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

Comparing AnyEvent/Changes (file contents):
Revision 1.346 by root, Thu Sep 3 12:45:35 2009 UTC vs.
Revision 1.462 by root, Sat Aug 13 02:35:32 2011 UTC

1Revision history for Perl extension AnyEvent. 1Revision history for Perl extension AnyEvent.
2 2
3TODO: packagename for ae types 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?
4 7
8TODO: redesign Impl
9TODO: make ae::run mroe robust against eintr, Andrew Suffield
10TODO: IO::Async sometimes cancels the wrong timer when you call cancel_timer -wait for IO::Async release.
11TODO: wrap
12TODO: shell on env
13TODO: parse_unix???
14TODO: debug shell pid?
15 - INCOMPATIBLE CHANGE: AnyEvent::DNS resource records now include
16 the TTL as fourth element - this affects ->resolve directly,
17 and ->request and ->dns_unpack indirectly. Convenience functions
18 such as AnyEvent::DNS::{a,txt,mx,srv} are not affected.
19 - INCOMPATIBLE CHANGE: backend authors now should not implement
20 one_event or loop, but instead the AnyEvent::CondVar::_wait and _poll
21 methods.
22 - DEPRECIATION: please change
23 "use AnyEvent::Impl::Perl" to "use AnyEvent::Loop" and
24 "AnyEvent::Impl::Perl::loop" to "AnyEvent::Loop::run"
25 in your programs, for future compatibility.
26 - *blush* idle watcher emulation was borked.
27 - AnyEvent::Strict now checks the AE::xxx functions as well.
28 - querying too long domain names would barf deep inside AnyEvent::DNS - since
29 this is hard to avoid for applications, AnyEvent:.DNS::request now checks for
30 for too long domain names explicitly and passes an undef to the callback
31 instead (reported by James Bromberger).
32 - new function: AnyEvent::postpone.
33 - IO::Async sometimes cancels the wrong timer when you call cancel_timer.
34 Invest extra overhead to work around this peculiar behaviour.
35 - split the pure perl backend into a loop module and an impl
36 module (AnyEvent::Loop and AnyEvent::Impl::Perl).
37 - some read types (regex, netstring, json, storable) could
38 errornously keep an AE::Handle object alive on their own
39 (reported by Mohammad Toossi).
40 - add more workarounds around child watcher bugs in IO::Async
41 and (newly introduced) in POE.
42 - use glib child watchers and try to work around its limitations,
43 if possible (based on a patch by Kevin Ryde).
44 - add support for FLTK (fltk2), via AnyEvent::Impl::FLTK2.
45 - many read types in AnyEvent::Handle would malfunction during
46 transitions to TLS because of a "delete $self->{rbuf}".
47 (reported by Gerald Galster).
48 - resolve AnyEvent::Util::guard at first-use-time, not at load time,
49 for further speed and memory savings when not used.
50 - improve condvar blocking wait performance (probably only noticable
51 with faster backends), also make it easier for event loops that
52 do not support blocking waits.
53 - PERL_ANYEVENT_MODEL now also allows full module names, although
54 the usefulness of this feature is not yet clear.
55 - add tests for all supported event loops, conditional on PERL_ANYEVENT_LOOP_TESTS.
56 - add t/80_ssltext.t.
57 - no longer use AUTOLOAD for the initial loading (for no very good reason).
58
595.34 Thu May 12 09:59:41 CEST 2011
60 - AE::TLS didn't support both cert_file and key at the same time,
61 due to some copy&paste bug (reported and analyzed by vti).
62 - improve AE::Handle timeout documentation slightly.
63
645.33 Mon Feb 7 21:51:41 CET 2011
65 - never appeared on CPAN, for unknown reasons.
66 - AnyEvent::Handle on_connect_error - do not try to destroy a handle
67 object when it's already gone (analyzed by Ryan Bullock).
68 - update idna algorithm(s) to UTS#46 version 6.0.0.
69 - preserve trailing dot in idn_to_ascii - testsuite says to preserve,
70 uts46 says it's an error, and testsuite makes more sense.
71 - remove leading dots in idn_to_ascii - testsuite says to remove,
72 uts46 says it's an error, and testsuite makes more sense.
73 - fixed a bug in idn ignored character processing.
74 - passes the uts46 6.0.0 testsuite except for three cases which
75 apparently are bugs in the testsuite itself.
