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

Comparing AnyEvent/Changes (file contents):
Revision 1.295 by root, Wed Jul 22 05:37:32 2009 UTC vs.
Revision 1.630 by root, Tue Feb 26 01:35:48 2019 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
17TODO: Олег Г <verdrehung@gmail.com> test nameserver
18
19 - INCOMPATIBLE CHANGE: tls_detect documentation gave separate major and
20 minor versions, while code passed only a single value. This version
21 follows the documentation and now passes separate major and minor values.
22 - work around Net::SSLeay not having been ported to openssl 1.1, but many
23 distributions compiling it against openssl 1.1, which unfortunately
24 succeeds and results in a very brfoken module.
25 - AnyEvent::Socket::parse_ipv6 would NOT, as advertised, accept ipv4
26 addresses. It now does and converts them to ipv4 mapped addresses.
27 - support CAA records, based on patch by Steve Atkins.
28 - add freenom and cloudflare nameservers as dns fallback.
29 - AnyEvent::Strict would not properly ward against io watchers
30 on files when the handle passed was a file descriptor.
31
327.14 Sat Jun 24 01:44:19 CEST 2017
33 - fix a crash bug in AnyEvent::Handle with openssl 1.1.0 (patched
34 by Paul Howarth and apparently tracked down by Mike McCauley).
35 - AnyEvent::Handle->keepalive was documented (and defined) twice
36 (found by Matt Merhar).
37 - AnyEvent::Socket::tcp_bind/tcp_server would immediately unlink a unix
38 listening socket unless a guard is used. Change this so that no
39 cleanup will be performed unless a guard is used and document this more
40 clearly.
41 - make tcp_bind/tcp_server error messages more regular.
42 - work around latest perl upstream breakage - this time, perl5porters were
43 apparently bullied by a bogus CVE (CVE-2016-1238), and since distros
44 stupidly apply anything that has "security fix" stamped on it, it's likely
45 going to be a problem in practise (and working around it is trivial on
46 windows and unix).
47 - add TCP_FASTOPEN/MSG_FASTOPEN and MSG_NOSIGNAL/MSG_MORE/MSG_DONTWAIT
48 to constants.
49 - update warnings to common::sense 3.74 standards.
50
517.13 Sat Sep 17 04:31:49 CEST 2016
52 - Only call tlsext_host_name for non-empty common names (reported
53 by Maxime Soulé).
54 - log a (single) notice message if SNI is not supported.
55 - upgrade to UTS-46:9.0.0 draft and switch to non-transitional
56 behaviour, beating thunderbird, ie, edge, chrome and safari to it :)
57 (see also https://bugzilla.mozilla.org/show_bug.cgi?id=1218179)
58 - turns out the UTS-46 IDNA testcase failures were indeed bugs in the
59 testcases and the specification and not in the code - the post-9.0.0
60 unicode files have all known problems fixed, so finally the AnyEvent
61 IDNA implementation can pass the full IDNA testsuite - without needing
62 a single fix :)
63 - guarantee (and document) that condvar callbacks will be removed
64 on invocation - important to avoid circular references.
65
667.12 Wed Jan 27 19:12:26 CET 2016
67 - use common name as hostname for tls connects, if Net::SSLeay
68 supports SNI.
69 - fix documentation of tls_autostart read type in AnyEvent::Handle,
70 analyzed by Felix Ostmann.
71
727.11 Thu Jul 16 14:36:00 CEST 2015
73 - AnyEvent::Socket::parse_ipv6 could accept malformed ipv6
74 addresses (extra "::" at end and similar cases).
75 - add a more explicit warning to AnyEvent::Handle that it doesn't
76 work on files, people keep getting confused.
77 - new function AnyEvent::Socket::tcp_bind.
78 - new functions AnyEvent::fh_block and AnyEvent::fh_unblock.
79 - aligned ipv6 address formatting with RFC 5952 (by not shortening
80 a single :0: to ::).
81 - added stability canary support.
82
837.09 Sat May 2 16:38:53 CEST 2015
84 - AnyEvent::Debug called an internal function (AnyEvent::Log::ft)
85 that was renamed to AnyEvent::Log:format_time. uinder its old name
86 (analyzed by Michael Stovenour).
87 - update AnyEvent::DNS fallback resolver addresses:
88 seems google effectively killed most other free dns resolvers,
89 so remove them, but add cable and wireless (ecrc) since it was
90 stable for 20 years or so, official or not, and there should be
91 an alternative to google.
92 - perl5porters broke windows error codes in 5.20, and mapped
93 WSAEWOULDBLOCK on the (different) EWOULDBLOCK error code, and
94 WSAEINPROGRESS into the incompatible ERINPROGRESS code, probably
95 because they were so cool. They probably broke other error codes
96 for no reason, too, but I didn't care for checking, it's too
97 depressing. This version only works around the WSAEWOULDBLOCK
98 issue, because I don't have a nice way to work around the
99 WSAEINPROGRESS bug.
100
1017.08 Wed Dec 10 05:27:17 CET 2014
102 - work around a newly introduced bug in Socket 2.011 (an
103 errornous sun_length check) (analyzed by Maxime Soulé).
