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

Comparing AnyEvent/Changes (file contents):
Revision 1.207 by root, Thu Nov 6 16:16:44 2008 UTC vs.
Revision 1.626 by root, Sat Sep 2 00:21:32 2017 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)
12TODO: inet_aton maybe pack C4? no, add a comment why not
13TODO: socket prepare callbacks, can they easily signal failure as well?
14
15TODO: stuff dns packet into decoded structure
16TODO: allow client programs to extend the dns encoder/decoder list
17
18 - AnyEvent::Socket::parse_ipv6 would NOT, as advertised, accept ipv4
19 addresses. It now does and converts them to ipv4 mapped addresses.
20 - support CAA records, based on patch by Steve Atkins.
21
227.14 Sat Jun 24 01:44:19 CEST 2017
23 - fix a crash bug in AnyEvent::Handle with openssl 1.1.0 (patched
24 by Paul Howarth and apparently tracked down by Mike McCauley).
25 - AnyEvent::Handle->keepalive was documented (and defined) twice
26 (found by Matt Merhar).
27 - AnyEvent::Socket::tcp_bind/tcp_server would immediately unlink a unix
28 listening socket unless a guard is used. Change this so that no
29 cleanup will be performed unless a guard is used and document this more
30 clearly.
31 - make tcp_bind/tcp_server error messages more regular.
32 - work around latest perl upstream breakage - this time, perl5porters were
33 apparently bullied by a bogus CVE (CVE-2016-1238), and since distros
34 stupidly apply anything that has "security fix" stamped on it, it's likely
35 going to be a problem in practise (and working around it is trivial on
36 windows and unix).
37 - add TCP_FASTOPEN/MSG_FASTOPEN and MSG_NOSIGNAL/MSG_MORE/MSG_DONTWAIT
38 to constants.
39 - update warnings to common::sense 3.74 standards.
40
417.13 Sat Sep 17 04:31:49 CEST 2016
42 - Only call tlsext_host_name for non-empty common names (reported
43 by Maxime Soulé).
44 - log a (single) notice message if SNI is not supported.
45 - upgrade to UTS-46:9.0.0 draft and switch to non-transitional
46 behaviour, beating thunderbird, ie, edge, chrome and safari to it :)
47 (see also https://bugzilla.mozilla.org/show_bug.cgi?id=1218179)
48 - turns out the UTS-46 IDNA testcase failures were indeed bugs in the
49 testcases and the specification and not in the code - the post-9.0.0
50 unicode files have all known problems fixed, so finally the AnyEvent
51 IDNA implementation can pass the full IDNA testsuite - without needing
52 a single fix :)
53 - guarantee (and document) that condvar callbacks will be removed
54 on invocation - important to avoid circular references.
55
567.12 Wed Jan 27 19:12:26 CET 2016
57 - use common name as hostname for tls connects, if Net::SSLeay
58 supports SNI.
59 - fix documentation of tls_autostart read type in AnyEvent::Handle,
60 analyzed by Felix Ostmann.
61
627.11 Thu Jul 16 14:36:00 CEST 2015
63 - AnyEvent::Socket::parse_ipv6 could accept malformed ipv6
64 addresses (extra "::" at end and similar cases).
65 - add a more explicit warning to AnyEvent::Handle that it doesn't
66 work on files, people keep getting confused.
67 - new function AnyEvent::Socket::tcp_bind.
68 - new functions AnyEvent::fh_block and AnyEvent::fh_unblock.
69 - aligned ipv6 address formatting with RFC 5952 (by not shortening
70 a single :0: to ::).
71 - added stability canary support.
72
737.09 Sat May 2 16:38:53 CEST 2015
74 - AnyEvent::Debug called an internal function (AnyEvent::Log::ft)
75 that was renamed to AnyEvent::Log:format_time. uinder its old name
76 (analyzed by Michael Stovenour).
77 - update AnyEvent::DNS fallback resolver addresses:
78 seems google effectively killed most other free dns resolvers,
79 so remove them, but add cable and wireless (ecrc) since it was
80 stable for 20 years or so, official or not, and there should be
81 an alternative to google.
82 - perl5porters broke windows error codes in 5.20, and mapped
83 WSAEWOULDBLOCK on the (different) EWOULDBLOCK error code, and
84 WSAEINPROGRESS into the incompatible ERINPROGRESS code, probably
85 because they were so cool. They probably broke other error codes
86 for no reason, too, but I didn't care for checking, it's too
87 depressing. This version only works around the WSAEWOULDBLOCK
88 issue, because I don't have a nice way to work around the
89 WSAEINPROGRESS bug.
90
917.08 Wed Dec 10 05:27:17 CET 2014
92 - work around a newly introduced bug in Socket 2.011 (an
93 errornous sun_length check) (analyzed by Maxime Soulé).
94 - AnyEvent::TLS didn't load (but refer to) AnyEvent::Socket
95 (analyzed by Ben Magistro).
96 - AnyEvent::Strict will now confess, not croak. This is in line with
97 it being a development/debugging tool.
98 - work around a number of libglib bugs (debug builds of libglib enforce
99 certain undocumented behaviour patterns such as not being able to
100 remove a child watch source after it has fired, which we will try
101 to emulate to avoid "criticals". what where they thinking...).
102 - mention json security issues in AnyEvent::Handle, now that Douglas
103 Crockford has foolishly and incompatibly changed JSON.
104 - changed default dns resolver "max_outstanding" value from 1 to 10,
105 the latter beinfg the intended value all along
106 (reported by Ilya Chesnokov).
107 - added new "AnyEvent::Impl::UV" interface module to the UV event lib
108 (written by Mike Lowell).
109
1107.07 Tue Dec 17 17:45:02 CET 2013
111 - the documentation for custom tls verify schemes was wrong. make it agree
112 with the code (reported by Maxime Soulé).
113 - added cbor read and write types to AnyEvent::Handle (using CBOR::XS).