76 - passes the uts46 6.0.1 testsuite except for eleven cases which
77 apparently are bugs in the testsuite itself.
78
795.31 Sun Jan 23 11:49:19 CET 2011
80 - work around a POE bug where POE would first poll for new events and
81 invoke handlers BEFORE handling the message queue.
82 - fix parsing of "domain" statement in resolv.conf (Johannes Plunien).
83 - stop_read now does just that in TLS mode. see its
84 documentation for caveats.
85 - update to common::sense 3.4 warning set (some warnings in perl impose
86 an arbitrary 32kb limit on regex matches, which affects
87 AnyEvent::Handle).
88 - AE::Handle had a broken rbuf_max method and missing wbuf_max method.
89 - new experimental AE::Handle::resettls method.
90
915.3 Fri Dec 31 05:48:13 CET 2010
92 - major IOAsync patch by Paul Evans - 0.33 is required, as the major
93 issues with IO::Async have been solved in that version, including
94 the availabilty of a default loop!
95 - implement new wbuf_max parameter for AnyEvent::Handle.
96 - added Cocoa::EventLoop backend by Daisuke Murase.
97 - the default testsuite now honors PERL_ANYEVENT_MODEL (Paul Evans).
98 - clarify handle on_prepare documentation.
99 - set CLOEXEC flag on parent-side fd's in AnyEvent::Util::run_cmd,
100 to avoid leaking them to any child processes.
101 - try to force files into the arch-dependend lib directory, for the
102 half-"turly obsessive compulsive person"(s) that depend on it.
103
1045.29 Sun Dec 5 10:49:21 CET 2010
105 - convert EV backend to EV 4.00 API (so better upgrade EV too).
106 - AE::Handle: implement dynamic read_size adjustment, add
107 max_read_size parameter, reduce default read_size to 2048.
108 - add some FAQ entries for servers and/or on_eof vs. on_error.
109 - work around OS bugs (cygwin again) when lingering in AE::Handle.
110 - reduce memory usage slightly when connect was used in AE::Handle.
111 - wrok around more segfaults in Tk.
112 - document tls_ctx => undef as valid.
113 - detect WNOHANG value at compiletime.
114 - include some probably linux-only support for building AnyEvent
115 as part of the perl core.
116 - improve compatibility of t/08_idna.t to perl 5.8.
117 - make AnyEvent::Strict truly optional (in case somebody builds a
118 minimal perl...).
119
1205.28 Wed Oct 13 04:14:23 CEST 2010
121 - due to a glitch, AnyEvent's internal getprotobyname wasn't
122 actually used in tcp_server.
123 - implement AnyEvent::CondVar->new method, which was documented but
124 missing.
125 - support autoloading of read types in unshift_read.
126 - AnyEvent::Strict: do the test for invalid fh arguments differently,
127 resulting in more informative error messages.
128 - new AnyEvent::FAQ manpage.
129 - clarify that you can actually call push_read etc. in on_read.
130 - put keys and certs in /tmp, not into the current directory,
131 and document this in the SECURITY CONSIDERATIONS sections
132 of AnyEvent::TLS.
133 - support ";" as resolv.conf comment character in AnyEvent::DNS, as
134 requested by Ron Isaacson.
135 - document $AnyEvent::DNS::RESOLVER variable.
136 - incorporated major typo patches by Abhijit Menon-Sen.
137 - AnyEvent::Handle now croaks when negative timeouts are passed.
138 - add a shitload of TCP_xxx sockopt name constants to AnyEvent::Util,
139 undocumented, but who knows when they might come in handy.
140
1415.271 Tue Jun 8 12:05:46 CEST 2010
142 - backport to perl 5.8.x.
143
1445.27 Sun Jun 6 12:12:05 CEST 2010
145 - postpone differently in AnyEvent::Socket now, as
146 when not, canceling the connection attempt might fail
147 (found by Felix Antonius Wilhelm Ostmann).
148 - explicitly check for non-stream sockets in AE::Handle, too many
149 clueless people fell into the trap of this somehow working.
150 - simplified and reworked the "OTHER MODULES" section.
151 - better/more condvar examples.
152
1535.261 Wed Apr 28 16:13:36 CEST 2010
154 - AF_INET6 was not properly used from Socket6 during configuration
155 time (found by Andy Grundman).