104 - AnyEvent::TLS didn't load (but refer to) AnyEvent::Socket
105 (analyzed by Ben Magistro).
106 - AnyEvent::Strict will now confess, not croak. This is in line with
107 it being a development/debugging tool.
108 - work around a number of libglib bugs (debug builds of libglib enforce
109 certain undocumented behaviour patterns such as not being able to
110 remove a child watch source after it has fired, which we will try
111 to emulate to avoid "criticals". what where they thinking...).
112 - mention json security issues in AnyEvent::Handle, now that Douglas
113 Crockford has foolishly and incompatibly changed JSON.
114 - changed default dns resolver "max_outstanding" value from 1 to 10,
115 the latter beinfg the intended value all along
116 (reported by Ilya Chesnokov).
117 - added new "AnyEvent::Impl::UV" interface module to the UV event lib
118 (written by Mike Lowell).
119
1207.07 Tue Dec 17 17:45:02 CET 2013
121 - the documentation for custom tls verify schemes was wrong. make it agree
122 with the code (reported by Maxime Soulé).
123 - added cbor read and write types to AnyEvent::Handle (using CBOR::XS).
124 - work around an API change in openssl that could cause wrong tls connection
125 aborts, likely on windows only (analyzed by sten).
126 - calling AnyEvent->now_update with AnyEvent::Impl::Perl caused an
127 endless loop (reported by Dietrich Rebmann).
128 - add tlsv1_1 and tlsv1_2 protocols to AnyEvent::TLS
129 (patch by Maxime Soulé).
130 - document AnyEvent::Impl::IOAsync::set_loop and
131 $AnyEvent::Impl::IOAsync::LOOP. Though only documented now, this
132 functionality has _always_ been available.
133 - force a toplevel domain name in t/81_hosts.t (analyzed by
134 David Jack Wange Olrik).
135 - document that AnyEvent::Log uses AnyEvent::IO.
136 - warn about AnyEvent::Filesys::Notify performance.
137 - praise the joys of AnyEvent::Fork::*.
138 - time for an =encoding directive.
139 - do no longer use JSON to create a default json coder, use
140 JSON::XS or JSON::PP directly.
141
1427.05 Wed Aug 21 10:38:08 CEST 2013
143 - uts46data.pl couldn't be found due to wrong naming of the file
144 (reported by Fulko Hew).
145 - handle lone \015's properly in AE::Handle's default line read
146 (reported by various people).
147 - untaint ip addresses found in /etc/hosts (patch by José Micó).
148 - the memleak fix in 7.03 caused resolving via /etc/hosts to always
149 fail on first use (reported and testcase by Andrew Whatson).
150 - expose AnyEvent::Log::format_time, and allow users to redefine it.
151 - expose AnyEvent::Log::default_format, and allow redefinition.
152 - expose AnyEvent::Log::fatal_exit, to allow redefinition.
153 - AnyEvent::Debug shell can now run coro shell commands, if available.
154 - t/63* tests were wrongly in MANIFEST.
155 - kernel.org's finger server went MIA, switch to freebsd.org and
156 icculus.org.
157 - clarify that IO::AIO and AnyEvent::AIO are needed for AnyEvent::IO
158 to function asynchronously (reported by Daniel Carosone).
159 - hard-disable $^W in most tests, it generates too much garbage output.
160 - use a (hopefully) more future-proof method to emulate common::sense.
161 - upgrade to UTS-46:6.2.0.
162 - switch to INSTLIB from INSTLIBDIR, as INSTLIBDIR was wrongly
163 documented. should not affect anything.
164
1657.04 Thu Nov 15 02:23:58 CET 2012
166 - debugging code left in.
167
1687.03 Thu Nov 15 02:16:11 CET 2012
169 - AnyEvent::Socket::inet_aton did not work when dns resolution was
170 used to find the addresses (analyzed and patch by Maxime Soulé).
171 - fix a memory leak in the /etc/hosts lookup code when hosts don't
172 resolve and are not in hosts (analyzed by sten).
173
1747.02 Tue Aug 14 04:27:58 CEST 2012
175 - AnyEvent::Util::run_cmd could block indefinitely (analyzed and test
176 program by Yorhel).
177 - verified that AnyEvent::Socket follows RFC5952.
178 - try to parse "ADDR#PORT" in addition to "ADDR PORT".
179
1807.01 Sun May 13 01:03:17 CEST 2012
181 - fail with EPROTO in AnyEvent::Handle wqhen TLS is requested but
182 not available, instead of throwing an exception.
183 - use File::Spec to get the tmpdir in t/*, to avoid needless
184 failures on (most, not mine :) windows boxes.
185 - new handle read types: tls_detect and tls_autostart.
186
1877.0 Fri Apr 13 06:33:30 CEST 2012
188 - child watchers are broken in POE 1.352 (also many earlier
189 versions) and there seems to be no way to work around it, as POE
190 itself is inherently racy. Document this regression and add a delay
191 in t/68_poe_03_child.t for the time being.
