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

Comparing AnyEvent/Changes (file contents):
Revision 1.158 by root, Sun Jun 22 12:17:47 2008 UTC vs.
Revision 1.646 by root, Sat Jul 25 17:58:50 2020 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?
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.
17TODO: AnyEvent::Loop should not use POSIIX when ti isn't used.
18
19TODO: AnyEvent::Log contexts should have a way to inject log messages received form elsewhere, using timestamp etc.
20TODO: add some cheta-sheet to log synopsis regarding AE_LOG=AnyEvent::MP=+
21TODO: tcp_connect option for AnyEvent::Handle?
22TODO: hosts files on windows probably contain \r, which is not stripped
23 - AnyEvent::Loop no longer loads the POSIX module if Time::HiRes
24 has monotonic clock support (suggested by Felipe Gasper).
25 - since Net::SSLeay (up to 1.88-1) is _still not ported to openssl 1.1
26 and many distributions therefore will have a completely broken
27 Net::SSLeay module for the forseeable future, AnyEvent::TLS
28 will now log a warning (that is not seen by default unless you
29 increase log verbosity).
30 - add tlsv1_3 disable switch to AnyEvent::TLS.
31 - add experimental "tls" method support for AnyEvent::TLS.
32 - tls methods were lowercase, but documented to be uppercase.
33 fix documentation.
34 - EV backend now uses EV::break when waiting for condvars,
35 making waiting for condvars more efficient. This might
36 "break" already-broken programs that recursively enter
37 the event loop even more.
38
397.17 Wed Sep 18 03:04:49 CEST 2019
40 - work around antique openssl version in RHEL 7 by formatting
41 dh parameters differently (reported by several people).
42 - add t/13_weaken.t.
43
447.16 Fri Jul 19 18:00:21 CEST 2019
45 - add ffdhe group dh parameters from RFC 7919, and use ffdhe3072
46 as new default, instead of schmorp1539.
47 - AnyEvent::Log did not reassess logging status of
48 AnyEvent::Log::logger's when contexts were changed with
49 ->attach/detach/slaves, causing them to not log even though a
50 recent attach should have caused them to log.
51 - added some more logging configuration examples.
52 - mention RFC 8482 in AnyEvent::DNS.
53
547.15 Tue Feb 26 03:07:42 CET 2019
55 - INCOMPATIBLE CHANGE: AnyEvent::Handle's tls_detect documentation gave
56 separate major and minor versions, while code passed only a single
57 value. This version follows the documentation and now passes separate
58 major and minor values.
59 - work around Net::SSLeay not having been ported to openssl 1.1, but many
60 distributions compiling it against openssl 1.1, which unfortunately
61 succeeds and results in a very broken module.
62 - AnyEvent::DNS::dns_unpack now stores the original DNS packet
63 in the __ member, to allow decoding of undecodable resouce records
64 containing compressed domian names.
65 - AnyEvent::Socket::parse_ipv6 would NOT, as advertised, accept ipv4
66 addresses. It now does and converts them to ipv4 mapped addresses.
67 - support CAA records, based on patch by Steve Atkins.
68 - add freenom and cloudflare nameservers as dns fallback.
69 - AnyEvent::Strict would not properly ward against io watchers
70 on files when the handle passed was a file descriptor.
71 - document "internal" variables used by the dns en-/decoder to allow
72 enterprising users to extend them in a semi-official way.
73
747.14 Sat Jun 24 01:44:19 CEST 2017
75 - fix a crash bug in AnyEvent::Handle with openssl 1.1.0 (patched
76 by Paul Howarth and apparently tracked down by Mike McCauley).
77 - AnyEvent::Handle->keepalive was documented (and defined) twice
78 (found by Matt Merhar).
79 - AnyEvent::Socket::tcp_bind/tcp_server would immediately unlink a unix
80 listening socket unless a guard is used. Change this so that no
81 cleanup will be performed unless a guard is used and document this more
82 clearly.
83 - make tcp_bind/tcp_server error messages more regular.
84 - work around latest perl upstream breakage - this time, perl5porters were
85 apparently bullied by a bogus CVE (CVE-2016-1238), and since distros
86 stupidly apply anything that has "security fix" stamped on it, it's likely
87 going to be a problem in practise (and working around it is trivial on
88 windows and unix).
89 - add TCP_FASTOPEN/MSG_FASTOPEN and MSG_NOSIGNAL/MSG_MORE/MSG_DONTWAIT
90 to constants.
91 - update warnings to common::sense 3.74 standards.
92
937.13 Sat Sep 17 04:31:49 CEST 2016
94 - Only call tlsext_host_name for non-empty common names (reported
95 by Maxime Soulé).
96 - log a (single) notice message if SNI is not supported.
97 - upgrade to UTS-46:9.0.0 draft and switch to non-transitional
98 behaviour, beating thunderbird, ie, edge, chrome and safari to it :)
99 (see also https://bugzilla.mozilla.org/show_bug.cgi?id=1218179)
100 - turns out the UTS-46 IDNA testcase failures were indeed bugs in the
101 testcases and the specification and not in the code - the post-9.0.0
102 unicode files have all known problems fixed, so finally the AnyEvent
103 IDNA implementation can pass the full IDNA testsuite - without needing
104 a single fix :)
105 - guarantee (and document) that condvar callbacks will be removed
106 on invocation - important to avoid circular references.
107
1087.12 Wed Jan 27 19:12:26 CET 2016
109 - use common name as hostname for tls connects, if Net::SSLeay
110 supports SNI.
111 - fix documentation of tls_autostart read type in AnyEvent::Handle,
112 analyzed by Felix Ostmann.
113
1147.11 Thu Jul 16 14:36:00 CEST 2015
115 - AnyEvent::Socket::parse_ipv6 could accept malformed ipv6
116 addresses (extra "::" at end and similar cases).
117 - add a more explicit warning to AnyEvent::Handle that it doesn't
118 work on files, people keep getting confused.
119 - new function AnyEvent::Socket::tcp_bind.
120 - new functions AnyEvent::fh_block and AnyEvent::fh_unblock.
121 - aligned ipv6 address formatting with RFC 5952 (by not shortening
122 a single :0: to ::).
123 - added stability canary support.
124
1257.09 Sat May 2 16:38:53 CEST 2015
126 - AnyEvent::Debug called an internal function (AnyEvent::Log::ft)
127 that was renamed to AnyEvent::Log:format_time. uinder its old name
128 (analyzed by Michael Stovenour).
129 - update AnyEvent::DNS fallback resolver addresses:
130 seems google effectively killed most other free dns resolvers,
131 so remove them, but add cable and wireless (ecrc) since it was
132 stable for 20 years or so, official or not, and there should be
133 an alternative to google.
