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

Comparing AnyEvent/Changes (file contents):
Revision 1.229 by root, Tue Apr 14 04:27:31 2009 UTC vs.
Revision 1.610 by root, Sun Mar 13 04:48:43 2016 UTC

1Revision history for Perl extension AnyEvent. 1Revision history for Perl extension AnyEvent.
2 2
3TODO: document TCP_*** constants
4TODO: add debug/trace logging to some modules
5TODO: use the eval-with-cb-call trick also to speed up JSON?
6TODO: FAQ: common pitfalls? invoke-before-return
7TODO: FAQ: scope-ids? hosts file order?
8TODO: AnyEvent::Handle with AnyEvent::IO?
9TODO: invalid. and localhost. specialcasing inside AEDNS and not AESocket (rfc6761)
10TODO: maybe implement env variable to give hosts precedence
11TODO: more formats for parse_ipv6 (single ipv6 address without port, ...p80, ...#80)
12
13TODO: inet_aton maybe pack C4? no, add a comment why not
14TODO: hosts read multiple times for initial concurrent tcp_connect's?
15
16 - Only call tlsext_host_name for non-empty common names (reported
17 by Maxime Soulé).
18 - log a (single) notice message if SNI is not supported.
19
207.12 Wed Jan 27 19:12:26 CET 2016
21 - use common name as hostname for tls connects, if Net::SSLeay
22 supports SNI.
23 - fix documentation of tls_autostart read type in AnyEvent::Handle,
24 analyzed by Felix Ostmann.
25
267.11 Thu Jul 16 14:36:00 CEST 2015
27 - AnyEvent::Socket::parse_ipv6 could accept malformed ipv6
28 addresses (extra "::" at end and similar cases).
29 - add a more explicit warning to AnyEvent::Handle that it doesn't
30 work on files, people keep getting confused.
31 - new function AnyEvent::Socket::tcp_bind.
32 - new functions AnyEvent::fh_block and AnyEvent::fh_unblock.
33 - aligned ipv6 address formatting with RFC 5952 (by not shortening
34 a single :0: to ::).
35 - added stability canary support.
36
377.09 Sat May 2 16:38:53 CEST 2015
38 - AnyEvent::Debug called an internal function (AnyEvent::Log::ft)
39 that was renamed to AnyEvent::Log:format_time. uinder its old name
40 (analyzed by Michael Stovenour).
41 - update AnyEvent::DNS fallback resolver addresses:
42 seems google effectively killed most other free dns resolvers,
43 so remove them, but add cable and wireless (ecrc) since it was
44 stable for 20 years or so, official or not, and there should be
45 an alternative to google.
46 - perl5porters broke windows error codes in 5.20, and mapped
47 WSAEWOULDBLOCK on the (different) EWOULDBLOCK error code, and
48 WSAEINPROGRESS into the incompatible ERINPROGRESS code, probably
49 because they were so cool. They probably broke other error codes
50 for no reason, too, but I didn't care for checking, it's too
51 depressing. This version only works around the WSAEWOULDBLOCK
52 issue, because I don't have a nice way to work around the
53 WSAEINPROGRESS bug.
54
557.08 Wed Dec 10 05:27:17 CET 2014
56 - work around a newly introduced bug in Socket 2.011 (an
57 errornous sun_length check) (analyzed by Maxime Soulé).
58 - AnyEvent::TLS didn't load (but refer to) AnyEvent::Socket
59 (analyzed by Ben Magistro).
60 - AnyEvent::Strict will now confess, not croak. This is in line with
61 it being a development/debugging tool.
62 - work around a number of libglib bugs (debug builds of libglib enforce
63 certain undocumented behaviour patterns such as not being able to
64 remove a child watch source after it has fired, which we will try
65 to emulate to avoid "criticals". what where they thinking...).
66 - mention json security issues in AnyEvent::Handle, now that Douglas
67 Crockford has foolishly and incompatibly changed JSON.
68 - changed default dns resolver "max_outstanding" value from 1 to 10,
69 the latter beinfg the intended value all along
70 (reported by Ilya Chesnokov).
71 - added new "AnyEvent::Impl::UV" interface module to the UV event lib
72 (written by Mike Lowell).
73
747.07 Tue Dec 17 17:45:02 CET 2013
75 - the documentation for custom tls verify schemes was wrong. make it agree
76 with the code (reported by Maxime Soulé).
77 - added cbor read and write types to AnyEvent::Handle (using CBOR::XS).
78 - work around an API change in openssl that could cause wrong tls connection
79 aborts, likely on windows only (analyzed by sten).
80 - calling AnyEvent->now_update with AnyEvent::Impl::Perl caused an
81 endless loop (reported by Dietrich Rebmann).
82 - add tlsv1_1 and tlsv1_2 protocols to AnyEvent::TLS
83 (patch by Maxime Soulé).
84 - document AnyEvent::Impl::IOAsync::set_loop and
85 $AnyEvent::Impl::IOAsync::LOOP. Though only documented now, this
86 functionality has _always_ been available.
87 - force a toplevel domain name in t/81_hosts.t (analyzed by
88 David Jack Wange Olrik).
89 - document that AnyEvent::Log uses AnyEvent::IO.
90 - warn about AnyEvent::Filesys::Notify performance.
91 - praise the joys of AnyEvent::Fork::*.
92 - time for an =encoding directive.