114 - work around an API change in openssl that could cause wrong tls connection
115 aborts, likely on windows only (analyzed by sten).
116 - calling AnyEvent->now_update with AnyEvent::Impl::Perl caused an
117 endless loop (reported by Dietrich Rebmann).
118 - add tlsv1_1 and tlsv1_2 protocols to AnyEvent::TLS
119 (patch by Maxime Soulé).
120 - document AnyEvent::Impl::IOAsync::set_loop and
121 $AnyEvent::Impl::IOAsync::LOOP. Though only documented now, this
122 functionality has _always_ been available.
123 - force a toplevel domain name in t/81_hosts.t (analyzed by
124 David Jack Wange Olrik).
125 - document that AnyEvent::Log uses AnyEvent::IO.
126 - warn about AnyEvent::Filesys::Notify performance.
127 - praise the joys of AnyEvent::Fork::*.
128 - time for an =encoding directive.
129 - do no longer use JSON to create a default json coder, use
130 JSON::XS or JSON::PP directly.
131
1327.05 Wed Aug 21 10:38:08 CEST 2013
133 - uts46data.pl couldn't be found due to wrong naming of the file
134 (reported by Fulko Hew).
135 - handle lone \015's properly in AE::Handle's default line read
136 (reported by various people).
137 - untaint ip addresses found in /etc/hosts (patch by José Micó).
138 - the memleak fix in 7.03 caused resolving via /etc/hosts to always
139 fail on first use (reported and testcase by Andrew Whatson).
140 - expose AnyEvent::Log::format_time, and allow users to redefine it.
141 - expose AnyEvent::Log::default_format, and allow redefinition.
142 - expose AnyEvent::Log::fatal_exit, to allow redefinition.
143 - AnyEvent::Debug shell can now run coro shell commands, if available.
144 - t/63* tests were wrongly in MANIFEST.
145 - kernel.org's finger server went MIA, switch to freebsd.org and
146 icculus.org.
147 - clarify that IO::AIO and AnyEvent::AIO are needed for AnyEvent::IO
148 to function asynchronously (reported by Daniel Carosone).
149 - hard-disable $^W in most tests, it generates too much garbage output.
150 - use a (hopefully) more future-proof method to emulate common::sense.
151 - upgrade to UTS-46:6.2.0.
152 - switch to INSTLIB from INSTLIBDIR, as INSTLIBDIR was wrongly
153 documented. should not affect anything.
154
1557.04 Thu Nov 15 02:23:58 CET 2012
156 - debugging code left in.
157
1587.03 Thu Nov 15 02:16:11 CET 2012
159 - AnyEvent::Socket::inet_aton did not work when dns resolution was
160 used to find the addresses (analyzed and patch by Maxime Soulé).
161 - fix a memory leak in the /etc/hosts lookup code when hosts don't
162 resolve and are not in hosts (analyzed by sten).
163
1647.02 Tue Aug 14 04:27:58 CEST 2012
165 - AnyEvent::Util::run_cmd could block indefinitely (analyzed and test
166 program by Yorhel).
167 - verified that AnyEvent::Socket follows RFC5952.
168 - try to parse "ADDR#PORT" in addition to "ADDR PORT".
169
1707.01 Sun May 13 01:03:17 CEST 2012
171 - fail with EPROTO in AnyEvent::Handle wqhen TLS is requested but
172 not available, instead of throwing an exception.
173 - use File::Spec to get the tmpdir in t/*, to avoid needless
174 failures on (most, not mine :) windows boxes.
175 - new handle read types: tls_detect and tls_autostart.
176
1777.0 Fri Apr 13 06:33:30 CEST 2012
178 - child watchers are broken in POE 1.352 (also many earlier
179 versions) and there seems to be no way to work around it, as POE
180 itself is inherently racy. Document this regression and add a delay
181 in t/68_poe_03_child.t for the time being.
182 - new module AnyEvent::IO, that is a frontend to either a pure-perl
183 synchronous I/O implementation (AnyEvent::IO::Perl), or to the
184 asynchronous I/O extension IO::AIO (AnyEvent::IO::IOAIO).
185 - load /etc/hosts only when DNS has no answer.
186 - stat /etc/hosts on every access and reload it if it changed.
187 - load /etc/hosts via AnyEvent::IO - potentially asynchronous.
188 - fix a buggy croak in the dh parameter reading in AnyEvent::TLS.
189 - AnyEvent::Log log_to_file and log_to_path now use AnyEvent::IO. As
190 a side effect, they now use true appending as opposed to libc
191 appending, and the file might not have been opened when the function
192 returns.
193 - the default logging level was not properly documented in a variety of
194 places, this has hopefully been rectified.
195 - updated uts46data.pl for unicode 6.1.0.
196 - made log messages generated by AnyEvent submodules not
197 include the package name anymore, as it will be logged
198 by default already.
199 - upgrade to the trick used by common-sense 3.5 to work
200 around extra warning torture/breakage under perl 5.15.x.
201 - log messages by anyevent are now ucfirst, are usually full
202 sentences and do no longer include the package name.
203 - the storable read type would sometimes throw an exception instead
204 of causing EBADMSG (analyzed by Maxime Soulé).
205 - work around a bug in openssl 1.0.1 which enforces a minimum rsa
206 keysize (reported by Paul Howarth).
207 - documented Rocco fabricating statements about AnyEvent and me.
208
2096.14 Tue Jan 31 20:00:24 CET 2012
210 - AnyEvent::Impl::Tk was broken due to a mysterious "die" inside,
211 probably an editing mistake (reported by Darin McBride).
212
2136.13 Thu Jan 12 07:27:01 CET 2012
214 - AnyEvent::Util::fork_call checked for POSIX availability in the wrong way,
215 causing it to fail if POSIX wasn't loaded already (analysed by Rock Power).
