with overloaded objects.
488 - try to disable wearnings inside IO::Async, it's just too buggy.
489 - do no longer use Event by default, only EV or Perl, Event is too buggy.
490 - hackishly provide fast AE:: interface when EV or Perl is the backend.
491 - inherit VERSION from AnyEvent::VERSION in most modules.
492
4934.91 Thu Aug 6 15:42:45 CEST 2009
494 - AE::Handle::starttls could get out of sync when the read buffer
495 already contains some TLS handshake.
496 - AE::Handle did not properly free the TLS session, causing bigger
497 memory leaks in Net::SSLeay than expected (reported by toaster).
498 - AE::Socket::tcp_connect will now artificially delay invoking
499 the callback to avoid returning after invoking the callback.
500 - convert many internal watcher uses to AE API.
501
5024.9 Sat Aug 1 11:07:01 CEST 2009
503 - Glib silently fails when registering a timeout with
504 a negative value, so avoid that.
505 - call condvar callback immediately when it is set
506 after the condition is already signalled.
507 - check rbuf_max condition only after trying to consume
508 data.
509 - normalise signal numbers to names when using the ae's
510 signal handling, but do not document this (yet).
511 - pure perl signal emulation did not properly set nonblocking
512 mode on the signal pipe on win32 (this is unlikely to have caused
513 any issues).
514 - new module: AnyEvent::Debug.
515 - AnyEvent::Strict now checks that a signal was specified
516 by name (not name or number).
517 - reduce memory footprint in typical cases by ~50kb by
518 only compiling signal code when necessary.
519 - add AnyEvent::Handle->rbuf_max.
520 - grab the AE namespace for future new API, implement stubs for
521 said future/alternative API.
522 - new function: AnyEvent::Socket::format_hostport.
523
5244.881 Tue Jul 28 12:51:53 CEST 2009
525 - work around a bug in local in pre-5.10 perls, causing
526 AnyEvent::Handle to recurse when it should not
527 (analysed by elmex).
528
5294.88 Tue Jul 28 04:04:37 CEST 2009
530 - re-bless the handle into a dummy package after calling
531 AnyEvent::Handle::destroy, so the user does not need to check
532 for errors after every push_write etc.
533 - do not attempt to run t/02_signals.t on obviously broken
534 platforms.
535
5364.87 Sun Jul 26 02:06:16 CEST 2009
537 - do not attempt to linger when there is no longer a valid fh.
538 - enforce tls mode to be either accept or connect, do not simply
539 segfault in Net::SSLeay.
540 - AnyEvent::Handle can now call tcp_connect itself (new parameters
541 connect, on_prepare, on_connect and on_connect_error). Updated
542 tutorial accordingly.
543 - add AnyEvent::Impl::Irssi backend.
544
5454.86 Mon Jul 20 23:52:29 CEST 2009
546 - since the verbose warning is not enough, explicitly document
547 that versions before 1.33 of Net::SSLeay are not secure.
548 - work around signal handling races in Event and (...) Event::Lib.
549 - try to align signal-race timer to full-second boundaries.
550 - work around Tk not liking negative timeouts.
551 - don't complain of different grades of environmental unfriendlyness
552 in IO::Async.
553
5544.85 Sat Jul 18 06:16:14 CEST 2009
555 - nail the signal race problem in perl once and for all
556 (see $AnyEvent::MAX_SIGNAL_LATENCY).
557 - take advantage of Async::Interrupt if it is available.
558 - load Time::HiRes and Guard modules on demand only.
559 - add optional/recommended modules section to AnyEvent
560 documentation.
561 - reduce memory usage considerably (and reduce startup penalty)
562 by not using "strict", "warnings" and "overload" modules.
563 - work around buggy windows/openbsd perls and provide EBADMSG
564 and EPROTO ourselves when missing.
565 - improve perl 5.6 compatibility of the core event loop.
566 - made Net::SSLeay version 1.33 a soft requirement.
567
5684.83 Fri Jul 17 16:56:26 CEST 2009
569 - implement AnyEvent::Socket::getprotobyname.
570 - AnyEvent::CondVar's will now detect recursive blocking
571 waits and will croak, as too many people fall into
572 this trap.
573 - AnyEvent::Handle will now call ->destroy on itself after
574 executing the on_error callback, instead of doing some
575 half-baked internal shutdown, for fatal errors.