93 - do no longer use JSON to create a default json coder, use
94 JSON::XS or JSON::PP directly.
95
967.05 Wed Aug 21 10:38:08 CEST 2013
97 - uts46data.pl couldn't be found due to wrong naming of the file
98 (reported by Fulko Hew).
99 - handle lone \015's properly in AE::Handle's default line read
100 (reported by various people).
101 - untaint ip addresses found in /etc/hosts (patch by José Micó).
102 - the memleak fix in 7.03 caused resolving via /etc/hosts to always
103 fail on first use (reported and testcase by Andrew Whatson).
104 - expose AnyEvent::Log::format_time, and allow users to redefine it.
105 - expose AnyEvent::Log::default_format, and allow redefinition.
106 - expose AnyEvent::Log::fatal_exit, to allow redefinition.
107 - AnyEvent::Debug shell can now run coro shell commands, if available.
108 - t/63* tests were wrongly in MANIFEST.
109 - kernel.org's finger server went MIA, switch to freebsd.org and
110 icculus.org.
111 - clarify that IO::AIO and AnyEvent::AIO are needed for AnyEvent::IO
112 to function asynchronously (reported by Daniel Carosone).
113 - hard-disable $^W in most tests, it generates too much garbage output.
114 - use a (hopefully) more future-proof method to emulate common::sense.
115 - upgrade to UTS-46:6.2.0.
116 - switch to INSTLIB from INSTLIBDIR, as INSTLIBDIR was wrongly
117 documented. should not affect anything.
118
1197.04 Thu Nov 15 02:23:58 CET 2012
120 - debugging code left in.
121
1227.03 Thu Nov 15 02:16:11 CET 2012
123 - AnyEvent::Socket::inet_aton did not work when dns resolution was
124 used to find the addresses (analyzed and patch by Maxime Soulé).
125 - fix a memory leak in the /etc/hosts lookup code when hosts don't
126 resolve and are not in hosts (analyzed by sten).
127
1287.02 Tue Aug 14 04:27:58 CEST 2012
129 - AnyEvent::Util::run_cmd could block indefinitely (analyzed and test
130 program by Yorhel).
131 - verified that AnyEvent::Socket follows RFC5952.
132 - try to parse "ADDR#PORT" in addition to "ADDR PORT".
133
1347.01 Sun May 13 01:03:17 CEST 2012
135 - fail with EPROTO in AnyEvent::Handle wqhen TLS is requested but
136 not available, instead of throwing an exception.
137 - use File::Spec to get the tmpdir in t/*, to avoid needless
138 failures on (most, not mine :) windows boxes.
139 - new handle read types: tls_detect and tls_autostart.
140
1417.0 Fri Apr 13 06:33:30 CEST 2012
142 - child watchers are broken in POE 1.352 (also many earlier
143 versions) and there seems to be no way to work around it, as POE
144 itself is inherently racy. Document this regression and add a delay
145 in t/68_poe_03_child.t for the time being.
146 - new module AnyEvent::IO, that is a frontend to either a pure-perl
147 synchronous I/O implementation (AnyEvent::IO::Perl), or to the
148 asynchronous I/O extension IO::AIO (AnyEvent::IO::IOAIO).
149 - load /etc/hosts only when DNS has no answer.
150 - stat /etc/hosts on every access and reload it if it changed.
151 - load /etc/hosts via AnyEvent::IO - potentially asynchronous.
152 - fix a buggy croak in the dh parameter reading in AnyEvent::TLS.
153 - AnyEvent::Log log_to_file and log_to_path now use AnyEvent::IO. As
154 a side effect, they now use true appending as opposed to libc
155 appending, and the file might not have been opened when the function
156 returns.
157 - the default logging level was not properly documented in a variety of
158 places, this has hopefully been rectified.
159 - updated uts46data.pl for unicode 6.1.0.
160 - made log messages generated by AnyEvent submodules not
161 include the package name anymore, as it will be logged
162 by default already.
163 - upgrade to the trick used by common-sense 3.5 to work
164 around extra warning torture/breakage under perl 5.15.x.
165 - log messages by anyevent are now ucfirst, are usually full
166 sentences and do no longer include the package name.
167 - the storable read type would sometimes throw an exception instead
168 of causing EBADMSG (analyzed by Maxime Soulé).
169 - work around a bug in openssl 1.0.1 which enforces a minimum rsa
170 keysize (reported by Paul Howarth).
171 - documented Rocco fabricating statements about AnyEvent and me.
172
1736.14 Tue Jan 31 20:00:24 CET 2012
174 - AnyEvent::Impl::Tk was broken due to a mysterious "die" inside,
175 probably an editing mistake (reported by Darin McBride).
176
1776.13 Thu Jan 12 07:27:01 CET 2012
178 - AnyEvent::Util::fork_call checked for POSIX availability in the wrong way,
179 causing it to fail if POSIX wasn't loaded already (analysed by Rock Power).
180 - AnyEvent::Handle::push_read (line => did pass $1 directly, so
181 regex-matching inside the callback would change the parameter, despite
182 $1 being dynamically scoped per-block. this perl bug is now being
183 worked around (testcase by Cindy Wang).
184
1856.12 Mon Dec 12 13:21:10 CET 2011
186 - $! was clobbered by subsequent calls in tcp_connect, due to
187 the postpone not saving/restoring it (analyzed by Richard Garnish).
188
1896.11 Tue Nov 22 10:36:05 CET 2011
190 - Tk cannot create windows when tainted unless you set a title, so
191 set a dummy title for AnyEvent's dummy mainwindow. How dummy.