156
1575.26 Mon Apr 12 04:49:35 CEST 2010
158 - don't generate spurious readyness notifications when select
159 returns EINTR in the pure perl backend (analysed and testcase
160 by Bas Denissen).
161 - give same examples in SYNOPSIS sections of AnyEvent and AE.
162 - provider faster implementations for AE::time/now/now_update for
163 EV and Perl backends.
164 - provide faster AE::cv and AE::time implementations for all
165 backends.
166 - no longer support lower/mixed-case signal names (this was
167 never documented nor universally supported).
168 - some more memory ยต-optimisations, and somewhat less messy
169 code to cope with both AnyEvent and AE APIs.
170
1715.251 Sat Mar 13 00:58:21 CET 2010
172 - make test failed if it was run before make install on perl <=
173 5.8.8, and older versions of perl make our live very difficult,
174 so write constants.pl during Makefile.PL time.
175
1765.25 Sat Mar 13 00:23:14 CET 2010
177 - fix a race condition in AnyEvent::Handle that would cause
178 a "bio_read: unsupported method" error (in _another_
179 ssl connection) after user code threw an exception.
180 - added AnyEvent::Handle->destroyed method.
181 - speed up AnyEvent::Socket::format_address and ::format_ipv6.
182 - the AnyEvent::Util::fh_nonblocking function worked only
183 by lucky accident on win32.
184 - smaller and faster AnyEvent::Util::fh_nonblocking.
185 - when the (required!) Time::HiRes module is missing, AnyEvent
186 did not fall back to built-in time properly.
187 - do not load Fcntl at runtime, saving memory and loading time.
188 - precompile a number of constants and use them instead of runtime
189 detection and eval.
190 - free detection code after detection and similar memory optimisations.
191 - Perl backend timer interval best effort drift has been improved
192 (same algorithm as EV).
193 - update unicode idna mapping table.
194
1955.24 Tue Jan 5 11:39:43 CET 2010
196 - cygwin never reports errors from failed connects. 1.5 just gives
197 you continous readyness and EAGAIN, 1.7 is even more broken and
198 just hangs. work around both issues in a cygwin-specific hack.
199 - improve idle watcher documentation slightly.
200
2015.23 Sun Dec 20 23:48:00 CET 2009
202 - support IDNs in resolve_sockaddr, and therefore in tcp_connect.
203 - implement punycode_encode/decode, idn_nameprep,
204 idn_to_ascii and idn_to_unicode operations in AnyEvent::Util.
205 - provide $AE::VERSION.
206 - removed traces of "no strict 'refs'".
207
2085.22 Sat Dec 5 03:51:13 CET 2009
209 - downgrade-or-fail in AnyEvent::Handle::push_write, to
210 diagnose encoding failures earlier and more succinctly.
211 (this works around bugs in perl, throwing away encoding info
212 when passing scalar data to extensions).
213 - add more examples to AnyEvent::Socket manpage.
214 - upgrade internal warning set to the same as common::sense 2.03.
215 - use pack "n/a*" for pre-5.8.9 perl compatibility in AnyEvent::DNS
216 (John Beppu).
217 - AnyEvent::Socket::inet_aton now properly supports ipv6, as documented.
218 - add google public dns servers to fallback server set and make sure
219 we load-balance properly between the three sets. also add all
220 fallback dns servers, not just a random one, to each dns config.
221
2225.21 Thu Nov 19 02:48:47 CET 2009
223 - fix a problem where socket constants were called with parameters
224 (spotted by David Friedland).
225 - fork_call never use'd POSIX (reported by Daisuke Maki).
226 - improve perl 5.6 compatibility further (but it still won't work
227 unless you rip out everything but the core).
228 - prefer Net::DNS::Resolver over ipconfig, if installed, on win32.
229 uses 10MB of RAM, but doesn't open a console window. *sigh*.
230
2315.202 Wed Oct 14 22:35:44 CEST 2009
232 - AnyEvent::DNS would unexpectedly clobber $_ under windows
233 (analysed by Matthias Waldorf).
234 - AnyEvent::Handle::run_cmd can now pass the PID of the
235 newly-created process, which is much less useful than it might
236 sound (based on patch by Yann Kerherve).
237
2385.201 Tue Sep 29 12:09:25 CEST 2009
239 - AnyEvent:Handle::on_starttls/on_stoptls methods were broken
240 (reported by Torsten Foertsch).
