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

Comparing AnyEvent/Changes (file contents):
Revision 1.366 by root, Sat Dec 5 02:58:30 2009 UTC vs.
Revision 1.593 by root, Thu Aug 28 16:51:20 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: improve confess skipping
18 - work around a newly introduced bug in Socket 2.011 (an
19 errornous sun_length check) (analyzed by Maxime Soulé).
20 - AnyEvent::TLS didn't load (but refer to) AnyEvent::Socket
21 (analyzed by Ben Magistro).
22 - AnyEvent::Strict will now confess, not croak. This is in line with
23 it being a development/debugging tool.
24 - work around a number of libglib bugs (debug builds of libglib enforce
25 certain undocumented behaviour patterns such as not being able to
26 remove a child watch source after it has fired, which we will try
27 to emulate to avoid "criticals". what where they thinking...).
28 - changed default dns resolver "max_outstanding" value from 1 to 10,
29 the latter beinfg the intended value all along
30 (reported by Ilya Chesnokov).
31 - added new "AnyEvent::Impl::UV" interface module to the UV event lib
32 (written by Mike Lowell).
33
347.07 Tue Dec 17 17:45:02 CET 2013
35 - the documentation for custom tls verify schemes was wrong. make it agree
36 with the code (reported by Maxime Soulé).
37 - added cbor read and write types to AnyEvent::Handle (using CBOR::XS).
38 - work around an API change in openssl that could cause wrong tls connection
39 aborts, likely on windows only (analyzed by sten).
40 - calling AnyEvent->now_update with AnyEvent::Impl::Perl caused an
41 endless loop (reported by Dietrich Rebmann).
42 - add tlsv1_1 and tlsv1_2 protocols to AnyEvent::TLS
43 (patch by Maxime Soulé).
44 - document AnyEvent::Impl::IOAsync::set_loop and
45 $AnyEvent::Impl::IOAsync::LOOP. Though only documented now, this
46 functionality has _always_ been available.
47 - force a toplevel domain name in t/81_hosts.t (analyzed by
48 David Jack Wange Olrik).
49 - document that AnyEvent::Log uses AnyEvent::IO.
50 - warn about AnyEvent::Filesys::Notify performance.
51 - praise the joys of AnyEvent::Fork::*.
52 - time for an =encoding directive.
53 - do no longer use JSON to create a default json coder, use
54 JSON::XS or JSON::PP directly.
55
567.05 Wed Aug 21 10:38:08 CEST 2013
57 - uts46data.pl couldn't be found due to wrong naming of the file
58 (reported by Fulko Hew).
59 - handle lone \015's properly in AE::Handle's default line read
60 (reported by various people).
61 - untaint ip addresses found in /etc/hosts (patch by José Micó).
62 - the memleak fix in 7.03 caused resolving via /etc/hosts to always
63 fail on first use (reported and testcase by Andrew Whatson).
64 - expose AnyEvent::Log::format_time, and allow users to redefine it.
65 - expose AnyEvent::Log::default_format, and allow redefinition.
66 - expose AnyEvent::Log::fatal_exit, to allow redefinition.
67 - AnyEvent::Debug shell can now run coro shell commands, if available.
68 - t/63* tests were wrongly in MANIFEST.
69 - kernel.org's finger server went MIA, switch to freebsd.org and
70 icculus.org.
71 - clarify that IO::AIO and AnyEvent::AIO are needed for AnyEvent::IO
72 to function asynchronously (reported by Daniel Carosone).
73 - hard-disable $^W in most tests, it generates too much garbage output.
74 - use a (hopefully) more future-proof method to emulate common::sense.
75 - upgrade to UTS-46:6.2.0.
76 - switch to INSTLIB from INSTLIBDIR, as INSTLIBDIR was wrongly
77 documented. should not affect anything.
78
797.04 Thu Nov 15 02:23:58 CET 2012
80 - debugging code left in.
81
827.03 Thu Nov 15 02:16:11 CET 2012
83 - AnyEvent::Socket::inet_aton did not work when dns resolution was
84 used to find the addresses (analyzed and patch by Maxime Soulé).