134 - perl5porters broke windows error codes in 5.20, and mapped
135 WSAEWOULDBLOCK on the (different) EWOULDBLOCK error code, and
136 WSAEINPROGRESS into the incompatible ERINPROGRESS code, probably
137 because they were so cool. They probably broke other error codes
138 for no reason, too, but I didn't care for checking, it's too
139 depressing. This version only works around the WSAEWOULDBLOCK
140 issue, because I don't have a nice way to work around the
141 WSAEINPROGRESS bug.
142
1437.08 Wed Dec 10 05:27:17 CET 2014
144 - work around a newly introduced bug in Socket 2.011 (an
145 errornous sun_length check) (analyzed by Maxime Soulé).
146 - AnyEvent::TLS didn't load (but refer to) AnyEvent::Socket
147 (analyzed by Ben Magistro).
148 - AnyEvent::Strict will now confess, not croak. This is in line with
149 it being a development/debugging tool.
150 - work around a number of libglib bugs (debug builds of libglib enforce
151 certain undocumented behaviour patterns such as not being able to
152 remove a child watch source after it has fired, which we will try
153 to emulate to avoid "criticals". what where they thinking...).
154 - mention json security issues in AnyEvent::Handle, now that Douglas
155 Crockford has foolishly and incompatibly changed JSON.
156 - changed default dns resolver "max_outstanding" value from 1 to 10,
157 the latter beinfg the intended value all along
158 (reported by Ilya Chesnokov).
159 - added new "AnyEvent::Impl::UV" interface module to the UV event lib
160 (written by Mike Lowell).
161
1627.07 Tue Dec 17 17:45:02 CET 2013
163 - the documentation for custom tls verify schemes was wrong. make it agree
164 with the code (reported by Maxime Soulé).
165 - added cbor read and write types to AnyEvent::Handle (using CBOR::XS).
166 - work around an API change in openssl that could cause wrong tls connection
167 aborts, likely on windows only (analyzed by sten).
168 - calling AnyEvent->now_update with AnyEvent::Impl::Perl caused an
169 endless loop (reported by Dietrich Rebmann).
170 - add tlsv1_1 and tlsv1_2 protocols to AnyEvent::TLS
171 (patch by Maxime Soulé).
172 - document AnyEvent::Impl::IOAsync::set_loop and
173 $AnyEvent::Impl::IOAsync::LOOP. Though only documented now, this
174 functionality has _always_ been available.
175 - force a toplevel domain name in t/81_hosts.t (analyzed by
176 David Jack Wange Olrik).
177 - document that AnyEvent::Log uses AnyEvent::IO.
178 - warn about AnyEvent::Filesys::Notify performance.
179 - praise the joys of AnyEvent::Fork::*.
180 - time for an =encoding directive.
181 - do no longer use JSON to create a default json coder, use
182 JSON::XS or JSON::PP directly.
183
1847.05 Wed Aug 21 10:38:08 CEST 2013
185 - uts46data.pl couldn't be found due to wrong naming of the file
186 (reported by Fulko Hew).
187 - handle lone \015's properly in AE::Handle's default line read
188 (reported by various people).
189 - untaint ip addresses found in /etc/hosts (patch by José Micó).
190 - the memleak fix in 7.03 caused resolving via /etc/hosts to always
191 fail on first use (reported and testcase by Andrew Whatson).
192 - expose AnyEvent::Log::format_time, and allow users to redefine it.
193 - expose AnyEvent::Log::default_format, and allow redefinition.
194 - expose AnyEvent::Log::fatal_exit, to allow redefinition.
195 - AnyEvent::Debug shell can now run coro shell commands, if available.
196 - t/63* tests were wrongly in MANIFEST.
197 - kernel.org's finger server went MIA, switch to freebsd.org and
198 icculus.org.
199 - clarify that IO::AIO and AnyEvent::AIO are needed for AnyEvent::IO
200 to function asynchronously (reported by Daniel Carosone).
201 - hard-disable $^W in most tests, it generates too much garbage output.
202 - use a (hopefully) more future-proof method to emulate common::sense.
203 - upgrade to UTS-46:6.2.0.
204 - switch to INSTLIB from INSTLIBDIR, as INSTLIBDIR was wrongly
205 documented. should not affect anything.
206
2077.04 Thu Nov 15 02:23:58 CET 2012
208 - debugging code left in.
209
2107.03 Thu Nov 15 02:16:11 CET 2012
211 - AnyEvent::Socket::inet_aton did not work when dns resolution was
212 used to find the addresses (analyzed and patch by Maxime Soulé).
213 - fix a memory leak in the /etc/hosts lookup code when hosts don't
214 resolve and are not in hosts (analyzed by sten).
215
2167.02 Tue Aug 14 04:27:58 CEST 2012
217 - AnyEvent::Util::run_cmd could block indefinitely (analyzed and test
218 program by Yorhel).
219 - verified that AnyEvent::Socket follows RFC5952.
220 - try to parse "ADDR#PORT" in addition to "ADDR PORT".
221
2227.01 Sun May 13 01:03:17 CEST 2012
223 - fail with EPROTO in AnyEvent::Handle wqhen TLS is requested but
224 not available, instead of throwing an exception.
225 - use File::Spec to get the tmpdir in t/*, to avoid needless
226 failures on (most, not mine :) windows boxes.
227 - new handle read types: tls_detect and tls_autostart.
228
2297.0 Fri Apr 13 06:33:30 CEST 2012
230 - child watchers are broken in POE 1.352 (also many earlier
231 versions) and there seems to be no way to work around it, as POE
232 itself is inherently racy. Document this regression and add a delay
233 in t/68_poe_03_child.t for the time being.
234 - new module AnyEvent::IO, that is a frontend to either a pure-perl
235 synchronous I/O implementation (AnyEvent::IO::Perl), or to the
236 asynchronous I/O extension IO::AIO (AnyEvent::IO::IOAIO).
237 - load /etc/hosts only when DNS has no answer.
238 - stat /etc/hosts on every access and reload it if it changed.
239 - load /etc/hosts via AnyEvent::IO - potentially asynchronous.
240 - fix a buggy croak in the dh parameter reading in AnyEvent::TLS.
241 - AnyEvent::Log log_to_file and log_to_path now use AnyEvent::IO. As
242 a side effect, they now use true appending as opposed to libc
243 appending, and the file might not have been opened when the function
244 returns.
245 - the default logging level was not properly documented in a variety of
246 places, this has hopefully been rectified.
247 - updated uts46data.pl for unicode 6.1.0.
248 - made log messages generated by AnyEvent submodules not
249 include the package name anymore, as it will be logged
250 by default already.
251 - upgrade to the trick used by common-sense 3.5 to work
252 around extra warning torture/breakage under perl 5.15.x.
253 - log messages by anyevent are now ucfirst, are usually full
254 sentences and do no longer include the package name.
255 - the storable read type would sometimes throw an exception instead
256 of causing EBADMSG (analyzed by Maxime Soulé).
257 - work around a bug in openssl 1.0.1 which enforces a minimum rsa
258 keysize (reported by Paul Howarth).
259 - documented Rocco fabricating statements about AnyEvent and me.
260
2616.14 Tue Jan 31 20:00:24 CET 2012
262 - AnyEvent::Impl::Tk was broken due to a mysterious "die" inside,
263 probably an editing mistake (reported by Darin McBride).