241 - common::sense 2.0 could cause tcp_server to throw an exception
242 (analysed by elmex).
243
2445.2 Mon Sep 14 07:04:49 CEST 2009
245 - INCOMPATIBLE CHANGE: do no longer support register_read_type
246 and register_write_type in AnyEvent::Handle, instead support
247 package names (the facility was mostly abused).
248 - implement "packagename-as-read/write type" support in
249 AnyEvent::Handle.
5 - AnyEvent::Handle: new options "keepalive" and "oobinline". 250 - AnyEvent::Handle: new options "keepalive" and "oobinline".
6 - oobinline set by default to avoid security issues. 251 - oobinline set by default to avoid security issues.
252 - the pure-perl event loop backend wrongly detected times() fallback
253 support (spotted by Pavel Boldin).
7 254
85.12 Tue Sep 1 20:26:50 CEST 2009 2555.12 Tue Sep 1 20:26:50 CEST 2009
9 - be more lenient when parsing resolv.conf files, as some people 256 - be more lenient when parsing resolv.conf files, as some people
10 use hashmarks as comment indicator inside directives (reported 257 use hashmarks as comment indicator inside directives (reported
11 by Michael S. Fischer). 258 by Michael S. Fischer).
… …
95 - new function: AnyEvent::Socket::format_hostport. 342 - new function: AnyEvent::Socket::format_hostport.
96 343
974.881 Tue Jul 28 12:51:53 CEST 2009 3444.881 Tue Jul 28 12:51:53 CEST 2009
98 - work around a bug in local in pre-5.10 perls, causing 345 - work around a bug in local in pre-5.10 perls, causing
99 AnyEvent::Handle to recurse when it should not 346 AnyEvent::Handle to recurse when it should not
100 (analyzed by elmex). 347 (analysed by elmex).
101 348
1024.88 Tue Jul 28 04:04:37 CEST 2009 3494.88 Tue Jul 28 04:04:37 CEST 2009
103 - re-bless the handle into a dummy package after calling 350 - re-bless the handle into a dummy package after calling
104 AnyEvent::Handle::destroy, so the user does not need to check 351 AnyEvent::Handle::destroy, so the user does not need to check
105 for errors after every push_write etc. 352 for errors after every push_write etc.
… …
3274.3 Fri Oct 3 09:18:43 CEST 2008 5744.3 Fri Oct 3 09:18:43 CEST 2008
328 - AnyEvent will now install a no-op signal handler for SIGPIPE, 575 - AnyEvent will now install a no-op signal handler for SIGPIPE,
329 unless one has been installed already. 576 unless one has been installed already.
330 - warn about not loaidng AnyEvent::Impl::POE early enough 577 - warn about not loaidng AnyEvent::Impl::POE early enough
331 (patch by Adam Rosenstein). 578 (patch by Adam Rosenstein).
332 - fixed a great number of bugs and corner cases in AnyEvent::Handle: 579 - fixed a great number of bugs and corner cases in AnyEvent::Handle.
333 - fix a bug in where in SSL connect mode, the 580 - fix a bug in where in SSL connect mode, the
334 client would first wait for some data by the server and 581 client would first wait for some data by the server and
335 otherwise hang. (reported and analyzed in an absolutely 582 otherwise hang. (reported and analysed in an absolutely
336 exemplary manner by Adam Rosenstein). 583 exemplary manner by Adam Rosenstein).
337 - fix a bug in where SSL EOF would not be treated 584 - fix a bug in where SSL EOF would not be treated
338 as stream EOF, putting the connection into a hung state. 585 as stream EOF, putting the connection into a hung state.
339 (reported and analyzed in an absolutely exemplary manner by Adam 586 (reported and analysed in an absolutely exemplary manner by Adam
340 Rosenstein). 587 Rosenstein).
341 - fix a potential segfault when the TLS context 588 - fix a potential segfault when the TLS context
342 would go missing in a read callback (e.g. due to stoptls) - 589 would go missing in a read callback (e.g. due to stoptls) -
343 Net::SSLeay of course makes no type checking whatsoever. 590 Net::SSLeay of course makes no type checking whatsoever.
344 - AnyEvent::Handle will not stop the read watcher in TLS mode. 591 - AnyEvent::Handle will not stop the read watcher in TLS mode.

Diff Legend

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