85 - fix a memory leak in the /etc/hosts lookup code when hosts don't
86 resolve and are not in hosts (analyzed by sten).
87
887.02 Tue Aug 14 04:27:58 CEST 2012
89 - AnyEvent::Util::run_cmd could block indefinitely (analyzed and test
90 program by Yorhel).
91 - verified that AnyEvent::Socket follows RFC5952.
92 - try to parse "ADDR#PORT" in addition to "ADDR PORT".
93
947.01 Sun May 13 01:03:17 CEST 2012
95 - fail with EPROTO in AnyEvent::Handle wqhen TLS is requested but
96 not available, instead of throwing an exception.
97 - use File::Spec to get the tmpdir in t/*, to avoid needless
98 failures on (most, not mine :) windows boxes.
99 - new handle read types: tls_detect and tls_autostart.
100
1017.0 Fri Apr 13 06:33:30 CEST 2012
102 - child watchers are broken in POE 1.352 (also many earlier
103 versions) and there seems to be no way to work around it, as POE
104 itself is inherently racy. Document this regression and add a delay
105 in t/68_poe_03_child.t for the time being.
106 - new module AnyEvent::IO, that is a frontend to either a pure-perl
107 synchronous I/O implementation (AnyEvent::IO::Perl), or to the
108 asynchronous I/O extension IO::AIO (AnyEvent::IO::IOAIO).
109 - load /etc/hosts only when DNS has no answer.
110 - stat /etc/hosts on every access and reload it if it changed.
111 - load /etc/hosts via AnyEvent::IO - potentially asynchronous.
112 - fix a buggy croak in the dh parameter reading in AnyEvent::TLS.
113 - AnyEvent::Log log_to_file and log_to_path now use AnyEvent::IO. As
114 a side effect, they now use true appending as opposed to libc
115 appending, and the file might not have been opened when the function
116 returns.
117 - the default logging level was not properly documented in a variety of
118 places, this has hopefully been rectified.
119 - updated uts46data.pl for unicode 6.1.0.
120 - made log messages generated by AnyEvent submodules not
121 include the package name anymore, as it will be logged
122 by default already.
123 - upgrade to the trick used by common-sense 3.5 to work
124 around extra warning torture/breakage under perl 5.15.x.
125 - log messages by anyevent are now ucfirst, are usually full
126 sentences and do no longer include the package name.
127 - the storable read type would sometimes throw an exception instead
128 of causing EBADMSG (analyzed by Maxime Soulé).
129 - work around a bug in openssl 1.0.1 which enforces a minimum rsa
130 keysize (reported by Paul Howarth).
131 - documented Rocco fabricating statements about AnyEvent and me.
132
1336.14 Tue Jan 31 20:00:24 CET 2012
134 - AnyEvent::Impl::Tk was broken due to a mysterious "die" inside,
135 probably an editing mistake (reported by Darin McBride).
136
1376.13 Thu Jan 12 07:27:01 CET 2012
138 - AnyEvent::Util::fork_call checked for POSIX availability in the wrong way,
139 causing it to fail if POSIX wasn't loaded already (analysed by Rock Power).
140 - AnyEvent::Handle::push_read (line => did pass $1 directly, so
141 regex-matching inside the callback would change the parameter, despite
142 $1 being dynamically scoped per-block. this perl bug is now being
143 worked around (testcase by Cindy Wang).
144
1456.12 Mon Dec 12 13:21:10 CET 2011
146 - $! was clobbered by subsequent calls in tcp_connect, due to
147 the postpone not saving/restoring it (analyzed by Richard Garnish).
148
1496.11 Tue Nov 22 10:36:05 CET 2011
150 - Tk cannot create windows when tainted unless you set a title, so
151 set a dummy title for AnyEvent's dummy mainwindow. How dummy.
152 - escape any nonprintable/nonascii characters when stringifying
153 backtraces.
154 - log the reason loading an interface module fails to load at level debug.
155
1566.1 Tue Oct 4 19:44:30 CEST 2011
157 - INCOMPATIBLE CHANGE: the default log level is now "4" (error
158 and above), and some messages inside AnyEvent have been elevated
159 to higher log levels to print by default.