216 - AnyEvent::Handle::push_read (line => did pass $1 directly, so
217 regex-matching inside the callback would change the parameter, despite
218 $1 being dynamically scoped per-block. this perl bug is now being
219 worked around (testcase by Cindy Wang).
220
2216.12 Mon Dec 12 13:21:10 CET 2011
222 - $! was clobbered by subsequent calls in tcp_connect, due to
223 the postpone not saving/restoring it (analyzed by Richard Garnish).
224
2256.11 Tue Nov 22 10:36:05 CET 2011
226 - Tk cannot create windows when tainted unless you set a title, so
227 set a dummy title for AnyEvent's dummy mainwindow. How dummy.
228 - escape any nonprintable/nonascii characters when stringifying
229 backtraces.
230 - log the reason loading an interface module fails to load at level debug.
231
2326.1 Tue Oct 4 19:44:30 CEST 2011
233 - INCOMPATIBLE CHANGE: the default log level is now "4" (error
234 and above), and some messages inside AnyEvent have been elevated
235 to higher log levels to print by default.
236 - AnyEvent::log, unlike AnyEvent::Log::log, did not always exit on fatal.
237 - work around yet another signal race bug, newly introduced by POE. My
238 how that module sucks.
239 - add google public ipv6 dns as fallback servers, so AE::DNS can finally
240 support ipv6-only operation even with defaults (also use ::1 instead
241 of 127.0.0.1).
242 - give better guidelines on how to chose a logging level.
243 - AnyEvent::Log can now cap (limit) log message priority by context.
244 - AE::log will now save and restore $! and $@, so callers don't have to.
245 - AE::Strict now checks that registered fds are still valid at regular
246 intervals.
247 - skip frequently called 'require' calls if the module appears loaded.
248 - increase signal latency test timer considerably for poor little osx,
249 which can't be bothered with basic posix functionality such as signal
250 delivery.
251 - mention some debugging aids in the tutorial.
252 - new env variable PERL_ANYEVENT_MAX_SIGNAL_LATENCY.
253 - private (%namespace) log contexts in PERL_ANYEVENT_LOG now have the name
254 of the context as title.
255 - AnyEvent::Strict would not restore read-write access to $_ when a callback
256 died - this is a user bug, but since AE::Strict exists to help dbeugging...
257 - detect the extremely evil IO::Async::Loop::AnyEvent module and refuse
258 to work with it - the author is well aware of the issues but refuses
259 to fix them. This extreme step has been taken because the module actively
260 attacks the goals of AnyEvent and makes it harder for module authors to use
261 AnyEvent.
262
2636.02 Fri Aug 26 20:08:31 CEST 2011
264 - log_to_syslog now accepts facility strings.
265 - log_to_syslog would not allow facilities - "$facility|$level"
266 *literally* means that string. sheesh.
267 - the CPAN parser got confused and thought we provide Sys::Syslog, try
268 to work around.
269 - renamed AnyEvent::Impl::FLTK2 to FLTK.
270
2716.01 Fri Aug 26 07:04:11 CEST 2011
272 - INCOMPATIBLE CHANGE: AnyEvent::DNS resource records now include
273 the TTL as fourth element - this affects ->resolve, ->request
274 and ->dns_unpack indirectly. Convenience functions such as
275 AnyEvent::DNS::{a,txt,mx,srv} are not affected, neither is
276 AnyEvent::Socket.
277 - INCOMPATIBLE CHANGE: backend authors now should not implement
278 one_event or loop, but instead the AnyEvent::CondVar::_wait and _poll
279 methods.
280 - INCOMPATIBLE CHANGE: verbosity levels have been redefined, see
281 AnyEvent (PERL_ANYEVENT_VERBOSE) and AnyEvent::Log for details.
282 - DEPRECIATION: please change
283 "use AnyEvent::Impl::Perl" to "use AnyEvent::Loop" and
284 "AnyEvent::Impl::Perl::loop" to "AnyEvent::Loop::run"
285 in your programs, for future compatibility.
286 - *blush* idle watcher emulation was borked.
287 - AnyEvent::Strict now checks the AE::xxx functions as well.
288 - querying too long domain names would barf deep inside AnyEvent::DNS - since
289 this is hard to avoid for applications, AnyEvent:.DNS::request now checks for
290 for too long domain names explicitly and passes an undef to the callback
291 instead (reported by James Bromberger).
292 - make AnyEvent::Util more robust against EINTR (analyzed by Andrew Suffield).
293 - new functions: AnyEvent::postpone, AnyEvent::log.
294 - new module: AnyEvent::Log, for simple logging needs.
295 - new env variable, PERL_ANYEVENT_LOG, for fine-grained logging config.
296 - new env variable, PERL_ANYEVENT_DEBUG_SHELL, to automatically start
297 a debug shell.
298 - new env variable, PERL_ANYEVENT_DEBUG_WRAP, to automatically enable
299 watcher instrumentation/wrapping (see AnyEvent::Debug::wrap).
300 - new env variable, PERL_ANYEVENT_HOSTS, to override the place where
301 /etc/hosts can be found.
302 - an empty PERL_ANYEVENT_RESOLV_CONF now selects the default
303 configuration.
304 - env variables can now be specified using AE_xyz instead of
305 PERL_ANYEVENT_xyz, subject to some rules.
306 - lots of new AnyEvent::Debug functionality.
307 - AnyEvent::Strict now checks for modifications of $_ by the callback
308 (by setting it to readonly while executing the callback, if possible).
309 - IO::Async sometimes cancels the wrong timer when you call cancel_timer.
310 Invest extra overhead to work around this peculiar behaviour.
311 - split the pure perl backend into a loop module and an impl
312 module (AnyEvent::Loop and AnyEvent::Impl::Perl).
313 - some read types (regex, netstring, json, storable) could
314 errornously keep an AE::Handle object alive on their own
315 (reported by Mohammad Toossi).
316 - AnyEvent::Socket::resolve_sockaddr and all functions using it now
317 supports /etc/hosts overrides.
