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

Comparing AnyEvent/Changes (file contents):
Revision 1.166 by root, Tue Jul 8 19:50:25 2008 UTC vs.
Revision 1.592 by root, Thu Aug 28 15:17:57 2014 UTC

1Revision history for Perl extension AnyEvent. 1Revision history for Perl extension AnyEvent.
2 2
34.2 303:55:22 <Sadrak> Doku zu new: It is possible to push requests on the read and write queues, and modify properties of the stream, even
4 while
503:55:23 <Sadrak> AnyEvent::Handle is connecting.
603:55:39 <Sadrak> doku zu on_connect: It is not allowed to use the read or write queues while the handle object is connecting.
7
8TODO: document TCP_*** constants
9TODO: add debug/trace logging to some modules
10TODO: use the eval-with-cb-call trick also to speed up JSON?
11TODO: FAQ: common pitfalls? invoke-before-return
12TODO: FAQ: scope-ids? hosts file order?
13TODO: AnyEvent::Handle with AnyEvent::IO?
14TODO: invalid. and localhost. specialcasing inside AEDNS and not AESocket (rfc6761)
15TODO: maybe implement env variable to give hosts precedence
16
17TODO: Ilya Chesnokov, resoolver requests, limit.
18TODO: improve confess skipping
19TODO: Mike Lowell AnyEvent + libuv
20 - work around a newly introduced bug in Socket 2.011 (an
21 errornous sun_length check) (analyzed by Maxime Soulé).
22 - AnyEvent::TLS didn't load (but refer to) AnyEvent::Socket
23 (analyzed by Ben Magistro).
24 - AnyEvent::Strict will now confess, not croak. This is in line with
25 it being a development/debugging tool.
26 - work around a number of libglib bugs (debug builds of libglib enforce
27 certain undocumented behaviour patterns such as not being able to
28 remove a child watch source after it has fired, which we will try
29 to emulate to avoid "criticals". what where they thinking...).
30 - changed default dns resolver "max_outstanding" value from 1 to 10,
31 the latter beinfg the intended value all along
32 (reported by Ilya Chesnokov).
33 - added new "AnyEvent::Impl::UV" interface module to the UV event lib
34 (written by Mike Lowell).
35
367.07 Tue Dec 17 17:45:02 CET 2013
37 - the documentation for custom tls verify schemes was wrong. make it agree
38 with the code (reported by Maxime Soulé).
39 - added cbor read and write types to AnyEvent::Handle (using CBOR::XS).
40 - work around an API change in openssl that could cause wrong tls connection
41 aborts, likely on windows only (analyzed by sten).
42 - calling AnyEvent->now_update with AnyEvent::Impl::Perl caused an
43 endless loop (reported by Dietrich Rebmann).
44 - add tlsv1_1 and tlsv1_2 protocols to AnyEvent::TLS
45 (patch by Maxime Soulé).
46 - document AnyEvent::Impl::IOAsync::set_loop and
47 $AnyEvent::Impl::IOAsync::LOOP. Though only documented now, this
48 functionality has _always_ been available.
49 - force a toplevel domain name in t/81_hosts.t (analyzed by
50 David Jack Wange Olrik).
51 - document that AnyEvent::Log uses AnyEvent::IO.
52 - warn about AnyEvent::Filesys::Notify performance.
53 - praise the joys of AnyEvent::Fork::*.
54 - time for an =encoding directive.
55 - do no longer use JSON to create a default json coder, use
56 JSON::XS or JSON::PP directly.
57
587.05 Wed Aug 21 10:38:08 CEST 2013
59 - uts46data.pl couldn't be found due to wrong naming of the file
60 (reported by Fulko Hew).
61 - handle lone \015's properly in AE::Handle's default line read
62 (reported by various people).
63 - untaint ip addresses found in /etc/hosts (patch by José Micó).
64 - the memleak fix in 7.03 caused resolving via /etc/hosts to always
65 fail on first use (reported and testcase by Andrew Whatson).
66 - expose AnyEvent::Log::format_time, and allow users to redefine it.
67 - expose AnyEvent::Log::default_format, and allow redefinition.
68 - expose AnyEvent::Log::fatal_exit, to allow redefinition.
69 - AnyEvent::Debug shell can now run coro shell commands, if available.
70 - t/63* tests were wrongly in MANIFEST.
71 - kernel.org's finger server went MIA, switch to freebsd.org and
72 icculus.org.
73 - clarify that IO::AIO and AnyEvent::AIO are needed for AnyEvent::IO
74 to function asynchronously (reported by Daniel Carosone).
75 - hard-disable $^W in most tests, it generates too much garbage output.
76 - use a (hopefully) more future-proof method to emulate common::sense.
77 - upgrade to UTS-46:6.2.0.
78 - switch to INSTLIB from INSTLIBDIR, as INSTLIBDIR was wrongly
79 documented. should not affect anything.
80
817.04 Thu Nov 15 02:23:58 CET 2012
82 - debugging code left in.
83
847.03 Thu Nov 15 02:16:11 CET 2012
85 - AnyEvent::Socket::inet_aton did not work when dns resolution was
86 used to find the addresses (analyzed and patch by Maxime Soulé).
87 - fix a memory leak in the /etc/hosts lookup code when hosts don't
88 resolve and are not in hosts (analyzed by sten).
89
907.02 Tue Aug 14 04:27:58 CEST 2012
91 - AnyEvent::Util::run_cmd could block indefinitely (analyzed and test
92 program by Yorhel).
93 - verified that AnyEvent::Socket follows RFC5952.
94 - try to parse "ADDR#PORT" in addition to "ADDR PORT".
95
967.01 Sun May 13 01:03:17 CEST 2012
97 - fail with EPROTO in AnyEvent::Handle wqhen TLS is requested but
98 not available, instead of throwing an exception.
99 - use File::Spec to get the tmpdir in t/*, to avoid needless
100 failures on (most, not mine :) windows boxes.
101 - new handle read types: tls_detect and tls_autostart.
102
1037.0 Fri Apr 13 06:33:30 CEST 2012
104 - child watchers are broken in POE 1.352 (also many earlier
105 versions) and there seems to be no way to work around it, as POE
106 itself is inherently racy. Document this regression and add a delay
107 in t/68_poe_03_child.t for the time being.