576 - clarify on_eof behaviour w.r.t. the read queue and
577 on_read callbacks.
578 - ignore some possible spurious wake-ups in tcp_connect.
579
5804.82 Sat Jul 11 00:34:55 CEST 2009
581 - POE and Event backends didn't accept some callable objects as
582 callbacks.
583 - use Config module instead of POSIX module to detect signal names
584 in AnyEvent::Strict and AnyEvent::Impl::EventLib, as the POSIX
585 module doesn't even have all POSIX signals :/.
586 - use more workarounds around the many refcnt/corruption bugs in
587 Event::Lib.
588 - work around a race condition in perl's select, causing t/03_child.t
589 to rarely fail.
590
5914.81 Thu Jul 9 10:30:30 CEST 2009
592 - AnyEvent::Handle didn't properly diagnose write errors
593 (it expected -1 from syswrite, how lame... :).
594 - support file descriptors in addition to file handles
595 in AnyEvent->io.
596 - new env variables: PERL_ANYEVENT_RESOLV_CONF,
597 PERL_ANYEVENT_MAX_OUTSTANDING_DNS, PERL_ANYEVENT_CA_FILE
598 and PERL_ANYEVENT_CA_PATH.
599 - provide a sensible synopsis section for AnyEvent::TLS.
600 - add a "supported backends" section to the manpage.
601 - added simple io watcher test to testsuite, using a
602 portable_socketpair.
603 - tried to improve the stability of the Event::Lib backend,
604 YMMV.
605
6064.8 Mon Jul 6 23:45:16 CEST 2009
607 - AnyEvent::DNS did not properly follow CNAME records with
608 uppercase targets.
609 - AnyEvent::DNS would errornously return AAAA records
610 with v4 mapped addresses (a faulty record) as ipv4 addresses,
611 causing AnyEvent::Socket to throw an exception.
612 - added new module AnyEvent::TLS for easier SSL/TLS context
613 creation, with many options including hostname verification,
614 secure default configuration, lots of documentation and,
615 predefined diffie-hellman keys for perfect forward security
616 and much more. get it while it's still fresh!
617 - use AnyEvent::TLS in AnyEvent::Handle for context management.
618 - load AnyEvent::Handle only on demand in AnyEvent::DNS,
619 so AnyEvent::Socket users have smaller memory footprint
620 in the common case.
621 - add AnyEvent::Handle->push_shutdown method.
622 - add an additional error message parameter to AnyEvent::Handle's
623 on_error callback (for TLS, $! is still available).
624 - add AnyEvent::Handle on_starttls/on_stoptls callbacks.
625 - make AnyEvent::Handle more robust against early conenction
626 failures (during new), and return C<undef> in that case
627 from the constructor.
628 - AnyEvent::Handle will now try to load only JSON::XS first,
629 then fall back to JSON.
630 - format_ipv4/format_ipv6 are now exported by default, for symmetry,
631 and because it was documented that way.
632
6334.451 Fri Jul 3 00:28:58 CEST 2009
634 - do not clear rbuf when shutting down an AnyEvent::Handle
635 object - doing so breaks AnyEvent::HTTP.
636
6374.45 Mon Jun 29 22:59:26 CEST 2009
638 - a write error could cause AnyEvent::Handle to create
639 an I/O watcher with an undefined $fh.
640 - special-case mapped ipv4 addresses in both
641 AnyEvent::Socket::format_address and parse_address,
642 to treat them just like ipv4 addresses.
643 - updated and overhauled the AnyEvent::Intro doc.
644 - implement AnyEvent::Socket::format_ipv4/ipv6.
645 - slightly speed up portable_pipe/socketpair.
646 - expand condvar begin/end documentation.
647
6484.42 Fri Jun 26 08:32:18 CEST 2009
649 - preliminary and neccesarily incomplete support for IO::Async.
650 - reset SIGCHLD to DEFAULT when AnyEvent is loaded, in
651 case it was set to IGNORE, to ensure we can catch child statuses
652 even when the calling env acts stupidly.
653 - updated benchmarks with IO::Async, which performs very well.
654
6554.412 Wed Jun 24 01:35:57 CEST 2009
656 - support an "untaint" attribute for AnyEvent::DNS
657 and set it on the default resolver.
658 - implement and document AnyEvent::Impl::Perl::loop.
659 - remove all anyevent-env variables from %ENV
660 when running in tainted mode.
661 - mention and extend the IO::Lambda benchmark.