192 - new module AnyEvent::IO, that is a frontend to either a pure-perl
193 synchronous I/O implementation (AnyEvent::IO::Perl), or to the
194 asynchronous I/O extension IO::AIO (AnyEvent::IO::IOAIO).
195 - load /etc/hosts only when DNS has no answer.
196 - stat /etc/hosts on every access and reload it if it changed.
197 - load /etc/hosts via AnyEvent::IO - potentially asynchronous.
198 - fix a buggy croak in the dh parameter reading in AnyEvent::TLS.
199 - AnyEvent::Log log_to_file and log_to_path now use AnyEvent::IO. As
200 a side effect, they now use true appending as opposed to libc
201 appending, and the file might not have been opened when the function
202 returns.
203 - the default logging level was not properly documented in a variety of
204 places, this has hopefully been rectified.
205 - updated uts46data.pl for unicode 6.1.0.
206 - made log messages generated by AnyEvent submodules not
207 include the package name anymore, as it will be logged
208 by default already.
209 - upgrade to the trick used by common-sense 3.5 to work
210 around extra warning torture/breakage under perl 5.15.x.
211 - log messages by anyevent are now ucfirst, are usually full
212 sentences and do no longer include the package name.
213 - the storable read type would sometimes throw an exception instead
214 of causing EBADMSG (analyzed by Maxime Soulé).
215 - work around a bug in openssl 1.0.1 which enforces a minimum rsa
216 keysize (reported by Paul Howarth).
217 - documented Rocco fabricating statements about AnyEvent and me.
218
2196.14 Tue Jan 31 20:00:24 CET 2012
220 - AnyEvent::Impl::Tk was broken due to a mysterious "die" inside,
221 probably an editing mistake (reported by Darin McBride).
222
2236.13 Thu Jan 12 07:27:01 CET 2012
224 - AnyEvent::Util::fork_call checked for POSIX availability in the wrong way,
225 causing it to fail if POSIX wasn't loaded already (analysed by Rock Power).
226 - AnyEvent::Handle::push_read (line => did pass $1 directly, so
227 regex-matching inside the callback would change the parameter, despite
228 $1 being dynamically scoped per-block. this perl bug is now being
229 worked around (testcase by Cindy Wang).
230
2316.12 Mon Dec 12 13:21:10 CET 2011
232 - $! was clobbered by subsequent calls in tcp_connect, due to
233 the postpone not saving/restoring it (analyzed by Richard Garnish).
234
2356.11 Tue Nov 22 10:36:05 CET 2011
236 - Tk cannot create windows when tainted unless you set a title, so
237 set a dummy title for AnyEvent's dummy mainwindow. How dummy.
238 - escape any nonprintable/nonascii characters when stringifying
239 backtraces.
240 - log the reason loading an interface module fails to load at level debug.
241
2426.1 Tue Oct 4 19:44:30 CEST 2011
243 - INCOMPATIBLE CHANGE: the default log level is now "4" (error
244 and above), and some messages inside AnyEvent have been elevated
245 to higher log levels to print by default.
246 - AnyEvent::log, unlike AnyEvent::Log::log, did not always exit on fatal.
247 - work around yet another signal race bug, newly introduced by POE. My
248 how that module sucks.
249 - add google public ipv6 dns as fallback servers, so AE::DNS can finally
250 support ipv6-only operation even with defaults (also use ::1 instead
251 of 127.0.0.1).
252 - give better guidelines on how to chose a logging level.
253 - AnyEvent::Log can now cap (limit) log message priority by context.
254 - AE::log will now save and restore $! and $@, so callers don't have to.
255 - AE::Strict now checks that registered fds are still valid at regular
256 intervals.
257 - skip frequently called 'require' calls if the module appears loaded.
258 - increase signal latency test timer considerably for poor little osx,
259 which can't be bothered with basic posix functionality such as signal
260 delivery.
261 - mention some debugging aids in the tutorial.
262 - new env variable PERL_ANYEVENT_MAX_SIGNAL_LATENCY.
263 - private (%namespace) log contexts in PERL_ANYEVENT_LOG now have the name
264 of the context as title.
265 - AnyEvent::Strict would not restore read-write access to $_ when a callback
266 died - this is a user bug, but since AE::Strict exists to help dbeugging...
267 - detect the extremely evil IO::Async::Loop::AnyEvent module and refuse
268 to work with it - the author is well aware of the issues but refuses
269 to fix them. This extreme step has been taken because the module actively
270 attacks the goals of AnyEvent and makes it harder for module authors to use
271 AnyEvent.
272
2736.02 Fri Aug 26 20:08:31 CEST 2011
274 - log_to_syslog now accepts facility strings.
275 - log_to_syslog would not allow facilities - "$facility|$level"
276 *literally* means that string. sheesh.
277 - the CPAN parser got confused and thought we provide Sys::Syslog, try
278 to work around.
