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

Comparing AnyEvent/Changes (file contents):
Revision 1.270 by root, Wed Jul 8 05:30:55 2009 UTC vs.
Revision 1.638 by root, Fri Oct 18 20:12:15 2019 UTC

1Revision history for Perl extension AnyEvent. 1Revision history for Perl extension AnyEvent.
2 2
34.801 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?
14TODO: Олег Г <verdrehung@gmail.com>, empty A record response causes
15AnyEvent::resolve_sockaddr to eventually call pack_sockaddr with undef -
16broken dns server, but maybe improve the response.
17
187.17 Wed Sep 18 03:04:49 CEST 2019
19 - work around antique openssl version in RHEL 7 by formatting
20 dh parameters differently (reported by several people).
21 - add t/13_weaken.t.
22
237.16 Fri Jul 19 18:00:21 CEST 2019
24 - add ffdhe group dh parameters from RFC 7919, and use ffdhe3072
25 as new default, instead of schmorp1539.
26 - AnyEvent::Log did not reassess logging status of
27 AnyEvent::Log::logger's when contexts were changed with
28 ->attach/detach/slaves, causing them to not log even though a
29 recent attach should have caused them to log.
30 - added some more logging configuration examples.
31 - mention RFC 8482 in AnyEvent::DNS.
32
337.15 Tue Feb 26 03:07:42 CET 2019
34 - INCOMPATIBLE CHANGE: AnyEvent::Handle's tls_detect documentation gave
35 separate major and minor versions, while code passed only a single
36 value. This version follows the documentation and now passes separate
37 major and minor values.
38 - work around Net::SSLeay not having been ported to openssl 1.1, but many
39 distributions compiling it against openssl 1.1, which unfortunately
40 succeeds and results in a very broken module.
41 - AnyEvent::DNS::dns_unpack now stores the original DNS packet
42 in the __ member, to allow decoding of undecodable resouce records
43 containing compressed domian names.
44 - AnyEvent::Socket::parse_ipv6 would NOT, as advertised, accept ipv4
45 addresses. It now does and converts them to ipv4 mapped addresses.
46 - support CAA records, based on patch by Steve Atkins.
47 - add freenom and cloudflare nameservers as dns fallback.
48 - AnyEvent::Strict would not properly ward against io watchers
49 on files when the handle passed was a file descriptor.
50 - document "internal" variables used by the dns en-/decoder to allow
51 enterprising users to extend them in a semi-official way.
52
537.14 Sat Jun 24 01:44:19 CEST 2017
54 - fix a crash bug in AnyEvent::Handle with openssl 1.1.0 (patched
55 by Paul Howarth and apparently tracked down by Mike McCauley).
56 - AnyEvent::Handle->keepalive was documented (and defined) twice
57 (found by Matt Merhar).
58 - AnyEvent::Socket::tcp_bind/tcp_server would immediately unlink a unix
59 listening socket unless a guard is used. Change this so that no
60 cleanup will be performed unless a guard is used and document this more
61 clearly.
62 - make tcp_bind/tcp_server error messages more regular.
63 - work around latest perl upstream breakage - this time, perl5porters were
64 apparently bullied by a bogus CVE (CVE-2016-1238), and since distros
65 stupidly apply anything that has "security fix" stamped on it, it's likely
66 going to be a problem in practise (and working around it is trivial on
67 windows and unix).
68 - add TCP_FASTOPEN/MSG_FASTOPEN and MSG_NOSIGNAL/MSG_MORE/MSG_DONTWAIT
69 to constants.
70 - update warnings to common::sense 3.74 standards.
71
727.13 Sat Sep 17 04:31:49 CEST 2016
73 - Only call tlsext_host_name for non-empty common names (reported
74 by Maxime Soulé).
75 - log a (single) notice message if SNI is not supported.
76 - upgrade to UTS-46:9.0.0 draft and switch to non-transitional
77 behaviour, beating thunderbird, ie, edge, chrome and safari to it :)
78 (see also https://bugzilla.mozilla.org/show_bug.cgi?id=1218179)
79 - turns out the UTS-46 IDNA testcase failures were indeed bugs in the
80 testcases and the specification and not in the code - the post-9.0.0
81 unicode files have all known problems fixed, so finally the AnyEvent
82 IDNA implementation can pass the full IDNA testsuite - without needing
83 a single fix :)
84 - guarantee (and document) that condvar callbacks will be removed
85 on invocation - important to avoid circular references.
86
877.12 Wed Jan 27 19:12:26 CET 2016
88 - use common name as hostname for tls connects, if Net::SSLeay
89 supports SNI.
90 - fix documentation of tls_autostart read type in AnyEvent::Handle,
91 analyzed by Felix Ostmann.
92
937.11 Thu Jul 16 14:36:00 CEST 2015
94 - AnyEvent::Socket::parse_ipv6 could accept malformed ipv6
95 addresses (extra "::" at end and similar cases).
96 - add a more explicit warning to AnyEvent::Handle that it doesn't
97 work on files, people keep getting confused.
98 - new function AnyEvent::Socket::tcp_bind.
99 - new functions AnyEvent::fh_block and AnyEvent::fh_unblock.
100 - aligned ipv6 address formatting with RFC 5952 (by not shortening
101 a single :0: to ::).
102 - added stability canary support.
103
1047.09 Sat May 2 16:38:53 CEST 2015
105 - AnyEvent::Debug called an internal function (AnyEvent::Log::ft)
106 that was renamed to AnyEvent::Log:format_time. uinder its old name
107 (analyzed by Michael Stovenour).