264
2656.13 Thu Jan 12 07:27:01 CET 2012
266 - AnyEvent::Util::fork_call checked for POSIX availability in the wrong way,
267 causing it to fail if POSIX wasn't loaded already (analysed by Rock Power).
268 - AnyEvent::Handle::push_read (line => did pass $1 directly, so
269 regex-matching inside the callback would change the parameter, despite
270 $1 being dynamically scoped per-block. this perl bug is now being
271 worked around (testcase by Cindy Wang).
272
2736.12 Mon Dec 12 13:21:10 CET 2011
274 - $! was clobbered by subsequent calls in tcp_connect, due to
275 the postpone not saving/restoring it (analyzed by Richard Garnish).
276
2776.11 Tue Nov 22 10:36:05 CET 2011
278 - Tk cannot create windows when tainted unless you set a title, so
279 set a dummy title for AnyEvent's dummy mainwindow. How dummy.
280 - escape any nonprintable/nonascii characters when stringifying
281 backtraces.
282 - log the reason loading an interface module fails to load at level debug.
283
2846.1 Tue Oct 4 19:44:30 CEST 2011
285 - INCOMPATIBLE CHANGE: the default log level is now "4" (error
286 and above), and some messages inside AnyEvent have been elevated
287 to higher log levels to print by default.
288 - AnyEvent::log, unlike AnyEvent::Log::log, did not always exit on fatal.
289 - work around yet another signal race bug, newly introduced by POE. My
290 how that module sucks.
291 - add google public ipv6 dns as fallback servers, so AE::DNS can finally
292 support ipv6-only operation even with defaults (also use ::1 instead
293 of 127.0.0.1).
294 - give better guidelines on how to chose a logging level.
295 - AnyEvent::Log can now cap (limit) log message priority by context.
296 - AE::log will now save and restore $! and $@, so callers don't have to.
297 - AE::Strict now checks that registered fds are still valid at regular
298 intervals.
299 - skip frequently called 'require' calls if the module appears loaded.
300 - increase signal latency test timer considerably for poor little osx,
301 which can't be bothered with basic posix functionality such as signal
302 delivery.
303 - mention some debugging aids in the tutorial.
304 - new env variable PERL_ANYEVENT_MAX_SIGNAL_LATENCY.
305 - private (%namespace) log contexts in PERL_ANYEVENT_LOG now have the name
306 of the context as title.
307 - AnyEvent::Strict would not restore read-write access to $_ when a callback
308 died - this is a user bug, but since AE::Strict exists to help dbeugging...
309 - detect the extremely evil IO::Async::Loop::AnyEvent module and refuse
310 to work with it - the author is well aware of the issues but refuses
311 to fix them. This extreme step has been taken because the module actively
312 attacks the goals of AnyEvent and makes it harder for module authors to use
313 AnyEvent.
314
3156.02 Fri Aug 26 20:08:31 CEST 2011
316 - log_to_syslog now accepts facility strings.
317 - log_to_syslog would not allow facilities - "$facility|$level"
318 *literally* means that string. sheesh.
319 - the CPAN parser got confused and thought we provide Sys::Syslog, try
320 to work around.
321 - renamed AnyEvent::Impl::FLTK2 to FLTK.
322
3236.01 Fri Aug 26 07:04:11 CEST 2011
324 - INCOMPATIBLE CHANGE: AnyEvent::DNS resource records now include
325 the TTL as fourth element - this affects ->resolve, ->request
326 and ->dns_unpack indirectly. Convenience functions such as
327 AnyEvent::DNS::{a,txt,mx,srv} are not affected, neither is
328 AnyEvent::Socket.
329 - INCOMPATIBLE CHANGE: backend authors now should not implement
330 one_event or loop, but instead the AnyEvent::CondVar::_wait and _poll
331 methods.
332 - INCOMPATIBLE CHANGE: verbosity levels have been redefined, see
333 AnyEvent (PERL_ANYEVENT_VERBOSE) and AnyEvent::Log for details.
334 - DEPRECIATION: please change
335 "use AnyEvent::Impl::Perl" to "use AnyEvent::Loop" and
336 "AnyEvent::Impl::Perl::loop" to "AnyEvent::Loop::run"
337 in your programs, for future compatibility.
338 - *blush* idle watcher emulation was borked.
339 - AnyEvent::Strict now checks the AE::xxx functions as well.
340 - querying too long domain names would barf deep inside AnyEvent::DNS - since
341 this is hard to avoid for applications, AnyEvent:.DNS::request now checks for
342 for too long domain names explicitly and passes an undef to the callback
343 instead (reported by James Bromberger).
344 - make AnyEvent::Util more robust against EINTR (analyzed by Andrew Suffield).
345 - new functions: AnyEvent::postpone, AnyEvent::log.
346 - new module: AnyEvent::Log, for simple logging needs.
347 - new env variable, PERL_ANYEVENT_LOG, for fine-grained logging config.
348 - new env variable, PERL_ANYEVENT_DEBUG_SHELL, to automatically start
349 a debug shell.
350 - new env variable, PERL_ANYEVENT_DEBUG_WRAP, to automatically enable
351 watcher instrumentation/wrapping (see AnyEvent::Debug::wrap).
352 - new env variable, PERL_ANYEVENT_HOSTS, to override the place where
353 /etc/hosts can be found.
354 - an empty PERL_ANYEVENT_RESOLV_CONF now selects the default
355 configuration.
356 - env variables can now be specified using AE_xyz instead of
357 PERL_ANYEVENT_xyz, subject to some rules.
358 - lots of new AnyEvent::Debug functionality.
359 - AnyEvent::Strict now checks for modifications of $_ by the callback
360 (by setting it to readonly while executing the callback, if possible).
361 - IO::Async sometimes cancels the wrong timer when you call cancel_timer.
362 Invest extra overhead to work around this peculiar behaviour.
363 - split the pure perl backend into a loop module and an impl
364 module (AnyEvent::Loop and AnyEvent::Impl::Perl).
365 - some read types (regex, netstring, json, storable) could
366 errornously keep an AE::Handle object alive on their own
367 (reported by Mohammad Toossi).
368 - AnyEvent::Socket::resolve_sockaddr and all functions using it now
369 supports /etc/hosts overrides.
370 - add more workarounds around child watcher bugs in IO::Async
371 and (newly introduced) in POE.
372 - use glib child watchers and try to work around its limitations,
373 if possible (based on a patch by Kevin Ryde).
374 - add support for FLTK (fltk2), via AnyEvent::Impl::FLTK2.
375 - many read types in AnyEvent::Handle would malfunction during
376 transitions to TLS because of a "delete $self->{rbuf}".
377 (reported by Gerald Galster).
378 - improve condvar blocking wait performance (probably only noticable
379 with faster backends), also make it easier for event loops that
380 do not support blocking waits.
381 - PERL_ANYEVENT_MODEL now also allows full module names, although
382 the usefulness of this feature is not yet clear.