279 - renamed AnyEvent::Impl::FLTK2 to FLTK.
280
2816.01 Fri Aug 26 07:04:11 CEST 2011
282 - INCOMPATIBLE CHANGE: AnyEvent::DNS resource records now include
283 the TTL as fourth element - this affects ->resolve, ->request
284 and ->dns_unpack indirectly. Convenience functions such as
285 AnyEvent::DNS::{a,txt,mx,srv} are not affected, neither is
286 AnyEvent::Socket.
287 - INCOMPATIBLE CHANGE: backend authors now should not implement
288 one_event or loop, but instead the AnyEvent::CondVar::_wait and _poll
289 methods.
290 - INCOMPATIBLE CHANGE: verbosity levels have been redefined, see
291 AnyEvent (PERL_ANYEVENT_VERBOSE) and AnyEvent::Log for details.
292 - DEPRECIATION: please change
293 "use AnyEvent::Impl::Perl" to "use AnyEvent::Loop" and
294 "AnyEvent::Impl::Perl::loop" to "AnyEvent::Loop::run"
295 in your programs, for future compatibility.
296 - *blush* idle watcher emulation was borked.
297 - AnyEvent::Strict now checks the AE::xxx functions as well.
298 - querying too long domain names would barf deep inside AnyEvent::DNS - since
299 this is hard to avoid for applications, AnyEvent:.DNS::request now checks for
300 for too long domain names explicitly and passes an undef to the callback
301 instead (reported by James Bromberger).
302 - make AnyEvent::Util more robust against EINTR (analyzed by Andrew Suffield).
303 - new functions: AnyEvent::postpone, AnyEvent::log.
304 - new module: AnyEvent::Log, for simple logging needs.
305 - new env variable, PERL_ANYEVENT_LOG, for fine-grained logging config.
306 - new env variable, PERL_ANYEVENT_DEBUG_SHELL, to automatically start
307 a debug shell.
308 - new env variable, PERL_ANYEVENT_DEBUG_WRAP, to automatically enable
309 watcher instrumentation/wrapping (see AnyEvent::Debug::wrap).
310 - new env variable, PERL_ANYEVENT_HOSTS, to override the place where
311 /etc/hosts can be found.
312 - an empty PERL_ANYEVENT_RESOLV_CONF now selects the default
313 configuration.
314 - env variables can now be specified using AE_xyz instead of
315 PERL_ANYEVENT_xyz, subject to some rules.
316 - lots of new AnyEvent::Debug functionality.
317 - AnyEvent::Strict now checks for modifications of $_ by the callback
318 (by setting it to readonly while executing the callback, if possible).
319 - IO::Async sometimes cancels the wrong timer when you call cancel_timer.
320 Invest extra overhead to work around this peculiar behaviour.
321 - split the pure perl backend into a loop module and an impl
322 module (AnyEvent::Loop and AnyEvent::Impl::Perl).
323 - some read types (regex, netstring, json, storable) could
324 errornously keep an AE::Handle object alive on their own
325 (reported by Mohammad Toossi).
326 - AnyEvent::Socket::resolve_sockaddr and all functions using it now
327 supports /etc/hosts overrides.
328 - add more workarounds around child watcher bugs in IO::Async
329 and (newly introduced) in POE.
330 - use glib child watchers and try to work around its limitations,
331 if possible (based on a patch by Kevin Ryde).
332 - add support for FLTK (fltk2), via AnyEvent::Impl::FLTK2.
333 - many read types in AnyEvent::Handle would malfunction during
334 transitions to TLS because of a "delete $self->{rbuf}".
335 (reported by Gerald Galster).
336 - improve condvar blocking wait performance (probably only noticable
337 with faster backends), also make it easier for event loops that
338 do not support blocking waits.
339 - PERL_ANYEVENT_MODEL now also allows full module names, although
340 the usefulness of this feature is not yet clear.
341 - add tests for all supported event loops, conditional on PERL_ANYEVENT_LOOP_TESTS.
342 - add t/80_ssltext.t.
343 - no longer use AUTOLOAD for the initial loading (for no very good reason).
344 - tcp_server now tries to unlink unix domain sockets when it gets destroyed.
345 - do not run event loop detection multiple times when programs call their cached
346 copies of AnyEvent::detect.
347 - parse_hostport now accepts absolute paths als unix domain sockets.
348 - suppress the idiotic warning inside POE by patching the POE::Kernel
349 run flags instead of asking the user to jump through contortions.
350 - add Task::Weaken as dependency (as recommends only at this time).
351 - replace PF_UNSPEC by 0, for good old minix.
352
3536.0 (not officially released, but was on CPAN for about 15 minutes :/)
354
3555.34 Thu May 12 09:59:41 CEST 2011
356 - AE::TLS didn't support both cert_file and key at the same time,
357 due to some copy&paste bug (reported and analyzed by vti).
358 - improve AE::Handle timeout documentation slightly.
359
3605.33 Mon Feb 7 21:51:41 CET 2011
361 - never appeared on CPAN, for unknown reasons.
