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

Comparing AnyEvent/Changes (file contents):
Revision 1.341 by root, Tue Sep 1 10:40:05 2009 UTC vs.
Revision 1.430 by root, Wed Dec 29 05:04:51 2010 UTC

1Revision history for Perl extension AnyEvent. 1Revision history for Perl extension AnyEvent.
2 2
3TODO: pass values to anyevent condvar callback, make this count as recv. 3TODO: document TCP_*** constants
4TODO: resolv.conf nameserver blabla # comment does not work 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
8 - major IOAsync patch by Paul Evans - 0.33 is required, as the major
9 issues with IO::Async have been solved in that version, including
10 the availabilty of a default loop!
11 - implement new wbuf_max parameter for AnyEvent::Handle.
12 - added Cocoa::EventLoop backend by Daisuke Murase.
13 - the default testsuite now honors PERL_ANYEVENT_MODEL (Paul Evans).
14 - set CLOEXEC flag on parent-side fd's in AnyEvent::Util::run_cmd,
15 to avoid leaking them to any child processes.
16 - try to force files into the arch-dependend lib directory, for the
17 half-"turly obsessive compulsive person"(s) that depend on it.
18
195.29 Sun Dec 5 10:49:21 CET 2010
20 - convert EV backend to EV 4.00 API (so better upgrade EV too).
21 - AE::Handle: implement dynamic read_size adjustment, add
22 max_read_size parameter, reduce default read_size to 2048.
23 - add some FAQ entries for servers and/or on_eof vs. on_error.
24 - work around OS bugs (cygwin again) when lingering in AE::Handle.
25 - reduce memory usage slightly when connect was used in AE::Handle.
26 - wrok around more segfaults in Tk.
27 - document tls_ctx => undef as valid.
28 - detect WNOHANG value at compiletime.
29 - include some probably linux-only support for building AnyEvent
30 as part of the perl core.
31 - improve compatibility of t/08_idna.t to perl 5.8.
32 - make AnyEvent::Strict truly optional (in case somebody builds a
33 minimal perl...).
34
355.28 Wed Oct 13 04:14:23 CEST 2010
36 - due to a glitch, AnyEvent's internal getprotobyname wasn't
37 actually used in tcp_server.
38 - implement AnyEvent::CondVar->new method, which was documented but
39 missing.
40 - support autoloading of read types in unshift_read.
41 - AnyEvent::Strict: do the test for invalid fh arguments differently,
42 resulting in more informative error messages.
43 - new AnyEvent::FAQ manpage.
44 - clarify that you can actually call push_read etc. in on_read.
45 - put keys and certs in /tmp, not into the current directory,
46 and document this in the SECURITY CONSIDERATIONS sections
47 of AnyEvent::TLS.
48 - support ";" as resolv.conf comment character in AnyEvent::DNS, as
49 requested by Ron Isaacson.
50 - document $AnyEvent::DNS::RESOLVER variable.
51 - incorporated major typo patches by Abhijit Menon-Sen.
52 - AnyEvent::Handle now croaks when negative timeouts are passed.
53 - add a shitload of TCP_xxx sockopt name constants to AnyEvent::Util,
54 undocumented, but who knows when they might come in handy.
55
565.271 Tue Jun 8 12:05:46 CEST 2010
57 - backport to perl 5.8.x.
58
595.27 Sun Jun 6 12:12:05 CEST 2010
60 - postpone differently in AnyEvent::Socket now, as
61 when not, canceling the connection attempt might fail
62 (found by Felix Antonius Wilhelm Ostmann).
63 - explicitly check for non-stream sockets in AE::Handle, too many
64 clueless people fell into the trap of this somehow working.
65 - simplified and reworked the "OTHER MODULES" section.
66 - better/more condvar examples.
67
685.261 Wed Apr 28 16:13:36 CEST 2010
69 - AF_INET6 was not properly used from Socket6 during configuration
70 time (found by Andy Grundman).