108 - new module AnyEvent::IO, that is a frontend to either a pure-perl
109 synchronous I/O implementation (AnyEvent::IO::Perl), or to the
110 asynchronous I/O extension IO::AIO (AnyEvent::IO::IOAIO).
111 - load /etc/hosts only when DNS has no answer.
112 - stat /etc/hosts on every access and reload it if it changed.
113 - load /etc/hosts via AnyEvent::IO - potentially asynchronous.
114 - fix a buggy croak in the dh parameter reading in AnyEvent::TLS.
115 - AnyEvent::Log log_to_file and log_to_path now use AnyEvent::IO. As
116 a side effect, they now use true appending as opposed to libc
117 appending, and the file might not have been opened when the function
118 returns.
119 - the default logging level was not properly documented in a variety of
120 places, this has hopefully been rectified.
121 - updated uts46data.pl for unicode 6.1.0.
122 - made log messages generated by AnyEvent submodules not
123 include the package name anymore, as it will be logged
124 by default already.
125 - upgrade to the trick used by common-sense 3.5 to work
126 around extra warning torture/breakage under perl 5.15.x.
127 - log messages by anyevent are now ucfirst, are usually full
128 sentences and do no longer include the package name.
129 - the storable read type would sometimes throw an exception instead
130 of causing EBADMSG (analyzed by Maxime Soulé).
131 - work around a bug in openssl 1.0.1 which enforces a minimum rsa
132 keysize (reported by Paul Howarth).
133 - documented Rocco fabricating statements about AnyEvent and me.
134
1356.14 Tue Jan 31 20:00:24 CET 2012
136 - AnyEvent::Impl::Tk was broken due to a mysterious "die" inside,
137 probably an editing mistake (reported by Darin McBride).
138
1396.13 Thu Jan 12 07:27:01 CET 2012
140 - AnyEvent::Util::fork_call checked for POSIX availability in the wrong way,
141 causing it to fail if POSIX wasn't loaded already (analysed by Rock Power).
142 - AnyEvent::Handle::push_read (line => did pass $1 directly, so
143 regex-matching inside the callback would change the parameter, despite
144 $1 being dynamically scoped per-block. this perl bug is now being
145 worked around (testcase by Cindy Wang).
146
1476.12 Mon Dec 12 13:21:10 CET 2011
148 - $! was clobbered by subsequent calls in tcp_connect, due to
149 the postpone not saving/restoring it (analyzed by Richard Garnish).
150
1516.11 Tue Nov 22 10:36:05 CET 2011
152 - Tk cannot create windows when tainted unless you set a title, so
153 set a dummy title for AnyEvent's dummy mainwindow. How dummy.
154 - escape any nonprintable/nonascii characters when stringifying
155 backtraces.
156 - log the reason loading an interface module fails to load at level debug.
157
1586.1 Tue Oct 4 19:44:30 CEST 2011
159 - INCOMPATIBLE CHANGE: the default log level is now "4" (error
160 and above), and some messages inside AnyEvent have been elevated
161 to higher log levels to print by default.
162 - AnyEvent::log, unlike AnyEvent::Log::log, did not always exit on fatal.
163 - work around yet another signal race bug, newly introduced by POE. My
164 how that module sucks.
165 - add google public ipv6 dns as fallback servers, so AE::DNS can finally
166 support ipv6-only operation even with defaults (also use ::1 instead
167 of 127.0.0.1).
168 - give better guidelines on how to chose a logging level.
169 - AnyEvent::Log can now cap (limit) log message priority by context.
170 - AE::log will now save and restore $! and $@, so callers don't have to.
171 - AE::Strict now checks that registered fds are still valid at regular
172 intervals.
173 - skip frequently called 'require' calls if the module appears loaded.
174 - increase signal latency test timer considerably for poor little osx,
175 which can't be bothered with basic posix functionality such as signal
176 delivery.
177 - mention some debugging aids in the tutorial.
178 - new env variable PERL_ANYEVENT_MAX_SIGNAL_LATENCY.
179 - private (%namespace) log contexts in PERL_ANYEVENT_LOG now have the name
180 of the context as title.
181 - AnyEvent::Strict would not restore read-write access to $_ when a callback
182 died - this is a user bug, but since AE::Strict exists to help dbeugging...
183 - detect the extremely evil IO::Async::Loop::AnyEvent module and refuse
184 to work with it - the author is well aware of the issues but refuses
185 to fix them. This extreme step has been taken because the module actively
186 attacks the goals of AnyEvent and makes it harder for module authors to use
187 AnyEvent.
188
1896.02 Fri Aug 26 20:08:31 CEST 2011
190 - log_to_syslog now accepts facility strings.
191 - log_to_syslog would not allow facilities - "$facility|$level"
192 *literally* means that string. sheesh.
193 - the CPAN parser got confused and thought we provide Sys::Syslog, try
194 to work around.
195 - renamed AnyEvent::Impl::FLTK2 to FLTK.
196
1976.01 Fri Aug 26 07:04:11 CEST 2011
198 - INCOMPATIBLE CHANGE: AnyEvent::DNS resource records now include
199 the TTL as fourth element - this affects ->resolve, ->request
200 and ->dns_unpack indirectly. Convenience functions such as
201 AnyEvent::DNS::{a,txt,mx,srv} are not affected, neither is
202 AnyEvent::Socket.
203 - INCOMPATIBLE CHANGE: backend authors now should not implement
204 one_event or loop, but instead the AnyEvent::CondVar::_wait and _poll
205 methods.
206 - INCOMPATIBLE CHANGE: verbosity levels have been redefined, see
207 AnyEvent (PERL_ANYEVENT_VERBOSE) and AnyEvent::Log for details.
208 - DEPRECIATION: please change
209 "use AnyEvent::Impl::Perl" to "use AnyEvent::Loop" and
210 "AnyEvent::Impl::Perl::loop" to "AnyEvent::Loop::run"
211 in your programs, for future compatibility.
212 - *blush* idle watcher emulation was borked.
213 - AnyEvent::Strict now checks the AE::xxx functions as well.