362 - AnyEvent::Handle on_connect_error - do not try to destroy a handle
363 object when it's already gone (analyzed by Ryan Bullock).
364 - update idna algorithm(s) to UTS#46 version 6.0.0.
365 - preserve trailing dot in idn_to_ascii - testsuite says to preserve,
366 uts46 says it's an error, and testsuite makes more sense.
367 - remove leading dots in idn_to_ascii - testsuite says to remove,
368 uts46 says it's an error, and testsuite makes more sense.
369 - fixed a bug in idn ignored character processing.
370 - passes the uts46 6.0.0 testsuite except for three cases which
371 apparently are bugs in the testsuite itself.
372 - passes the uts46 6.0.1 testsuite except for eleven cases which
373 apparently are bugs in the testsuite itself.
374
3755.31 Sun Jan 23 11:49:19 CET 2011
376 - work around a POE bug where POE would first poll for new events and
377 invoke handlers BEFORE handling the message queue.
378 - fix parsing of "domain" statement in resolv.conf (Johannes Plunien).
379 - stop_read now does just that in TLS mode. see its
380 documentation for caveats.
381 - update to common::sense 3.4 warning set (some warnings in perl impose
382 an arbitrary 32kb limit on regex matches, which affects
383 AnyEvent::Handle).
384 - AE::Handle had a broken rbuf_max method and missing wbuf_max method.
385 - new experimental AE::Handle::resettls method.
386
3875.3 Fri Dec 31 05:48:13 CET 2010
388 - major IOAsync patch by Paul Evans - 0.33 is required, as the major
389 issues with IO::Async have been solved in that version, including
390 the availabilty of a default loop!
391 - implement new wbuf_max parameter for AnyEvent::Handle.
392 - added Cocoa::EventLoop backend by Daisuke Murase.
393 - the default testsuite now honors PERL_ANYEVENT_MODEL (Paul Evans).
394 - clarify handle on_prepare documentation.
395 - set CLOEXEC flag on parent-side fd's in AnyEvent::Util::run_cmd,
396 to avoid leaking them to any child processes.
397 - try to force files into the arch-dependend lib directory, for the
398 half-"turly obsessive compulsive person"(s) that depend on it.
399
4005.29 Sun Dec 5 10:49:21 CET 2010
401 - convert EV backend to EV 4.00 API (so better upgrade EV too).
402 - AE::Handle: implement dynamic read_size adjustment, add
403 max_read_size parameter, reduce default read_size to 2048.
404 - add some FAQ entries for servers and/or on_eof vs. on_error.
405 - work around OS bugs (cygwin again) when lingering in AE::Handle.
406 - reduce memory usage slightly when connect was used in AE::Handle.
407 - wrok around more segfaults in Tk.
408 - document tls_ctx => undef as valid.
409 - detect WNOHANG value at compiletime.
410 - include some probably linux-only support for building AnyEvent
411 as part of the perl core.
412 - improve compatibility of t/08_idna.t to perl 5.8.
413 - make AnyEvent::Strict truly optional (in case somebody builds a
414 minimal perl...).
415
4165.28 Wed Oct 13 04:14:23 CEST 2010
417 - due to a glitch, AnyEvent's internal getprotobyname wasn't
418 actually used in tcp_server.
419 - implement AnyEvent::CondVar->new method, which was documented but
420 missing.
421 - support autoloading of read types in unshift_read.
422 - AnyEvent::Strict: do the test for invalid fh arguments differently,
423 resulting in more informative error messages.
424 - new AnyEvent::FAQ manpage.
425 - clarify that you can actually call push_read etc. in on_read.
426 - put keys and certs in /tmp, not into the current directory,
427 and document this in the SECURITY CONSIDERATIONS sections
428 of AnyEvent::TLS.
429 - support ";" as resolv.conf comment character in AnyEvent::DNS, as
430 requested by Ron Isaacson.
431 - document $AnyEvent::DNS::RESOLVER variable.
432 - incorporated major typo patches by Abhijit Menon-Sen.
433 - AnyEvent::Handle now croaks when negative timeouts are passed.
434 - add a shitload of TCP_xxx sockopt name constants to AnyEvent::Util,
435 undocumented, but who knows when they might come in handy.
436
4375.271 Tue Jun 8 12:05:46 CEST 2010
438 - backport to perl 5.8.x.
439
4405.27 Sun Jun 6 12:12:05 CEST 2010
441 - postpone differently in AnyEvent::Socket now, as
442 when not, canceling the connection attempt might fail
443 (found by Felix Antonius Wilhelm Ostmann).
444 - explicitly check for non-stream sockets in AE::Handle, too many
445 clueless people fell into the trap of this somehow working.
446 - simplified and reworked the "OTHER MODULES" section.
447 - better/more condvar examples.
448
4495.261 Wed Apr 28 16:13:36 CEST 2010
450 - AF_INET6 was not properly used from Socket6 during configuration
451 time (found by Andy Grundman).