160 - AnyEvent::log, unlike AnyEvent::Log::log, did not always exit on fatal.
161 - work around yet another signal race bug, newly introduced by POE. My
162 how that module sucks.
163 - add google public ipv6 dns as fallback servers, so AE::DNS can finally
164 support ipv6-only operation even with defaults (also use ::1 instead
165 of 127.0.0.1).
166 - give better guidelines on how to chose a logging level.
167 - AnyEvent::Log can now cap (limit) log message priority by context.
168 - AE::log will now save and restore $! and $@, so callers don't have to.
169 - AE::Strict now checks that registered fds are still valid at regular
170 intervals.
171 - skip frequently called 'require' calls if the module appears loaded.
172 - increase signal latency test timer considerably for poor little osx,
173 which can't be bothered with basic posix functionality such as signal
174 delivery.
175 - mention some debugging aids in the tutorial.
176 - new env variable PERL_ANYEVENT_MAX_SIGNAL_LATENCY.
177 - private (%namespace) log contexts in PERL_ANYEVENT_LOG now have the name
178 of the context as title.
179 - AnyEvent::Strict would not restore read-write access to $_ when a callback
180 died - this is a user bug, but since AE::Strict exists to help dbeugging...
181 - detect the extremely evil IO::Async::Loop::AnyEvent module and refuse
182 to work with it - the author is well aware of the issues but refuses
183 to fix them. This extreme step has been taken because the module actively
184 attacks the goals of AnyEvent and makes it harder for module authors to use
185 AnyEvent.
186
1876.02 Fri Aug 26 20:08:31 CEST 2011
188 - log_to_syslog now accepts facility strings.
189 - log_to_syslog would not allow facilities - "$facility|$level"
190 *literally* means that string. sheesh.
191 - the CPAN parser got confused and thought we provide Sys::Syslog, try
192 to work around.
193 - renamed AnyEvent::Impl::FLTK2 to FLTK.
194
1956.01 Fri Aug 26 07:04:11 CEST 2011
196 - INCOMPATIBLE CHANGE: AnyEvent::DNS resource records now include
197 the TTL as fourth element - this affects ->resolve, ->request
198 and ->dns_unpack indirectly. Convenience functions such as
199 AnyEvent::DNS::{a,txt,mx,srv} are not affected, neither is
200 AnyEvent::Socket.
201 - INCOMPATIBLE CHANGE: backend authors now should not implement
202 one_event or loop, but instead the AnyEvent::CondVar::_wait and _poll
203 methods.
204 - INCOMPATIBLE CHANGE: verbosity levels have been redefined, see
205 AnyEvent (PERL_ANYEVENT_VERBOSE) and AnyEvent::Log for details.
206 - DEPRECIATION: please change
207 "use AnyEvent::Impl::Perl" to "use AnyEvent::Loop" and
208 "AnyEvent::Impl::Perl::loop" to "AnyEvent::Loop::run"
209 in your programs, for future compatibility.
210 - *blush* idle watcher emulation was borked.
211 - AnyEvent::Strict now checks the AE::xxx functions as well.
212 - querying too long domain names would barf deep inside AnyEvent::DNS - since
213 this is hard to avoid for applications, AnyEvent:.DNS::request now checks for
214 for too long domain names explicitly and passes an undef to the callback
215 instead (reported by James Bromberger).
216 - make AnyEvent::Util more robust against EINTR (analyzed by Andrew Suffield).
217 - new functions: AnyEvent::postpone, AnyEvent::log.
218 - new module: AnyEvent::Log, for simple logging needs.
219 - new env variable, PERL_ANYEVENT_LOG, for fine-grained logging config.
220 - new env variable, PERL_ANYEVENT_DEBUG_SHELL, to automatically start
221 a debug shell.
222 - new env variable, PERL_ANYEVENT_DEBUG_WRAP, to automatically enable
223 watcher instrumentation/wrapping (see AnyEvent::Debug::wrap).