214 - querying too long domain names would barf deep inside AnyEvent::DNS - since
215 this is hard to avoid for applications, AnyEvent:.DNS::request now checks for
216 for too long domain names explicitly and passes an undef to the callback
217 instead (reported by James Bromberger).
218 - make AnyEvent::Util more robust against EINTR (analyzed by Andrew Suffield).
219 - new functions: AnyEvent::postpone, AnyEvent::log.
220 - new module: AnyEvent::Log, for simple logging needs.
221 - new env variable, PERL_ANYEVENT_LOG, for fine-grained logging config.
222 - new env variable, PERL_ANYEVENT_DEBUG_SHELL, to automatically start
223 a debug shell.
224 - new env variable, PERL_ANYEVENT_DEBUG_WRAP, to automatically enable
225 watcher instrumentation/wrapping (see AnyEvent::Debug::wrap).
226 - new env variable, PERL_ANYEVENT_HOSTS, to override the place where
227 /etc/hosts can be found.
228 - an empty PERL_ANYEVENT_RESOLV_CONF now selects the default
229 configuration.
230 - env variables can now be specified using AE_xyz instead of
231 PERL_ANYEVENT_xyz, subject to some rules.
232 - lots of new AnyEvent::Debug functionality.
233 - AnyEvent::Strict now checks for modifications of $_ by the callback
234 (by setting it to readonly while executing the callback, if possible).
235 - IO::Async sometimes cancels the wrong timer when you call cancel_timer.
236 Invest extra overhead to work around this peculiar behaviour.
237 - split the pure perl backend into a loop module and an impl
238 module (AnyEvent::Loop and AnyEvent::Impl::Perl).
239 - some read types (regex, netstring, json, storable) could
240 errornously keep an AE::Handle object alive on their own
241 (reported by Mohammad Toossi).
242 - AnyEvent::Socket::resolve_sockaddr and all functions using it now
243 supports /etc/hosts overrides.
244 - add more workarounds around child watcher bugs in IO::Async
245 and (newly introduced) in POE.
246 - use glib child watchers and try to work around its limitations,
247 if possible (based on a patch by Kevin Ryde).
248 - add support for FLTK (fltk2), via AnyEvent::Impl::FLTK2.
249 - many read types in AnyEvent::Handle would malfunction during
250 transitions to TLS because of a "delete $self->{rbuf}".
251 (reported by Gerald Galster).
252 - improve condvar blocking wait performance (probably only noticable
253 with faster backends), also make it easier for event loops that
254 do not support blocking waits.
255 - PERL_ANYEVENT_MODEL now also allows full module names, although
256 the usefulness of this feature is not yet clear.
257 - add tests for all supported event loops, conditional on PERL_ANYEVENT_LOOP_TESTS.
258 - add t/80_ssltext.t.
259 - no longer use AUTOLOAD for the initial loading (for no very good reason).
260 - tcp_server now tries to unlink unix domain sockets when it gets destroyed.
261 - do not run event loop detection multiple times when programs call their cached
262 copies of AnyEvent::detect.
263 - parse_hostport now accepts absolute paths als unix domain sockets.
264 - suppress the idiotic warning inside POE by patching the POE::Kernel
265 run flags instead of asking the user to jump through contortions.
266 - add Task::Weaken as dependency (as recommends only at this time).
267 - replace PF_UNSPEC by 0, for good old minix.
268
2696.0 (not officially released, but was on CPAN for about 15 minutes :/)
270
2715.34 Thu May 12 09:59:41 CEST 2011
272 - AE::TLS didn't support both cert_file and key at the same time,
273 due to some copy&paste bug (reported and analyzed by vti).
274 - improve AE::Handle timeout documentation slightly.
275
2765.33 Mon Feb 7 21:51:41 CET 2011
277 - never appeared on CPAN, for unknown reasons.
278 - AnyEvent::Handle on_connect_error - do not try to destroy a handle
279 object when it's already gone (analyzed by Ryan Bullock).
280 - update idna algorithm(s) to UTS#46 version 6.0.0.
281 - preserve trailing dot in idn_to_ascii - testsuite says to preserve,
282 uts46 says it's an error, and testsuite makes more sense.
283 - remove leading dots in idn_to_ascii - testsuite says to remove,
284 uts46 says it's an error, and testsuite makes more sense.
285 - fixed a bug in idn ignored character processing.
286 - passes the uts46 6.0.0 testsuite except for three cases which
287 apparently are bugs in the testsuite itself.
288 - passes the uts46 6.0.1 testsuite except for eleven cases which
289 apparently are bugs in the testsuite itself.
290
2915.31 Sun Jan 23 11:49:19 CET 2011
292 - work around a POE bug where POE would first poll for new events and
293 invoke handlers BEFORE handling the message queue.
294 - fix parsing of "domain" statement in resolv.conf (Johannes Plunien).
295 - stop_read now does just that in TLS mode. see its
296 documentation for caveats.
297 - update to common::sense 3.4 warning set (some warnings in perl impose
298 an arbitrary 32kb limit on regex matches, which affects
299 AnyEvent::Handle).
300 - AE::Handle had a broken rbuf_max method and missing wbuf_max method.
301 - new experimental AE::Handle::resettls method.
302
3035.3 Fri Dec 31 05:48:13 CET 2010
304 - major IOAsync patch by Paul Evans - 0.33 is required, as the major
305 issues with IO::Async have been solved in that version, including
306 the availabilty of a default loop!
307 - implement new wbuf_max parameter for AnyEvent::Handle.
308 - added Cocoa::EventLoop backend by Daisuke Murase.
309 - the default testsuite now honors PERL_ANYEVENT_MODEL (Paul Evans).
310 - clarify handle on_prepare documentation.
311 - set CLOEXEC flag on parent-side fd's in AnyEvent::Util::run_cmd,
312 to avoid leaking them to any child processes.
313 - try to force files into the arch-dependend lib directory, for the
314 half-"turly obsessive compulsive person"(s) that depend on it.
315
3165.29 Sun Dec 5 10:49:21 CET 2010
317 - convert EV backend to EV 4.00 API (so better upgrade EV too).