662
6634.411 Sun Jun 7 18:48:02 CEST 2009
664 - do not try to use F_SETFD on windows.
665
6664.41 Thu May 14 06:40:11 CEST 2009
667 - work around issues in older perls (5.8.0?) when
668 a signal handler is deleted from the %SIG hash.
669 - use POSIX::_exit in child test, to avoid running
670 destructors.
671 - speed up CHLD handling by relying on SIGCHLD being
672 synchronously delivered, even when we roll our own
673 implementation.
674 - AnyEvent::DNS: add the "dname" resource record name
675 for cosmetic reasons.
676
6774.4 Sun Apr 26 20:12:33 CEST 2009
678 - implemented idle watchers, where applicable.
679 - AnyEvent->time died when Event backend was in use.
680 - fix a memleak in the Tk backend.
681 - sped up Tk timer handling.
682 - clip negative "after" values to 0 in AnyEvent::Impl::Event
683 to avoid spamming out warning messages.
684 - fix Qt timers without interval.
685 - avoid Qt zero-timeout special case, in old libqt's.
686
6874.352 Mon Apr 20 16:31:11 CEST 2009
688 - fix AnyEvent::Strict error messages for child watchers.
689 - fix/update Makefile.PL "recommends" versions.
690 - add AnyEvent->now_update.
691
6924.351 Sat Apr 11 07:56:14 CEST 2009
693 - actually make the signal pipe work under win32.
694 - localise $! in signal handler to avoid changing $!, although
695 perl itself does not seem to save/restore errno either.
696 - set the cloexec flag on the signal pipe (normally set by perl too).
697
6984.35 Fri Mar 27 11:48:20 CET 2009
699 - event models relying on AnyEvent's signal watcher emulation did
700 invoke the callback asynchronously, contradicting documentation
701 and causing signals to get lost (this includes AnyEvent's own
702 event loop). AnyEvent now uses the standard pipe trick to make
703 callback execution synchronous to the event loop.
704 - AnyEvent::Handle didn't free TLS context data on DESTROY
705 (patch by Pavel Shaydo).
706 - work around the ever-incompatibly-changing API of MakeMaker.
707 - document that changing global variables without restoring them
708 is a bad idea in Perl (noted by Adam Rosenstein).
709 - AnyEvent::Strict now barfs if ->io is passed a file.
710
7114.34 Thu Feb 12 18:32:45 CET 2009
712 - separately buffer TLS read data, as otherwise the read queue could
713 deadlock as receiving data is not expected while draining the
714 read queue (which cna only happen with TLS).
715 - raise EBADMSG error on JSON decoding errors.
716 - fix some minor manpage bugs (reported by Maximilian Gaß).
717 - speed up select bitmask parsing quite a bit in the pure perl
718 backend.
719 - use CORE::select instead of just select in AnyEvent::Impl::Perl.
720
7214.331 Tue Jan 6 21:07:25 CET 2009
722 - socketpair fails on many vista machines because vista has
723 completely broken accept/getpeername and getsockname functions,
724 so we provide our own socketpair emulation that kind of works
725 (AnyEvent::Util::portable_pipe).
726 - new function: AnyEvent::Util::portable_socketpair.
727 - take advantage of the Guard module if it exists.
728
7294.33 Fri Nov 21 02:35:40 CET 2008
730 - AnyEvent::Strict did errornously flag a fileno of 0 as
731 illegal.
732 - reduce memory usage and slightly speed up the pure perl backend
733 by only storing the file descriptor, not the file handle.
734 - add missing autocork method to AnyEvent::Handle (reported
735 by Adam Rosenstein).
736 - AnyEvent::DNS->resolve errornously documented an $rcode result
737 argument, but there isn't (spotted by Henrik Krohns).
738 - the naked truth about Tk - it's basically unmaintained.
739
7404.32 Mon Nov 3 22:46:32 CET 2008
741 - fix AnyEvent::Socket::resolve_sockaddr to properly support
742 unix sockets again and choose a proper default for
743 the socket type.
744 - fix call to resolve_sockaddr in tcp_connect.
745
7464.31 Thu Oct 30 04:41:48 CET 2008
747 - implemented AnyEvent::Handle->destroy method.
748 - hint about unexpected effects in TLS mode.
749 - speed up AnyEvent::Impl::Perl by using more arrays
750 instead of hashes without the slightest loss of readability :)
751 - work around a perl argument refcounting bug.