192 - escape any nonprintable/nonascii characters when stringifying
193 backtraces.
194 - log the reason loading an interface module fails to load at level debug.
195
1966.1 Tue Oct 4 19:44:30 CEST 2011
197 - INCOMPATIBLE CHANGE: the default log level is now "4" (error
198 and above), and some messages inside AnyEvent have been elevated
199 to higher log levels to print by default.
200 - AnyEvent::log, unlike AnyEvent::Log::log, did not always exit on fatal.
201 - work around yet another signal race bug, newly introduced by POE. My
202 how that module sucks.
203 - add google public ipv6 dns as fallback servers, so AE::DNS can finally
204 support ipv6-only operation even with defaults (also use ::1 instead
205 of 127.0.0.1).
206 - give better guidelines on how to chose a logging level.
207 - AnyEvent::Log can now cap (limit) log message priority by context.
208 - AE::log will now save and restore $! and $@, so callers don't have to.
209 - AE::Strict now checks that registered fds are still valid at regular
210 intervals.
211 - skip frequently called 'require' calls if the module appears loaded.
212 - increase signal latency test timer considerably for poor little osx,
213 which can't be bothered with basic posix functionality such as signal
214 delivery.
215 - mention some debugging aids in the tutorial.
216 - new env variable PERL_ANYEVENT_MAX_SIGNAL_LATENCY.
217 - private (%namespace) log contexts in PERL_ANYEVENT_LOG now have the name
218 of the context as title.
219 - AnyEvent::Strict would not restore read-write access to $_ when a callback
220 died - this is a user bug, but since AE::Strict exists to help dbeugging...
221 - detect the extremely evil IO::Async::Loop::AnyEvent module and refuse
222 to work with it - the author is well aware of the issues but refuses
223 to fix them. This extreme step has been taken because the module actively
224 attacks the goals of AnyEvent and makes it harder for module authors to use
225 AnyEvent.
226
2276.02 Fri Aug 26 20:08:31 CEST 2011
228 - log_to_syslog now accepts facility strings.
229 - log_to_syslog would not allow facilities - "$facility|$level"
230 *literally* means that string. sheesh.
231 - the CPAN parser got confused and thought we provide Sys::Syslog, try
232 to work around.
233 - renamed AnyEvent::Impl::FLTK2 to FLTK.
234
2356.01 Fri Aug 26 07:04:11 CEST 2011
236 - INCOMPATIBLE CHANGE: AnyEvent::DNS resource records now include
237 the TTL as fourth element - this affects ->resolve, ->request
238 and ->dns_unpack indirectly. Convenience functions such as
239 AnyEvent::DNS::{a,txt,mx,srv} are not affected, neither is
240 AnyEvent::Socket.
241 - INCOMPATIBLE CHANGE: backend authors now should not implement
242 one_event or loop, but instead the AnyEvent::CondVar::_wait and _poll
243 methods.
244 - INCOMPATIBLE CHANGE: verbosity levels have been redefined, see
245 AnyEvent (PERL_ANYEVENT_VERBOSE) and AnyEvent::Log for details.
246 - DEPRECIATION: please change
247 "use AnyEvent::Impl::Perl" to "use AnyEvent::Loop" and
248 "AnyEvent::Impl::Perl::loop" to "AnyEvent::Loop::run"
249 in your programs, for future compatibility.
250 - *blush* idle watcher emulation was borked.
251 - AnyEvent::Strict now checks the AE::xxx functions as well.
252 - querying too long domain names would barf deep inside AnyEvent::DNS - since
253 this is hard to avoid for applications, AnyEvent:.DNS::request now checks for
254 for too long domain names explicitly and passes an undef to the callback
255 instead (reported by James Bromberger).
256 - make AnyEvent::Util more robust against EINTR (analyzed by Andrew Suffield).
257 - new functions: AnyEvent::postpone, AnyEvent::log.
258 - new module: AnyEvent::Log, for simple logging needs.
259 - new env variable, PERL_ANYEVENT_LOG, for fine-grained logging config.
260 - new env variable, PERL_ANYEVENT_DEBUG_SHELL, to automatically start
261 a debug shell.
262 - new env variable, PERL_ANYEVENT_DEBUG_WRAP, to automatically enable
263 watcher instrumentation/wrapping (see AnyEvent::Debug::wrap).
264 - new env variable, PERL_ANYEVENT_HOSTS, to override the place where
265 /etc/hosts can be found.
266 - an empty PERL_ANYEVENT_RESOLV_CONF now selects the default
267 configuration.
268 - env variables can now be specified using AE_xyz instead of
269 PERL_ANYEVENT_xyz, subject to some rules.
270 - lots of new AnyEvent::Debug functionality.
271 - AnyEvent::Strict now checks for modifications of $_ by the callback
272 (by setting it to readonly while executing the callback, if possible).
273 - IO::Async sometimes cancels the wrong timer when you call cancel_timer.
274 Invest extra overhead to work around this peculiar behaviour.
275 - split the pure perl backend into a loop module and an impl
276 module (AnyEvent::Loop and AnyEvent::Impl::Perl).
277 - some read types (regex, netstring, json, storable) could
278 errornously keep an AE::Handle object alive on their own
279 (reported by Mohammad Toossi).
280 - AnyEvent::Socket::resolve_sockaddr and all functions using it now
281 supports /etc/hosts overrides.