71
725.26 Mon Apr 12 04:49:35 CEST 2010
73 - don't generate spurious readyness notifications when select
74 returns EINTR in the pure perl backend (analysed and testcase
75 by Bas Denissen).
76 - give same examples in SYNOPSIS sections of AnyEvent and AE.
77 - provider faster implementations for AE::time/now/now_update for
78 EV and Perl backends.
79 - provide faster AE::cv and AE::time implementations for all
80 backends.
81 - no longer support lower/mixed-case signal names (this was
82 never documented nor universally supported).
83 - some more memory ยต-optimisations, and somewhat less messy
84 code to cope with both AnyEvent and AE APIs.
85
865.251 Sat Mar 13 00:58:21 CET 2010
87 - make test failed if it was run before make install on perl <=
88 5.8.8, and older versions of perl make our live very difficult,
89 so write constants.pl during Makefile.PL time.
90
915.25 Sat Mar 13 00:23:14 CET 2010
92 - fix a race condition in AnyEvent::Handle that would cause
93 a "bio_read: unsupported method" error (in _another_
94 ssl connection) after user code threw an exception.
95 - added AnyEvent::Handle->destroyed method.
96 - speed up AnyEvent::Socket::format_address and ::format_ipv6.
97 - the AnyEvent::Util::fh_nonblocking function worked only
98 by lucky accident on win32.
99 - smaller and faster AnyEvent::Util::fh_nonblocking.
100 - when the (required!) Time::HiRes module is missing, AnyEvent
101 did not fall back to built-in time properly.
102 - do not load Fcntl at runtime, saving memory and loading time.
103 - precompile a number of constants and use them instead of runtime
104 detection and eval.
105 - free detection code after detection and similar memory optimisations.
106 - Perl backend timer interval best effort drift has been improved
107 (same algorithm as EV).
108 - update unicode idna mapping table.
109
1105.24 Tue Jan 5 11:39:43 CET 2010
111 - cygwin never reports errors from failed connects. 1.5 just gives
112 you continous readyness and EAGAIN, 1.7 is even more broken and
113 just hangs. work around both issues in a cygwin-specific hack.
114 - improve idle watcher documentation slightly.
115
1165.23 Sun Dec 20 23:48:00 CET 2009
117 - support IDNs in resolve_sockaddr, and therefore in tcp_connect.
118 - implement punycode_encode/decode, idn_nameprep,
119 idn_to_ascii and idn_to_unicode operations in AnyEvent::Util.
120 - provide $AE::VERSION.
121 - removed traces of "no strict 'refs'".
122
1235.22 Sat Dec 5 03:51:13 CET 2009
124 - downgrade-or-fail in AnyEvent::Handle::push_write, to
125 diagnose encoding failures earlier and more succinctly.
126 (this works around bugs in perl, throwing away encoding info
127 when passing scalar data to extensions).
128 - add more examples to AnyEvent::Socket manpage.
129 - upgrade internal warning set to the same as common::sense 2.03.
130 - use pack "n/a*" for pre-5.8.9 perl compatibility in AnyEvent::DNS
131 (John Beppu).
132 - AnyEvent::Socket::inet_aton now properly supports ipv6, as documented.
133 - add google public dns servers to fallback server set and make sure
134 we load-balance properly between the three sets. also add all
135 fallback dns servers, not just a random one, to each dns config.
136
1375.21 Thu Nov 19 02:48:47 CET 2009
138 - fix a problem where socket constants were called with parameters
139 (spotted by David Friedland).
140 - fork_call never use'd POSIX (reported by Daisuke Maki).
141 - improve perl 5.6 compatibility further (but it still won't work
142 unless you rip out everything but the core).
143 - prefer Net::DNS::Resolver over ipconfig, if installed, on win32.
144 uses 10MB of RAM, but doesn't open a console window. *sigh*.
145
1465.202 Wed Oct 14 22:35:44 CEST 2009
147 - AnyEvent::DNS would unexpectedly clobber $_ under windows
148 (analysed by Matthias Waldorf).