752 - with some perl patches applied, AnyEvent now seems leak-free.
753
7544.3 Fri Oct 3 09:18:43 CEST 2008
755 - AnyEvent will now install a no-op signal handler for SIGPIPE,
756 unless one has been installed already.
757 - warn about not loaidng AnyEvent::Impl::POE early enough
758 (patch by Adam Rosenstein).
759 - fixed a great number of bugs and corner cases in AnyEvent::Handle.
760 - fix a bug in where in SSL connect mode, the
761 client would first wait for some data by the server and
762 otherwise hang. (reported and analysed in an absolutely
763 exemplary manner by Adam Rosenstein).
764 - fix a bug in where SSL EOF would not be treated
765 as stream EOF, putting the connection into a hung state.
766 (reported and analysed in an absolutely exemplary manner by Adam
767 Rosenstein).
768 - fix a potential segfault when the TLS context
769 would go missing in a read callback (e.g. due to stoptls) -
770 Net::SSLeay of course makes no type checking whatsoever.
771 - AnyEvent::Handle will not stop the read watcher in TLS mode.
772 - AnyEvent::Handle->stoptls will send a TLS close notify
773 instead of simply destroying the stream now.
774 - fix a bug where large blocks of data written in TLS mode
775 would not be sent unless triggered by receives.
776 - on_drain will now take the tls write buffer into account.
777 - SSL operations have been streamlined a bit and should be faster
778 now, more to come, though.
779 - added a FAQ section, will document nontrivial issues in there.
780
7814.234 Mon Sep 29 04:08:13 CEST 2008
782 - fix child watcher documentation: only child exits will be reported,
783 no trace events.
784 - mention SIGPIPE in AnyEvent::Handle.
785 - perl backend will now use POSIX::times as monotonic clock
786 when available and the monotonic clock option is not.
787 - run even without Time::HiRes in some cases now, but do not
788 provide sub-second accuracy (all *supported* perl versions have
789 Time::HiRes).
790
7914.233 Fri Aug 22 01:48:05 CEST 2008
792 - fix a bug in the testsuite, causing a hang.
793 - clarified lots of AnyEvent::Handle constructor arguments.
794
7954.232 Thu Aug 21 20:44:25 CEST 2008
796 - fix a bug in AnyEvent::Handle that could cause two fatal error
797 callbacks on EOF, which caused AnyEvent::HTTP to signal success twice
798 (testcase by Leon Brocard).
799 - clarified on_eof/on_error documentation a bit.
800
8014.231 Tue Jul 29 13:12:15 CEST 2008
802 - remove some debugging code left in AnyEvent::Util::fork_call
803 (and no, it's impossible to implement with the broken windows
804 perls without resource leaks or worse).
805
8064.23 Tue Jul 29 12:19:59 CEST 2008
807 - document the first parameter passed to condvar callbacks to be
808 the callback.
809 - add AnyEvent::Socket::{ntoa,aton} aliases.
810 - optimize the AE::Handle->push_read (line) for the default
811 eol marker.
812 - optimize push_read (packstring|storable) for small packets.
813 - invoke on_error callback when no on_eof callback is set.
814 - fix a bug in push_read (storable) of unknown impact.
815
8164.22 Sun Jul 20 16:34:13 CEST 2008
817 - new function AnyEvent::Socket::parse_hostport.
818 - as the bulkheads at microsoft can't even get getprotobyname reliably
819 working on their shitty fucking broken os we need to hardcode
820 some common protocol numbers in AnyEvent::Socket. How can
821 people even bother with such a pile of shit as windows.
822
8234.21 Thu Jul 17 14:40:05 CEST 2008
824 - INCOMPATIBLE CHANGE: AnyEvent::Util::fork_call now has a prototype
825 which is incompatible to previous invocation syntax.
826 - work around more windows perl bugs in fork_call. I cannot imagine
827 why anybody would bother implementing fork in such an obviously
828 unusable way.
829 - avoid calling $do_retry if the request has already finished in
830 AnyEvent::DNS, thus avoiding an exception (reported by Anatoly
831 K. Sharifulin).
832
8334.2 Sat Jul 12 22:42:11 CEST 2008
834 - support an "interval" argument for repeating timers.
835 - fix a bug in the pure perl event loop that caused timers to
836 be delayed under some circumstances.
837 - implement optional argument checking via PERL_ANYEVENT_STRICT.