452
4535.26 Mon Apr 12 04:49:35 CEST 2010
454 - don't generate spurious readyness notifications when select
455 returns EINTR in the pure perl backend (analysed and testcase
456 by Bas Denissen).
457 - give same examples in SYNOPSIS sections of AnyEvent and AE.
458 - provider faster implementations for AE::time/now/now_update for
459 EV and Perl backends.
460 - provide faster AE::cv and AE::time implementations for all
461 backends.
462 - no longer support lower/mixed-case signal names (this was
463 never documented nor universally supported).
464 - some more memory µ-optimisations, and somewhat less messy
465 code to cope with both AnyEvent and AE APIs.
466
4675.251 Sat Mar 13 00:58:21 CET 2010
468 - make test failed if it was run before make install on perl <=
469 5.8.8, and older versions of perl make our live very difficult,
470 so write constants.pl during Makefile.PL time.
471
4725.25 Sat Mar 13 00:23:14 CET 2010
473 - fix a race condition in AnyEvent::Handle that would cause
474 a "bio_read: unsupported method" error (in _another_
475 ssl connection) after user code threw an exception.
476 - added AnyEvent::Handle->destroyed method.
477 - speed up AnyEvent::Socket::format_address and ::format_ipv6.
478 - the AnyEvent::Util::fh_nonblocking function worked only
479 by lucky accident on win32.
480 - smaller and faster AnyEvent::Util::fh_nonblocking.
481 - when the (required!) Time::HiRes module is missing, AnyEvent
482 did not fall back to built-in time properly.
483 - do not load Fcntl at runtime, saving memory and loading time.
484 - precompile a number of constants and use them instead of runtime
485 detection and eval.
486 - free detection code after detection and similar memory optimisations.
487 - Perl backend timer interval best effort drift has been improved
488 (same algorithm as EV).
489 - update unicode idna mapping table.
490
4915.24 Tue Jan 5 11:39:43 CET 2010
492 - cygwin never reports errors from failed connects. 1.5 just gives
493 you continous readyness and EAGAIN, 1.7 is even more broken and
494 just hangs. work around both issues in a cygwin-specific hack.
495 - improve idle watcher documentation slightly.
496
4975.23 Sun Dec 20 23:48:00 CET 2009
498 - support IDNs in resolve_sockaddr, and therefore in tcp_connect.
499 - implement punycode_encode/decode, idn_nameprep,
500 idn_to_ascii and idn_to_unicode operations in AnyEvent::Util.
501 - provide $AE::VERSION.
502 - removed traces of "no strict 'refs'".
503
5045.22 Sat Dec 5 03:51:13 CET 2009
505 - downgrade-or-fail in AnyEvent::Handle::push_write, to
506 diagnose encoding failures earlier and more succinctly.
507 (this works around bugs in perl, throwing away encoding info
508 when passing scalar data to extensions).
509 - add more examples to AnyEvent::Socket manpage.
510 - upgrade internal warning set to the same as common::sense 2.03.
511 - use pack "n/a*" for pre-5.8.9 perl compatibility in AnyEvent::DNS
512 (John Beppu).
513 - AnyEvent::Socket::inet_aton now properly supports ipv6, as documented.
514 - add google public dns servers to fallback server set and make sure
515 we load-balance properly between the three sets. also add all
516 fallback dns servers, not just a random one, to each dns config.
517
5185.21 Thu Nov 19 02:48:47 CET 2009
519 - fix a problem where socket constants were called with parameters
520 (spotted by David Friedland).
521 - fork_call never use'd POSIX (reported by Daisuke Maki).
522 - improve perl 5.6 compatibility further (but it still won't work
523 unless you rip out everything but the core).
524 - prefer Net::DNS::Resolver over ipconfig, if installed, on win32.
525 uses 10MB of RAM, but doesn't open a console window. *sigh*.
526
5275.202 Wed Oct 14 22:35:44 CEST 2009
528 - AnyEvent::DNS would unexpectedly clobber $_ under windows
529 (analysed by Matthias Waldorf).
530 - AnyEvent::Handle::run_cmd can now pass the PID of the
531 newly-created process, which is much less useful than it might
532 sound (based on patch by Yann Kerherve).
533
5345.201 Tue Sep 29 12:09:25 CEST 2009
535 - AnyEvent:Handle::on_starttls/on_stoptls methods were broken
536 (reported by Torsten Foertsch).
537 - common::sense 2.0 could cause tcp_server to throw an exception
538 (analysed by elmex).
539
5405.2 Mon Sep 14 07:04:49 CEST 2009
541 - INCOMPATIBLE CHANGE: do no longer support register_read_type
542 and register_write_type in AnyEvent::Handle, instead support
543 package names (the facility was mostly abused).
544 - implement "packagename-as-read/write type" support in
545 AnyEvent::Handle.
546 - AnyEvent::Handle: new options "keepalive" and "oobinline".
547 - oobinline set by default to avoid security issues.
548 - the pure-perl event loop backend wrongly detected times() fallback
549 support (spotted by Pavel Boldin).