282 - add more workarounds around child watcher bugs in IO::Async
283 and (newly introduced) in POE.
284 - use glib child watchers and try to work around its limitations,
285 if possible (based on a patch by Kevin Ryde).
286 - add support for FLTK (fltk2), via AnyEvent::Impl::FLTK2.
287 - many read types in AnyEvent::Handle would malfunction during
288 transitions to TLS because of a "delete $self->{rbuf}".
289 (reported by Gerald Galster).
290 - improve condvar blocking wait performance (probably only noticable
291 with faster backends), also make it easier for event loops that
292 do not support blocking waits.
293 - PERL_ANYEVENT_MODEL now also allows full module names, although
294 the usefulness of this feature is not yet clear.
295 - add tests for all supported event loops, conditional on PERL_ANYEVENT_LOOP_TESTS.
296 - add t/80_ssltext.t.
297 - no longer use AUTOLOAD for the initial loading (for no very good reason).
298 - tcp_server now tries to unlink unix domain sockets when it gets destroyed.
299 - do not run event loop detection multiple times when programs call their cached
300 copies of AnyEvent::detect.
301 - parse_hostport now accepts absolute paths als unix domain sockets.
302 - suppress the idiotic warning inside POE by patching the POE::Kernel
303 run flags instead of asking the user to jump through contortions.
304 - add Task::Weaken as dependency (as recommends only at this time).
305 - replace PF_UNSPEC by 0, for good old minix.
306
3076.0 (not officially released, but was on CPAN for about 15 minutes :/)
308
3095.34 Thu May 12 09:59:41 CEST 2011
310 - AE::TLS didn't support both cert_file and key at the same time,
311 due to some copy&paste bug (reported and analyzed by vti).
312 - improve AE::Handle timeout documentation slightly.
313
3145.33 Mon Feb 7 21:51:41 CET 2011
315 - never appeared on CPAN, for unknown reasons.
316 - AnyEvent::Handle on_connect_error - do not try to destroy a handle
317 object when it's already gone (analyzed by Ryan Bullock).
318 - update idna algorithm(s) to UTS#46 version 6.0.0.
319 - preserve trailing dot in idn_to_ascii - testsuite says to preserve,
320 uts46 says it's an error, and testsuite makes more sense.
321 - remove leading dots in idn_to_ascii - testsuite says to remove,
322 uts46 says it's an error, and testsuite makes more sense.
323 - fixed a bug in idn ignored character processing.
324 - passes the uts46 6.0.0 testsuite except for three cases which
325 apparently are bugs in the testsuite itself.
326 - passes the uts46 6.0.1 testsuite except for eleven cases which
327 apparently are bugs in the testsuite itself.
328
3295.31 Sun Jan 23 11:49:19 CET 2011
330 - work around a POE bug where POE would first poll for new events and
331 invoke handlers BEFORE handling the message queue.
332 - fix parsing of "domain" statement in resolv.conf (Johannes Plunien).
333 - stop_read now does just that in TLS mode. see its
334 documentation for caveats.
335 - update to common::sense 3.4 warning set (some warnings in perl impose
336 an arbitrary 32kb limit on regex matches, which affects
337 AnyEvent::Handle).
338 - AE::Handle had a broken rbuf_max method and missing wbuf_max method.
339 - new experimental AE::Handle::resettls method.
340
3415.3 Fri Dec 31 05:48:13 CET 2010
342 - major IOAsync patch by Paul Evans - 0.33 is required, as the major
343 issues with IO::Async have been solved in that version, including
344 the availabilty of a default loop!
345 - implement new wbuf_max parameter for AnyEvent::Handle.
346 - added Cocoa::EventLoop backend by Daisuke Murase.
347 - the default testsuite now honors PERL_ANYEVENT_MODEL (Paul Evans).
348 - clarify handle on_prepare documentation.
349 - set CLOEXEC flag on parent-side fd's in AnyEvent::Util::run_cmd,
350 to avoid leaking them to any child processes.
351 - try to force files into the arch-dependend lib directory, for the
352 half-"turly obsessive compulsive person"(s) that depend on it.
353
3545.29 Sun Dec 5 10:49:21 CET 2010
355 - convert EV backend to EV 4.00 API (so better upgrade EV too).
356 - AE::Handle: implement dynamic read_size adjustment, add
357 max_read_size parameter, reduce default read_size to 2048.
358 - add some FAQ entries for servers and/or on_eof vs. on_error.
359 - work around OS bugs (cygwin again) when lingering in AE::Handle.
360 - reduce memory usage slightly when connect was used in AE::Handle.
361 - wrok around more segfaults in Tk.
362 - document tls_ctx => undef as valid.
363 - detect WNOHANG value at compiletime.
364 - include some probably linux-only support for building AnyEvent
365 as part of the perl core.
366 - improve compatibility of t/08_idna.t to perl 5.8.
367 - make AnyEvent::Strict truly optional (in case somebody builds a
368 minimal perl...).
369
3705.28 Wed Oct 13 04:14:23 CEST 2010
371 - due to a glitch, AnyEvent's internal getprotobyname wasn't
372 actually used in tcp_server.
373 - implement AnyEvent::CondVar->new method, which was documented but
374 missing.
375 - support autoloading of read types in unshift_read.
376 - AnyEvent::Strict: do the test for invalid fh arguments differently,
377 resulting in more informative error messages.