318 - AE::Handle: implement dynamic read_size adjustment, add
319 max_read_size parameter, reduce default read_size to 2048.
320 - add some FAQ entries for servers and/or on_eof vs. on_error.
321 - work around OS bugs (cygwin again) when lingering in AE::Handle.
322 - reduce memory usage slightly when connect was used in AE::Handle.
323 - wrok around more segfaults in Tk.
324 - document tls_ctx => undef as valid.
325 - detect WNOHANG value at compiletime.
326 - include some probably linux-only support for building AnyEvent
327 as part of the perl core.
328 - improve compatibility of t/08_idna.t to perl 5.8.
329 - make AnyEvent::Strict truly optional (in case somebody builds a
330 minimal perl...).
331
3325.28 Wed Oct 13 04:14:23 CEST 2010
333 - due to a glitch, AnyEvent's internal getprotobyname wasn't
334 actually used in tcp_server.
335 - implement AnyEvent::CondVar->new method, which was documented but
336 missing.
337 - support autoloading of read types in unshift_read.
338 - AnyEvent::Strict: do the test for invalid fh arguments differently,
339 resulting in more informative error messages.
340 - new AnyEvent::FAQ manpage.
341 - clarify that you can actually call push_read etc. in on_read.
342 - put keys and certs in /tmp, not into the current directory,
343 and document this in the SECURITY CONSIDERATIONS sections
344 of AnyEvent::TLS.
345 - support ";" as resolv.conf comment character in AnyEvent::DNS, as
346 requested by Ron Isaacson.
347 - document $AnyEvent::DNS::RESOLVER variable.
348 - incorporated major typo patches by Abhijit Menon-Sen.
349 - AnyEvent::Handle now croaks when negative timeouts are passed.
350 - add a shitload of TCP_xxx sockopt name constants to AnyEvent::Util,
351 undocumented, but who knows when they might come in handy.
352
3535.271 Tue Jun 8 12:05:46 CEST 2010
354 - backport to perl 5.8.x.
355
3565.27 Sun Jun 6 12:12:05 CEST 2010
357 - postpone differently in AnyEvent::Socket now, as
358 when not, canceling the connection attempt might fail
359 (found by Felix Antonius Wilhelm Ostmann).
360 - explicitly check for non-stream sockets in AE::Handle, too many
361 clueless people fell into the trap of this somehow working.
362 - simplified and reworked the "OTHER MODULES" section.
363 - better/more condvar examples.
364
3655.261 Wed Apr 28 16:13:36 CEST 2010
366 - AF_INET6 was not properly used from Socket6 during configuration
367 time (found by Andy Grundman).
368
3695.26 Mon Apr 12 04:49:35 CEST 2010
370 - don't generate spurious readyness notifications when select
371 returns EINTR in the pure perl backend (analysed and testcase
372 by Bas Denissen).
373 - give same examples in SYNOPSIS sections of AnyEvent and AE.
374 - provider faster implementations for AE::time/now/now_update for
375 EV and Perl backends.
376 - provide faster AE::cv and AE::time implementations for all
377 backends.
378 - no longer support lower/mixed-case signal names (this was
379 never documented nor universally supported).
380 - some more memory µ-optimisations, and somewhat less messy
381 code to cope with both AnyEvent and AE APIs.
382
3835.251 Sat Mar 13 00:58:21 CET 2010
384 - make test failed if it was run before make install on perl <=
385 5.8.8, and older versions of perl make our live very difficult,
386 so write constants.pl during Makefile.PL time.
387
3885.25 Sat Mar 13 00:23:14 CET 2010
389 - fix a race condition in AnyEvent::Handle that would cause
390 a "bio_read: unsupported method" error (in _another_
391 ssl connection) after user code threw an exception.
392 - added AnyEvent::Handle->destroyed method.
393 - speed up AnyEvent::Socket::format_address and ::format_ipv6.
394 - the AnyEvent::Util::fh_nonblocking function worked only
395 by lucky accident on win32.
396 - smaller and faster AnyEvent::Util::fh_nonblocking.
397 - when the (required!) Time::HiRes module is missing, AnyEvent
398 did not fall back to built-in time properly.
399 - do not load Fcntl at runtime, saving memory and loading time.
400 - precompile a number of constants and use them instead of runtime
401 detection and eval.
402 - free detection code after detection and similar memory optimisations.
403 - Perl backend timer interval best effort drift has been improved
404 (same algorithm as EV).
405 - update unicode idna mapping table.
406
4075.24 Tue Jan 5 11:39:43 CET 2010
408 - cygwin never reports errors from failed connects. 1.5 just gives
409 you continous readyness and EAGAIN, 1.7 is even more broken and
410 just hangs. work around both issues in a cygwin-specific hack.
411 - improve idle watcher documentation slightly.
412
4135.23 Sun Dec 20 23:48:00 CET 2009
414 - support IDNs in resolve_sockaddr, and therefore in tcp_connect.
415 - implement punycode_encode/decode, idn_nameprep,
416 idn_to_ascii and idn_to_unicode operations in AnyEvent::Util.
417 - provide $AE::VERSION.
418 - removed traces of "no strict 'refs'".
419
4205.22 Sat Dec 5 03:51:13 CET 2009
421 - downgrade-or-fail in AnyEvent::Handle::push_write, to
422 diagnose encoding failures earlier and more succinctly.
423 (this works around bugs in perl, throwing away encoding info
424 when passing scalar data to extensions).
425 - add more examples to AnyEvent::Socket manpage.
426 - upgrade internal warning set to the same as common::sense 2.03.
427 - use pack "n/a*" for pre-5.8.9 perl compatibility in AnyEvent::DNS
428 (John Beppu).
429 - AnyEvent::Socket::inet_aton now properly supports ipv6, as documented.
430 - add google public dns servers to fallback server set and make sure
431 we load-balance properly between the three sets. also add all
432 fallback dns servers, not just a random one, to each dns config.
433
4345.21 Thu Nov 19 02:48:47 CET 2009
435 - fix a problem where socket constants were called with parameters
436 (spotted by David Friedland).
437 - fork_call never use'd POSIX (reported by Daisuke Maki).
438 - improve perl 5.6 compatibility further (but it still won't work
439 unless you rip out everything but the core).