550
5515.12 Tue Sep 1 20:26:50 CEST 2009
552 - be more lenient when parsing resolv.conf files, as some people
553 use hashmarks as comment indicator inside directives (reported
554 by Michael S. Fischer).
555 - use same set of warnings as common::sense 2.0.
556 - fix a potential 32 bit overflow issue due to perl having problems
557 with large hex constants in the Perl backend.
558 - do not use the slower 5.6 version fo the enc_name code in AnyEvent::DNS
559 by default.
560 - fix some prototypes.
561 - (optionally) require Async::Interrupt 1.02.
562 - replace opendns fallbacks by verizon/level3 public dns servers.
563
5645.112 Fri Aug 21 13:59:04 CEST 2009
565 - AnyEvent::Handle could cause hard crashes in Net::SSLeay when
566 using starttls with an illegal/uncreatable tls context.
567 - port to broken win32 perls.
568
5695.111 Sun Aug 16 18:44:36 CEST 2009
570 - perl errornously requires sockaddr_un structures returned by the
571 kernel to have a specific length. work around this bug by padding
572 sockaddr_un structures (found to fail on OS X, as perl only uses
573 the correct code for GNU/Linux, but likely to fail on other
574 OSes as well).
575
5765.11 Wed Aug 12 17:49:37 CEST 2009
577 - fixed a bug in run_cmd with input from filename.
578 - create a json coder object when none is given in write direction
579 to reduce dependency on JSON.
580
5815.1 Tue Aug 11 03:17:32 CEST 2009
582 - tcp_connect now reports the correct error in $!, not an unrelated
583 one (was broken since 4.91).
584 - AE::Handle did not properly clear rtimeout/wtimeout watchers
585 when retrying a tcp_connect.
586 - new functions: AnyEvent::Util::run_cmd and close_all_fds_except.
587 - fixed wrong documentation in AE::Socket.
588
5895.01 Mon Aug 10 03:16:32 CEST 2009
590 - last release broke EV child watchers :/.
591
5925.0 Sun Aug 9 17:32:09 CEST 2009
593 - officially introduce and document the AE API.
594 - lowering the timeout at runtime did not have immediate effect
595 in AnyEvent::Handle.
596 - AE::Handle now has separate and independent read and write timeout
597 settings.
598 - probe for Guard module when AnyEvent::Util is loaded, not at
599 runtime (Event doesn't like eval "use" from callbacks, and it avoids
600 exporting the probe function).
601 - Impl::IOAsync's io watchers did not use the filehandle emulation
602 code to work around it's watcher limits.
603 - work around IO::Async's problems with overloaded objects.
604 - try to disable wearnings inside IO::Async, it's just too buggy.
605 - do no longer use Event by default, only EV or Perl, Event is too buggy.
606 - hackishly provide fast AE:: interface when EV or Perl is the backend.
607 - inherit VERSION from AnyEvent::VERSION in most modules.
608
6094.91 Thu Aug 6 15:42:45 CEST 2009
610 - AE::Handle::starttls could get out of sync when the read buffer
611 already contains some TLS handshake.
612 - AE::Handle did not properly free the TLS session, causing bigger
613 memory leaks in Net::SSLeay than expected (reported by toaster).
614 - AE::Socket::tcp_connect will now artificially delay invoking
615 the callback to avoid returning after invoking the callback.
616 - convert many internal watcher uses to AE API.
617
6184.9 Sat Aug 1 11:07:01 CEST 2009
619 - Glib silently fails when registering a timeout with
620 a negative value, so avoid that.
621 - call condvar callback immediately when it is set
622 after the condition is already signalled.
623 - check rbuf_max condition only after trying to consume
624 data.
625 - normalise signal numbers to names when using the ae's
626 signal handling, but do not document this (yet).
627 - pure perl signal emulation did not properly set nonblocking
628 mode on the signal pipe on win32 (this is unlikely to have caused
629 any issues).
630 - new module: AnyEvent::Debug.
631 - AnyEvent::Strict now checks that a signal was specified
632 by name (not name or number).
633 - reduce memory footprint in typical cases by ~50kb by
634 only compiling signal code when necessary.
635 - add AnyEvent::Handle->rbuf_max.
636 - grab the AE namespace for future new API, implement stubs for
637 said future/alternative API.
638 - new function: AnyEvent::Socket::format_hostport.
639
6404.881 Tue Jul 28 12:51:53 CEST 2009
641 - work around a bug in local in pre-5.10 perls, causing
642 AnyEvent::Handle to recurse when it should not
643 (analysed by elmex).
644
6454.88 Tue Jul 28 04:04:37 CEST 2009
646 - re-bless the handle into a dummy package after calling
647 AnyEvent::Handle::destroy, so the user does not need to check
648 for errors after every push_write etc.
649 - do not attempt to run t/02_signals.t on obviously broken
650 platforms.
651
6524.87 Sun Jul 26 02:06:16 CEST 2009
3 - do not attempt to linger when there is no longer a valid fh. 653 - do not attempt to linger when there is no longer a valid fh.