224 - new env variable, PERL_ANYEVENT_HOSTS, to override the place where
225 /etc/hosts can be found.
226 - an empty PERL_ANYEVENT_RESOLV_CONF now selects the default
227 configuration.
228 - env variables can now be specified using AE_xyz instead of
229 PERL_ANYEVENT_xyz, subject to some rules.
230 - lots of new AnyEvent::Debug functionality.
231 - AnyEvent::Strict now checks for modifications of $_ by the callback
232 (by setting it to readonly while executing the callback, if possible).
233 - IO::Async sometimes cancels the wrong timer when you call cancel_timer.
234 Invest extra overhead to work around this peculiar behaviour.
235 - split the pure perl backend into a loop module and an impl
236 module (AnyEvent::Loop and AnyEvent::Impl::Perl).
237 - some read types (regex, netstring, json, storable) could
238 errornously keep an AE::Handle object alive on their own
239 (reported by Mohammad Toossi).
240 - AnyEvent::Socket::resolve_sockaddr and all functions using it now
241 supports /etc/hosts overrides.
242 - add more workarounds around child watcher bugs in IO::Async
243 and (newly introduced) in POE.
244 - use glib child watchers and try to work around its limitations,
245 if possible (based on a patch by Kevin Ryde).
246 - add support for FLTK (fltk2), via AnyEvent::Impl::FLTK2.
247 - many read types in AnyEvent::Handle would malfunction during
248 transitions to TLS because of a "delete $self->{rbuf}".
249 (reported by Gerald Galster).
250 - improve condvar blocking wait performance (probably only noticable
251 with faster backends), also make it easier for event loops that
252 do not support blocking waits.
253 - PERL_ANYEVENT_MODEL now also allows full module names, although
254 the usefulness of this feature is not yet clear.
255 - add tests for all supported event loops, conditional on PERL_ANYEVENT_LOOP_TESTS.
256 - add t/80_ssltext.t.
257 - no longer use AUTOLOAD for the initial loading (for no very good reason).
258 - tcp_server now tries to unlink unix domain sockets when it gets destroyed.
259 - do not run event loop detection multiple times when programs call their cached
260 copies of AnyEvent::detect.
261 - parse_hostport now accepts absolute paths als unix domain sockets.
262 - suppress the idiotic warning inside POE by patching the POE::Kernel
263 run flags instead of asking the user to jump through contortions.
264 - add Task::Weaken as dependency (as recommends only at this time).
265 - replace PF_UNSPEC by 0, for good old minix.
266
2676.0 (not officially released, but was on CPAN for about 15 minutes :/)
268
2695.34 Thu May 12 09:59:41 CEST 2011
270 - AE::TLS didn't support both cert_file and key at the same time,
271 due to some copy&paste bug (reported and analyzed by vti).
272 - improve AE::Handle timeout documentation slightly.
273
2745.33 Mon Feb 7 21:51:41 CET 2011
275 - never appeared on CPAN, for unknown reasons.
276 - AnyEvent::Handle on_connect_error - do not try to destroy a handle
277 object when it's already gone (analyzed by Ryan Bullock).
278 - update idna algorithm(s) to UTS#46 version 6.0.0.
279 - preserve trailing dot in idn_to_ascii - testsuite says to preserve,
280 uts46 says it's an error, and testsuite makes more sense.
281 - remove leading dots in idn_to_ascii - testsuite says to remove,
282 uts46 says it's an error, and testsuite makes more sense.
283 - fixed a bug in idn ignored character processing.
284 - passes the uts46 6.0.0 testsuite except for three cases which
285 apparently are bugs in the testsuite itself.
286 - passes the uts46 6.0.1 testsuite except for eleven cases which
287 apparently are bugs in the testsuite itself.
288
2895.31 Sun Jan 23 11:49:19 CET 2011
290 - work around a POE bug where POE would first poll for new events and
291 invoke handlers BEFORE handling the message queue.
292 - fix parsing of "domain" statement in resolv.conf (Johannes Plunien).
293 - stop_read now does just that in TLS mode. see its
294 documentation for caveats.