440 - prefer Net::DNS::Resolver over ipconfig, if installed, on win32.
441 uses 10MB of RAM, but doesn't open a console window. *sigh*.
442
4435.202 Wed Oct 14 22:35:44 CEST 2009
444 - AnyEvent::DNS would unexpectedly clobber $_ under windows
445 (analysed by Matthias Waldorf).
446 - AnyEvent::Handle::run_cmd can now pass the PID of the
447 newly-created process, which is much less useful than it might
448 sound (based on patch by Yann Kerherve).
449
4505.201 Tue Sep 29 12:09:25 CEST 2009
451 - AnyEvent:Handle::on_starttls/on_stoptls methods were broken
452 (reported by Torsten Foertsch).
453 - common::sense 2.0 could cause tcp_server to throw an exception
454 (analysed by elmex).
455
4565.2 Mon Sep 14 07:04:49 CEST 2009
457 - INCOMPATIBLE CHANGE: do no longer support register_read_type
458 and register_write_type in AnyEvent::Handle, instead support
459 package names (the facility was mostly abused).
460 - implement "packagename-as-read/write type" support in
461 AnyEvent::Handle.
462 - AnyEvent::Handle: new options "keepalive" and "oobinline".
463 - oobinline set by default to avoid security issues.
464 - the pure-perl event loop backend wrongly detected times() fallback
465 support (spotted by Pavel Boldin).
466
4675.12 Tue Sep 1 20:26:50 CEST 2009
468 - be more lenient when parsing resolv.conf files, as some people
469 use hashmarks as comment indicator inside directives (reported
470 by Michael S. Fischer).
471 - use same set of warnings as common::sense 2.0.
472 - fix a potential 32 bit overflow issue due to perl having problems
473 with large hex constants in the Perl backend.
474 - do not use the slower 5.6 version fo the enc_name code in AnyEvent::DNS
475 by default.
476 - fix some prototypes.
477 - (optionally) require Async::Interrupt 1.02.
478 - replace opendns fallbacks by verizon/level3 public dns servers.
479
4805.112 Fri Aug 21 13:59:04 CEST 2009
481 - AnyEvent::Handle could cause hard crashes in Net::SSLeay when
482 using starttls with an illegal/uncreatable tls context.
483 - port to broken win32 perls.
484
4855.111 Sun Aug 16 18:44:36 CEST 2009
486 - perl errornously requires sockaddr_un structures returned by the
487 kernel to have a specific length. work around this bug by padding
488 sockaddr_un structures (found to fail on OS X, as perl only uses
489 the correct code for GNU/Linux, but likely to fail on other
490 OSes as well).
491
4925.11 Wed Aug 12 17:49:37 CEST 2009
493 - fixed a bug in run_cmd with input from filename.
494 - create a json coder object when none is given in write direction
495 to reduce dependency on JSON.
496
4975.1 Tue Aug 11 03:17:32 CEST 2009
498 - tcp_connect now reports the correct error in $!, not an unrelated
499 one (was broken since 4.91).
500 - AE::Handle did not properly clear rtimeout/wtimeout watchers
501 when retrying a tcp_connect.
502 - new functions: AnyEvent::Util::run_cmd and close_all_fds_except.
503 - fixed wrong documentation in AE::Socket.
504
5055.01 Mon Aug 10 03:16:32 CEST 2009
506 - last release broke EV child watchers :/.
507
5085.0 Sun Aug 9 17:32:09 CEST 2009
509 - officially introduce and document the AE API.
510 - lowering the timeout at runtime did not have immediate effect
511 in AnyEvent::Handle.
512 - AE::Handle now has separate and independent read and write timeout
513 settings.
514 - probe for Guard module when AnyEvent::Util is loaded, not at
515 runtime (Event doesn't like eval "use" from callbacks, and it avoids
516 exporting the probe function).
517 - Impl::IOAsync's io watchers did not use the filehandle emulation
518 code to work around it's watcher limits.
519 - work around IO::Async's problems with overloaded objects.
520 - try to disable wearnings inside IO::Async, it's just too buggy.
521 - do no longer use Event by default, only EV or Perl, Event is too buggy.
522 - hackishly provide fast AE:: interface when EV or Perl is the backend.
523 - inherit VERSION from AnyEvent::VERSION in most modules.
524
5254.91 Thu Aug 6 15:42:45 CEST 2009
526 - AE::Handle::starttls could get out of sync when the read buffer
527 already contains some TLS handshake.
528 - AE::Handle did not properly free the TLS session, causing bigger
529 memory leaks in Net::SSLeay than expected (reported by toaster).
530 - AE::Socket::tcp_connect will now artificially delay invoking
531 the callback to avoid returning after invoking the callback.
532 - convert many internal watcher uses to AE API.
533
5344.9 Sat Aug 1 11:07:01 CEST 2009
535 - Glib silently fails when registering a timeout with
536 a negative value, so avoid that.
537 - call condvar callback immediately when it is set
538 after the condition is already signalled.
539 - check rbuf_max condition only after trying to consume
540 data.
541 - normalise signal numbers to names when using the ae's
542 signal handling, but do not document this (yet).
543 - pure perl signal emulation did not properly set nonblocking
544 mode on the signal pipe on win32 (this is unlikely to have caused
545 any issues).
546 - new module: AnyEvent::Debug.
547 - AnyEvent::Strict now checks that a signal was specified
548 by name (not name or number).
549 - reduce memory footprint in typical cases by ~50kb by
550 only compiling signal code when necessary.
551 - add AnyEvent::Handle->rbuf_max.
552 - grab the AE namespace for future new API, implement stubs for
553 said future/alternative API.
554 - new function: AnyEvent::Socket::format_hostport.
555
5564.881 Tue Jul 28 12:51:53 CEST 2009
557 - work around a bug in local in pre-5.10 perls, causing
558 AnyEvent::Handle to recurse when it should not
559 (analysed by elmex).
560
5614.88 Tue Jul 28 04:04:37 CEST 2009
562 - re-bless the handle into a dummy package after calling
563 AnyEvent::Handle::destroy, so the user does not need to check
564 for errors after every push_write etc.