108 - update AnyEvent::DNS fallback resolver addresses:
109 seems google effectively killed most other free dns resolvers,
110 so remove them, but add cable and wireless (ecrc) since it was
111 stable for 20 years or so, official or not, and there should be
112 an alternative to google.
113 - perl5porters broke windows error codes in 5.20, and mapped
114 WSAEWOULDBLOCK on the (different) EWOULDBLOCK error code, and
115 WSAEINPROGRESS into the incompatible ERINPROGRESS code, probably
116 because they were so cool. They probably broke other error codes
117 for no reason, too, but I didn't care for checking, it's too
118 depressing. This version only works around the WSAEWOULDBLOCK
119 issue, because I don't have a nice way to work around the
120 WSAEINPROGRESS bug.
121
1227.08 Wed Dec 10 05:27:17 CET 2014
123 - work around a newly introduced bug in Socket 2.011 (an
124 errornous sun_length check) (analyzed by Maxime Soulé).
125 - AnyEvent::TLS didn't load (but refer to) AnyEvent::Socket
126 (analyzed by Ben Magistro).
127 - AnyEvent::Strict will now confess, not croak. This is in line with
128 it being a development/debugging tool.
129 - work around a number of libglib bugs (debug builds of libglib enforce
130 certain undocumented behaviour patterns such as not being able to
131 remove a child watch source after it has fired, which we will try
132 to emulate to avoid "criticals". what where they thinking...).
133 - mention json security issues in AnyEvent::Handle, now that Douglas
134 Crockford has foolishly and incompatibly changed JSON.
135 - changed default dns resolver "max_outstanding" value from 1 to 10,
136 the latter beinfg the intended value all along
137 (reported by Ilya Chesnokov).
138 - added new "AnyEvent::Impl::UV" interface module to the UV event lib
139 (written by Mike Lowell).
140
1417.07 Tue Dec 17 17:45:02 CET 2013
142 - the documentation for custom tls verify schemes was wrong. make it agree
143 with the code (reported by Maxime Soulé).
144 - added cbor read and write types to AnyEvent::Handle (using CBOR::XS).
145 - work around an API change in openssl that could cause wrong tls connection
146 aborts, likely on windows only (analyzed by sten).
147 - calling AnyEvent->now_update with AnyEvent::Impl::Perl caused an
148 endless loop (reported by Dietrich Rebmann).
149 - add tlsv1_1 and tlsv1_2 protocols to AnyEvent::TLS
150 (patch by Maxime Soulé).
151 - document AnyEvent::Impl::IOAsync::set_loop and
152 $AnyEvent::Impl::IOAsync::LOOP. Though only documented now, this
153 functionality has _always_ been available.
154 - force a toplevel domain name in t/81_hosts.t (analyzed by
155 David Jack Wange Olrik).
156 - document that AnyEvent::Log uses AnyEvent::IO.
157 - warn about AnyEvent::Filesys::Notify performance.
158 - praise the joys of AnyEvent::Fork::*.
159 - time for an =encoding directive.
160 - do no longer use JSON to create a default json coder, use
161 JSON::XS or JSON::PP directly.
162
1637.05 Wed Aug 21 10:38:08 CEST 2013
164 - uts46data.pl couldn't be found due to wrong naming of the file
165 (reported by Fulko Hew).
166 - handle lone \015's properly in AE::Handle's default line read
167 (reported by various people).
168 - untaint ip addresses found in /etc/hosts (patch by José Micó).
169 - the memleak fix in 7.03 caused resolving via /etc/hosts to always
170 fail on first use (reported and testcase by Andrew Whatson).
171 - expose AnyEvent::Log::format_time, and allow users to redefine it.
172 - expose AnyEvent::Log::default_format, and allow redefinition.
173 - expose AnyEvent::Log::fatal_exit, to allow redefinition.
174 - AnyEvent::Debug shell can now run coro shell commands, if available.
175 - t/63* tests were wrongly in MANIFEST.
176 - kernel.org's finger server went MIA, switch to freebsd.org and
177 icculus.org.
178 - clarify that IO::AIO and AnyEvent::AIO are needed for AnyEvent::IO
179 to function asynchronously (reported by Daniel Carosone).
180 - hard-disable $^W in most tests, it generates too much garbage output.
181 - use a (hopefully) more future-proof method to emulate common::sense.
182 - upgrade to UTS-46:6.2.0.
183 - switch to INSTLIB from INSTLIBDIR, as INSTLIBDIR was wrongly
184 documented. should not affect anything.
185
1867.04 Thu Nov 15 02:23:58 CET 2012
187 - debugging code left in.
188
1897.03 Thu Nov 15 02:16:11 CET 2012
190 - AnyEvent::Socket::inet_aton did not work when dns resolution was
191 used to find the addresses (analyzed and patch by Maxime Soulé).
192 - fix a memory leak in the /etc/hosts lookup code when hosts don't
193 resolve and are not in hosts (analyzed by sten).
194
1957.02 Tue Aug 14 04:27:58 CEST 2012
196 - AnyEvent::Util::run_cmd could block indefinitely (analyzed and test
197 program by Yorhel).
198 - verified that AnyEvent::Socket follows RFC5952.
199 - try to parse "ADDR#PORT" in addition to "ADDR PORT".
200
2017.01 Sun May 13 01:03:17 CEST 2012
202 - fail with EPROTO in AnyEvent::Handle wqhen TLS is requested but
203 not available, instead of throwing an exception.
