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

Comparing AnyEvent/Changes (file contents):
Revision 1.376 by root, Mon Jan 11 17:41:59 2010 UTC vs.
Revision 1.590 by root, Thu Aug 28 10:43:02 2014 UTC

1Revision history for Perl extension AnyEvent. 1Revision history for Perl extension AnyEvent.
2 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 - changed default dns resolver "max_outstanding" value from 1 to 10,
27 the latter beinfg the intended value all along
28 (reported by Ilya Chesnokov).
29
307.07 Tue Dec 17 17:45:02 CET 2013
31 - the documentation for custom tls verify schemes was wrong. make it agree
32 with the code (reported by Maxime Soulé).
33 - added cbor read and write types to AnyEvent::Handle (using CBOR::XS).
34 - work around an API change in openssl that could cause wrong tls connection
35 aborts, likely on windows only (analyzed by sten).
36 - calling AnyEvent->now_update with AnyEvent::Impl::Perl caused an
37 endless loop (reported by Dietrich Rebmann).
38 - add tlsv1_1 and tlsv1_2 protocols to AnyEvent::TLS
39 (patch by Maxime Soulé).
40 - document AnyEvent::Impl::IOAsync::set_loop and
41 $AnyEvent::Impl::IOAsync::LOOP. Though only documented now, this
42 functionality has _always_ been available.
43 - force a toplevel domain name in t/81_hosts.t (analyzed by
44 David Jack Wange Olrik).
45 - document that AnyEvent::Log uses AnyEvent::IO.
46 - warn about AnyEvent::Filesys::Notify performance.
47 - praise the joys of AnyEvent::Fork::*.
48 - time for an =encoding directive.
49 - do no longer use JSON to create a default json coder, use
50 JSON::XS or JSON::PP directly.
51
527.05 Wed Aug 21 10:38:08 CEST 2013
53 - uts46data.pl couldn't be found due to wrong naming of the file
54 (reported by Fulko Hew).
55 - handle lone \015's properly in AE::Handle's default line read
56 (reported by various people).
57 - untaint ip addresses found in /etc/hosts (patch by José Micó).
58 - the memleak fix in 7.03 caused resolving via /etc/hosts to always
59 fail on first use (reported and testcase by Andrew Whatson).
60 - expose AnyEvent::Log::format_time, and allow users to redefine it.
61 - expose AnyEvent::Log::default_format, and allow redefinition.
62 - expose AnyEvent::Log::fatal_exit, to allow redefinition.
63 - AnyEvent::Debug shell can now run coro shell commands, if available.
64 - t/63* tests were wrongly in MANIFEST.
65 - kernel.org's finger server went MIA, switch to freebsd.org and
66 icculus.org.
67 - clarify that IO::AIO and AnyEvent::AIO are needed for AnyEvent::IO
68 to function asynchronously (reported by Daniel Carosone).
69 - hard-disable $^W in most tests, it generates too much garbage output.
70 - use a (hopefully) more future-proof method to emulate common::sense.
71 - upgrade to UTS-46:6.2.0.
72 - switch to INSTLIB from INSTLIBDIR, as INSTLIBDIR was wrongly
73 documented. should not affect anything.
74
757.04 Thu Nov 15 02:23:58 CET 2012
76 - debugging code left in.
77
787.03 Thu Nov 15 02:16:11 CET 2012
79 - AnyEvent::Socket::inet_aton did not work when dns resolution was
80 used to find the addresses (analyzed and patch by Maxime Soulé).
81 - fix a memory leak in the /etc/hosts lookup code when hosts don't
82 resolve and are not in hosts (analyzed by sten).
83
847.02 Tue Aug 14 04:27:58 CEST 2012
85 - AnyEvent::Util::run_cmd could block indefinitely (analyzed and test
86 program by Yorhel).
87 - verified that AnyEvent::Socket follows RFC5952.