378 - new AnyEvent::FAQ manpage.
379 - clarify that you can actually call push_read etc. in on_read.
380 - put keys and certs in /tmp, not into the current directory,
381 and document this in the SECURITY CONSIDERATIONS sections
382 of AnyEvent::TLS.
383 - support ";" as resolv.conf comment character in AnyEvent::DNS, as
384 requested by Ron Isaacson.
385 - document $AnyEvent::DNS::RESOLVER variable.
386 - incorporated major typo patches by Abhijit Menon-Sen.
387 - AnyEvent::Handle now croaks when negative timeouts are passed.
388 - add a shitload of TCP_xxx sockopt name constants to AnyEvent::Util,
389 undocumented, but who knows when they might come in handy.
390
3915.271 Tue Jun 8 12:05:46 CEST 2010
392 - backport to perl 5.8.x.
393
3945.27 Sun Jun 6 12:12:05 CEST 2010
395 - postpone differently in AnyEvent::Socket now, as
396 when not, canceling the connection attempt might fail
397 (found by Felix Antonius Wilhelm Ostmann).
398 - explicitly check for non-stream sockets in AE::Handle, too many
399 clueless people fell into the trap of this somehow working.
400 - simplified and reworked the "OTHER MODULES" section.
401 - better/more condvar examples.
402
4035.261 Wed Apr 28 16:13:36 CEST 2010
404 - AF_INET6 was not properly used from Socket6 during configuration
405 time (found by Andy Grundman).
406
4075.26 Mon Apr 12 04:49:35 CEST 2010
408 - don't generate spurious readyness notifications when select
409 returns EINTR in the pure perl backend (analysed and testcase
410 by Bas Denissen).
411 - give same examples in SYNOPSIS sections of AnyEvent and AE.
412 - provider faster implementations for AE::time/now/now_update for
413 EV and Perl backends.
414 - provide faster AE::cv and AE::time implementations for all
415 backends.
416 - no longer support lower/mixed-case signal names (this was
417 never documented nor universally supported).
418 - some more memory µ-optimisations, and somewhat less messy
419 code to cope with both AnyEvent and AE APIs.
420
4215.251 Sat Mar 13 00:58:21 CET 2010
422 - make test failed if it was run before make install on perl <=
423 5.8.8, and older versions of perl make our live very difficult,
424 so write constants.pl during Makefile.PL time.
425
4265.25 Sat Mar 13 00:23:14 CET 2010
427 - fix a race condition in AnyEvent::Handle that would cause
428 a "bio_read: unsupported method" error (in _another_
429 ssl connection) after user code threw an exception.
430 - added AnyEvent::Handle->destroyed method.
431 - speed up AnyEvent::Socket::format_address and ::format_ipv6.
432 - the AnyEvent::Util::fh_nonblocking function worked only
433 by lucky accident on win32.
434 - smaller and faster AnyEvent::Util::fh_nonblocking.
435 - when the (required!) Time::HiRes module is missing, AnyEvent
436 did not fall back to built-in time properly.
437 - do not load Fcntl at runtime, saving memory and loading time.
438 - precompile a number of constants and use them instead of runtime
439 detection and eval.
440 - free detection code after detection and similar memory optimisations.
441 - Perl backend timer interval best effort drift has been improved
442 (same algorithm as EV).
443 - update unicode idna mapping table.
444
4455.24 Tue Jan 5 11:39:43 CET 2010
446 - cygwin never reports errors from failed connects. 1.5 just gives
447 you continous readyness and EAGAIN, 1.7 is even more broken and
448 just hangs. work around both issues in a cygwin-specific hack.
449 - improve idle watcher documentation slightly.
450
4515.23 Sun Dec 20 23:48:00 CET 2009
452 - support IDNs in resolve_sockaddr, and therefore in tcp_connect.
453 - implement punycode_encode/decode, idn_nameprep,
454 idn_to_ascii and idn_to_unicode operations in AnyEvent::Util.
455 - provide $AE::VERSION.
456 - removed traces of "no strict 'refs'".
457
4585.22 Sat Dec 5 03:51:13 CET 2009
459 - downgrade-or-fail in AnyEvent::Handle::push_write, to
460 diagnose encoding failures earlier and more succinctly.
461 (this works around bugs in perl, throwing away encoding info
462 when passing scalar data to extensions).
463 - add more examples to AnyEvent::Socket manpage.
464 - upgrade internal warning set to the same as common::sense 2.03.
465 - use pack "n/a*" for pre-5.8.9 perl compatibility in AnyEvent::DNS
466 (John Beppu).
467 - AnyEvent::Socket::inet_aton now properly supports ipv6, as documented.
468 - add google public dns servers to fallback server set and make sure
469 we load-balance properly between the three sets. also add all
470 fallback dns servers, not just a random one, to each dns config.
471
4725.21 Thu Nov 19 02:48:47 CET 2009
473 - fix a problem where socket constants were called with parameters
474 (spotted by David Friedland).
475 - fork_call never use'd POSIX (reported by Daisuke Maki).
476 - improve perl 5.6 compatibility further (but it still won't work
477 unless you rip out everything but the core).
478 - prefer Net::DNS::Resolver over ipconfig, if installed, on win32.
479 uses 10MB of RAM, but doesn't open a console window. *sigh*.
480
4815.202 Wed Oct 14 22:35:44 CEST 2009
482 - AnyEvent::DNS would unexpectedly clobber $_ under windows
483 (analysed by Matthias Waldorf).