318 - add more workarounds around child watcher bugs in IO::Async
319 and (newly introduced) in POE.
320 - use glib child watchers and try to work around its limitations,
321 if possible (based on a patch by Kevin Ryde).
322 - add support for FLTK (fltk2), via AnyEvent::Impl::FLTK2.
323 - many read types in AnyEvent::Handle would malfunction during
324 transitions to TLS because of a "delete $self->{rbuf}".
325 (reported by Gerald Galster).
326 - improve condvar blocking wait performance (probably only noticable
327 with faster backends), also make it easier for event loops that
328 do not support blocking waits.
329 - PERL_ANYEVENT_MODEL now also allows full module names, although
330 the usefulness of this feature is not yet clear.
331 - add tests for all supported event loops, conditional on PERL_ANYEVENT_LOOP_TESTS.
332 - add t/80_ssltext.t.
333 - no longer use AUTOLOAD for the initial loading (for no very good reason).
334 - tcp_server now tries to unlink unix domain sockets when it gets destroyed.
335 - do not run event loop detection multiple times when programs call their cached
336 copies of AnyEvent::detect.
337 - parse_hostport now accepts absolute paths als unix domain sockets.
338 - suppress the idiotic warning inside POE by patching the POE::Kernel
339 run flags instead of asking the user to jump through contortions.
340 - add Task::Weaken as dependency (as recommends only at this time).
341 - replace PF_UNSPEC by 0, for good old minix.
342
3436.0 (not officially released, but was on CPAN for about 15 minutes :/)
344
3455.34 Thu May 12 09:59:41 CEST 2011
346 - AE::TLS didn't support both cert_file and key at the same time,
347 due to some copy&paste bug (reported and analyzed by vti).
348 - improve AE::Handle timeout documentation slightly.
349
3505.33 Mon Feb 7 21:51:41 CET 2011
351 - never appeared on CPAN, for unknown reasons.
352 - AnyEvent::Handle on_connect_error - do not try to destroy a handle
353 object when it's already gone (analyzed by Ryan Bullock).
354 - update idna algorithm(s) to UTS#46 version 6.0.0.
355 - preserve trailing dot in idn_to_ascii - testsuite says to preserve,
356 uts46 says it's an error, and testsuite makes more sense.
357 - remove leading dots in idn_to_ascii - testsuite says to remove,
358 uts46 says it's an error, and testsuite makes more sense.
359 - fixed a bug in idn ignored character processing.
360 - passes the uts46 6.0.0 testsuite except for three cases which
361 apparently are bugs in the testsuite itself.
362 - passes the uts46 6.0.1 testsuite except for eleven cases which
363 apparently are bugs in the testsuite itself.
364
3655.31 Sun Jan 23 11:49:19 CET 2011
366 - work around a POE bug where POE would first poll for new events and
367 invoke handlers BEFORE handling the message queue.
368 - fix parsing of "domain" statement in resolv.conf (Johannes Plunien).
369 - stop_read now does just that in TLS mode. see its
370 documentation for caveats.
371 - update to common::sense 3.4 warning set (some warnings in perl impose
372 an arbitrary 32kb limit on regex matches, which affects
373 AnyEvent::Handle).
374 - AE::Handle had a broken rbuf_max method and missing wbuf_max method.
375 - new experimental AE::Handle::resettls method.
376
3775.3 Fri Dec 31 05:48:13 CET 2010
378 - major IOAsync patch by Paul Evans - 0.33 is required, as the major
379 issues with IO::Async have been solved in that version, including
380 the availabilty of a default loop!
381 - implement new wbuf_max parameter for AnyEvent::Handle.
382 - added Cocoa::EventLoop backend by Daisuke Murase.
383 - the default testsuite now honors PERL_ANYEVENT_MODEL (Paul Evans).
384 - clarify handle on_prepare documentation.
385 - set CLOEXEC flag on parent-side fd's in AnyEvent::Util::run_cmd,
386 to avoid leaking them to any child processes.
387 - try to force files into the arch-dependend lib directory, for the
388 half-"turly obsessive compulsive person"(s) that depend on it.
389
3905.29 Sun Dec 5 10:49:21 CET 2010
391 - convert EV backend to EV 4.00 API (so better upgrade EV too).
392 - AE::Handle: implement dynamic read_size adjustment, add
393 max_read_size parameter, reduce default read_size to 2048.
394 - add some FAQ entries for servers and/or on_eof vs. on_error.
395 - work around OS bugs (cygwin again) when lingering in AE::Handle.
396 - reduce memory usage slightly when connect was used in AE::Handle.
397 - wrok around more segfaults in Tk.
398 - document tls_ctx => undef as valid.
399 - detect WNOHANG value at compiletime.
400 - include some probably linux-only support for building AnyEvent
401 as part of the perl core.
402 - improve compatibility of t/08_idna.t to perl 5.8.
403 - make AnyEvent::Strict truly optional (in case somebody builds a
404 minimal perl...).
405
4065.28 Wed Oct 13 04:14:23 CEST 2010
407 - due to a glitch, AnyEvent's internal getprotobyname wasn't
408 actually used in tcp_server.
409 - implement AnyEvent::CondVar->new method, which was documented but
410 missing.
411 - support autoloading of read types in unshift_read.
412 - AnyEvent::Strict: do the test for invalid fh arguments differently,
413 resulting in more informative error messages.
414 - new AnyEvent::FAQ manpage.
415 - clarify that you can actually call push_read etc. in on_read.
416 - put keys and certs in /tmp, not into the current directory,
417 and document this in the SECURITY CONSIDERATIONS sections
418 of AnyEvent::TLS.
419 - support ";" as resolv.conf comment character in AnyEvent::DNS, as
420 requested by Ron Isaacson.
421 - document $AnyEvent::DNS::RESOLVER variable.
422 - incorporated major typo patches by Abhijit Menon-Sen.
423 - AnyEvent::Handle now croaks when negative timeouts are passed.