88 - try to parse "ADDR#PORT" in addition to "ADDR PORT".
89
907.01 Sun May 13 01:03:17 CEST 2012
91 - fail with EPROTO in AnyEvent::Handle wqhen TLS is requested but
92 not available, instead of throwing an exception.
93 - use File::Spec to get the tmpdir in t/*, to avoid needless
94 failures on (most, not mine :) windows boxes.
95 - new handle read types: tls_detect and tls_autostart.
96
977.0 Fri Apr 13 06:33:30 CEST 2012
98 - child watchers are broken in POE 1.352 (also many earlier
99 versions) and there seems to be no way to work around it, as POE
100 itself is inherently racy. Document this regression and add a delay
101 in t/68_poe_03_child.t for the time being.
102 - new module AnyEvent::IO, that is a frontend to either a pure-perl
103 synchronous I/O implementation (AnyEvent::IO::Perl), or to the
104 asynchronous I/O extension IO::AIO (AnyEvent::IO::IOAIO).
105 - load /etc/hosts only when DNS has no answer.
106 - stat /etc/hosts on every access and reload it if it changed.
107 - load /etc/hosts via AnyEvent::IO - potentially asynchronous.
108 - fix a buggy croak in the dh parameter reading in AnyEvent::TLS.
109 - AnyEvent::Log log_to_file and log_to_path now use AnyEvent::IO. As
110 a side effect, they now use true appending as opposed to libc
111 appending, and the file might not have been opened when the function
112 returns.
113 - the default logging level was not properly documented in a variety of
114 places, this has hopefully been rectified.
115 - updated uts46data.pl for unicode 6.1.0.
116 - made log messages generated by AnyEvent submodules not
117 include the package name anymore, as it will be logged
118 by default already.
119 - upgrade to the trick used by common-sense 3.5 to work
120 around extra warning torture/breakage under perl 5.15.x.
121 - log messages by anyevent are now ucfirst, are usually full
122 sentences and do no longer include the package name.
123 - the storable read type would sometimes throw an exception instead
124 of causing EBADMSG (analyzed by Maxime Soulé).
125 - work around a bug in openssl 1.0.1 which enforces a minimum rsa
126 keysize (reported by Paul Howarth).
127 - documented Rocco fabricating statements about AnyEvent and me.
128
1296.14 Tue Jan 31 20:00:24 CET 2012
130 - AnyEvent::Impl::Tk was broken due to a mysterious "die" inside,
131 probably an editing mistake (reported by Darin McBride).
132
1336.13 Thu Jan 12 07:27:01 CET 2012
134 - AnyEvent::Util::fork_call checked for POSIX availability in the wrong way,
135 causing it to fail if POSIX wasn't loaded already (analysed by Rock Power).
136 - AnyEvent::Handle::push_read (line => did pass $1 directly, so
137 regex-matching inside the callback would change the parameter, despite
138 $1 being dynamically scoped per-block. this perl bug is now being
139 worked around (testcase by Cindy Wang).
140
1416.12 Mon Dec 12 13:21:10 CET 2011
142 - $! was clobbered by subsequent calls in tcp_connect, due to
143 the postpone not saving/restoring it (analyzed by Richard Garnish).
144
1456.11 Tue Nov 22 10:36:05 CET 2011
146 - Tk cannot create windows when tainted unless you set a title, so
147 set a dummy title for AnyEvent's dummy mainwindow. How dummy.
148 - escape any nonprintable/nonascii characters when stringifying
149 backtraces.
150 - log the reason loading an interface module fails to load at level debug.
151
1526.1 Tue Oct 4 19:44:30 CEST 2011
153 - INCOMPATIBLE CHANGE: the default log level is now "4" (error
154 and above), and some messages inside AnyEvent have been elevated
155 to higher log levels to print by default.
156 - AnyEvent::log, unlike AnyEvent::Log::log, did not always exit on fatal.