204 - use File::Spec to get the tmpdir in t/*, to avoid needless
205 failures on (most, not mine :) windows boxes.
206 - new handle read types: tls_detect and tls_autostart.
207
2087.0 Fri Apr 13 06:33:30 CEST 2012
209 - child watchers are broken in POE 1.352 (also many earlier
210 versions) and there seems to be no way to work around it, as POE
211 itself is inherently racy. Document this regression and add a delay
212 in t/68_poe_03_child.t for the time being.
213 - new module AnyEvent::IO, that is a frontend to either a pure-perl
214 synchronous I/O implementation (AnyEvent::IO::Perl), or to the
215 asynchronous I/O extension IO::AIO (AnyEvent::IO::IOAIO).
216 - load /etc/hosts only when DNS has no answer.
217 - stat /etc/hosts on every access and reload it if it changed.
218 - load /etc/hosts via AnyEvent::IO - potentially asynchronous.
219 - fix a buggy croak in the dh parameter reading in AnyEvent::TLS.
220 - AnyEvent::Log log_to_file and log_to_path now use AnyEvent::IO. As
221 a side effect, they now use true appending as opposed to libc
222 appending, and the file might not have been opened when the function
223 returns.
224 - the default logging level was not properly documented in a variety of
225 places, this has hopefully been rectified.
226 - updated uts46data.pl for unicode 6.1.0.
227 - made log messages generated by AnyEvent submodules not
228 include the package name anymore, as it will be logged
229 by default already.
230 - upgrade to the trick used by common-sense 3.5 to work
231 around extra warning torture/breakage under perl 5.15.x.
232 - log messages by anyevent are now ucfirst, are usually full
233 sentences and do no longer include the package name.
234 - the storable read type would sometimes throw an exception instead
235 of causing EBADMSG (analyzed by Maxime Soulé).
236 - work around a bug in openssl 1.0.1 which enforces a minimum rsa
237 keysize (reported by Paul Howarth).
238 - documented Rocco fabricating statements about AnyEvent and me.
239
2406.14 Tue Jan 31 20:00:24 CET 2012
241 - AnyEvent::Impl::Tk was broken due to a mysterious "die" inside,
242 probably an editing mistake (reported by Darin McBride).
243
2446.13 Thu Jan 12 07:27:01 CET 2012
245 - AnyEvent::Util::fork_call checked for POSIX availability in the wrong way,
246 causing it to fail if POSIX wasn't loaded already (analysed by Rock Power).
247 - AnyEvent::Handle::push_read (line => did pass $1 directly, so
248 regex-matching inside the callback would change the parameter, despite
249 $1 being dynamically scoped per-block. this perl bug is now being
250 worked around (testcase by Cindy Wang).
251
2526.12 Mon Dec 12 13:21:10 CET 2011
253 - $! was clobbered by subsequent calls in tcp_connect, due to
254 the postpone not saving/restoring it (analyzed by Richard Garnish).
255
2566.11 Tue Nov 22 10:36:05 CET 2011
257 - Tk cannot create windows when tainted unless you set a title, so
258 set a dummy title for AnyEvent's dummy mainwindow. How dummy.
259 - escape any nonprintable/nonascii characters when stringifying
260 backtraces.
261 - log the reason loading an interface module fails to load at level debug.
262
2636.1 Tue Oct 4 19:44:30 CEST 2011
264 - INCOMPATIBLE CHANGE: the default log level is now "4" (error
265 and above), and some messages inside AnyEvent have been elevated
266 to higher log levels to print by default.
267 - AnyEvent::log, unlike AnyEvent::Log::log, did not always exit on fatal.
268 - work around yet another signal race bug, newly introduced by POE. My
269 how that module sucks.
270 - add google public ipv6 dns as fallback servers, so AE::DNS can finally
271 support ipv6-only operation even with defaults (also use ::1 instead
272 of 127.0.0.1).
273 - give better guidelines on how to chose a logging level.
274 - AnyEvent::Log can now cap (limit) log message priority by context.
275 - AE::log will now save and restore $! and $@, so callers don't have to.
276 - AE::Strict now checks that registered fds are still valid at regular
277 intervals.
278 - skip frequently called 'require' calls if the module appears loaded.
279 - increase signal latency test timer considerably for poor little osx,
280 which can't be bothered with basic posix functionality such as signal
281 delivery.
282 - mention some debugging aids in the tutorial.
283 - new env variable PERL_ANYEVENT_MAX_SIGNAL_LATENCY.
284 - private (%namespace) log contexts in PERL_ANYEVENT_LOG now have the name
285 of the context as title.
286 - AnyEvent::Strict would not restore read-write access to $_ when a callback
287 died - this is a user bug, but since AE::Strict exists to help dbeugging...
288 - detect the extremely evil IO::Async::Loop::AnyEvent module and refuse
289 to work with it - the author is well aware of the issues but refuses
290 to fix them. This extreme step has been taken because the module actively
291 attacks the goals of AnyEvent and makes it harder for module authors to use
292 AnyEvent.
293
2946.02 Fri Aug 26 20:08:31 CEST 2011
295 - log_to_syslog now accepts facility strings.
296 - log_to_syslog would not allow facilities - "$facility|$level"
297 *literally* means that string. sheesh.
298 - the CPAN parser got confused and thought we provide Sys::Syslog, try
299 to work around.
300 - renamed AnyEvent::Impl::FLTK2 to FLTK.