484 - AnyEvent::Handle::run_cmd can now pass the PID of the
485 newly-created process, which is much less useful than it might
486 sound (based on patch by Yann Kerherve).
487
4885.201 Tue Sep 29 12:09:25 CEST 2009
489 - AnyEvent:Handle::on_starttls/on_stoptls methods were broken
490 (reported by Torsten Foertsch).
491 - common::sense 2.0 could cause tcp_server to throw an exception
492 (analysed by elmex).
493
4945.2 Mon Sep 14 07:04:49 CEST 2009
495 - INCOMPATIBLE CHANGE: do no longer support register_read_type
496 and register_write_type in AnyEvent::Handle, instead support
497 package names (the facility was mostly abused).
498 - implement "packagename-as-read/write type" support in
499 AnyEvent::Handle.
500 - AnyEvent::Handle: new options "keepalive" and "oobinline".
501 - oobinline set by default to avoid security issues.
502 - the pure-perl event loop backend wrongly detected times() fallback
503 support (spotted by Pavel Boldin).
504
5055.12 Tue Sep 1 20:26:50 CEST 2009
506 - be more lenient when parsing resolv.conf files, as some people
507 use hashmarks as comment indicator inside directives (reported
508 by Michael S. Fischer).
509 - use same set of warnings as common::sense 2.0.
510 - fix a potential 32 bit overflow issue due to perl having problems
511 with large hex constants in the Perl backend.
512 - do not use the slower 5.6 version fo the enc_name code in AnyEvent::DNS
513 by default.
514 - fix some prototypes.
515 - (optionally) require Async::Interrupt 1.02.
516 - replace opendns fallbacks by verizon/level3 public dns servers.
517
5185.112 Fri Aug 21 13:59:04 CEST 2009
519 - AnyEvent::Handle could cause hard crashes in Net::SSLeay when
520 using starttls with an illegal/uncreatable tls context.
521 - port to broken win32 perls.
522
5235.111 Sun Aug 16 18:44:36 CEST 2009
524 - perl errornously requires sockaddr_un structures returned by the
525 kernel to have a specific length. work around this bug by padding
526 sockaddr_un structures (found to fail on OS X, as perl only uses
527 the correct code for GNU/Linux, but likely to fail on other
528 OSes as well).
529
5305.11 Wed Aug 12 17:49:37 CEST 2009
531 - fixed a bug in run_cmd with input from filename.
532 - create a json coder object when none is given in write direction
533 to reduce dependency on JSON.
534
5355.1 Tue Aug 11 03:17:32 CEST 2009
536 - tcp_connect now reports the correct error in $!, not an unrelated
537 one (was broken since 4.91).
538 - AE::Handle did not properly clear rtimeout/wtimeout watchers
539 when retrying a tcp_connect.
540 - new functions: AnyEvent::Util::run_cmd and close_all_fds_except.
541 - fixed wrong documentation in AE::Socket.
542
5435.01 Mon Aug 10 03:16:32 CEST 2009
544 - last release broke EV child watchers :/.
545
5465.0 Sun Aug 9 17:32:09 CEST 2009
547 - officially introduce and document the AE API.
548 - lowering the timeout at runtime did not have immediate effect
549 in AnyEvent::Handle.
550 - AE::Handle now has separate and independent read and write timeout
551 settings.
552 - probe for Guard module when AnyEvent::Util is loaded, not at
553 runtime (Event doesn't like eval "use" from callbacks, and it avoids
554 exporting the probe function).
555 - Impl::IOAsync's io watchers did not use the filehandle emulation
556 code to work around it's watcher limits.
557 - work around IO::Async's problems with overloaded objects.
558 - try to disable wearnings inside IO::Async, it's just too buggy.
559 - do no longer use Event by default, only EV or Perl, Event is too buggy.
560 - hackishly provide fast AE:: interface when EV or Perl is the backend.
561 - inherit VERSION from AnyEvent::VERSION in most modules.
562
5634.91 Thu Aug 6 15:42:45 CEST 2009
564 - AE::Handle::starttls could get out of sync when the read buffer
565 already contains some TLS handshake.
566 - AE::Handle did not properly free the TLS session, causing bigger
567 memory leaks in Net::SSLeay than expected (reported by toaster).
568 - AE::Socket::tcp_connect will now artificially delay invoking
569 the callback to avoid returning after invoking the callback.
570 - convert many internal watcher uses to AE API.
571
5724.9 Sat Aug 1 11:07:01 CEST 2009
573 - Glib silently fails when registering a timeout with
574 a negative value, so avoid that.
575 - call condvar callback immediately when it is set
576 after the condition is already signalled.
577 - check rbuf_max condition only after trying to consume
578 data.
579 - normalise signal numbers to names when using the ae's
580 signal handling, but do not document this (yet).
581 - pure perl signal emulation did not properly set nonblocking
582 mode on the signal pipe on win32 (this is unlikely to have caused
583 any issues).
584 - new module: AnyEvent::Debug.
585 - AnyEvent::Strict now checks that a signal was specified
586 by name (not name or number).
587 - reduce memory footprint in typical cases by ~50kb by
588 only compiling signal code when necessary.
589 - add AnyEvent::Handle->rbuf_max.
590 - grab the AE namespace for future new API, implement stubs for
591 said future/alternative API.
592 - new function: AnyEvent::Socket::format_hostport.