157 - work around yet another signal race bug, newly introduced by POE. My
158 how that module sucks.
159 - add google public ipv6 dns as fallback servers, so AE::DNS can finally
160 support ipv6-only operation even with defaults (also use ::1 instead
161 of 127.0.0.1).
162 - give better guidelines on how to chose a logging level.
163 - AnyEvent::Log can now cap (limit) log message priority by context.
164 - AE::log will now save and restore $! and $@, so callers don't have to.
165 - AE::Strict now checks that registered fds are still valid at regular
166 intervals.
167 - skip frequently called 'require' calls if the module appears loaded.
168 - increase signal latency test timer considerably for poor little osx,
169 which can't be bothered with basic posix functionality such as signal
170 delivery.
171 - mention some debugging aids in the tutorial.
172 - new env variable PERL_ANYEVENT_MAX_SIGNAL_LATENCY.
173 - private (%namespace) log contexts in PERL_ANYEVENT_LOG now have the name
174 of the context as title.
175 - AnyEvent::Strict would not restore read-write access to $_ when a callback
176 died - this is a user bug, but since AE::Strict exists to help dbeugging...
177 - detect the extremely evil IO::Async::Loop::AnyEvent module and refuse
178 to work with it - the author is well aware of the issues but refuses
179 to fix them. This extreme step has been taken because the module actively
180 attacks the goals of AnyEvent and makes it harder for module authors to use
181 AnyEvent.
182
1836.02 Fri Aug 26 20:08:31 CEST 2011
184 - log_to_syslog now accepts facility strings.
185 - log_to_syslog would not allow facilities - "$facility|$level"
186 *literally* means that string. sheesh.
187 - the CPAN parser got confused and thought we provide Sys::Syslog, try
188 to work around.
189 - renamed AnyEvent::Impl::FLTK2 to FLTK.
190
1916.01 Fri Aug 26 07:04:11 CEST 2011
192 - INCOMPATIBLE CHANGE: AnyEvent::DNS resource records now include
193 the TTL as fourth element - this affects ->resolve, ->request
194 and ->dns_unpack indirectly. Convenience functions such as
195 AnyEvent::DNS::{a,txt,mx,srv} are not affected, neither is
196 AnyEvent::Socket.
197 - INCOMPATIBLE CHANGE: backend authors now should not implement
198 one_event or loop, but instead the AnyEvent::CondVar::_wait and _poll
199 methods.
200 - INCOMPATIBLE CHANGE: verbosity levels have been redefined, see
201 AnyEvent (PERL_ANYEVENT_VERBOSE) and AnyEvent::Log for details.
202 - DEPRECIATION: please change
203 "use AnyEvent::Impl::Perl" to "use AnyEvent::Loop" and
204 "AnyEvent::Impl::Perl::loop" to "AnyEvent::Loop::run"
205 in your programs, for future compatibility.
206 - *blush* idle watcher emulation was borked.
207 - AnyEvent::Strict now checks the AE::xxx functions as well.
208 - querying too long domain names would barf deep inside AnyEvent::DNS - since
209 this is hard to avoid for applications, AnyEvent:.DNS::request now checks for
210 for too long domain names explicitly and passes an undef to the callback
211 instead (reported by James Bromberger).
212 - make AnyEvent::Util more robust against EINTR (analyzed by Andrew Suffield).
213 - new functions: AnyEvent::postpone, AnyEvent::log.
214 - new module: AnyEvent::Log, for simple logging needs.
215 - new env variable, PERL_ANYEVENT_LOG, for fine-grained logging config.
216 - new env variable, PERL_ANYEVENT_DEBUG_SHELL, to automatically start
217 a debug shell.
218 - new env variable, PERL_ANYEVENT_DEBUG_WRAP, to automatically enable
219 watcher instrumentation/wrapping (see AnyEvent::Debug::wrap).