565 - do not attempt to run t/02_signals.t on obviously broken
566 platforms.
567
5684.87 Sun Jul 26 02:06:16 CEST 2009
569 - do not attempt to linger when there is no longer a valid fh.
570 - enforce tls mode to be either accept or connect, do not simply
571 segfault in Net::SSLeay.
572 - AnyEvent::Handle can now call tcp_connect itself (new parameters
573 connect, on_prepare, on_connect and on_connect_error). Updated
574 tutorial accordingly.
575 - add AnyEvent::Impl::Irssi backend.
576
5774.86 Mon Jul 20 23:52:29 CEST 2009
578 - since the verbose warning is not enough, explicitly document
579 that versions before 1.33 of Net::SSLeay are not secure.
580 - work around signal handling races in Event and (...) Event::Lib.
581 - try to align signal-race timer to full-second boundaries.
582 - work around Tk not liking negative timeouts.
583 - don't complain of different grades of environmental unfriendlyness
584 in IO::Async.
585
5864.85 Sat Jul 18 06:16:14 CEST 2009
587 - nail the signal race problem in perl once and for all
588 (see $AnyEvent::MAX_SIGNAL_LATENCY).
589 - take advantage of Async::Interrupt if it is available.
590 - load Time::HiRes and Guard modules on demand only.
591 - add optional/recommended modules section to AnyEvent
592 documentation.
593 - reduce memory usage considerably (and reduce startup penalty)
594 by not using "strict", "warnings" and "overload" modules.
595 - work around buggy windows/openbsd perls and provide EBADMSG
596 and EPROTO ourselves when missing.
597 - improve perl 5.6 compatibility of the core event loop.
598 - made Net::SSLeay version 1.33 a soft requirement.
599
6004.83 Fri Jul 17 16:56:26 CEST 2009
601 - implement AnyEvent::Socket::getprotobyname.
602 - AnyEvent::CondVar's will now detect recursive blocking
603 waits and will croak, as too many people fall into
604 this trap.
605 - AnyEvent::Handle will now call ->destroy on itself after
606 executing the on_error callback, instead of doing some
607 half-baked internal shutdown, for fatal errors.
608 - clarify on_eof behaviour w.r.t. the read queue and
609 on_read callbacks.
610 - ignore some possible spurious wake-ups in tcp_connect.
611
6124.82 Sat Jul 11 00:34:55 CEST 2009
613 - POE and Event backends didn't accept some callable objects as
614 callbacks.
615 - use Config module instead of POSIX module to detect signal names
616 in AnyEvent::Strict and AnyEvent::Impl::EventLib, as the POSIX
617 module doesn't even have all POSIX signals :/.
618 - use more workarounds around the many refcnt/corruption bugs in
619 Event::Lib.
620 - work around a race condition in perl's select, causing t/03_child.t
621 to rarely fail.
622
6234.81 Thu Jul 9 10:30:30 CEST 2009
624 - AnyEvent::Handle didn't properly diagnose write errors
625 (it expected -1 from syswrite, how lame... :).
626 - support file descriptors in addition to file handles
627 in AnyEvent->io.
628 - new env variables: PERL_ANYEVENT_RESOLV_CONF,
629 PERL_ANYEVENT_MAX_OUTSTANDING_DNS, PERL_ANYEVENT_CA_FILE
630 and PERL_ANYEVENT_CA_PATH.
631 - provide a sensible synopsis section for AnyEvent::TLS.
632 - add a "supported backends" section to the manpage.
633 - added simple io watcher test to testsuite, using a
634 portable_socketpair.
635 - tried to improve the stability of the Event::Lib backend,
636 YMMV.
637
6384.8 Mon Jul 6 23:45:16 CEST 2009
639 - AnyEvent::DNS did not properly follow CNAME records with
640 uppercase targets.
641 - AnyEvent::DNS would errornously return AAAA records
642 with v4 mapped addresses (a faulty record) as ipv4 addresses,
643 causing AnyEvent::Socket to throw an exception.
644 - added new module AnyEvent::TLS for easier SSL/TLS context
645 creation, with many options including hostname verification,
646 secure default configuration, lots of documentation and,
647 predefined diffie-hellman keys for perfect forward security
648 and much more. get it while it's still fresh!
649 - use AnyEvent::TLS in AnyEvent::Handle for context management.
650 - load AnyEvent::Handle only on demand in AnyEvent::DNS,
651 so AnyEvent::Socket users have smaller memory footprint
652 in the common case.
653 - add AnyEvent::Handle->push_shutdown method.
654 - add an additional error message parameter to AnyEvent::Handle's
655 on_error callback (for TLS, $! is still available).
656 - add AnyEvent::Handle on_starttls/on_stoptls callbacks.
657 - make AnyEvent::Handle more robust against early conenction
658 failures (during new), and return C<undef> in that case
659 from the constructor.
660 - AnyEvent::Handle will now try to load only JSON::XS first,
661 then fall back to JSON.
662 - format_ipv4/format_ipv6 are now exported by default, for symmetry,
663 and because it was documented that way.
664
6654.451 Fri Jul 3 00:28:58 CEST 2009
666 - do not clear rbuf when shutting down an AnyEvent::Handle
667 object - doing so breaks AnyEvent::HTTP.
668
6694.45 Mon Jun 29 22:59:26 CEST 2009
670 - a write error could cause AnyEvent::Handle to create
671 an I/O watcher with an undefined $fh.
672 - special-case mapped ipv4 addresses in both
673 AnyEvent::Socket::format_address and parse_address,
674 to treat them just like ipv4 addresses.
675 - updated and overhauled the AnyEvent::Intro doc.
676 - implement AnyEvent::Socket::format_ipv4/ipv6.
677 - slightly speed up portable_pipe/socketpair.
678 - expand condvar begin/end documentation.
679
6804.42 Fri Jun 26 08:32:18 CEST 2009
681 - preliminary and neccesarily incomplete support for IO::Async.
682 - reset SIGCHLD to DEFAULT when AnyEvent is loaded, in
683 case it was set to IGNORE, to ensure we can catch child statuses
684 even when the calling env acts stupidly.
685 - updated benchmarks with IO::Async, which performs very well.