295 - update to common::sense 3.4 warning set (some warnings in perl impose
296 an arbitrary 32kb limit on regex matches, which affects
297 AnyEvent::Handle).
298 - AE::Handle had a broken rbuf_max method and missing wbuf_max method.
299 - new experimental AE::Handle::resettls method.
300
3015.3 Fri Dec 31 05:48:13 CET 2010
302 - major IOAsync patch by Paul Evans - 0.33 is required, as the major
303 issues with IO::Async have been solved in that version, including
304 the availabilty of a default loop!
305 - implement new wbuf_max parameter for AnyEvent::Handle.
306 - added Cocoa::EventLoop backend by Daisuke Murase.
307 - the default testsuite now honors PERL_ANYEVENT_MODEL (Paul Evans).
308 - clarify handle on_prepare documentation.
309 - set CLOEXEC flag on parent-side fd's in AnyEvent::Util::run_cmd,
310 to avoid leaking them to any child processes.
311 - try to force files into the arch-dependend lib directory, for the
312 half-"turly obsessive compulsive person"(s) that depend on it.
313
3145.29 Sun Dec 5 10:49:21 CET 2010
315 - convert EV backend to EV 4.00 API (so better upgrade EV too).
316 - AE::Handle: implement dynamic read_size adjustment, add
317 max_read_size parameter, reduce default read_size to 2048.
318 - add some FAQ entries for servers and/or on_eof vs. on_error.
319 - work around OS bugs (cygwin again) when lingering in AE::Handle.
320 - reduce memory usage slightly when connect was used in AE::Handle.
321 - wrok around more segfaults in Tk.
322 - document tls_ctx => undef as valid.
323 - detect WNOHANG value at compiletime.
324 - include some probably linux-only support for building AnyEvent
325 as part of the perl core.
326 - improve compatibility of t/08_idna.t to perl 5.8.
327 - make AnyEvent::Strict truly optional (in case somebody builds a
328 minimal perl...).
329
3305.28 Wed Oct 13 04:14:23 CEST 2010
331 - due to a glitch, AnyEvent's internal getprotobyname wasn't
332 actually used in tcp_server.
333 - implement AnyEvent::CondVar->new method, which was documented but
334 missing.
335 - support autoloading of read types in unshift_read.
336 - AnyEvent::Strict: do the test for invalid fh arguments differently,
337 resulting in more informative error messages.
338 - new AnyEvent::FAQ manpage.
339 - clarify that you can actually call push_read etc. in on_read.
340 - put keys and certs in /tmp, not into the current directory,
341 and document this in the SECURITY CONSIDERATIONS sections
342 of AnyEvent::TLS.
343 - support ";" as resolv.conf comment character in AnyEvent::DNS, as
344 requested by Ron Isaacson.
345 - document $AnyEvent::DNS::RESOLVER variable.
346 - incorporated major typo patches by Abhijit Menon-Sen.
347 - AnyEvent::Handle now croaks when negative timeouts are passed.
348 - add a shitload of TCP_xxx sockopt name constants to AnyEvent::Util,
349 undocumented, but who knows when they might come in handy.
350
3515.271 Tue Jun 8 12:05:46 CEST 2010
352 - backport to perl 5.8.x.
353
3545.27 Sun Jun 6 12:12:05 CEST 2010
355 - postpone differently in AnyEvent::Socket now, as
356 when not, canceling the connection attempt might fail
357 (found by Felix Antonius Wilhelm Ostmann).
358 - explicitly check for non-stream sockets in AE::Handle, too many
359 clueless people fell into the trap of this somehow working.
360 - simplified and reworked the "OTHER MODULES" section.
361 - better/more condvar examples.
362
3635.261 Wed Apr 28 16:13:36 CEST 2010
364 - AF_INET6 was not properly used from Socket6 during configuration
365 time (found by Andy Grundman).
366
3675.26 Mon Apr 12 04:49:35 CEST 2010
368 - don't generate spurious readyness notifications when select
369 returns EINTR in the pure perl backend (analysed and testcase
370 by Bas Denissen).
371 - give same examples in SYNOPSIS sections of AnyEvent and AE.