424 - add a shitload of TCP_xxx sockopt name constants to AnyEvent::Util,
425 undocumented, but who knows when they might come in handy.
426
4275.271 Tue Jun 8 12:05:46 CEST 2010
428 - backport to perl 5.8.x.
429
4305.27 Sun Jun 6 12:12:05 CEST 2010
431 - postpone differently in AnyEvent::Socket now, as
432 when not, canceling the connection attempt might fail
433 (found by Felix Antonius Wilhelm Ostmann).
434 - explicitly check for non-stream sockets in AE::Handle, too many
435 clueless people fell into the trap of this somehow working.
436 - simplified and reworked the "OTHER MODULES" section.
437 - better/more condvar examples.
438
4395.261 Wed Apr 28 16:13:36 CEST 2010
440 - AF_INET6 was not properly used from Socket6 during configuration
441 time (found by Andy Grundman).
442
4435.26 Mon Apr 12 04:49:35 CEST 2010
444 - don't generate spurious readyness notifications when select
445 returns EINTR in the pure perl backend (analysed and testcase
446 by Bas Denissen).
447 - give same examples in SYNOPSIS sections of AnyEvent and AE.
448 - provider faster implementations for AE::time/now/now_update for
449 EV and Perl backends.
450 - provide faster AE::cv and AE::time implementations for all
451 backends.
452 - no longer support lower/mixed-case signal names (this was
453 never documented nor universally supported).
454 - some more memory µ-optimisations, and somewhat less messy
455 code to cope with both AnyEvent and AE APIs.
456
4575.251 Sat Mar 13 00:58:21 CET 2010
458 - make test failed if it was run before make install on perl <=
459 5.8.8, and older versions of perl make our live very difficult,
460 so write constants.pl during Makefile.PL time.
461
4625.25 Sat Mar 13 00:23:14 CET 2010
463 - fix a race condition in AnyEvent::Handle that would cause
464 a "bio_read: unsupported method" error (in _another_
465 ssl connection) after user code threw an exception.
466 - added AnyEvent::Handle->destroyed method.
467 - speed up AnyEvent::Socket::format_address and ::format_ipv6.
468 - the AnyEvent::Util::fh_nonblocking function worked only
469 by lucky accident on win32.
470 - smaller and faster AnyEvent::Util::fh_nonblocking.
471 - when the (required!) Time::HiRes module is missing, AnyEvent
472 did not fall back to built-in time properly.
473 - do not load Fcntl at runtime, saving memory and loading time.
474 - precompile a number of constants and use them instead of runtime
475 detection and eval.
476 - free detection code after detection and similar memory optimisations.
477 - Perl backend timer interval best effort drift has been improved
478 (same algorithm as EV).
479 - update unicode idna mapping table.
480
4815.24 Tue Jan 5 11:39:43 CET 2010
482 - cygwin never reports errors from failed connects. 1.5 just gives
483 you continous readyness and EAGAIN, 1.7 is even more broken and
484 just hangs. work around both issues in a cygwin-specific hack.
485 - improve idle watcher documentation slightly.
486
4875.23 Sun Dec 20 23:48:00 CET 2009
488 - support IDNs in resolve_sockaddr, and therefore in tcp_connect.
489 - implement punycode_encode/decode, idn_nameprep,
490 idn_to_ascii and idn_to_unicode operations in AnyEvent::Util.
491 - provide $AE::VERSION.
492 - removed traces of "no strict 'refs'".
493
4945.22 Sat Dec 5 03:51:13 CET 2009
495 - downgrade-or-fail in AnyEvent::Handle::push_write, to
496 diagnose encoding failures earlier and more succinctly.
497 (this works around bugs in perl, throwing away encoding info
498 when passing scalar data to extensions).
499 - add more examples to AnyEvent::Socket manpage.
500 - upgrade internal warning set to the same as common::sense 2.03.
501 - use pack "n/a*" for pre-5.8.9 perl compatibility in AnyEvent::DNS
502 (John Beppu).
503 - AnyEvent::Socket::inet_aton now properly supports ipv6, as documented.
504 - add google public dns servers to fallback server set and make sure
505 we load-balance properly between the three sets. also add all
506 fallback dns servers, not just a random one, to each dns config.
507
5085.21 Thu Nov 19 02:48:47 CET 2009
509 - fix a problem where socket constants were called with parameters
510 (spotted by David Friedland).
511 - fork_call never use'd POSIX (reported by Daisuke Maki).
512 - improve perl 5.6 compatibility further (but it still won't work
513 unless you rip out everything but the core).
514 - prefer Net::DNS::Resolver over ipconfig, if installed, on win32.
515 uses 10MB of RAM, but doesn't open a console window. *sigh*.
516
5175.202 Wed Oct 14 22:35:44 CEST 2009
518 - AnyEvent::DNS would unexpectedly clobber $_ under windows
519 (analysed by Matthias Waldorf).
520 - AnyEvent::Handle::run_cmd can now pass the PID of the
521 newly-created process, which is much less useful than it might
522 sound (based on patch by Yann Kerherve).
523
5245.201 Tue Sep 29 12:09:25 CEST 2009
525 - AnyEvent:Handle::on_starttls/on_stoptls methods were broken
526 (reported by Torsten Foertsch).
527 - common::sense 2.0 could cause tcp_server to throw an exception
528 (analysed by elmex).
529
5305.2 Mon Sep 14 07:04:49 CEST 2009
531 - INCOMPATIBLE CHANGE: do no longer support register_read_type
532 and register_write_type in AnyEvent::Handle, instead support
533 package names (the facility was mostly abused).
534 - implement "packagename-as-read/write type" support in
535 AnyEvent::Handle.
536 - AnyEvent::Handle: new options "keepalive" and "oobinline".
537 - oobinline set by default to avoid security issues.
538 - the pure-perl event loop backend wrongly detected times() fallback
539 support (spotted by Pavel Boldin).