220 - new env variable, PERL_ANYEVENT_HOSTS, to override the place where
221 /etc/hosts can be found.
222 - an empty PERL_ANYEVENT_RESOLV_CONF now selects the default
223 configuration.
224 - env variables can now be specified using AE_xyz instead of
225 PERL_ANYEVENT_xyz, subject to some rules.
226 - lots of new AnyEvent::Debug functionality.
227 - AnyEvent::Strict now checks for modifications of $_ by the callback
228 (by setting it to readonly while executing the callback, if possible).
229 - IO::Async sometimes cancels the wrong timer when you call cancel_timer.
230 Invest extra overhead to work around this peculiar behaviour.
231 - split the pure perl backend into a loop module and an impl
232 module (AnyEvent::Loop and AnyEvent::Impl::Perl).
233 - some read types (regex, netstring, json, storable) could
234 errornously keep an AE::Handle object alive on their own
235 (reported by Mohammad Toossi).
236 - AnyEvent::Socket::resolve_sockaddr and all functions using it now
237 supports /etc/hosts overrides.
238 - add more workarounds around child watcher bugs in IO::Async
239 and (newly introduced) in POE.
240 - use glib child watchers and try to work around its limitations,
241 if possible (based on a patch by Kevin Ryde).
242 - add support for FLTK (fltk2), via AnyEvent::Impl::FLTK2.
243 - many read types in AnyEvent::Handle would malfunction during
244 transitions to TLS because of a "delete $self->{rbuf}".
245 (reported by Gerald Galster).
246 - improve condvar blocking wait performance (probably only noticable
247 with faster backends), also make it easier for event loops that
248 do not support blocking waits.
249 - PERL_ANYEVENT_MODEL now also allows full module names, although
250 the usefulness of this feature is not yet clear.
251 - add tests for all supported event loops, conditional on PERL_ANYEVENT_LOOP_TESTS.
252 - add t/80_ssltext.t.
253 - no longer use AUTOLOAD for the initial loading (for no very good reason).
254 - tcp_server now tries to unlink unix domain sockets when it gets destroyed.
255 - do not run event loop detection multiple times when programs call their cached
256 copies of AnyEvent::detect.
257 - parse_hostport now accepts absolute paths als unix domain sockets.
258 - suppress the idiotic warning inside POE by patching the POE::Kernel
259 run flags instead of asking the user to jump through contortions.
260 - add Task::Weaken as dependency (as recommends only at this time).
261 - replace PF_UNSPEC by 0, for good old minix.
262
2636.0 (not officially released, but was on CPAN for about 15 minutes :/)
264
2655.34 Thu May 12 09:59:41 CEST 2011
266 - AE::TLS didn't support both cert_file and key at the same time,
267 due to some copy&paste bug (reported and analyzed by vti).
268 - improve AE::Handle timeout documentation slightly.
269
2705.33 Mon Feb 7 21:51:41 CET 2011
271 - never appeared on CPAN, for unknown reasons.
272 - AnyEvent::Handle on_connect_error - do not try to destroy a handle
273 object when it's already gone (analyzed by Ryan Bullock).
274 - update idna algorithm(s) to UTS#46 version 6.0.0.
275 - preserve trailing dot in idn_to_ascii - testsuite says to preserve,
276 uts46 says it's an error, and testsuite makes more sense.
277 - remove leading dots in idn_to_ascii - testsuite says to remove,
278 uts46 says it's an error, and testsuite makes more sense.
279 - fixed a bug in idn ignored character processing.
280 - passes the uts46 6.0.0 testsuite except for three cases which
281 apparently are bugs in the testsuite itself.
282 - passes the uts46 6.0.1 testsuite except for eleven cases which
283 apparently are bugs in the testsuite itself.
284
2855.31 Sun Jan 23 11:49:19 CET 2011
286 - work around a POE bug where POE would first poll for new events and
287 invoke handlers BEFORE handling the message queue.