686
6874.412 Wed Jun 24 01:35:57 CEST 2009
688 - support an "untaint" attribute for AnyEvent::DNS
689 and set it on the default resolver.
690 - implement and document AnyEvent::Impl::Perl::loop.
691 - remove all anyevent-env variables from %ENV
692 when running in tainted mode.
693 - mention and extend the IO::Lambda benchmark.
694
6954.411 Sun Jun 7 18:48:02 CEST 2009
696 - do not try to use F_SETFD on windows.
697
6984.41 Thu May 14 06:40:11 CEST 2009
699 - work around issues in older perls (5.8.0?) when
700 a signal handler is deleted from the %SIG hash.
701 - use POSIX::_exit in child test, to avoid running
702 destructors.
703 - speed up CHLD handling by relying on SIGCHLD being
704 synchronously delivered, even when we roll our own
705 implementation.
706 - AnyEvent::DNS: add the "dname" resource record name
707 for cosmetic reasons.
708
7094.4 Sun Apr 26 20:12:33 CEST 2009
710 - implemented idle watchers, where applicable.
711 - AnyEvent->time died when Event backend was in use.
712 - fix a memleak in the Tk backend.
713 - sped up Tk timer handling.
714 - clip negative "after" values to 0 in AnyEvent::Impl::Event
715 to avoid spamming out warning messages.
716 - fix Qt timers without interval.
717 - avoid Qt zero-timeout special case, in old libqt's.
718
7194.352 Mon Apr 20 16:31:11 CEST 2009
720 - fix AnyEvent::Strict error messages for child watchers.
721 - fix/update Makefile.PL "recommends" versions.
722 - add AnyEvent->now_update.
723
7244.351 Sat Apr 11 07:56:14 CEST 2009
725 - actually make the signal pipe work under win32.
726 - localise $! in signal handler to avoid changing $!, although
727 perl itself does not seem to save/restore errno either.
728 - set the cloexec flag on the signal pipe (normally set by perl too).
729
7304.35 Fri Mar 27 11:48:20 CET 2009
731 - event models relying on AnyEvent's signal watcher emulation did
732 invoke the callback asynchronously, contradicting documentation
733 and causing signals to get lost (this includes AnyEvent's own
734 event loop). AnyEvent now uses the standard pipe trick to make
735 callback execution synchronous to the event loop.
736 - AnyEvent::Handle didn't free TLS context data on DESTROY
737 (patch by Pavel Shaydo).
738 - work around the ever-incompatibly-changing API of MakeMaker.
739 - document that changing global variables without restoring them
740 is a bad idea in Perl (noted by Adam Rosenstein).
741 - AnyEvent::Strict now barfs if ->io is passed a file.
742
7434.34 Thu Feb 12 18:32:45 CET 2009
744 - separately buffer TLS read data, as otherwise the read queue could
745 deadlock as receiving data is not expected while draining the
746 read queue (which cna only happen with TLS).
747 - raise EBADMSG error on JSON decoding errors.
748 - fix some minor manpage bugs (reported by Maximilian Gaß).
749 - speed up select bitmask parsing quite a bit in the pure perl
750 backend.
751 - use CORE::select instead of just select in AnyEvent::Impl::Perl.
752
7534.331 Tue Jan 6 21:07:25 CET 2009
754 - socketpair fails on many vista machines because vista has
755 completely broken accept/getpeername and getsockname functions,
756 so we provide our own socketpair emulation that kind of works
757 (AnyEvent::Util::portable_pipe).
758 - new function: AnyEvent::Util::portable_socketpair.
759 - take advantage of the Guard module if it exists.
760
7614.33 Fri Nov 21 02:35:40 CET 2008
762 - AnyEvent::Strict did errornously flag a fileno of 0 as
763 illegal.
764 - reduce memory usage and slightly speed up the pure perl backend
765 by only storing the file descriptor, not the file handle.
766 - add missing autocork method to AnyEvent::Handle (reported
767 by Adam Rosenstein).
768 - AnyEvent::DNS->resolve errornously documented an $rcode result
769 argument, but there isn't (spotted by Henrik Krohns).
770 - the naked truth about Tk - it's basically unmaintained.
771
7724.32 Mon Nov 3 22:46:32 CET 2008
773 - fix AnyEvent::Socket::resolve_sockaddr to properly support
774 unix sockets again and choose a proper default for
775 the socket type.
776 - fix call to resolve_sockaddr in tcp_connect.
777
7784.31 Thu Oct 30 04:41:48 CET 2008
779 - implemented AnyEvent::Handle->destroy method.
780 - hint about unexpected effects in TLS mode.
781 - speed up AnyEvent::Impl::Perl by using more arrays
782 instead of hashes without the slightest loss of readability :)
783 - work around a perl argument refcounting bug.
784 - with some perl patches applied, AnyEvent now seems leak-free.
785
7864.3 Fri Oct 3 09:18:43 CEST 2008
787 - AnyEvent will now install a no-op signal handler for SIGPIPE,
788 unless one has been installed already.
789 - warn about not loaidng AnyEvent::Impl::POE early enough
790 (patch by Adam Rosenstein).
791 - fixed a great number of bugs and corner cases in AnyEvent::Handle.
792 - fix a bug in where in SSL connect mode, the
793 client would first wait for some data by the server and
794 otherwise hang. (reported and analysed in an absolutely
795 exemplary manner by Adam Rosenstein).
796 - fix a bug in where SSL EOF would not be treated
797 as stream EOF, putting the connection into a hung state.
798 (reported and analysed in an absolutely exemplary manner by Adam
799 Rosenstein).
800 - fix a potential segfault when the TLS context
801 would go missing in a read callback (e.g. due to stoptls) -
802 Net::SSLeay of course makes no type checking whatsoever.
803 - AnyEvent::Handle will not stop the read watcher in TLS mode.
804 - AnyEvent::Handle->stoptls will send a TLS close notify
805 instead of simply destroying the stream now.
806 - fix a bug where large blocks of data written in TLS mode
807 would not be sent unless triggered by receives.
808 - on_drain will now take the tls write buffer into account.
809 - SSL operations have been streamlined a bit and should be faster
810 now, more to come, though.