301
3026.01 Fri Aug 26 07:04:11 CEST 2011
303 - INCOMPATIBLE CHANGE: AnyEvent::DNS resource records now include
304 the TTL as fourth element - this affects ->resolve, ->request
305 and ->dns_unpack indirectly. Convenience functions such as
306 AnyEvent::DNS::{a,txt,mx,srv} are not affected, neither is
307 AnyEvent::Socket.
308 - INCOMPATIBLE CHANGE: backend authors now should not implement
309 one_event or loop, but instead the AnyEvent::CondVar::_wait and _poll
310 methods.
311 - INCOMPATIBLE CHANGE: verbosity levels have been redefined, see
312 AnyEvent (PERL_ANYEVENT_VERBOSE) and AnyEvent::Log for details.
313 - DEPRECIATION: please change
314 "use AnyEvent::Impl::Perl" to "use AnyEvent::Loop" and
315 "AnyEvent::Impl::Perl::loop" to "AnyEvent::Loop::run"
316 in your programs, for future compatibility.
317 - *blush* idle watcher emulation was borked.
318 - AnyEvent::Strict now checks the AE::xxx functions as well.
319 - querying too long domain names would barf deep inside AnyEvent::DNS - since
320 this is hard to avoid for applications, AnyEvent:.DNS::request now checks for
321 for too long domain names explicitly and passes an undef to the callback
322 instead (reported by James Bromberger).
323 - make AnyEvent::Util more robust against EINTR (analyzed by Andrew Suffield).
324 - new functions: AnyEvent::postpone, AnyEvent::log.
325 - new module: AnyEvent::Log, for simple logging needs.
326 - new env variable, PERL_ANYEVENT_LOG, for fine-grained logging config.
327 - new env variable, PERL_ANYEVENT_DEBUG_SHELL, to automatically start
328 a debug shell.
329 - new env variable, PERL_ANYEVENT_DEBUG_WRAP, to automatically enable
330 watcher instrumentation/wrapping (see AnyEvent::Debug::wrap).
331 - new env variable, PERL_ANYEVENT_HOSTS, to override the place where
332 /etc/hosts can be found.
333 - an empty PERL_ANYEVENT_RESOLV_CONF now selects the default
334 configuration.
335 - env variables can now be specified using AE_xyz instead of
336 PERL_ANYEVENT_xyz, subject to some rules.
337 - lots of new AnyEvent::Debug functionality.
338 - AnyEvent::Strict now checks for modifications of $_ by the callback
339 (by setting it to readonly while executing the callback, if possible).
340 - IO::Async sometimes cancels the wrong timer when you call cancel_timer.
341 Invest extra overhead to work around this peculiar behaviour.
342 - split the pure perl backend into a loop module and an impl
343 module (AnyEvent::Loop and AnyEvent::Impl::Perl).
344 - some read types (regex, netstring, json, storable) could
345 errornously keep an AE::Handle object alive on their own
346 (reported by Mohammad Toossi).
347 - AnyEvent::Socket::resolve_sockaddr and all functions using it now
348 supports /etc/hosts overrides.
349 - add more workarounds around child watcher bugs in IO::Async
350 and (newly introduced) in POE.
351 - use glib child watchers and try to work around its limitations,
352 if possible (based on a patch by Kevin Ryde).
353 - add support for FLTK (fltk2), via AnyEvent::Impl::FLTK2.
354 - many read types in AnyEvent::Handle would malfunction during
355 transitions to TLS because of a "delete $self->{rbuf}".
356 (reported by Gerald Galster).
357 - improve condvar blocking wait performance (probably only noticable
358 with faster backends), also make it easier for event loops that
359 do not support blocking waits.
360 - PERL_ANYEVENT_MODEL now also allows full module names, although
361 the usefulness of this feature is not yet clear.
362 - add tests for all supported event loops, conditional on PERL_ANYEVENT_LOOP_TESTS.
363 - add t/80_ssltext.t.
364 - no longer use AUTOLOAD for the initial loading (for no very good reason).
365 - tcp_server now tries to unlink unix domain sockets when it gets destroyed.
366 - do not run event loop detection multiple times when programs call their cached
367 copies of AnyEvent::detect.
368 - parse_hostport now accepts absolute paths als unix domain sockets.
369 - suppress the idiotic warning inside POE by patching the POE::Kernel
370 run flags instead of asking the user to jump through contortions.
371 - add Task::Weaken as dependency (as recommends only at this time).
372 - replace PF_UNSPEC by 0, for good old minix.
373
3746.0 (not officially released, but was on CPAN for about 15 minutes :/)
375
3765.34 Thu May 12 09:59:41 CEST 2011
377 - AE::TLS didn't support both cert_file and key at the same time,
378 due to some copy&paste bug (reported and analyzed by vti).
379 - improve AE::Handle timeout documentation slightly.
380
3815.33 Mon Feb 7 21:51:41 CET 2011
382 - never appeared on CPAN, for unknown reasons.
383 - AnyEvent::Handle on_connect_error - do not try to destroy a handle
384 object when it's already gone (analyzed by Ryan Bullock).
385 - update idna algorithm(s) to UTS#46 version 6.0.0.
386 - preserve trailing dot in idn_to_ascii - testsuite says to preserve,
387 uts46 says it's an error, and testsuite makes more sense.
388 - remove leading dots in idn_to_ascii - testsuite says to remove,
389 uts46 says it's an error, and testsuite makes more sense.
390 - fixed a bug in idn ignored character processing.