383 - add tests for all supported event loops, conditional on PERL_ANYEVENT_LOOP_TESTS.
384 - add t/80_ssltext.t.
385 - no longer use AUTOLOAD for the initial loading (for no very good reason).
386 - tcp_server now tries to unlink unix domain sockets when it gets destroyed.
387 - do not run event loop detection multiple times when programs call their cached
388 copies of AnyEvent::detect.
389 - parse_hostport now accepts absolute paths als unix domain sockets.
390 - suppress the idiotic warning inside POE by patching the POE::Kernel
391 run flags instead of asking the user to jump through contortions.
392 - add Task::Weaken as dependency (as recommends only at this time).
393 - replace PF_UNSPEC by 0, for good old minix.
394
3956.0 (not officially released, but was on CPAN for about 15 minutes :/)
396
3975.34 Thu May 12 09:59:41 CEST 2011
398 - AE::TLS didn't support both cert_file and key at the same time,
399 due to some copy&paste bug (reported and analyzed by vti).
400 - improve AE::Handle timeout documentation slightly.
401
4025.33 Mon Feb 7 21:51:41 CET 2011
403 - never appeared on CPAN, for unknown reasons.
404 - AnyEvent::Handle on_connect_error - do not try to destroy a handle
405 object when it's already gone (analyzed by Ryan Bullock).
406 - update idna algorithm(s) to UTS#46 version 6.0.0.
407 - preserve trailing dot in idn_to_ascii - testsuite says to preserve,
408 uts46 says it's an error, and testsuite makes more sense.
409 - remove leading dots in idn_to_ascii - testsuite says to remove,
410 uts46 says it's an error, and testsuite makes more sense.
411 - fixed a bug in idn ignored character processing.
412 - passes the uts46 6.0.0 testsuite except for three cases which
413 apparently are bugs in the testsuite itself.
414 - passes the uts46 6.0.1 testsuite except for eleven cases which
415 apparently are bugs in the testsuite itself.
416
4175.31 Sun Jan 23 11:49:19 CET 2011
418 - work around a POE bug where POE would first poll for new events and
419 invoke handlers BEFORE handling the message queue.
420 - fix parsing of "domain" statement in resolv.conf (Johannes Plunien).
421 - stop_read now does just that in TLS mode. see its
422 documentation for caveats.
423 - update to common::sense 3.4 warning set (some warnings in perl impose
424 an arbitrary 32kb limit on regex matches, which affects
425 AnyEvent::Handle).
426 - AE::Handle had a broken rbuf_max method and missing wbuf_max method.
427 - new experimental AE::Handle::resettls method.
428
4295.3 Fri Dec 31 05:48:13 CET 2010
430 - major IOAsync patch by Paul Evans - 0.33 is required, as the major
431 issues with IO::Async have been solved in that version, including
432 the availabilty of a default loop!
433 - implement new wbuf_max parameter for AnyEvent::Handle.
434 - added Cocoa::EventLoop backend by Daisuke Murase.
435 - the default testsuite now honors PERL_ANYEVENT_MODEL (Paul Evans).
436 - clarify handle on_prepare documentation.
437 - set CLOEXEC flag on parent-side fd's in AnyEvent::Util::run_cmd,
438 to avoid leaking them to any child processes.
439 - try to force files into the arch-dependend lib directory, for the
440 half-"turly obsessive compulsive person"(s) that depend on it.
441
4425.29 Sun Dec 5 10:49:21 CET 2010
443 - convert EV backend to EV 4.00 API (so better upgrade EV too).
444 - AE::Handle: implement dynamic read_size adjustment, add
445 max_read_size parameter, reduce default read_size to 2048.
446 - add some FAQ entries for servers and/or on_eof vs. on_error.
447 - work around OS bugs (cygwin again) when lingering in AE::Handle.
448 - reduce memory usage slightly when connect was used in AE::Handle.
449 - wrok around more segfaults in Tk.
450 - document tls_ctx => undef as valid.
451 - detect WNOHANG value at compiletime.
452 - include some probably linux-only support for building AnyEvent
453 as part of the perl core.
454 - improve compatibility of t/08_idna.t to perl 5.8.
455 - make AnyEvent::Strict truly optional (in case somebody builds a
456 minimal perl...).
457
4585.28 Wed Oct 13 04:14:23 CEST 2010
459 - due to a glitch, AnyEvent's internal getprotobyname wasn't
460 actually used in tcp_server.
461 - implement AnyEvent::CondVar->new method, which was documented but
462 missing.
463 - support autoloading of read types in unshift_read.
464 - AnyEvent::Strict: do the test for invalid fh arguments differently,
465 resulting in more informative error messages.
466 - new AnyEvent::FAQ manpage.
467 - clarify that you can actually call push_read etc. in on_read.
468 - put keys and certs in /tmp, not into the current directory,
469 and document this in the SECURITY CONSIDERATIONS sections
470 of AnyEvent::TLS.
471 - support ";" as resolv.conf comment character in AnyEvent::DNS, as
472 requested by Ron Isaacson.
473 - document $AnyEvent::DNS::RESOLVER variable.
474 - incorporated major typo patches by Abhijit Menon-Sen.
475 - AnyEvent::Handle now croaks when negative timeouts are passed.
476 - add a shitload of TCP_xxx sockopt name constants to AnyEvent::Util,
477 undocumented, but who knows when they might come in handy.
478
4795.271 Tue Jun 8 12:05:46 CEST 2010
480 - backport to perl 5.8.x.
481
4825.27 Sun Jun 6 12:12:05 CEST 2010
483 - postpone differently in AnyEvent::Socket now, as
484 when not, canceling the connection attempt might fail
485 (found by Felix Antonius Wilhelm Ostmann).
486 - explicitly check for non-stream sockets in AE::Handle, too many
487 clueless people fell into the trap of this somehow working.
488 - simplified and reworked the "OTHER MODULES" section.
489 - better/more condvar examples.
490
4915.261 Wed Apr 28 16:13:36 CEST 2010
492 - AF_INET6 was not properly used from Socket6 during configuration
493 time (found by Andy Grundman).
494
4955.26 Mon Apr 12 04:49:35 CEST 2010
496 - don't generate spurious readyness notifications when select
497 returns EINTR in the pure perl backend (analysed and testcase
498 by Bas Denissen).
499 - give same examples in SYNOPSIS sections of AnyEvent and AE.
500 - provider faster implementations for AE::time/now/now_update for
501 EV and Perl backends.
502 - provide faster AE::cv and AE::time implementations for all
503 backends.
504 - no longer support lower/mixed-case signal names (this was
505 never documented nor universally supported).
506 - some more memory µ-optimisations, and somewhat less messy
507 code to cope with both AnyEvent and AE APIs.
508
5095.251 Sat Mar 13 00:58:21 CET 2010
510 - make test failed if it was run before make install on perl <=
511 5.8.8, and older versions of perl make our live very difficult,
512 so write constants.pl during Makefile.PL time.