593
5944.881 Tue Jul 28 12:51:53 CEST 2009
595 - work around a bug in local in pre-5.10 perls, causing
596 AnyEvent::Handle to recurse when it should not
597 (analysed by elmex).
598
5994.88 Tue Jul 28 04:04:37 CEST 2009
600 - re-bless the handle into a dummy package after calling
601 AnyEvent::Handle::destroy, so the user does not need to check
602 for errors after every push_write etc.
603 - do not attempt to run t/02_signals.t on obviously broken
604 platforms.
605
6064.87 Sun Jul 26 02:06:16 CEST 2009
607 - do not attempt to linger when there is no longer a valid fh.
608 - enforce tls mode to be either accept or connect, do not simply
609 segfault in Net::SSLeay.
610 - AnyEvent::Handle can now call tcp_connect itself (new parameters
611 connect, on_prepare, on_connect and on_connect_error). Updated
612 tutorial accordingly.
613 - add AnyEvent::Impl::Irssi backend.
614
6154.86 Mon Jul 20 23:52:29 CEST 2009
616 - since the verbose warning is not enough, explicitly document
617 that versions before 1.33 of Net::SSLeay are not secure.
618 - work around signal handling races in Event and (...) Event::Lib.
619 - try to align signal-race timer to full-second boundaries.
620 - work around Tk not liking negative timeouts.
621 - don't complain of different grades of environmental unfriendlyness
622 in IO::Async.
623
6244.85 Sat Jul 18 06:16:14 CEST 2009
625 - nail the signal race problem in perl once and for all
626 (see $AnyEvent::MAX_SIGNAL_LATENCY).
627 - take advantage of Async::Interrupt if it is available.
628 - load Time::HiRes and Guard modules on demand only.
629 - add optional/recommended modules section to AnyEvent
630 documentation.
631 - reduce memory usage considerably (and reduce startup penalty)
632 by not using "strict", "warnings" and "overload" modules.
633 - work around buggy windows/openbsd perls and provide EBADMSG
634 and EPROTO ourselves when missing.
635 - improve perl 5.6 compatibility of the core event loop.
636 - made Net::SSLeay version 1.33 a soft requirement.
637
6384.83 Fri Jul 17 16:56:26 CEST 2009
639 - implement AnyEvent::Socket::getprotobyname.
640 - AnyEvent::CondVar's will now detect recursive blocking
641 waits and will croak, as too many people fall into
642 this trap.
643 - AnyEvent::Handle will now call ->destroy on itself after
644 executing the on_error callback, instead of doing some
645 half-baked internal shutdown, for fatal errors.
646 - clarify on_eof behaviour w.r.t. the read queue and
647 on_read callbacks.
648 - ignore some possible spurious wake-ups in tcp_connect.
649
6504.82 Sat Jul 11 00:34:55 CEST 2009
651 - POE and Event backends didn't accept some callable objects as
652 callbacks.
653 - use Config module instead of POSIX module to detect signal names
654 in AnyEvent::Strict and AnyEvent::Impl::EventLib, as the POSIX
655 module doesn't even have all POSIX signals :/.
656 - use more workarounds around the many refcnt/corruption bugs in
657 Event::Lib.
658 - work around a race condition in perl's select, causing t/03_child.t
659 to rarely fail.
660
6614.81 Thu Jul 9 10:30:30 CEST 2009
662 - AnyEvent::Handle didn't properly diagnose write errors
663 (it expected -1 from syswrite, how lame... :).
664 - support file descriptors in addition to file handles
665 in AnyEvent->io.
666 - new env variables: PERL_ANYEVENT_RESOLV_CONF,
667 PERL_ANYEVENT_MAX_OUTSTANDING_DNS, PERL_ANYEVENT_CA_FILE
668 and PERL_ANYEVENT_CA_PATH.
669 - provide a sensible synopsis section for AnyEvent::TLS.
670 - add a "supported backends" section to the manpage.
671 - added simple io watcher test to testsuite, using a
672 portable_socketpair.
673 - tried to improve the stability of the Event::Lib backend,
674 YMMV.
675
6764.8 Mon Jul 6 23:45:16 CEST 2009
677 - AnyEvent::DNS did not properly follow CNAME records with
678 uppercase targets.
679 - AnyEvent::DNS would errornously return AAAA records
680 with v4 mapped addresses (a faulty record) as ipv4 addresses,
681 causing AnyEvent::Socket to throw an exception.
682 - added new module AnyEvent::TLS for easier SSL/TLS context
683 creation, with many options including hostname verification,
684 secure default configuration, lots of documentation and,
685 predefined diffie-hellman keys for perfect forward security
686 and much more. get it while it's still fresh!
687 - use AnyEvent::TLS in AnyEvent::Handle for context management.
688 - load AnyEvent::Handle only on demand in AnyEvent::DNS,
689 so AnyEvent::Socket users have smaller memory footprint
690 in the common case.
691 - add AnyEvent::Handle->push_shutdown method.
692 - add an additional error message parameter to AnyEvent::Handle's
693 on_error callback (for TLS, $! is still available).
694 - add AnyEvent::Handle on_starttls/on_stoptls callbacks.
695 - make AnyEvent::Handle more robust against early connection
696 failures (during new), and return C<undef> in that case
697 from the constructor.
698 - AnyEvent::Handle will now try to load only JSON::XS first,
699 then fall back to JSON.
700 - format_ipv4/format_ipv6 are now exported by default, for symmetry,
701 and because it was documented that way.
