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

Comparing AnyEvent/Changes (file contents):
Revision 1.377 by root, Sat Jan 30 21:27:59 2010 UTC vs.
Revision 1.470 by root, Sun Aug 14 13:56:52 2011 UTC

1Revision history for Perl extension AnyEvent. 1Revision history for Perl extension AnyEvent.
2 2
3TODO: document TCP_*** constants
4TODO: catch AE::xxx for anyevent strict and anyevent debug
5TODO: anyevent::trace? or debug tracing all watchers?
6TODO: anyevent strict checking for $_ changes?
7
8TODO: IO::Async sometimes cancels the wrong timer when you call cancel_timer -wait for IO::Async release.
9TODO: wrap
10TODO: parse_unix???
11TODO: logging?
12TODO: check wrap for leaks
13TODO: check for $_ changes in strict?
14
156.01
16
176.0 (not officially released)
18 - INCOMPATIBLE CHANGE: AnyEvent::DNS resource records now include
19 the TTL as fourth element - this affects ->resolve directly,
20 and ->request and ->dns_unpack indirectly. Convenience functions
21 such as AnyEvent::DNS::{a,txt,mx,srv} are not affected, neither
22 is AnyEvent::Socket.
23 - INCOMPATIBLE CHANGE: backend authors now should not implement
24 one_event or loop, but instead the AnyEvent::CondVar::_wait and _poll
25 methods.
26 - DEPRECIATION: please change
27 "use AnyEvent::Impl::Perl" to "use AnyEvent::Loop" and
28 "AnyEvent::Impl::Perl::loop" to "AnyEvent::Loop::run"
29 in your programs, for future compatibility.
30 - *blush* idle watcher emulation was borked.
31 - AnyEvent::Strict now checks the AE::xxx functions as well.
32 - querying too long domain names would barf deep inside AnyEvent::DNS - since
33 this is hard to avoid for applications, AnyEvent:.DNS::request now checks for
34 for too long domain names explicitly and passes an undef to the callback
35 instead (reported by James Bromberger).
36 - make AnyEvent::Util more robust against EINTR (analyzed by Andrew Suffield).
37 - new function: AnyEvent::postpone.
38 - new env variable, PERL_ANYEVENT_DEBUG_SHELL, to automatically start
39 a debug shell.
40 - new env variable, PERL_ANYEVENT_DEBUG_WRAP, to automatically enable
41 watcher instrumentation/wrapping (see AnyEvent::Debug::wrap).
42 - IO::Async sometimes cancels the wrong timer when you call cancel_timer.
43 Invest extra overhead to work around this peculiar behaviour.
44 - split the pure perl backend into a loop module and an impl
45 module (AnyEvent::Loop and AnyEvent::Impl::Perl).
46 - some read types (regex, netstring, json, storable) could
47 errornously keep an AE::Handle object alive on their own
48 (reported by Mohammad Toossi).
49 - add more workarounds around child watcher bugs in IO::Async
50 and (newly introduced) in POE.
51 - use glib child watchers and try to work around its limitations,
52 if possible (based on a patch by Kevin Ryde).
53 - add support for FLTK (fltk2), via AnyEvent::Impl::FLTK2.
54 - many read types in AnyEvent::Handle would malfunction during
55 transitions to TLS because of a "delete $self->{rbuf}".
56 (reported by Gerald Galster).
57 - improve condvar blocking wait performance (probably only noticable
58 with faster backends), also make it easier for event loops that
59 do not support blocking waits.
60 - PERL_ANYEVENT_MODEL now also allows full module names, although
61 the usefulness of this feature is not yet clear.
62 - add tests for all supported event loops, conditional on PERL_ANYEVENT_LOOP_TESTS.
63 - add t/80_ssltext.t.
64 - no longer use AUTOLOAD for the initial loading (for no very good reason).
65 - tcp_server now tries to unlink unix domain sockets when it gets destroyed.
66 - do not run event loop detection multiple times when programs call their cached
67 copies of AnyEvent::detect.
68
695.34 Thu May 12 09:59:41 CEST 2011
70 - AE::TLS didn't support both cert_file and key at the same time,
71 due to some copy&paste bug (reported and analyzed by vti).
72 - improve AE::Handle timeout documentation slightly.
73
745.33 Mon Feb 7 21:51:41 CET 2011
75 - never appeared on CPAN, for unknown reasons.