513
5145.25 Sat Mar 13 00:23:14 CET 2010
515 - fix a race condition in AnyEvent::Handle that would cause
516 a "bio_read: unsupported method" error (in _another_
517 ssl connection) after user code threw an exception.
518 - added AnyEvent::Handle->destroyed method.
519 - speed up AnyEvent::Socket::format_address and ::format_ipv6.
520 - the AnyEvent::Util::fh_nonblocking function worked only
521 by lucky accident on win32.
522 - smaller and faster AnyEvent::Util::fh_nonblocking.
523 - when the (required!) Time::HiRes module is missing, AnyEvent
524 did not fall back to built-in time properly.
525 - do not load Fcntl at runtime, saving memory and loading time.
526 - precompile a number of constants and use them instead of runtime
527 detection and eval.
528 - free detection code after detection and similar memory optimisations.
529 - Perl backend timer interval best effort drift has been improved
530 (same algorithm as EV).
531 - update unicode idna mapping table.
532
5335.24 Tue Jan 5 11:39:43 CET 2010
534 - cygwin never reports errors from failed connects. 1.5 just gives
535 you continous readyness and EAGAIN, 1.7 is even more broken and
536 just hangs. work around both issues in a cygwin-specific hack.
537 - improve idle watcher documentation slightly.
538
5395.23 Sun Dec 20 23:48:00 CET 2009
540 - support IDNs in resolve_sockaddr, and therefore in tcp_connect.
541 - implement punycode_encode/decode, idn_nameprep,
542 idn_to_ascii and idn_to_unicode operations in AnyEvent::Util.
543 - provide $AE::VERSION.
544 - removed traces of "no strict 'refs'".
545
5465.22 Sat Dec 5 03:51:13 CET 2009
547 - downgrade-or-fail in AnyEvent::Handle::push_write, to
548 diagnose encoding failures earlier and more succinctly.
549 (this works around bugs in perl, throwing away encoding info
550 when passing scalar data to extensions).
551 - add more examples to AnyEvent::Socket manpage.
552 - upgrade internal warning set to the same as common::sense 2.03.
553 - use pack "n/a*" for pre-5.8.9 perl compatibility in AnyEvent::DNS
554 (John Beppu).
555 - AnyEvent::Socket::inet_aton now properly supports ipv6, as documented.
556 - add google public dns servers to fallback server set and make sure
557 we load-balance properly between the three sets. also add all
558 fallback dns servers, not just a random one, to each dns config.
559
5605.21 Thu Nov 19 02:48:47 CET 2009
561 - fix a problem where socket constants were called with parameters
562 (spotted by David Friedland).
563 - fork_call never use'd POSIX (reported by Daisuke Maki).
564 - improve perl 5.6 compatibility further (but it still won't work
565 unless you rip out everything but the core).
566 - prefer Net::DNS::Resolver over ipconfig, if installed, on win32.
567 uses 10MB of RAM, but doesn't open a console window. *sigh*.
568
5695.202 Wed Oct 14 22:35:44 CEST 2009
570 - AnyEvent::DNS would unexpectedly clobber $_ under windows
571 (analysed by Matthias Waldorf).
572 - AnyEvent::Handle::run_cmd can now pass the PID of the
573 newly-created process, which is much less useful than it might
574 sound (based on patch by Yann Kerherve).
575
5765.201 Tue Sep 29 12:09:25 CEST 2009
577 - AnyEvent:Handle::on_starttls/on_stoptls methods were broken
578 (reported by Torsten Foertsch).
579 - common::sense 2.0 could cause tcp_server to throw an exception
580 (analysed by elmex).
581
5825.2 Mon Sep 14 07:04:49 CEST 2009
583 - INCOMPATIBLE CHANGE: do no longer support register_read_type
584 and register_write_type in AnyEvent::Handle, instead support
585 package names (the facility was mostly abused).
586 - implement "packagename-as-read/write type" support in
587 AnyEvent::Handle.
588 - AnyEvent::Handle: new options "keepalive" and "oobinline".
589 - oobinline set by default to avoid security issues.
590 - the pure-perl event loop backend wrongly detected times() fallback
591 support (spotted by Pavel Boldin).
592
5935.12 Tue Sep 1 20:26:50 CEST 2009
594 - be more lenient when parsing resolv.conf files, as some people
595 use hashmarks as comment indicator inside directives (reported
596 by Michael S. Fischer).
597 - use same set of warnings as common::sense 2.0.
598 - fix a potential 32 bit overflow issue due to perl having problems
599 with large hex constants in the Perl backend.
600 - do not use the slower 5.6 version fo the enc_name code in AnyEvent::DNS
601 by default.
602 - fix some prototypes.
603 - (optionally) require Async::Interrupt 1.02.
604 - replace opendns fallbacks by verizon/level3 public dns servers.
605
6065.112 Fri Aug 21 13:59:04 CEST 2009
607 - AnyEvent::Handle could cause hard crashes in Net::SSLeay when
608 using starttls with an illegal/uncreatable tls context.
609 - port to broken win32 perls.
610
6115.111 Sun Aug 16 18:44:36 CEST 2009
612 - perl errornously requires sockaddr_un structures returned by the
613 kernel to have a specific length. work around this bug by padding
614 sockaddr_un structures (found to fail on OS X, as perl only uses
615 the correct code for GNU/Linux, but likely to fail on other
616 OSes as well).
617
6185.11 Wed Aug 12 17:49:37 CEST 2009
619 - fixed a bug in run_cmd with input from filename.
620 - create a json coder object when none is given in write direction
621 to reduce dependency on JSON.
622
6235.1 Tue Aug 11 03:17:32 CEST 2009
624 - tcp_connect now reports the correct error in $!, not an unrelated
625 one (was broken since 4.91).
626 - AE::Handle did not properly clear rtimeout/wtimeout watchers
627 when retrying a tcp_connect.
628 - new functions: AnyEvent::Util::run_cmd and close_all_fds_except.
629 - fixed wrong documentation in AE::Socket.
630
6315.01 Mon Aug 10 03:16:32 CEST 2009
632 - last release broke EV child watchers :/.
633
6345.0 Sun Aug 9 17:32:09 CEST 2009
635 - officially introduce and document the AE API.
636 - lowering the timeout at runtime did not have immediate effect
637 in AnyEvent::Handle.
638 - AE::Handle now has separate and independent read and write timeout
639 settings.
640 - probe for Guard module when AnyEvent::Util is loaded, not at
641 runtime (Event doesn't like eval "use" from callbacks, and it avoids
642 exporting the probe function).
643 - Impl::IOAsync's io watchers did not use the filehandle emulation
644 code to work around it's watcher limits.
645 - work around IO::Async's problems with overloaded objects.
646 - try to disable wearnings inside IO::Async, it's just too buggy.
647 - do no longer use Event by default, only EV or Perl, Event is too buggy.
648 - hackishly provide fast AE:: interface when EV or Perl is the backend.
649 - inherit VERSION from AnyEvent::VERSION in most modules.