288 - fix parsing of "domain" statement in resolv.conf (Johannes Plunien).
289 - stop_read now does just that in TLS mode. see its
290 documentation for caveats.
291 - update to common::sense 3.4 warning set (some warnings in perl impose
292 an arbitrary 32kb limit on regex matches, which affects
293 AnyEvent::Handle).
294 - AE::Handle had a broken rbuf_max method and missing wbuf_max method.
295 - new experimental AE::Handle::resettls method.
296
2975.3 Fri Dec 31 05:48:13 CET 2010
298 - major IOAsync patch by Paul Evans - 0.33 is required, as the major
299 issues with IO::Async have been solved in that version, including
300 the availabilty of a default loop!
301 - implement new wbuf_max parameter for AnyEvent::Handle.
302 - added Cocoa::EventLoop backend by Daisuke Murase.
303 - the default testsuite now honors PERL_ANYEVENT_MODEL (Paul Evans).
304 - clarify handle on_prepare documentation.
305 - set CLOEXEC flag on parent-side fd's in AnyEvent::Util::run_cmd,
306 to avoid leaking them to any child processes.
307 - try to force files into the arch-dependend lib directory, for the
308 half-"turly obsessive compulsive person"(s) that depend on it.
309
3105.29 Sun Dec 5 10:49:21 CET 2010
311 - convert EV backend to EV 4.00 API (so better upgrade EV too).
312 - AE::Handle: implement dynamic read_size adjustment, add
313 max_read_size parameter, reduce default read_size to 2048.
314 - add some FAQ entries for servers and/or on_eof vs. on_error.
315 - work around OS bugs (cygwin again) when lingering in AE::Handle.
316 - reduce memory usage slightly when connect was used in AE::Handle.
317 - wrok around more segfaults in Tk.
318 - document tls_ctx => undef as valid.
319 - detect WNOHANG value at compiletime.
320 - include some probably linux-only support for building AnyEvent
321 as part of the perl core.
322 - improve compatibility of t/08_idna.t to perl 5.8.
323 - make AnyEvent::Strict truly optional (in case somebody builds a
324 minimal perl...).
325
3265.28 Wed Oct 13 04:14:23 CEST 2010
327 - due to a glitch, AnyEvent's internal getprotobyname wasn't
328 actually used in tcp_server.
329 - implement AnyEvent::CondVar->new method, which was documented but
330 missing.
331 - support autoloading of read types in unshift_read.
332 - AnyEvent::Strict: do the test for invalid fh arguments differently,
333 resulting in more informative error messages.
334 - new AnyEvent::FAQ manpage.
335 - clarify that you can actually call push_read etc. in on_read.
336 - put keys and certs in /tmp, not into the current directory,
337 and document this in the SECURITY CONSIDERATIONS sections
338 of AnyEvent::TLS.
339 - support ";" as resolv.conf comment character in AnyEvent::DNS, as
340 requested by Ron Isaacson.
341 - document $AnyEvent::DNS::RESOLVER variable.
342 - incorporated major typo patches by Abhijit Menon-Sen.
343 - AnyEvent::Handle now croaks when negative timeouts are passed.
344 - add a shitload of TCP_xxx sockopt name constants to AnyEvent::Util,
345 undocumented, but who knows when they might come in handy.
346
3475.271 Tue Jun 8 12:05:46 CEST 2010
348 - backport to perl 5.8.x.
349
3505.27 Sun Jun 6 12:12:05 CEST 2010
351 - postpone differently in AnyEvent::Socket now, as
352 when not, canceling the connection attempt might fail
353 (found by Felix Antonius Wilhelm Ostmann).
354 - explicitly check for non-stream sockets in AE::Handle, too many
355 clueless people fell into the trap of this somehow working.
356 - simplified and reworked the "OTHER MODULES" section.
357 - better/more condvar examples.