76 - AnyEvent::Handle on_connect_error - do not try to destroy a handle
77 object when it's already gone (analyzed by Ryan Bullock).
78 - update idna algorithm(s) to UTS#46 version 6.0.0.
79 - preserve trailing dot in idn_to_ascii - testsuite says to preserve,
80 uts46 says it's an error, and testsuite makes more sense.
81 - remove leading dots in idn_to_ascii - testsuite says to remove,
82 uts46 says it's an error, and testsuite makes more sense.
83 - fixed a bug in idn ignored character processing.
84 - passes the uts46 6.0.0 testsuite except for three cases which
85 apparently are bugs in the testsuite itself.
86 - passes the uts46 6.0.1 testsuite except for eleven cases which
87 apparently are bugs in the testsuite itself.
88
895.31 Sun Jan 23 11:49:19 CET 2011
90 - work around a POE bug where POE would first poll for new events and
91 invoke handlers BEFORE handling the message queue.
92 - fix parsing of "domain" statement in resolv.conf (Johannes Plunien).
93 - stop_read now does just that in TLS mode. see its
94 documentation for caveats.
95 - update to common::sense 3.4 warning set (some warnings in perl impose
96 an arbitrary 32kb limit on regex matches, which affects
97 AnyEvent::Handle).
98 - AE::Handle had a broken rbuf_max method and missing wbuf_max method.
99 - new experimental AE::Handle::resettls method.
100
1015.3 Fri Dec 31 05:48:13 CET 2010
102 - major IOAsync patch by Paul Evans - 0.33 is required, as the major
103 issues with IO::Async have been solved in that version, including
104 the availabilty of a default loop!
105 - implement new wbuf_max parameter for AnyEvent::Handle.
106 - added Cocoa::EventLoop backend by Daisuke Murase.
107 - the default testsuite now honors PERL_ANYEVENT_MODEL (Paul Evans).
108 - clarify handle on_prepare documentation.
109 - set CLOEXEC flag on parent-side fd's in AnyEvent::Util::run_cmd,
110 to avoid leaking them to any child processes.
111 - try to force files into the arch-dependend lib directory, for the
112 half-"turly obsessive compulsive person"(s) that depend on it.
113
1145.29 Sun Dec 5 10:49:21 CET 2010
115 - convert EV backend to EV 4.00 API (so better upgrade EV too).
116 - AE::Handle: implement dynamic read_size adjustment, add
117 max_read_size parameter, reduce default read_size to 2048.
118 - add some FAQ entries for servers and/or on_eof vs. on_error.
119 - work around OS bugs (cygwin again) when lingering in AE::Handle.
120 - reduce memory usage slightly when connect was used in AE::Handle.
121 - wrok around more segfaults in Tk.
122 - document tls_ctx => undef as valid.
123 - detect WNOHANG value at compiletime.
124 - include some probably linux-only support for building AnyEvent
125 as part of the perl core.
126 - improve compatibility of t/08_idna.t to perl 5.8.
127 - make AnyEvent::Strict truly optional (in case somebody builds a
128 minimal perl...).
129
1305.28 Wed Oct 13 04:14:23 CEST 2010
131 - due to a glitch, AnyEvent's internal getprotobyname wasn't
132 actually used in tcp_server.
133 - implement AnyEvent::CondVar->new method, which was documented but
134 missing.
135 - support autoloading of read types in unshift_read.
136 - AnyEvent::Strict: do the test for invalid fh arguments differently,
137 resulting in more informative error messages.
138 - new AnyEvent::FAQ manpage.
139 - clarify that you can actually call push_read etc. in on_read.
140 - put keys and certs in /tmp, not into the current directory,
141 and document this in the SECURITY CONSIDERATIONS sections
142 of AnyEvent::TLS.
143 - support ";" as resolv.conf comment character in AnyEvent::DNS, as
144 requested by Ron Isaacson.
145 - document $AnyEvent::DNS::RESOLVER variable.
146 - incorporated major typo patches by Abhijit Menon-Sen.
147 - AnyEvent::Handle now croaks when negative timeouts are passed.
148 - add a shitload of TCP_xxx sockopt name constants to AnyEvent::Util,
149 undocumented, but who knows when they might come in handy.
150
1515.271 Tue Jun 8 12:05:46 CEST 2010
152 - backport to perl 5.8.x.