391 - passes the uts46 6.0.0 testsuite except for three cases which
392 apparently are bugs in the testsuite itself.
393 - passes the uts46 6.0.1 testsuite except for eleven cases which
394 apparently are bugs in the testsuite itself.
395
3965.31 Sun Jan 23 11:49:19 CET 2011
397 - work around a POE bug where POE would first poll for new events and
398 invoke handlers BEFORE handling the message queue.
399 - fix parsing of "domain" statement in resolv.conf (Johannes Plunien).
400 - stop_read now does just that in TLS mode. see its
401 documentation for caveats.
402 - update to common::sense 3.4 warning set (some warnings in perl impose
403 an arbitrary 32kb limit on regex matches, which affects
404 AnyEvent::Handle).
405 - AE::Handle had a broken rbuf_max method and missing wbuf_max method.
406 - new experimental AE::Handle::resettls method.
407
4085.3 Fri Dec 31 05:48:13 CET 2010
409 - major IOAsync patch by Paul Evans - 0.33 is required, as the major
410 issues with IO::Async have been solved in that version, including
411 the availabilty of a default loop!
412 - implement new wbuf_max parameter for AnyEvent::Handle.
413 - added Cocoa::EventLoop backend by Daisuke Murase.
414 - the default testsuite now honors PERL_ANYEVENT_MODEL (Paul Evans).
415 - clarify handle on_prepare documentation.
416 - set CLOEXEC flag on parent-side fd's in AnyEvent::Util::run_cmd,
417 to avoid leaking them to any child processes.
418 - try to force files into the arch-dependend lib directory, for the
419 half-"turly obsessive compulsive person"(s) that depend on it.
420
4215.29 Sun Dec 5 10:49:21 CET 2010
422 - convert EV backend to EV 4.00 API (so better upgrade EV too).
423 - AE::Handle: implement dynamic read_size adjustment, add
424 max_read_size parameter, reduce default read_size to 2048.
425 - add some FAQ entries for servers and/or on_eof vs. on_error.
426 - work around OS bugs (cygwin again) when lingering in AE::Handle.
427 - reduce memory usage slightly when connect was used in AE::Handle.
428 - wrok around more segfaults in Tk.
429 - document tls_ctx => undef as valid.
430 - detect WNOHANG value at compiletime.
431 - include some probably linux-only support for building AnyEvent
432 as part of the perl core.
433 - improve compatibility of t/08_idna.t to perl 5.8.
434 - make AnyEvent::Strict truly optional (in case somebody builds a
435 minimal perl...).
436
4375.28 Wed Oct 13 04:14:23 CEST 2010
438 - due to a glitch, AnyEvent's internal getprotobyname wasn't
439 actually used in tcp_server.
440 - implement AnyEvent::CondVar->new method, which was documented but
441 missing.
442 - support autoloading of read types in unshift_read.
443 - AnyEvent::Strict: do the test for invalid fh arguments differently,
444 resulting in more informative error messages.
445 - new AnyEvent::FAQ manpage.
446 - clarify that you can actually call push_read etc. in on_read.
447 - put keys and certs in /tmp, not into the current directory,
448 and document this in the SECURITY CONSIDERATIONS sections
449 of AnyEvent::TLS.
450 - support ";" as resolv.conf comment character in AnyEvent::DNS, as
451 requested by Ron Isaacson.
452 - document $AnyEvent::DNS::RESOLVER variable.
453 - incorporated major typo patches by Abhijit Menon-Sen.
454 - AnyEvent::Handle now croaks when negative timeouts are passed.
455 - add a shitload of TCP_xxx sockopt name constants to AnyEvent::Util,
456 undocumented, but who knows when they might come in handy.
457
4585.271 Tue Jun 8 12:05:46 CEST 2010
459 - backport to perl 5.8.x.
460
4615.27 Sun Jun 6 12:12:05 CEST 2010
462 - postpone differently in AnyEvent::Socket now, as
463 when not, canceling the connection attempt might fail
464 (found by Felix Antonius Wilhelm Ostmann).
465 - explicitly check for non-stream sockets in AE::Handle, too many
466 clueless people fell into the trap of this somehow working.
467 - simplified and reworked the "OTHER MODULES" section.
468 - better/more condvar examples.
469
4705.261 Wed Apr 28 16:13:36 CEST 2010
471 - AF_INET6 was not properly used from Socket6 during configuration
472 time (found by Andy Grundman).
473
4745.26 Mon Apr 12 04:49:35 CEST 2010
475 - don't generate spurious readyness notifications when select
476 returns EINTR in the pure perl backend (analysed and testcase
477 by Bas Denissen).
478 - give same examples in SYNOPSIS sections of AnyEvent and AE.
479 - provider faster implementations for AE::time/now/now_update for
480 EV and Perl backends.
481 - provide faster AE::cv and AE::time implementations for all
482 backends.
483 - no longer support lower/mixed-case signal names (this was
484 never documented nor universally supported).
485 - some more memory µ-optimisations, and somewhat less messy
486 code to cope with both AnyEvent and AE APIs.
487
4885.251 Sat Mar 13 00:58:21 CET 2010
489 - make test failed if it was run before make install on perl <=
490 5.8.8, and older versions of perl make our live very difficult,
491 so write constants.pl during Makefile.PL time.
492
4935.25 Sat Mar 13 00:23:14 CET 2010
494 - fix a race condition in AnyEvent::Handle that would cause
495 a "bio_read: unsupported method" error (in _another_
496 ssl connection) after user code threw an exception.