540
5415.12 Tue Sep 1 20:26:50 CEST 2009
542 - be more lenient when parsing resolv.conf files, as some people
543 use hashmarks as comment indicator inside directives (reported
544 by Michael S. Fischer).
545 - use same set of warnings as common::sense 2.0.
546 - fix a potential 32 bit overflow issue due to perl having problems
547 with large hex constants in the Perl backend.
548 - do not use the slower 5.6 version fo the enc_name code in AnyEvent::DNS
549 by default.
550 - fix some prototypes.
551 - (optionally) require Async::Interrupt 1.02.
552 - replace opendns fallbacks by verizon/level3 public dns servers.
553
5545.112 Fri Aug 21 13:59:04 CEST 2009
555 - AnyEvent::Handle could cause hard crashes in Net::SSLeay when
556 using starttls with an illegal/uncreatable tls context.
557 - port to broken win32 perls.
558
5595.111 Sun Aug 16 18:44:36 CEST 2009
560 - perl errornously requires sockaddr_un structures returned by the
561 kernel to have a specific length. work around this bug by padding
562 sockaddr_un structures (found to fail on OS X, as perl only uses
563 the correct code for GNU/Linux, but likely to fail on other
564 OSes as well).
565
5665.11 Wed Aug 12 17:49:37 CEST 2009
567 - fixed a bug in run_cmd with input from filename.
568 - create a json coder object when none is given in write direction
569 to reduce dependency on JSON.
570
5715.1 Tue Aug 11 03:17:32 CEST 2009
572 - tcp_connect now reports the correct error in $!, not an unrelated
573 one (was broken since 4.91).
574 - AE::Handle did not properly clear rtimeout/wtimeout watchers
575 when retrying a tcp_connect.
576 - new functions: AnyEvent::Util::run_cmd and close_all_fds_except.
577 - fixed wrong documentation in AE::Socket.
578
5795.01 Mon Aug 10 03:16:32 CEST 2009
580 - last release broke EV child watchers :/.
581
5825.0 Sun Aug 9 17:32:09 CEST 2009
583 - officially introduce and document the AE API.
584 - lowering the timeout at runtime did not have immediate effect
585 in AnyEvent::Handle.
586 - AE::Handle now has separate and independent read and write timeout
587 settings.
588 - probe for Guard module when AnyEvent::Util is loaded, not at
589 runtime (Event doesn't like eval "use" from callbacks, and it avoids
590 exporting the probe function).
591 - Impl::IOAsync's io watchers did not use the filehandle emulation
592 code to work around it's watcher limits.
593 - work around IO::Async's problems with overloaded objects.
594 - try to disable wearnings inside IO::Async, it's just too buggy.
595 - do no longer use Event by default, only EV or Perl, Event is too buggy.
596 - hackishly provide fast AE:: interface when EV or Perl is the backend.
597 - inherit VERSION from AnyEvent::VERSION in most modules.
598
5994.91 Thu Aug 6 15:42:45 CEST 2009
600 - AE::Handle::starttls could get out of sync when the read buffer
601 already contains some TLS handshake.
602 - AE::Handle did not properly free the TLS session, causing bigger
603 memory leaks in Net::SSLeay than expected (reported by toaster).
604 - AE::Socket::tcp_connect will now artificially delay invoking
605 the callback to avoid returning after invoking the callback.
606 - convert many internal watcher uses to AE API.
607
6084.9 Sat Aug 1 11:07:01 CEST 2009
609 - Glib silently fails when registering a timeout with
610 a negative value, so avoid that.
611 - call condvar callback immediately when it is set
612 after the condition is already signalled.
613 - check rbuf_max condition only after trying to consume
614 data.
615 - normalise signal numbers to names when using the ae's
616 signal handling, but do not document this (yet).
617 - pure perl signal emulation did not properly set nonblocking
618 mode on the signal pipe on win32 (this is unlikely to have caused
619 any issues).
620 - new module: AnyEvent::Debug.
621 - AnyEvent::Strict now checks that a signal was specified
622 by name (not name or number).
623 - reduce memory footprint in typical cases by ~50kb by
624 only compiling signal code when necessary.
625 - add AnyEvent::Handle->rbuf_max.
626 - grab the AE namespace for future new API, implement stubs for
627 said future/alternative API.
628 - new function: AnyEvent::Socket::format_hostport.
629
6304.881 Tue Jul 28 12:51:53 CEST 2009
631 - work around a bug in local in pre-5.10 perls, causing
632 AnyEvent::Handle to recurse when it should not
633 (analysed by elmex).
634
6354.88 Tue Jul 28 04:04:37 CEST 2009
636 - re-bless the handle into a dummy package after calling
637 AnyEvent::Handle::destroy, so the user does not need to check
638 for errors after every push_write etc.
639 - do not attempt to run t/02_signals.t on obviously broken
640 platforms.
641
6424.87 Sun Jul 26 02:06:16 CEST 2009
643 - do not attempt to linger when there is no longer a valid fh.
644 - enforce tls mode to be either accept or connect, do not simply
645 segfault in Net::SSLeay.
646 - AnyEvent::Handle can now call tcp_connect itself (new parameters
647 connect, on_prepare, on_connect and on_connect_error). Updated
648 tutorial accordingly.
649 - add AnyEvent::Impl::Irssi backend.
650
6514.86 Mon Jul 20 23:52:29 CEST 2009
652 - since the verbose warning is not enough, explicitly document
653 that versions before 1.33 of Net::SSLeay are not secure.
654 - work around signal handling races in Event and (...) Event::Lib.
655 - try to align signal-race timer to full-second boundaries.
656 - work around Tk not liking negative timeouts.
657 - don't complain of different grades of environmental unfriendlyness
658 in IO::Async.
659
6604.85 Sat Jul 18 06:16:14 CEST 2009
661 - nail the signal race problem in perl once and for all
662 (see $AnyEvent::MAX_SIGNAL_LATENCY).