702
7034.451 Fri Jul 3 00:28:58 CEST 2009
704 - do not clear rbuf when shutting down an AnyEvent::Handle
705 object - doing so breaks AnyEvent::HTTP.
706
7074.45 Mon Jun 29 22:59:26 CEST 2009
708 - a write error could cause AnyEvent::Handle to create
709 an I/O watcher with an undefined $fh.
710 - special-case mapped ipv4 addresses in both
711 AnyEvent::Socket::format_address and parse_address,
712 to treat them just like ipv4 addresses.
713 - updated and overhauled the AnyEvent::Intro doc.
714 - implement AnyEvent::Socket::format_ipv4/ipv6.
715 - slightly speed up portable_pipe/socketpair.
716 - expand condvar begin/end documentation.
717
7184.42 Fri Jun 26 08:32:18 CEST 2009
719 - preliminary and neccesarily incomplete support for IO::Async.
720 - reset SIGCHLD to DEFAULT when AnyEvent is loaded, in
721 case it was set to IGNORE, to ensure we can catch child statuses
722 even when the calling env acts stupidly.
723 - updated benchmarks with IO::Async, which performs very well.
724
7254.412 Wed Jun 24 01:35:57 CEST 2009
726 - support an "untaint" attribute for AnyEvent::DNS
727 and set it on the default resolver.
728 - implement and document AnyEvent::Impl::Perl::loop.
729 - remove all anyevent-env variables from %ENV
730 when running in tainted mode.
731 - mention and extend the IO::Lambda benchmark.
732
7334.411 Sun Jun 7 18:48:02 CEST 2009
734 - do not try to use F_SETFD on windows.
735
7364.41 Thu May 14 06:40:11 CEST 2009
737 - work around issues in older perls (5.8.0?) when
738 a signal handler is deleted from the %SIG hash.
739 - use POSIX::_exit in child test, to avoid running
740 destructors.
741 - speed up CHLD handling by relying on SIGCHLD being
742 synchronously delivered, even when we roll our own
743 implementation.
744 - AnyEvent::DNS: add the "dname" resource record name
745 for cosmetic reasons.
746
7474.4 Sun Apr 26 20:12:33 CEST 2009
748 - implemented idle watchers, where applicable.
749 - AnyEvent->time died when Event backend was in use.
750 - fix a memleak in the Tk backend.
751 - sped up Tk timer handling.
752 - clip negative "after" values to 0 in AnyEvent::Impl::Event
753 to avoid spamming out warning messages.
754 - fix Qt timers without interval.
755 - avoid Qt zero-timeout special case, in old libqt's.
756
7574.352 Mon Apr 20 16:31:11 CEST 2009
3 - fix AnyEvent::Strict error messages for child watchers. 758 - fix AnyEvent::Strict error messages for child watchers.
759 - fix/update Makefile.PL "recommends" versions.
760 - add AnyEvent->now_update.
4 761
54.351 Sat Apr 11 07:56:14 CEST 2009 7624.351 Sat Apr 11 07:56:14 CEST 2009
6 - actually make the signal pipe work under win32. 763 - actually make the signal pipe work under win32.
7 - localise $! in signal handler to avoid changing $!, although 764 - localise $! in signal handler to avoid changing $!, although
8 perl itself does not seem to save/restore errno either. 765 perl itself does not seem to save/restore errno either.
… …
674.3 Fri Oct 3 09:18:43 CEST 2008 8244.3 Fri Oct 3 09:18:43 CEST 2008
68 - AnyEvent will now install a no-op signal handler for SIGPIPE, 825 - AnyEvent will now install a no-op signal handler for SIGPIPE,
69 unless one has been installed already. 826 unless one has been installed already.
70 - warn about not loaidng AnyEvent::Impl::POE early enough 827 - warn about not loaidng AnyEvent::Impl::POE early enough
71 (patch by Adam Rosenstein). 828 (patch by Adam Rosenstein).
72 - fixed a great number of bugs and corner cases in AnyEvent::Handle: 829 - fixed a great number of bugs and corner cases in AnyEvent::Handle.
73 - fix a bug in where in SSL connect mode, the 830 - fix a bug in where in SSL connect mode, the
74 client would first wait for some data by the server and 831 client would first wait for some data by the server and
75 otherwise hang. (reported and analyzed in an absolutely 832 otherwise hang. (reported and analysed in an absolutely
76 exemplary manner by Adam Rosenstein). 833 exemplary manner by Adam Rosenstein).
77 - fix a bug in where SSL EOF would not be treated 834 - fix a bug in where SSL EOF would not be treated
78 as stream EOF, putting the connection into a hung state. 835 as stream EOF, putting the connection into a hung state.
79 (reported and analyzed in an absolutely exemplary manner by Adam 836 (reported and analysed in an absolutely exemplary manner by Adam
80 Rosenstein). 837 Rosenstein).
81 - fix a potential segfault when the TLS context 838 - fix a potential segfault when the TLS context
82 would go missing in a read callback (e.g. due to stoptls) - 839 would go missing in a read callback (e.g. due to stoptls) -
83 Net::SSLeay of course makes no type checking whatsoever. 840 Net::SSLeay of course makes no type checking whatsoever.
84 - AnyEvent::Handle will not stop the read watcher in TLS mode. 841 - AnyEvent::Handle will not stop the read watcher in TLS mode.

Diff Legend

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