650
6514.91 Thu Aug 6 15:42:45 CEST 2009
652 - AE::Handle::starttls could get out of sync when the read buffer
653 already contains some TLS handshake.
654 - AE::Handle did not properly free the TLS session, causing bigger
655 memory leaks in Net::SSLeay than expected (reported by toaster).
656 - AE::Socket::tcp_connect will now artificially delay invoking
657 the callback to avoid returning after invoking the callback.
658 - convert many internal watcher uses to AE API.
659
6604.9 Sat Aug 1 11:07:01 CEST 2009
661 - Glib silently fails when registering a timeout with
662 a negative value, so avoid that.
663 - call condvar callback immediately when it is set
664 after the condition is already signalled.
665 - check rbuf_max condition only after trying to consume
666 data.
667 - normalise signal numbers to names when using the ae's
668 signal handling, but do not document this (yet).
669 - pure perl signal emulation did not properly set nonblocking
670 mode on the signal pipe on win32 (this is unlikely to have caused
671 any issues).
672 - new module: AnyEvent::Debug.
673 - AnyEvent::Strict now checks that a signal was specified
674 by name (not name or number).
675 - reduce memory footprint in typical cases by ~50kb by
676 only compiling signal code when necessary.
677 - add AnyEvent::Handle->rbuf_max.
678 - grab the AE namespace for future new API, implement stubs for
679 said future/alternative API.
680 - new function: AnyEvent::Socket::format_hostport.
681
6824.881 Tue Jul 28 12:51:53 CEST 2009
683 - work around a bug in local in pre-5.10 perls, causing
684 AnyEvent::Handle to recurse when it should not
685 (analysed by elmex).
686
6874.88 Tue Jul 28 04:04:37 CEST 2009
688 - re-bless the handle into a dummy package after calling
689 AnyEvent::Handle::destroy, so the user does not need to check
690 for errors after every push_write etc.
691 - do not attempt to run t/02_signals.t on obviously broken
692 platforms.
693
6944.87 Sun Jul 26 02:06:16 CEST 2009
695 - do not attempt to linger when there is no longer a valid fh.
696 - enforce tls mode to be either accept or connect, do not simply
697 segfault in Net::SSLeay.
698 - AnyEvent::Handle can now call tcp_connect itself (new parameters
699 connect, on_prepare, on_connect and on_connect_error). Updated
700 tutorial accordingly.
701 - add AnyEvent::Impl::Irssi backend.
702
7034.86 Mon Jul 20 23:52:29 CEST 2009
704 - since the verbose warning is not enough, explicitly document
705 that versions before 1.33 of Net::SSLeay are not secure.
706 - work around signal handling races in Event and (...) Event::Lib.
707 - try to align signal-race timer to full-second boundaries.
708 - work around Tk not liking negative timeouts.
709 - don't complain of different grades of environmental unfriendlyness
710 in IO::Async.
711
7124.85 Sat Jul 18 06:16:14 CEST 2009
713 - nail the signal race problem in perl once and for all
714 (see $AnyEvent::MAX_SIGNAL_LATENCY).
715 - take advantage of Async::Interrupt if it is available.
716 - load Time::HiRes and Guard modules on demand only.
717 - add optional/recommended modules section to AnyEvent
718 documentation.
719 - reduce memory usage considerably (and reduce startup penalty)
720 by not using "strict", "warnings" and "overload" modules.
721 - work around buggy windows/openbsd perls and provide EBADMSG
722 and EPROTO ourselves when missing.
723 - improve perl 5.6 compatibility of the core event loop.
724 - made Net::SSLeay version 1.33 a soft requirement.
725
7264.83 Fri Jul 17 16:56:26 CEST 2009
727 - implement AnyEvent::Socket::getprotobyname.
728 - AnyEvent::CondVar's will now detect recursive blocking
729 waits and will croak, as too many people fall into
730 this trap.
731 - AnyEvent::Handle will now call ->destroy on itself after
732 executing the on_error callback, instead of doing some
733 half-baked internal shutdown, for fatal errors.
734 - clarify on_eof behaviour w.r.t. the read queue and
735 on_read callbacks.
736 - ignore some possible spurious wake-ups in tcp_connect.
737
7384.82 Sat Jul 11 00:34:55 CEST 2009
739 - POE and Event backends didn't accept some callable objects as
740 callbacks.
741 - use Config module instead of POSIX module to detect signal names
742 in AnyEvent::Strict and AnyEvent::Impl::EventLib, as the POSIX
743 module doesn't even have all POSIX signals :/.
744 - use more workarounds around the many refcnt/corruption bugs in
745 Event::Lib.
746 - work around a race condition in perl's select, causing t/03_child.t
747 to rarely fail.
748
7494.81 Thu Jul 9 10:30:30 CEST 2009
750 - AnyEvent::Handle didn't properly diagnose write errors
751 (it expected -1 from syswrite, how lame... :).
752 - support file descriptors in addition to file handles
753 in AnyEvent->io.
754 - new env variables: PERL_ANYEVENT_RESOLV_CONF,
755 PERL_ANYEVENT_MAX_OUTSTANDING_DNS, PERL_ANYEVENT_CA_FILE
756 and PERL_ANYEVENT_CA_PATH.
757 - provide a sensible synopsis section for AnyEvent::TLS.
758 - add a "supported backends" section to the manpage.
759 - added simple io watcher test to testsuite, using a
760 portable_socketpair.
761 - tried to improve the stability of the Event::Lib backend,
762 YMMV.
763
7644.8 Mon Jul 6 23:45:16 CEST 2009
765 - AnyEvent::DNS did not properly follow CNAME records with
766 uppercase targets.
767 - AnyEvent::DNS would errornously return AAAA records
768 with v4 mapped addresses (a faulty record) as ipv4 addresses,
769 causing AnyEvent::Socket to throw an exception.
770 - added new module AnyEvent::TLS for easier SSL/TLS context
771 creation, with many options including hostname verification,
772 secure default configuration, lots of documentation and,
773 predefined diffie-hellman keys for perfect forward security
774 and much more. get it while it's still fresh!
775 - use AnyEvent::TLS in AnyEvent::Handle for context management.
776 - load AnyEvent::Handle only on demand in AnyEvent::DNS,
777 so AnyEvent::Socket users have smaller memory footprint
778 in the common case.
779 - add AnyEvent::Handle->push_shutdown method.
780 - add an additional error message parameter to AnyEvent::Handle's
781 on_error callback (for TLS, $! is still available).
782 - add AnyEvent::Handle on_starttls/on_stoptls callbacks.
783 - make AnyEvent::Handle more robust against early connection
784 failures (during new), and return C<undef> in that case
785 from the constructor.
786 - AnyEvent::Handle will now try to load only JSON::XS first,
787 then fall back to JSON.
788 - format_ipv4/format_ipv6 are now exported by default, for symmetry,
789 and because it was documented that way.
790
7914.451 Fri Jul 3 00:28:58 CEST 2009
792 - do not clear rbuf when shutting down an AnyEvent::Handle
793 object - doing so breaks AnyEvent::HTTP.