663 - take advantage of Async::Interrupt if it is available.
664 - load Time::HiRes and Guard modules on demand only.
665 - add optional/recommended modules section to AnyEvent
666 documentation.
667 - reduce memory usage considerably (and reduce startup penalty)
668 by not using "strict", "warnings" and "overload" modules.
669 - work around buggy windows/openbsd perls and provide EBADMSG
670 and EPROTO ourselves when missing.
671 - improve perl 5.6 compatibility of the core event loop.
672 - made Net::SSLeay version 1.33 a soft requirement.
673
6744.83 Fri Jul 17 16:56:26 CEST 2009
675 - implement AnyEvent::Socket::getprotobyname.
676 - AnyEvent::CondVar's will now detect recursive blocking
677 waits and will croak, as too many people fall into
678 this trap.
679 - AnyEvent::Handle will now call ->destroy on itself after
680 executing the on_error callback, instead of doing some
681 half-baked internal shutdown, for fatal errors.
682 - clarify on_eof behaviour w.r.t. the read queue and
683 on_read callbacks.
684 - ignore some possible spurious wake-ups in tcp_connect.
685
6864.82 Sat Jul 11 00:34:55 CEST 2009
687 - POE and Event backends didn't accept some callable objects as
688 callbacks.
689 - use Config module instead of POSIX module to detect signal names
690 in AnyEvent::Strict and AnyEvent::Impl::EventLib, as the POSIX
691 module doesn't even have all POSIX signals :/.
692 - use more workarounds around the many refcnt/corruption bugs in
693 Event::Lib.
694 - work around a race condition in perl's select, causing t/03_child.t
695 to rarely fail.
696
6974.81 Thu Jul 9 10:30:30 CEST 2009
698 - AnyEvent::Handle didn't properly diagnose write errors
699 (it expected -1 from syswrite, how lame... :).
700 - support file descriptors in addition to file handles
701 in AnyEvent->io.
702 - new env variables: PERL_ANYEVENT_RESOLV_CONF,
703 PERL_ANYEVENT_MAX_OUTSTANDING_DNS, PERL_ANYEVENT_CA_FILE
704 and PERL_ANYEVENT_CA_PATH.
705 - provide a sensible synopsis section for AnyEvent::TLS.
706 - add a "supported backends" section to the manpage.
707 - added simple io watcher test to testsuite, using a
708 portable_socketpair.
709 - tried to improve the stability of the Event::Lib backend,
710 YMMV.
711
7124.8 Mon Jul 6 23:45:16 CEST 2009
713 - AnyEvent::DNS did not properly follow CNAME records with
714 uppercase targets.
715 - AnyEvent::DNS would errornously return AAAA records
716 with v4 mapped addresses (a faulty record) as ipv4 addresses,
717 causing AnyEvent::Socket to throw an exception.
718 - added new module AnyEvent::TLS for easier SSL/TLS context
719 creation, with many options including hostname verification,
720 secure default configuration, lots of documentation and,
721 predefined diffie-hellman keys for perfect forward security
722 and much more. get it while it's still fresh!
723 - use AnyEvent::TLS in AnyEvent::Handle for context management.
724 - load AnyEvent::Handle only on demand in AnyEvent::DNS,
725 so AnyEvent::Socket users have smaller memory footprint
726 in the common case.
727 - add AnyEvent::Handle->push_shutdown method.
728 - add an additional error message parameter to AnyEvent::Handle's
729 on_error callback (for TLS, $! is still available).
730 - add AnyEvent::Handle on_starttls/on_stoptls callbacks.
731 - make AnyEvent::Handle more robust against early connection
732 failures (during new), and return C<undef> in that case
733 from the constructor.
734 - AnyEvent::Handle will now try to load only JSON::XS first,
735 then fall back to JSON.
736 - format_ipv4/format_ipv6 are now exported by default, for symmetry,
737 and because it was documented that way.
738
7394.451 Fri Jul 3 00:28:58 CEST 2009
740 - do not clear rbuf when shutting down an AnyEvent::Handle
741 object - doing so breaks AnyEvent::HTTP.
742
7434.45 Mon Jun 29 22:59:26 CEST 2009
744 - a write error could cause AnyEvent::Handle to create
745 an I/O watcher with an undefined $fh.
746 - special-case mapped ipv4 addresses in both
747 AnyEvent::Socket::format_address and parse_address,
748 to treat them just like ipv4 addresses.
749 - updated and overhauled the AnyEvent::Intro doc.
750 - implement AnyEvent::Socket::format_ipv4/ipv6.
751 - slightly speed up portable_pipe/socketpair.
752 - expand condvar begin/end documentation.
753
7544.42 Fri Jun 26 08:32:18 CEST 2009
755 - preliminary and neccesarily incomplete support for IO::Async.
756 - reset SIGCHLD to DEFAULT when AnyEvent is loaded, in
757 case it was set to IGNORE, to ensure we can catch child statuses
758 even when the calling env acts stupidly.
759 - updated benchmarks with IO::Async, which performs very well.
760
7614.412 Wed Jun 24 01:35:57 CEST 2009
762 - support an "untaint" attribute for AnyEvent::DNS
763 and set it on the default resolver.
764 - implement and document AnyEvent::Impl::Perl::loop.
765 - remove all anyevent-env variables from %ENV
766 when running in tainted mode.
767 - mention and extend the IO::Lambda benchmark.
768
7694.411 Sun Jun 7 18:48:02 CEST 2009
770 - do not try to use F_SETFD on windows.
771
7724.41 Thu May 14 06:40:11 CEST 2009
773 - work around issues in older perls (5.8.0?) when
774 a signal handler is deleted from the %SIG hash.
775 - use POSIX::_exit in child test, to avoid running
776 destructors.
777 - speed up CHLD handling by relying on SIGCHLD being
778 synchronously delivered, even when we roll our own
779 implementation.