497 - added AnyEvent::Handle->destroyed method.
498 - speed up AnyEvent::Socket::format_address and ::format_ipv6.
499 - the AnyEvent::Util::fh_nonblocking function worked only
500 by lucky accident on win32.
501 - smaller and faster AnyEvent::Util::fh_nonblocking.
502 - when the (required!) Time::HiRes module is missing, AnyEvent
503 did not fall back to built-in time properly.
504 - do not load Fcntl at runtime, saving memory and loading time.
505 - precompile a number of constants and use them instead of runtime
506 detection and eval.
507 - free detection code after detection and similar memory optimisations.
508 - Perl backend timer interval best effort drift has been improved
509 (same algorithm as EV).
510 - update unicode idna mapping table.
511
5125.24 Tue Jan 5 11:39:43 CET 2010
513 - cygwin never reports errors from failed connects. 1.5 just gives
514 you continous readyness and EAGAIN, 1.7 is even more broken and
515 just hangs. work around both issues in a cygwin-specific hack.
516 - improve idle watcher documentation slightly.
517
5185.23 Sun Dec 20 23:48:00 CET 2009
519 - support IDNs in resolve_sockaddr, and therefore in tcp_connect.
520 - implement punycode_encode/decode, idn_nameprep,
521 idn_to_ascii and idn_to_unicode operations in AnyEvent::Util.
522 - provide $AE::VERSION.
523 - removed traces of "no strict 'refs'".
524
5255.22 Sat Dec 5 03:51:13 CET 2009
526 - downgrade-or-fail in AnyEvent::Handle::push_write, to
527 diagnose encoding failures earlier and more succinctly.
528 (this works around bugs in perl, throwing away encoding info
529 when passing scalar data to extensions).
530 - add more examples to AnyEvent::Socket manpage.
531 - upgrade internal warning set to the same as common::sense 2.03.
532 - use pack "n/a*" for pre-5.8.9 perl compatibility in AnyEvent::DNS
533 (John Beppu).
534 - AnyEvent::Socket::inet_aton now properly supports ipv6, as documented.
535 - add google public dns servers to fallback server set and make sure
536 we load-balance properly between the three sets. also add all
537 fallback dns servers, not just a random one, to each dns config.
538
5395.21 Thu Nov 19 02:48:47 CET 2009
540 - fix a problem where socket constants were called with parameters
541 (spotted by David Friedland).
542 - fork_call never use'd POSIX (reported by Daisuke Maki).
543 - improve perl 5.6 compatibility further (but it still won't work
544 unless you rip out everything but the core).
545 - prefer Net::DNS::Resolver over ipconfig, if installed, on win32.
546 uses 10MB of RAM, but doesn't open a console window. *sigh*.
547
5485.202 Wed Oct 14 22:35:44 CEST 2009
549 - AnyEvent::DNS would unexpectedly clobber $_ under windows
550 (analysed by Matthias Waldorf).
551 - AnyEvent::Handle::run_cmd can now pass the PID of the
552 newly-created process, which is much less useful than it might
553 sound (based on patch by Yann Kerherve).
554
5555.201 Tue Sep 29 12:09:25 CEST 2009
556 - AnyEvent:Handle::on_starttls/on_stoptls methods were broken
557 (reported by Torsten Foertsch).
558 - common::sense 2.0 could cause tcp_server to throw an exception
559 (analysed by elmex).
560
5615.2 Mon Sep 14 07:04:49 CEST 2009
562 - INCOMPATIBLE CHANGE: do no longer support register_read_type
563 and register_write_type in AnyEvent::Handle, instead support
564 package names (the facility was mostly abused).
565 - implement "packagename-as-read/write type" support in
566 AnyEvent::Handle.
567 - AnyEvent::Handle: new options "keepalive" and "oobinline".
568 - oobinline set by default to avoid security issues.
569 - the pure-perl event loop backend wrongly detected times() fallback
570 support (spotted by Pavel Boldin).
571
5725.12 Tue Sep 1 20:26:50 CEST 2009
573 - be more lenient when parsing resolv.conf files, as some people
574 use hashmarks as comment indicator inside directives (reported
575 by Michael S. Fischer).
576 - use same set of warnings as common::sense 2.0.
577 - fix a potential 32 bit overflow issue due to perl having problems
578 with large hex constants in the Perl backend.
579 - do not use the slower 5.6 version fo the enc_name code in AnyEvent::DNS
580 by default.
581 - fix some prototypes.
582 - (optionally) require Async::Interrupt 1.02.
583 - replace opendns fallbacks by verizon/level3 public dns servers.
584
5855.112 Fri Aug 21 13:59:04 CEST 2009
586 - AnyEvent::Handle could cause hard crashes in Net::SSLeay when
587 using starttls with an illegal/uncreatable tls context.
588 - port to broken win32 perls.
589
5905.111 Sun Aug 16 18:44:36 CEST 2009
591 - perl errornously requires sockaddr_un structures returned by the
592 kernel to have a specific length. work around this bug by padding
593 sockaddr_un structures (found to fail on OS X, as perl only uses
594 the correct code for GNU/Linux, but likely to fail on other
595 OSes as well).
596
5975.11 Wed Aug 12 17:49:37 CEST 2009
598 - fixed a bug in run_cmd with input from filename.
599 - create a json coder object when none is given in write direction
600 to reduce dependency on JSON.
601
6025.1 Tue Aug 11 03:17:32 CEST 2009
603 - tcp_connect now reports the correct error in $!, not an unrelated
604 one (was broken since 4.91).