794
7954.45 Mon Jun 29 22:59:26 CEST 2009
796 - a write error could cause AnyEvent::Handle to create
797 an I/O watcher with an undefined $fh.
798 - special-case mapped ipv4 addresses in both
799 AnyEvent::Socket::format_address and parse_address,
800 to treat them just like ipv4 addresses.
801 - updated and overhauled the AnyEvent::Intro doc.
802 - implement AnyEvent::Socket::format_ipv4/ipv6.
803 - slightly speed up portable_pipe/socketpair.
804 - expand condvar begin/end documentation.
805
8064.42 Fri Jun 26 08:32:18 CEST 2009
807 - preliminary and neccesarily incomplete support for IO::Async.
808 - reset SIGCHLD to DEFAULT when AnyEvent is loaded, in
809 case it was set to IGNORE, to ensure we can catch child statuses
810 even when the calling env acts stupidly.
811 - updated benchmarks with IO::Async, which performs very well.
812
8134.412 Wed Jun 24 01:35:57 CEST 2009
814 - support an "untaint" attribute for AnyEvent::DNS
815 and set it on the default resolver.
816 - implement and document AnyEvent::Impl::Perl::loop.
817 - remove all anyevent-env variables from %ENV
818 when running in tainted mode.
819 - mention and extend the IO::Lambda benchmark.
820
8214.411 Sun Jun 7 18:48:02 CEST 2009
822 - do not try to use F_SETFD on windows.
823
8244.41 Thu May 14 06:40:11 CEST 2009
825 - work around issues in older perls (5.8.0?) when
826 a signal handler is deleted from the %SIG hash.
827 - use POSIX::_exit in child test, to avoid running
828 destructors.
829 - speed up CHLD handling by relying on SIGCHLD being
830 synchronously delivered, even when we roll our own
831 implementation.
832 - AnyEvent::DNS: add the "dname" resource record name
833 for cosmetic reasons.
834
8354.4 Sun Apr 26 20:12:33 CEST 2009
836 - implemented idle watchers, where applicable.
837 - AnyEvent->time died when Event backend was in use.
838 - fix a memleak in the Tk backend.
839 - sped up Tk timer handling.
840 - clip negative "after" values to 0 in AnyEvent::Impl::Event
841 to avoid spamming out warning messages.
842 - fix Qt timers without interval.
843 - avoid Qt zero-timeout special case, in old libqt's.
844
8454.352 Mon Apr 20 16:31:11 CEST 2009
846 - fix AnyEvent::Strict error messages for child watchers.
847 - fix/update Makefile.PL "recommends" versions.
848 - add AnyEvent->now_update.
849
8504.351 Sat Apr 11 07:56:14 CEST 2009
851 - actually make the signal pipe work under win32.
852 - localise $! in signal handler to avoid changing $!, although
853 perl itself does not seem to save/restore errno either.
854 - set the cloexec flag on the signal pipe (normally set by perl too).
855
8564.35 Fri Mar 27 11:48:20 CET 2009
857 - event models relying on AnyEvent's signal watcher emulation did
858 invoke the callback asynchronously, contradicting documentation
859 and causing signals to get lost (this includes AnyEvent's own
860 event loop). AnyEvent now uses the standard pipe trick to make
861 callback execution synchronous to the event loop.
862 - AnyEvent::Handle didn't free TLS context data on DESTROY
863 (patch by Pavel Shaydo).
864 - work around the ever-incompatibly-changing API of MakeMaker.
865 - document that changing global variables without restoring them
866 is a bad idea in Perl (noted by Adam Rosenstein).
867 - AnyEvent::Strict now barfs if ->io is passed a file.
868
8694.34 Thu Feb 12 18:32:45 CET 2009
870 - separately buffer TLS read data, as otherwise the read queue could
871 deadlock as receiving data is not expected while draining the
872 read queue (which cna only happen with TLS).
873 - raise EBADMSG error on JSON decoding errors.
874 - fix some minor manpage bugs (reported by Maximilian Gaß).
875 - speed up select bitmask parsing quite a bit in the pure perl
876 backend.
877 - use CORE::select instead of just select in AnyEvent::Impl::Perl.
878
8794.331 Tue Jan 6 21:07:25 CET 2009
880 - socketpair fails on many vista machines because vista has
881 completely broken accept/getpeername and getsockname functions,
882 so we provide our own socketpair emulation that kind of works
883 (AnyEvent::Util::portable_pipe).
884 - new function: AnyEvent::Util::portable_socketpair.
885 - take advantage of the Guard module if it exists.
886
8874.33 Fri Nov 21 02:35:40 CET 2008
888 - AnyEvent::Strict did errornously flag a fileno of 0 as
889 illegal.
890 - reduce memory usage and slightly speed up the pure perl backend
891 by only storing the file descriptor, not the file handle.
892 - add missing autocork method to AnyEvent::Handle (reported
893 by Adam Rosenstein).
894 - AnyEvent::DNS->resolve errornously documented an $rcode result
895 argument, but there isn't (spotted by Henrik Krohns).
896 - the naked truth about Tk - it's basically unmaintained.
897
8984.32 Mon Nov 3 22:46:32 CET 2008
899 - fix AnyEvent::Socket::resolve_sockaddr to properly support
900 unix sockets again and choose a proper default for
901 the socket type.
902 - fix call to resolve_sockaddr in tcp_connect.
903
9044.31 Thu Oct 30 04:41:48 CET 2008
905 - implemented AnyEvent::Handle->destroy method.
906 - hint about unexpected effects in TLS mode.
907 - speed up AnyEvent::Impl::Perl by using more arrays
908 instead of hashes without the slightest loss of readability :)
909 - work around a perl argument refcounting bug.
910 - with some perl patches applied, AnyEvent now seems leak-free.
911
9124.3 Fri Oct 3 09:18:43 CEST 2008
913 - AnyEvent will now install a no-op signal handler for SIGPIPE,
914 unless one has been installed already.
915 - warn about not loaidng AnyEvent::Impl::POE early enough
916 (patch by Adam Rosenstein).
917 - fixed a great number of bugs and corner cases in AnyEvent::Handle.
918 - fix a bug in where in SSL connect mode, the
919 client would first wait for some data by the server and
920 otherwise hang. (reported and analysed in an absolutely
921 exemplary manner by Adam Rosenstein).
922 - fix a bug in where SSL EOF would not be treated
923 as stream EOF, putting the connection into a hung state.
924 (reported and analysed in an absolutely exemplary manner by Adam
925 Rosenstein).
926 - fix a potential segfault when the TLS context
927 would go missing in a read callback (e.g. due to stoptls) -
928 Net::SSLeay of course makes no type checking whatsoever.
929 - AnyEvent::Handle will not stop the read watcher in TLS mode.
930 - AnyEvent::Handle->stoptls will send a TLS close notify
931 instead of simply destroying the stream now.