780 - AnyEvent::DNS: add the "dname" resource record name
781 for cosmetic reasons.
782
7834.4 Sun Apr 26 20:12:33 CEST 2009
784 - implemented idle watchers, where applicable.
785 - AnyEvent->time died when Event backend was in use.
786 - fix a memleak in the Tk backend.
787 - sped up Tk timer handling.
788 - clip negative "after" values to 0 in AnyEvent::Impl::Event
789 to avoid spamming out warning messages.
790 - fix Qt timers without interval.
791 - avoid Qt zero-timeout special case, in old libqt's.
792
7934.352 Mon Apr 20 16:31:11 CEST 2009
794 - fix AnyEvent::Strict error messages for child watchers.
795 - fix/update Makefile.PL "recommends" versions.
796 - add AnyEvent->now_update.
797
7984.351 Sat Apr 11 07:56:14 CEST 2009
799 - actually make the signal pipe work under win32.
800 - localise $! in signal handler to avoid changing $!, although
801 perl itself does not seem to save/restore errno either.
802 - set the cloexec flag on the signal pipe (normally set by perl too).
803
8044.35 Fri Mar 27 11:48:20 CET 2009
805 - event models relying on AnyEvent's signal watcher emulation did
806 invoke the callback asynchronously, contradicting documentation
807 and causing signals to get lost (this includes AnyEvent's own
808 event loop). AnyEvent now uses the standard pipe trick to make
809 callback execution synchronous to the event loop.
810 - AnyEvent::Handle didn't free TLS context data on DESTROY
811 (patch by Pavel Shaydo).
812 - work around the ever-incompatibly-changing API of MakeMaker.
813 - document that changing global variables without restoring them
814 is a bad idea in Perl (noted by Adam Rosenstein).
815 - AnyEvent::Strict now barfs if ->io is passed a file.
816
8174.34 Thu Feb 12 18:32:45 CET 2009
818 - separately buffer TLS read data, as otherwise the read queue could
819 deadlock as receiving data is not expected while draining the
820 read queue (which cna only happen with TLS).
821 - raise EBADMSG error on JSON decoding errors.
822 - fix some minor manpage bugs (reported by Maximilian Gaß).
823 - speed up select bitmask parsing quite a bit in the pure perl
824 backend.
825 - use CORE::select instead of just select in AnyEvent::Impl::Perl.
826
8274.331 Tue Jan 6 21:07:25 CET 2009
828 - socketpair fails on many vista machines because vista has
829 completely broken accept/getpeername and getsockname functions,
830 so we provide our own socketpair emulation that kind of works
831 (AnyEvent::Util::portable_pipe).
832 - new function: AnyEvent::Util::portable_socketpair.
833 - take advantage of the Guard module if it exists.
834
8354.33 Fri Nov 21 02:35:40 CET 2008
836 - AnyEvent::Strict did errornously flag a fileno of 0 as
837 illegal.
3 - reduce memory usage and slightly speed up the pure perl backend 838 - reduce memory usage and slightly speed up the pure perl backend
4 by only storing the file descriptor, not the file handle. 839 by only storing the file descriptor, not the file handle.
5 - add missing autocork method to AnyEvent::Handle (reported 840 - add missing autocork method to AnyEvent::Handle (reported
6 by Adam Rosenstein). 841 by Adam Rosenstein).
842 - AnyEvent::DNS->resolve errornously documented an $rcode result
843 argument, but there isn't (spotted by Henrik Krohns).
844 - the naked truth about Tk - it's basically unmaintained.
7 845
84.32 Mon Nov 3 22:46:32 CET 2008 8464.32 Mon Nov 3 22:46:32 CET 2008
9 - fix AnyEvent::Socket::resolve_sockaddr to properly support 847 - fix AnyEvent::Socket::resolve_sockaddr to properly support
10 unix sockets again and choose a proper default for 848 unix sockets again and choose a proper default for
11 the socket type. 849 the socket type.
… …
224.3 Fri Oct 3 09:18:43 CEST 2008 8604.3 Fri Oct 3 09:18:43 CEST 2008
23 - AnyEvent will now install a no-op signal handler for SIGPIPE, 861 - AnyEvent will now install a no-op signal handler for SIGPIPE,
24 unless one has been installed already. 862 unless one has been installed already.
25 - warn about not loaidng AnyEvent::Impl::POE early enough 863 - warn about not loaidng AnyEvent::Impl::POE early enough
26 (patch by Adam Rosenstein). 864 (patch by Adam Rosenstein).
27 - fixed a great number of bugs and corner cases in AnyEvent::Handle: 865 - fixed a great number of bugs and corner cases in AnyEvent::Handle.
28 - fix a bug in where in SSL connect mode, the 866 - fix a bug in where in SSL connect mode, the
29 client would first wait for some data by the server and 867 client would first wait for some data by the server and
30 otherwise hang. (reported and analyzed in an absolutely 868 otherwise hang. (reported and analysed in an absolutely
31 exemplary manner by Adam Rosenstein). 869 exemplary manner by Adam Rosenstein).
32 - fix a bug in where SSL EOF would not be treated 870 - fix a bug in where SSL EOF would not be treated
33 as stream EOF, putting the connection into a hung state. 871 as stream EOF, putting the connection into a hung state.
34 (reported and analyzed in an absolutely exemplary manner by Adam 872 (reported and analysed in an absolutely exemplary manner by Adam
35 Rosenstein). 873 Rosenstein).
36 - fix a potential segfault when the TLS context 874 - fix a potential segfault when the TLS context
37 would go missing in a read callback (e.g. due to stoptls) - 875 would go missing in a read callback (e.g. due to stoptls) -
38 Net::SSLeay of course makes no type checking whatsoever. 876 Net::SSLeay of course makes no type checking whatsoever.
39 - AnyEvent::Handle will not stop the read watcher in TLS mode. 877 - AnyEvent::Handle will not stop the read watcher in TLS mode.

Diff Legend

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