605 - AE::Handle did not properly clear rtimeout/wtimeout watchers
606 when retrying a tcp_connect.
607 - new functions: AnyEvent::Util::run_cmd and close_all_fds_except.
608 - fixed wrong documentation in AE::Socket.
609
6105.01 Mon Aug 10 03:16:32 CEST 2009
611 - last release broke EV child watchers :/.
612
6135.0 Sun Aug 9 17:32:09 CEST 2009
614 - officially introduce and document the AE API.
615 - lowering the timeout at runtime did not have immediate effect
616 in AnyEvent::Handle.
617 - AE::Handle now has separate and independent read and write timeout
618 settings.
619 - probe for Guard module when AnyEvent::Util is loaded, not at
620 runtime (Event doesn't like eval "use" from callbacks, and it avoids
621 exporting the probe function).
622 - Impl::IOAsync's io watchers did not use the filehandle emulation
623 code to work around it's watcher limits.
624 - work around IO::Async's problems with overloaded objects.
625 - try to disable wearnings inside IO::Async, it's just too buggy.
626 - do no longer use Event by default, only EV or Perl, Event is too buggy.
627 - hackishly provide fast AE:: interface when EV or Perl is the backend.
628 - inherit VERSION from AnyEvent::VERSION in most modules.
629
6304.91 Thu Aug 6 15:42:45 CEST 2009
631 - AE::Handle::starttls could get out of sync when the read buffer
632 already contains some TLS handshake.
633 - AE::Handle did not properly free the TLS session, causing bigger
634 memory leaks in Net::SSLeay than expected (reported by toaster).
635 - AE::Socket::tcp_connect will now artificially delay invoking
636 the callback to avoid returning after invoking the callback.
637 - convert many internal watcher uses to AE API.
638
6394.9 Sat Aug 1 11:07:01 CEST 2009
640 - Glib silently fails when registering a timeout with
641 a negative value, so avoid that.
642 - call condvar callback immediately when it is set
643 after the condition is already signalled.
644 - check rbuf_max condition only after trying to consume
645 data.
646 - normalise signal numbers to names when using the ae's
647 signal handling, but do not document this (yet).
648 - pure perl signal emulation did not properly set nonblocking
649 mode on the signal pipe on win32 (this is unlikely to have caused
650 any issues).
651 - new module: AnyEvent::Debug.
652 - AnyEvent::Strict now checks that a signal was specified
653 by name (not name or number).
654 - reduce memory footprint in typical cases by ~50kb by
655 only compiling signal code when necessary.
656 - add AnyEvent::Handle->rbuf_max.
657 - grab the AE namespace for future new API, implement stubs for
658 said future/alternative API.
659 - new function: AnyEvent::Socket::format_hostport.
660
6614.881 Tue Jul 28 12:51:53 CEST 2009
662 - work around a bug in local in pre-5.10 perls, causing
663 AnyEvent::Handle to recurse when it should not
664 (analysed by elmex).
665
6664.88 Tue Jul 28 04:04:37 CEST 2009
667 - re-bless the handle into a dummy package after calling
668 AnyEvent::Handle::destroy, so the user does not need to check
669 for errors after every push_write etc.
670 - do not attempt to run t/02_signals.t on obviously broken
671 platforms.
672
6734.87 Sun Jul 26 02:06:16 CEST 2009
674 - do not attempt to linger when there is no longer a valid fh.
675 - enforce tls mode to be either accept or connect, do not simply
676 segfault in Net::SSLeay.
677 - AnyEvent::Handle can now call tcp_connect itself (new parameters
678 connect, on_prepare, on_connect and on_connect_error). Updated
679 tutorial accordingly.
680 - add AnyEvent::Impl::Irssi backend.
681
6824.86 Mon Jul 20 23:52:29 CEST 2009
683 - since the verbose warning is not enough, explicitly document
684 that versions before 1.33 of Net::SSLeay are not secure.
685 - work around signal handling races in Event and (...) Event::Lib.
686 - try to align signal-race timer to full-second boundaries.
687 - work around Tk not liking negative timeouts.
688 - don't complain of different grades of environmental unfriendlyness
689 in IO::Async.
690
6914.85 Sat Jul 18 06:16:14 CEST 2009
692 - nail the signal race problem in perl once and for all
693 (see $AnyEvent::MAX_SIGNAL_LATENCY).
694 - take advantage of Async::Interrupt if it is available.
695 - load Time::HiRes and Guard modules on demand only.
696 - add optional/recommended modules section to AnyEvent
697 documentation.
698 - reduce memory usage considerably (and reduce startup penalty)
699 by not using "strict", "warnings" and "overload" modules.
700 - work around buggy windows/openbsd perls and provide EBADMSG
701 and EPROTO ourselves when missing.
702 - improve perl 5.6 compatibility of the core event loop.
703 - made Net::SSLeay version 1.33 a soft requirement.
704
7054.83 Fri Jul 17 16:56:26 CEST 2009
706 - implement AnyEvent::Socket::getprotobyname.
707 - AnyEvent::CondVar's will now detect recursive blocking
708 waits and will croak, as too many people fall into
709 this trap.
710 - AnyEvent::Handle will now call ->destroy on itself after
711 executing the on_error callback, instead of doing some
712 half-baked internal shutdown, for fatal errors.
713 - clarify on_eof behaviour w.r.t. the read queue and
714 on_read callbacks.
715 - ignore some possible spurious wake-ups in tcp_connect.