153
1545.27 Sun Jun 6 12:12:05 CEST 2010
155 - postpone differently in AnyEvent::Socket now, as
156 when not, canceling the connection attempt might fail
157 (found by Felix Antonius Wilhelm Ostmann).
158 - explicitly check for non-stream sockets in AE::Handle, too many
159 clueless people fell into the trap of this somehow working.
160 - simplified and reworked the "OTHER MODULES" section.
161 - better/more condvar examples.
162
1635.261 Wed Apr 28 16:13:36 CEST 2010
164 - AF_INET6 was not properly used from Socket6 during configuration
165 time (found by Andy Grundman).
166
1675.26 Mon Apr 12 04:49:35 CEST 2010
168 - don't generate spurious readyness notifications when select
169 returns EINTR in the pure perl backend (analysed and testcase
170 by Bas Denissen).
171 - give same examples in SYNOPSIS sections of AnyEvent and AE.
172 - provider faster implementations for AE::time/now/now_update for
173 EV and Perl backends.
174 - provide faster AE::cv and AE::time implementations for all
175 backends.
176 - no longer support lower/mixed-case signal names (this was
177 never documented nor universally supported).
178 - some more memory ยต-optimisations, and somewhat less messy
179 code to cope with both AnyEvent and AE APIs.
180
1815.251 Sat Mar 13 00:58:21 CET 2010
182 - make test failed if it was run before make install on perl <=
183 5.8.8, and older versions of perl make our live very difficult,
184 so write constants.pl during Makefile.PL time.
185
1865.25 Sat Mar 13 00:23:14 CET 2010
187 - fix a race condition in AnyEvent::Handle that would cause
188 a "bio_read: unsupported method" error (in _another_
189 ssl connection) after user code threw an exception.
190 - added AnyEvent::Handle->destroyed method.
3 - speed up AnyEvent::Socket::format_address and ::format_ipv6. 191 - speed up AnyEvent::Socket::format_address and ::format_ipv6.
192 - the AnyEvent::Util::fh_nonblocking function worked only
193 by lucky accident on win32.
4 - smaller and faster AnyEvent::Util::fh_nonblocking. 194 - smaller and faster AnyEvent::Util::fh_nonblocking.
195 - when the (required!) Time::HiRes module is missing, AnyEvent
196 did not fall back to built-in time properly.
197 - do not load Fcntl at runtime, saving memory and loading time.
198 - precompile a number of constants and use them instead of runtime
199 detection and eval.
200 - free detection code after detection and similar memory optimisations.
201 - Perl backend timer interval best effort drift has been improved
202 (same algorithm as EV).
203 - update unicode idna mapping table.
5 204
65.24 Tue Jan 5 11:39:43 CET 2010 2055.24 Tue Jan 5 11:39:43 CET 2010
7 - cygwin never reports errors from failed connects. 1.5 just gives 206 - cygwin never reports errors from failed connects. 1.5 just gives
8 you continous readyness and EAGAIN, 1.7 is even more broken and 207 you continous readyness and EAGAIN, 1.7 is even more broken and
9 just hangs. work around both issues in a cygwin-specific hack. 208 just hangs. work around both issues in a cygwin-specific hack.
… …
3854.3 Fri Oct 3 09:18:43 CEST 2008 5844.3 Fri Oct 3 09:18:43 CEST 2008
386 - AnyEvent will now install a no-op signal handler for SIGPIPE, 585 - AnyEvent will now install a no-op signal handler for SIGPIPE,
387 unless one has been installed already. 586 unless one has been installed already.
388 - warn about not loaidng AnyEvent::Impl::POE early enough 587 - warn about not loaidng AnyEvent::Impl::POE early enough
389 (patch by Adam Rosenstein). 588 (patch by Adam Rosenstein).
390 - fixed a great number of bugs and corner cases in AnyEvent::Handle: 589 - fixed a great number of bugs and corner cases in AnyEvent::Handle.
391 - fix a bug in where in SSL connect mode, the 590 - fix a bug in where in SSL connect mode, the
392 client would first wait for some data by the server and 591 client would first wait for some data by the server and
393 otherwise hang. (reported and analysed in an absolutely 592 otherwise hang. (reported and analysed in an absolutely
394 exemplary manner by Adam Rosenstein). 593 exemplary manner by Adam Rosenstein).
395 - fix a bug in where SSL EOF would not be treated 594 - fix a bug in where SSL EOF would not be treated

Diff Legend

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