811 - added a FAQ section, will document nontrivial issues in there.
812
8134.234 Mon Sep 29 04:08:13 CEST 2008
814 - fix child watcher documentation: only child exits will be reported,
815 no trace events.
816 - mention SIGPIPE in AnyEvent::Handle.
817 - perl backend will now use POSIX::times as monotonic clock
818 when available and the monotonic clock option is not.
819 - run even without Time::HiRes in some cases now, but do not
820 provide sub-second accuracy (all *supported* perl versions have
821 Time::HiRes).
822
8234.233 Fri Aug 22 01:48:05 CEST 2008
824 - fix a bug in the testsuite, causing a hang.
825 - clarified lots of AnyEvent::Handle constructor arguments.
826
8274.232 Thu Aug 21 20:44:25 CEST 2008
828 - fix a bug in AnyEvent::Handle that could cause two fatal error
829 callbacks on EOF, which caused AnyEvent::HTTP to signal success twice
830 (testcase by Leon Brocard).
831 - clarified on_eof/on_error documentation a bit.
832
8334.231 Tue Jul 29 13:12:15 CEST 2008
834 - remove some debugging code left in AnyEvent::Util::fork_call
835 (and no, it's impossible to implement with the broken windows
836 perls without resource leaks or worse).
837
8384.23 Tue Jul 29 12:19:59 CEST 2008
839 - document the first parameter passed to condvar callbacks to be
840 the callback.
841 - add AnyEvent::Socket::{ntoa,aton} aliases.
842 - optimize the AE::Handle->push_read (line) for the default
843 eol marker.
844 - optimize push_read (packstring|storable) for small packets.
845 - invoke on_error callback when no on_eof callback is set.
846 - fix a bug in push_read (storable) of unknown impact.
847
8484.22 Sun Jul 20 16:34:13 CEST 2008
849 - new function AnyEvent::Socket::parse_hostport.
850 - as the bulkheads at microsoft can't even get getprotobyname reliably
851 working on their shitty fucking broken os we need to hardcode
852 some common protocol numbers in AnyEvent::Socket. How can
853 people even bother with such a pile of shit as windows.
854
8554.21 Thu Jul 17 14:40:05 CEST 2008
856 - INCOMPATIBLE CHANGE: AnyEvent::Util::fork_call now has a prototype
857 which is incompatible to previous invocation syntax.
858 - work around more windows perl bugs in fork_call. I cannot imagine
859 why anybody would bother implementing fork in such an obviously
860 unusable way.
861 - avoid calling $do_retry if the request has already finished in
862 AnyEvent::DNS, thus avoiding an exception (reported by Anatoly
863 K. Sharifulin).
864
8654.2 Sat Jul 12 22:42:11 CEST 2008
4 - support an "interval" argument for repeating timers. 866 - support an "interval" argument for repeating timers.
867 - fix a bug in the pure perl event loop that caused timers to
868 be delayed under some circumstances.
869 - implement optional argument checking via PERL_ANYEVENT_STRICT.
5 - not importing anything from Event causes it to malfunction 870 - not importing anything from Event causes it to malfunction
6 and not use Time::HiRes, work around this bug. 871 and not use Time::HiRes, work around this bug.
7 - fix a bug in the pure perl event loop that caused timers to 872 - more bug workarounds against the endlessly broken Tk module,
8 be delayed under some circumstances. 873 seems to work now on 32 bit machines, but less so on 64 bit.
9 - minor optimisations applied to most backends. 874 - minor optimisations applied to most backends.
10 875
114.161 Fri Jul 4 14:24:48 CEST 2008 8764.161 Fri Jul 4 14:24:48 CEST 2008
12 - fixed the t/handle/02_write.t, it read after write, with the 877 - fixed the t/handle/02_write.t, it read after write, with the
13 assumption that all data could be written before read 878 assumption that all data could be written before read
… …
167 - support service names in tcp_connect and tcp_server. 1032 - support service names in tcp_connect and tcp_server.
168 - work around netbsd/openbsd bug causing unix sockets not to 1033 - work around netbsd/openbsd bug causing unix sockets not to
169 become writable until empty. 1034 become writable until empty.
170 - only make a dummy read for error detection when getpeername 1035 - only make a dummy read for error detection when getpeername
171 returns ENOTCONN. 1036 returns ENOTCONN.
172 - defend AnyEvent::Util::inet_aton against Coro::LWP brutal 1037 - defend AnyEvent::Util::inet_aton against Coro::LWP's brutal
173 override. 1038 override.
174 - new AnyEvent::Guard::cancel. 1039 - new AnyEvent::Guard::cancel.
175 1040
1763.5 Sat May 17 23:17:14 CEST 2008 10413.5 Sat May 17 23:17:14 CEST 2008
177 - gracefully deal with systems not defining CLOCK_MONOTONIC 1042 - gracefully deal with systems not defining CLOCK_MONOTONIC
… …
191 1056
1923.41 Sun May 11 19:53:13 CEST 2008 10573.41 Sun May 11 19:53:13 CEST 2008
193 - work around a bug in perl 5.8.8's local. Reported by Yi Ma Mao. 1058 - work around a bug in perl 5.8.8's local. Reported by Yi Ma Mao.
194 1059
1953.4 Sun May 11 00:29:25 CEST 2008 10603.4 Sun May 11 00:29:25 CEST 2008
196 - complete re-implementation of condvars with a hopefully much mroe useful 1061 - complete re-implementation of condvars with a hopefully much more useful
197 API (backward compatible functions exist but are undocumented). 1062 API (backward compatible functions exist but are undocumented).
198 - AnyEvent::Handle fully rewritten with a hopefully more powerful API 1063 - AnyEvent::Handle fully rewritten with a hopefully more powerful API
199 (still subject to change). 1064 (still subject to change).
200 - take advantage of CLOCK_MONOTONIC, if available, in AnyEvent::Impl::Perl. 1065 - take advantage of CLOCK_MONOTONIC, if available, in AnyEvent::Impl::Perl.
201 - provide AnyEvent::post_detect and @AnyEvent::post_detect, which 1066 - provide AnyEvent::post_detect and @AnyEvent::post_detect, which

Diff Legend

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