838 - not importing anything from Event causes it to malfunction
839 and not use Time::HiRes, work around this bug.
840 - more bug workarounds against the endlessly broken Tk module,
841 seems to work now on 32 bit machines, but less so on 64 bit.
842 - minor optimisations applied to most backends.
843
8444.161 Fri Jul 4 14:24:48 CEST 2008
845 - fixed the t/handle/02_write.t, it read after write, with the
846 assumption that all data could be written before read
847 (reported by Jim Razmus).
848
8494.160 Thu Jul 3 04:02:21 CEST 2008
850 - re-registering signal watchers after unregistering one did
851 not work if a backend falls back on AnyEvent's default
852 implementation (also affected child watchers).
853 - new AnyEvent::Handle options: autocork and no_delay.
854
8554.152 Sun Jun 22 14:15:44 CEST 2008
856 - allow for 32-bit perls that implement shifts differently
857 on different architectures in parse_ipv4 (reported and
858 analysed by Keiichi DAIBA).
859
8604.151 Fri Jun 6 17:34:24 CEST 2008
861 - make sure specifying _only_ on_read and never pushing reads
862 works.
863
8644.15 Fri Jun 6 13:00:46 CEST 2008
865 - the pure perl backend would keep some watchers alive when more than
866 one watcher was registered for the same fd.
867 - new "packstring" and "storable" read and write types
868 for AnyEvent::Handle.
869 - allow on_eof handler to be called after on_error with EPIPE returns.
870 - do not immediately call on_read callback in handle constructor.
2 871
34.14 Thu Jun 5 20:29:31 CEST 2008 8724.14 Thu Jun 5 20:29:31 CEST 2008
4 - Fixed a bug in DNS SRV priority sorting. 873 - Fixed a bug in DNS SRV priority sorting.
5 - AnyEvent::Util::guard now reports runtime errors while 874 - AnyEvent::Util::guard now reports runtime errors while
6 executing the guard block as warnings. 875 executing the guard block as warnings.
… …
131 - support service names in tcp_connect and tcp_server. 1000 - support service names in tcp_connect and tcp_server.
132 - work around netbsd/openbsd bug causing unix sockets not to 1001 - work around netbsd/openbsd bug causing unix sockets not to
133 become writable until empty. 1002 become writable until empty.
134 - only make a dummy read for error detection when getpeername 1003 - only make a dummy read for error detection when getpeername
135 returns ENOTCONN. 1004 returns ENOTCONN.
136 - defend AnyEvent::Util::inet_aton against Coro::LWP brutal 1005 - defend AnyEvent::Util::inet_aton against Coro::LWP's brutal
137 override. 1006 override.
138 - new AnyEvent::Guard::cancel. 1007 - new AnyEvent::Guard::cancel.
139 1008
1403.5 Sat May 17 23:17:14 CEST 2008 10093.5 Sat May 17 23:17:14 CEST 2008
141 - gracefully deal with systems not defining CLOCK_MONOTONIC 1010 - gracefully deal with systems not defining CLOCK_MONOTONIC
… …
155 1024
1563.41 Sun May 11 19:53:13 CEST 2008 10253.41 Sun May 11 19:53:13 CEST 2008
157 - work around a bug in perl 5.8.8's local. Reported by Yi Ma Mao. 1026 - work around a bug in perl 5.8.8's local. Reported by Yi Ma Mao.
158 1027
1593.4 Sun May 11 00:29:25 CEST 2008 10283.4 Sun May 11 00:29:25 CEST 2008
160 - complete re-implementation of condvars with a hopefully much mroe useful 1029 - complete re-implementation of condvars with a hopefully much more useful
161 API (backward compatible functions exist but are undocumented). 1030 API (backward compatible functions exist but are undocumented).
162 - AnyEvent::Handle fully rewritten with a hopefully more powerful API 1031 - AnyEvent::Handle fully rewritten with a hopefully more powerful API
163 (still subject to change). 1032 (still subject to change).
164 - take advantage of CLOCK_MONOTONIC, if available, in AnyEvent::Impl::Perl. 1033 - take advantage of CLOCK_MONOTONIC, if available, in AnyEvent::Impl::Perl.
165 - provide AnyEvent::post_detect and @AnyEvent::post_detect, which 1034 - provide AnyEvent::post_detect and @AnyEvent::post_detect, which

Diff Legend

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