372 - provider faster implementations for AE::time/now/now_update for
373 EV and Perl backends.
374 - provide faster AE::cv and AE::time implementations for all
375 backends.
376 - no longer support lower/mixed-case signal names (this was
377 never documented nor universally supported).
378 - some more memory µ-optimisations, and somewhat less messy
379 code to cope with both AnyEvent and AE APIs.
380
3815.251 Sat Mar 13 00:58:21 CET 2010
382 - make test failed if it was run before make install on perl <=
383 5.8.8, and older versions of perl make our live very difficult,
384 so write constants.pl during Makefile.PL time.
385
3865.25 Sat Mar 13 00:23:14 CET 2010
387 - fix a race condition in AnyEvent::Handle that would cause
388 a "bio_read: unsupported method" error (in _another_
389 ssl connection) after user code threw an exception.
390 - added AnyEvent::Handle->destroyed method.
391 - speed up AnyEvent::Socket::format_address and ::format_ipv6.
392 - the AnyEvent::Util::fh_nonblocking function worked only
393 by lucky accident on win32.
394 - smaller and faster AnyEvent::Util::fh_nonblocking.
395 - when the (required!) Time::HiRes module is missing, AnyEvent
396 did not fall back to built-in time properly.
397 - do not load Fcntl at runtime, saving memory and loading time.
398 - precompile a number of constants and use them instead of runtime
399 detection and eval.
400 - free detection code after detection and similar memory optimisations.
401 - Perl backend timer interval best effort drift has been improved
402 (same algorithm as EV).
403 - update unicode idna mapping table.
404
4055.24 Tue Jan 5 11:39:43 CET 2010
406 - cygwin never reports errors from failed connects. 1.5 just gives
407 you continous readyness and EAGAIN, 1.7 is even more broken and
408 just hangs. work around both issues in a cygwin-specific hack.
409 - improve idle watcher documentation slightly.
410
4115.23 Sun Dec 20 23:48:00 CET 2009
412 - support IDNs in resolve_sockaddr, and therefore in tcp_connect.
413 - implement punycode_encode/decode, idn_nameprep,
414 idn_to_ascii and idn_to_unicode operations in AnyEvent::Util.
3 - provide $AE::VERSION. 415 - provide $AE::VERSION.
416 - removed traces of "no strict 'refs'".
4 417
55.22 Sat Dec 5 03:51:13 CET 2009 4185.22 Sat Dec 5 03:51:13 CET 2009
6 - downgrade-or-fail in AnyEvent::Handle::push_write, to 419 - downgrade-or-fail in AnyEvent::Handle::push_write, to
7 diagnose encoding failures earlier and more succinctly. 420 diagnose encoding failures earlier and more succinctly.
8 (this works around bugs in perl, throwing away encoding info 421 (this works around bugs in perl, throwing away encoding info
… …
3714.3 Fri Oct 3 09:18:43 CEST 2008 7844.3 Fri Oct 3 09:18:43 CEST 2008
372 - AnyEvent will now install a no-op signal handler for SIGPIPE, 785 - AnyEvent will now install a no-op signal handler for SIGPIPE,
373 unless one has been installed already. 786 unless one has been installed already.
374 - warn about not loaidng AnyEvent::Impl::POE early enough 787 - warn about not loaidng AnyEvent::Impl::POE early enough
375 (patch by Adam Rosenstein). 788 (patch by Adam Rosenstein).
376 - fixed a great number of bugs and corner cases in AnyEvent::Handle: 789 - fixed a great number of bugs and corner cases in AnyEvent::Handle.
377 - fix a bug in where in SSL connect mode, the 790 - fix a bug in where in SSL connect mode, the
378 client would first wait for some data by the server and 791 client would first wait for some data by the server and
379 otherwise hang. (reported and analysed in an absolutely 792 otherwise hang. (reported and analysed in an absolutely
380 exemplary manner by Adam Rosenstein). 793 exemplary manner by Adam Rosenstein).
381 - fix a bug in where SSL EOF would not be treated 794 - fix a bug in where SSL EOF would not be treated

Diff Legend

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