932 - fix a bug where large blocks of data written in TLS mode
933 would not be sent unless triggered by receives.
934 - on_drain will now take the tls write buffer into account.
935 - SSL operations have been streamlined a bit and should be faster
936 now, more to come, though.
937 - added a FAQ section, will document nontrivial issues in there.
938
9394.234 Mon Sep 29 04:08:13 CEST 2008
940 - fix child watcher documentation: only child exits will be reported,
941 no trace events.
942 - mention SIGPIPE in AnyEvent::Handle.
943 - perl backend will now use POSIX::times as monotonic clock
944 when available and the monotonic clock option is not.
945 - run even without Time::HiRes in some cases now, but do not
946 provide sub-second accuracy (all *supported* perl versions have
947 Time::HiRes).
948
9494.233 Fri Aug 22 01:48:05 CEST 2008
950 - fix a bug in the testsuite, causing a hang.
951 - clarified lots of AnyEvent::Handle constructor arguments.
952
9534.232 Thu Aug 21 20:44:25 CEST 2008
954 - fix a bug in AnyEvent::Handle that could cause two fatal error
955 callbacks on EOF, which caused AnyEvent::HTTP to signal success twice
956 (testcase by Leon Brocard).
957 - clarified on_eof/on_error documentation a bit.
958
9594.231 Tue Jul 29 13:12:15 CEST 2008
960 - remove some debugging code left in AnyEvent::Util::fork_call
961 (and no, it's impossible to implement with the broken windows
962 perls without resource leaks or worse).
963
9644.23 Tue Jul 29 12:19:59 CEST 2008
965 - document the first parameter passed to condvar callbacks to be
966 the callback.
967 - add AnyEvent::Socket::{ntoa,aton} aliases.
968 - optimize the AE::Handle->push_read (line) for the default
969 eol marker.
970 - optimize push_read (packstring|storable) for small packets.
971 - invoke on_error callback when no on_eof callback is set.
972 - fix a bug in push_read (storable) of unknown impact.
973
9744.22 Sun Jul 20 16:34:13 CEST 2008
975 - new function AnyEvent::Socket::parse_hostport.
976 - as the bulkheads at microsoft can't even get getprotobyname reliably
977 working on their shitty fucking broken os we need to hardcode
978 some common protocol numbers in AnyEvent::Socket. How can
979 people even bother with such a pile of shit as windows.
980
9814.21 Thu Jul 17 14:40:05 CEST 2008
982 - INCOMPATIBLE CHANGE: AnyEvent::Util::fork_call now has a prototype
983 which is incompatible to previous invocation syntax.
984 - work around more windows perl bugs in fork_call. I cannot imagine
985 why anybody would bother implementing fork in such an obviously
986 unusable way.
987 - avoid calling $do_retry if the request has already finished in
988 AnyEvent::DNS, thus avoiding an exception (reported by Anatoly
989 K. Sharifulin).
990
9914.2 Sat Jul 12 22:42:11 CEST 2008
992 - support an "interval" argument for repeating timers.
993 - fix a bug in the pure perl event loop that caused timers to
994 be delayed under some circumstances.
995 - implement optional argument checking via PERL_ANYEVENT_STRICT.
996 - not importing anything from Event causes it to malfunction
997 and not use Time::HiRes, work around this bug.
998 - more bug workarounds against the endlessly broken Tk module,
999 seems to work now on 32 bit machines, but less so on 64 bit.
1000 - minor optimisations applied to most backends.
1001
10024.161 Fri Jul 4 14:24:48 CEST 2008
1003 - fixed the t/handle/02_write.t, it read after write, with the
1004 assumption that all data could be written before read
1005 (reported by Jim Razmus).
1006
10074.160 Thu Jul 3 04:02:21 CEST 2008
1008 - re-registering signal watchers after unregistering one did
1009 not work if a backend falls back on AnyEvent's default
1010 implementation (also affected child watchers).
1011 - new AnyEvent::Handle options: autocork and no_delay.
1012
34.152 Sun Jun 22 14:15:44 CEST 2008 10134.152 Sun Jun 22 14:15:44 CEST 2008
4 - allow for 32-bit perls that implement shifts difefrently 1014 - allow for 32-bit perls that implement shifts differently
5 on diferent architectures in parse_ipv4 (reported and 1015 on different architectures in parse_ipv4 (reported and
6 analysed by Keiichi DAIBA). 1016 analysed by Keiichi DAIBA).
7 1017
84.151 Fri Jun 6 17:34:24 CEST 2008 10184.151 Fri Jun 6 17:34:24 CEST 2008
9 - make sure specifying _only_ on_read and never pushing reads 1019 - make sure specifying _only_ on_read and never pushing reads
10 works. 1020 works.
… …
148 - support service names in tcp_connect and tcp_server. 1158 - support service names in tcp_connect and tcp_server.
149 - work around netbsd/openbsd bug causing unix sockets not to 1159 - work around netbsd/openbsd bug causing unix sockets not to
150 become writable until empty. 1160 become writable until empty.
151 - only make a dummy read for error detection when getpeername 1161 - only make a dummy read for error detection when getpeername
152 returns ENOTCONN. 1162 returns ENOTCONN.
153 - defend AnyEvent::Util::inet_aton against Coro::LWP brutal 1163 - defend AnyEvent::Util::inet_aton against Coro::LWP's brutal
154 override. 1164 override.
155 - new AnyEvent::Guard::cancel. 1165 - new AnyEvent::Guard::cancel.
156 1166
1573.5 Sat May 17 23:17:14 CEST 2008 11673.5 Sat May 17 23:17:14 CEST 2008
158 - gracefully deal with systems not defining CLOCK_MONOTONIC 1168 - gracefully deal with systems not defining CLOCK_MONOTONIC
… …
172 1182
1733.41 Sun May 11 19:53:13 CEST 2008 11833.41 Sun May 11 19:53:13 CEST 2008
174 - work around a bug in perl 5.8.8's local. Reported by Yi Ma Mao. 1184 - work around a bug in perl 5.8.8's local. Reported by Yi Ma Mao.
175 1185
1763.4 Sun May 11 00:29:25 CEST 2008 11863.4 Sun May 11 00:29:25 CEST 2008
177 - complete re-implementation of condvars with a hopefully much mroe useful 1187 - complete re-implementation of condvars with a hopefully much more useful
178 API (backward compatible functions exist but are undocumented). 1188 API (backward compatible functions exist but are undocumented).
179 - AnyEvent::Handle fully rewritten with a hopefully more powerful API 1189 - AnyEvent::Handle fully rewritten with a hopefully more powerful API
180 (still subject to change). 1190 (still subject to change).
181 - take advantage of CLOCK_MONOTONIC, if available, in AnyEvent::Impl::Perl. 1191 - take advantage of CLOCK_MONOTONIC, if available, in AnyEvent::Impl::Perl.
182 - provide AnyEvent::post_detect and @AnyEvent::post_detect, which 1192 - provide AnyEvent::post_detect and @AnyEvent::post_detect, which

Diff Legend

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