654 - enforce tls mode to be either accept or connect, do not simply
655 segfault in Net::SSLeay.
656 - AnyEvent::Handle can now call tcp_connect itself (new parameters
657 connect, on_prepare, on_connect and on_connect_error). Updated
658 tutorial accordingly.
659 - add AnyEvent::Impl::Irssi backend.
4 660
54.86 Mon Jul 20 23:52:29 CEST 2009 6614.86 Mon Jul 20 23:52:29 CEST 2009
6 - since the verbose warning is not enough, explicitly document 662 - since the verbose warning is not enough, explicitly document
7 that versions before 1.33 of Net::SSLeay are not secure. 663 that versions before 1.33 of Net::SSLeay are not secure.
8 - work around signal handling races in Event and (...) Event::Lib. 664 - work around signal handling races in Event and (...) Event::Lib.
9 - try to align signal-race timer to full-second boundaries. 665 - try to align signal-race timer to full-second boundaries.
10 - work around Tk not liking negative timeouts. 666 - work around Tk not liking negative timeouts.
11 - don't complain of different grades of environmental unfriendlyness 667 - don't complain of different grades of environmental unfriendlyness
12 in IO::ASync. 668 in IO::Async.
13 669
144.85 Sat Jul 18 06:16:14 CEST 2009 6704.85 Sat Jul 18 06:16:14 CEST 2009
15 - nail the signal race problem in perl once and for all 671 - nail the signal race problem in perl once and for all
16 (see $AnyEvent::MAX_SIGNAL_LATENCY). 672 (see $AnyEvent::MAX_SIGNAL_LATENCY).
17 - take advantage of Async::Interrupt if it is available. 673 - take advantage of Async::Interrupt if it is available.
… …
80 in the common case. 736 in the common case.
81 - add AnyEvent::Handle->push_shutdown method. 737 - add AnyEvent::Handle->push_shutdown method.
82 - add an additional error message parameter to AnyEvent::Handle's 738 - add an additional error message parameter to AnyEvent::Handle's
83 on_error callback (for TLS, $! is still available). 739 on_error callback (for TLS, $! is still available).
84 - add AnyEvent::Handle on_starttls/on_stoptls callbacks. 740 - add AnyEvent::Handle on_starttls/on_stoptls callbacks.
85 - make AnyEvent::Handle more robust against early conenction 741 - make AnyEvent::Handle more robust against early connection
86 failures (during new), and return C<undef> in that case 742 failures (during new), and return C<undef> in that case
87 from the constructor. 743 from the constructor.
88 - AnyEvent::Handle will now try to load only JSON::XS first, 744 - AnyEvent::Handle will now try to load only JSON::XS first,
89 then fall back to JSON. 745 then fall back to JSON.
90 - format_ipv4/format_ipv6 are now exported by default, for symmetry, 746 - format_ipv4/format_ipv6 are now exported by default, for symmetry,
… …
2144.3 Fri Oct 3 09:18:43 CEST 2008 8704.3 Fri Oct 3 09:18:43 CEST 2008
215 - AnyEvent will now install a no-op signal handler for SIGPIPE, 871 - AnyEvent will now install a no-op signal handler for SIGPIPE,
216 unless one has been installed already. 872 unless one has been installed already.
217 - warn about not loaidng AnyEvent::Impl::POE early enough 873 - warn about not loaidng AnyEvent::Impl::POE early enough
218 (patch by Adam Rosenstein). 874 (patch by Adam Rosenstein).
219 - fixed a great number of bugs and corner cases in AnyEvent::Handle: 875 - fixed a great number of bugs and corner cases in AnyEvent::Handle.
220 - fix a bug in where in SSL connect mode, the 876 - fix a bug in where in SSL connect mode, the
221 client would first wait for some data by the server and 877 client would first wait for some data by the server and
222 otherwise hang. (reported and analyzed in an absolutely 878 otherwise hang. (reported and analysed in an absolutely
223 exemplary manner by Adam Rosenstein). 879 exemplary manner by Adam Rosenstein).
224 - fix a bug in where SSL EOF would not be treated 880 - fix a bug in where SSL EOF would not be treated
225 as stream EOF, putting the connection into a hung state. 881 as stream EOF, putting the connection into a hung state.
226 (reported and analyzed in an absolutely exemplary manner by Adam 882 (reported and analysed in an absolutely exemplary manner by Adam
227 Rosenstein). 883 Rosenstein).
228 - fix a potential segfault when the TLS context 884 - fix a potential segfault when the TLS context
229 would go missing in a read callback (e.g. due to stoptls) - 885 would go missing in a read callback (e.g. due to stoptls) -
230 Net::SSLeay of course makes no type checking whatsoever. 886 Net::SSLeay of course makes no type checking whatsoever.
231 - AnyEvent::Handle will not stop the read watcher in TLS mode. 887 - AnyEvent::Handle will not stop the read watcher in TLS mode.

Diff Legend

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