716
7174.82 Sat Jul 11 00:34:55 CEST 2009
718 - POE and Event backends didn't accept some callable objects as
719 callbacks.
720 - use Config module instead of POSIX module to detect signal names
721 in AnyEvent::Strict and AnyEvent::Impl::EventLib, as the POSIX
722 module doesn't even have all POSIX signals :/.
723 - use more workarounds around the many refcnt/corruption bugs in
724 Event::Lib.
725 - work around a race condition in perl's select, causing t/03_child.t
726 to rarely fail.
727
7284.81 Thu Jul 9 10:30:30 CEST 2009
729 - AnyEvent::Handle didn't properly diagnose write errors
730 (it expected -1 from syswrite, how lame... :).
4 - support file descriptors in addition to file handles 731 - support file descriptors in addition to file handles
5 in AnyEvent->io. 732 in AnyEvent->io.
6 - new env variables: PERL_ANYEVENT_RESOLV_CONF, 733 - new env variables: PERL_ANYEVENT_RESOLV_CONF,
7 PERL_ANYEVENT_MAX_OUTSTANDING_DNS, PERL_ANYEVENT_CA_FILE 734 PERL_ANYEVENT_MAX_OUTSTANDING_DNS, PERL_ANYEVENT_CA_FILE
8 and PERL_ANYEVENT_CA_PATH. 735 and PERL_ANYEVENT_CA_PATH.
9 - provide a sensible synopsis section for AnyEvent::TLS. 736 - provide a sensible synopsis section for AnyEvent::TLS.
737 - add a "supported backends" section to the manpage.
10 - added simple io watcher test to testsuite, using a 738 - added simple io watcher test to testsuite, using a
11 portable_socketpair. 739 portable_socketpair.
740 - tried to improve the stability of the Event::Lib backend,
741 YMMV.
12 742
134.8 Mon Jul 6 23:45:16 CEST 2009 7434.8 Mon Jul 6 23:45:16 CEST 2009
14 - AnyEvent::DNS did not properly follow CNAME records with 744 - AnyEvent::DNS did not properly follow CNAME records with
15 uppercase targets. 745 uppercase targets.
16 - AnyEvent::DNS would errornously return AAAA records 746 - AnyEvent::DNS would errornously return AAAA records
… …
27 in the common case. 757 in the common case.
28 - add AnyEvent::Handle->push_shutdown method. 758 - add AnyEvent::Handle->push_shutdown method.
29 - add an additional error message parameter to AnyEvent::Handle's 759 - add an additional error message parameter to AnyEvent::Handle's
30 on_error callback (for TLS, $! is still available). 760 on_error callback (for TLS, $! is still available).
31 - add AnyEvent::Handle on_starttls/on_stoptls callbacks. 761 - add AnyEvent::Handle on_starttls/on_stoptls callbacks.
32 - make AnyEvent::Handle more robust against early conenction 762 - make AnyEvent::Handle more robust against early connection
33 failures (during new), and return C<undef> in that case 763 failures (during new), and return C<undef> in that case
34 from the constructor. 764 from the constructor.
35 - AnyEvent::Handle will now try to load only JSON::XS first, 765 - AnyEvent::Handle will now try to load only JSON::XS first,
36 then fall back to JSON. 766 then fall back to JSON.
37 - format_ipv4/format_ipv6 are now exported by default, for symmetry, 767 - format_ipv4/format_ipv6 are now exported by default, for symmetry,
… …
1614.3 Fri Oct 3 09:18:43 CEST 2008 8914.3 Fri Oct 3 09:18:43 CEST 2008
162 - AnyEvent will now install a no-op signal handler for SIGPIPE, 892 - AnyEvent will now install a no-op signal handler for SIGPIPE,
163 unless one has been installed already. 893 unless one has been installed already.
164 - warn about not loaidng AnyEvent::Impl::POE early enough 894 - warn about not loaidng AnyEvent::Impl::POE early enough
165 (patch by Adam Rosenstein). 895 (patch by Adam Rosenstein).
166 - fixed a great number of bugs and corner cases in AnyEvent::Handle: 896 - fixed a great number of bugs and corner cases in AnyEvent::Handle.
167 - fix a bug in where in SSL connect mode, the 897 - fix a bug in where in SSL connect mode, the
168 client would first wait for some data by the server and 898 client would first wait for some data by the server and
169 otherwise hang. (reported and analyzed in an absolutely 899 otherwise hang. (reported and analysed in an absolutely
170 exemplary manner by Adam Rosenstein). 900 exemplary manner by Adam Rosenstein).
171 - fix a bug in where SSL EOF would not be treated 901 - fix a bug in where SSL EOF would not be treated
172 as stream EOF, putting the connection into a hung state. 902 as stream EOF, putting the connection into a hung state.
173 (reported and analyzed in an absolutely exemplary manner by Adam 903 (reported and analysed in an absolutely exemplary manner by Adam
174 Rosenstein). 904 Rosenstein).
175 - fix a potential segfault when the TLS context 905 - fix a potential segfault when the TLS context
176 would go missing in a read callback (e.g. due to stoptls) - 906 would go missing in a read callback (e.g. due to stoptls) -
177 Net::SSLeay of course makes no type checking whatsoever. 907 Net::SSLeay of course makes no type checking whatsoever.
178 - AnyEvent::Handle will not stop the read watcher in TLS mode. 908 - AnyEvent::Handle will not stop the read watcher in TLS mode.

Diff Legend

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