358
3595.261 Wed Apr 28 16:13:36 CEST 2010
360 - AF_INET6 was not properly used from Socket6 during configuration
361 time (found by Andy Grundman).
362
3635.26 Mon Apr 12 04:49:35 CEST 2010
364 - don't generate spurious readyness notifications when select
365 returns EINTR in the pure perl backend (analysed and testcase
366 by Bas Denissen).
367 - give same examples in SYNOPSIS sections of AnyEvent and AE.
368 - provider faster implementations for AE::time/now/now_update for
369 EV and Perl backends.
370 - provide faster AE::cv and AE::time implementations for all
371 backends.
372 - no longer support lower/mixed-case signal names (this was
373 never documented nor universally supported).
374 - some more memory µ-optimisations, and somewhat less messy
375 code to cope with both AnyEvent and AE APIs.
376
3775.251 Sat Mar 13 00:58:21 CET 2010
378 - make test failed if it was run before make install on perl <=
379 5.8.8, and older versions of perl make our live very difficult,
380 so write constants.pl during Makefile.PL time.
381
3825.25 Sat Mar 13 00:23:14 CET 2010
383 - fix a race condition in AnyEvent::Handle that would cause
384 a "bio_read: unsupported method" error (in _another_
385 ssl connection) after user code threw an exception.
386 - added AnyEvent::Handle->destroyed method.
3 - speed up AnyEvent::Socket::format_address and ::format_ipv6. 387 - speed up AnyEvent::Socket::format_address and ::format_ipv6.
388 - the AnyEvent::Util::fh_nonblocking function worked only
389 by lucky accident on win32.
390 - smaller and faster AnyEvent::Util::fh_nonblocking.
391 - when the (required!) Time::HiRes module is missing, AnyEvent
392 did not fall back to built-in time properly.
393 - do not load Fcntl at runtime, saving memory and loading time.
394 - precompile a number of constants and use them instead of runtime
395 detection and eval.
396 - free detection code after detection and similar memory optimisations.
397 - Perl backend timer interval best effort drift has been improved
398 (same algorithm as EV).
399 - update unicode idna mapping table.
4 400
55.24 Tue Jan 5 11:39:43 CET 2010 4015.24 Tue Jan 5 11:39:43 CET 2010
6 - cygwin never reports errors from failed connects. 1.5 just gives 402 - cygwin never reports errors from failed connects. 1.5 just gives
7 you continous readyness and EAGAIN, 1.7 is even more broken and 403 you continous readyness and EAGAIN, 1.7 is even more broken and
8 just hangs. work around both issues in a cygwin-specific hack. 404 just hangs. work around both issues in a cygwin-specific hack.
… …
3844.3 Fri Oct 3 09:18:43 CEST 2008 7804.3 Fri Oct 3 09:18:43 CEST 2008
385 - AnyEvent will now install a no-op signal handler for SIGPIPE, 781 - AnyEvent will now install a no-op signal handler for SIGPIPE,
386 unless one has been installed already. 782 unless one has been installed already.
387 - warn about not loaidng AnyEvent::Impl::POE early enough 783 - warn about not loaidng AnyEvent::Impl::POE early enough
388 (patch by Adam Rosenstein). 784 (patch by Adam Rosenstein).
389 - fixed a great number of bugs and corner cases in AnyEvent::Handle: 785 - fixed a great number of bugs and corner cases in AnyEvent::Handle.
390 - fix a bug in where in SSL connect mode, the 786 - fix a bug in where in SSL connect mode, the
391 client would first wait for some data by the server and 787 client would first wait for some data by the server and
392 otherwise hang. (reported and analysed in an absolutely 788 otherwise hang. (reported and analysed in an absolutely
393 exemplary manner by Adam Rosenstein). 789 exemplary manner by Adam Rosenstein).
394 - fix a bug in where SSL EOF would not be treated 790 - fix a bug in where SSL EOF would not be treated

Diff Legend

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