149 - AnyEvent::Handle::run_cmd can now pass the PID of the
150 newly-created process, which is much less useful than it might
151 sound (based on patch by Yann Kerherve).
152
1535.201 Tue Sep 29 12:09:25 CEST 2009
154 - AnyEvent:Handle::on_starttls/on_stoptls methods were broken
155 (reported by Torsten Foertsch).
156 - common::sense 2.0 could cause tcp_server to throw an exception
157 (analysed by elmex).
158
1595.2 Mon Sep 14 07:04:49 CEST 2009
160 - INCOMPATIBLE CHANGE: do no longer support register_read_type
161 and register_write_type in AnyEvent::Handle, instead support
162 package names (the facility was mostly abused).
163 - implement "packagename-as-read/write type" support in
164 AnyEvent::Handle.
165 - AnyEvent::Handle: new options "keepalive" and "oobinline".
166 - oobinline set by default to avoid security issues.
167 - the pure-perl event loop backend wrongly detected times() fallback
168 support (spotted by Pavel Boldin).
169
1705.12 Tue Sep 1 20:26:50 CEST 2009
171 - be more lenient when parsing resolv.conf files, as some people
172 use hashmarks as comment indicator inside directives (reported
173 by Michael S. Fischer).
174 - use same set of warnings as common::sense 2.0.
175 - fix a potential 32 bit overflow issue due to perl having problems
176 with large hex constants in the Perl backend.
177 - do not use the slower 5.6 version fo the enc_name code in AnyEvent::DNS
178 by default.
179 - fix some prototypes.
180 - (optionally) require Async::Interrupt 1.02.
181 - replace opendns fallbacks by verizon/level3 public dns servers.
182
55.112 Fri Aug 21 13:59:04 CEST 2009 1835.112 Fri Aug 21 13:59:04 CEST 2009
6 - AnyEvent::Handle could cause hard crashes in Net::SSLeay when 184 - AnyEvent::Handle could cause hard crashes in Net::SSLeay when
7 using starttls with an illegal/uncreatable tls context. 185 using starttls with an illegal/uncreatable tls context.
8 - port to broken win32 perls. 186 - port to broken win32 perls.
9 187
… …
79 - new function: AnyEvent::Socket::format_hostport. 257 - new function: AnyEvent::Socket::format_hostport.
80 258
814.881 Tue Jul 28 12:51:53 CEST 2009 2594.881 Tue Jul 28 12:51:53 CEST 2009
82 - work around a bug in local in pre-5.10 perls, causing 260 - work around a bug in local in pre-5.10 perls, causing
83 AnyEvent::Handle to recurse when it should not 261 AnyEvent::Handle to recurse when it should not
84 (analyzed by elmex). 262 (analysed by elmex).
85 263
864.88 Tue Jul 28 04:04:37 CEST 2009 2644.88 Tue Jul 28 04:04:37 CEST 2009
87 - re-bless the handle into a dummy package after calling 265 - re-bless the handle into a dummy package after calling
88 AnyEvent::Handle::destroy, so the user does not need to check 266 AnyEvent::Handle::destroy, so the user does not need to check
89 for errors after every push_write etc. 267 for errors after every push_write etc.
… …
3114.3 Fri Oct 3 09:18:43 CEST 2008 4894.3 Fri Oct 3 09:18:43 CEST 2008
312 - AnyEvent will now install a no-op signal handler for SIGPIPE, 490 - AnyEvent will now install a no-op signal handler for SIGPIPE,
313 unless one has been installed already. 491 unless one has been installed already.
314 - warn about not loaidng AnyEvent::Impl::POE early enough 492 - warn about not loaidng AnyEvent::Impl::POE early enough
315 (patch by Adam Rosenstein). 493 (patch by Adam Rosenstein).
316 - fixed a great number of bugs and corner cases in AnyEvent::Handle: 494 - fixed a great number of bugs and corner cases in AnyEvent::Handle.
317 - fix a bug in where in SSL connect mode, the 495 - fix a bug in where in SSL connect mode, the
318 client would first wait for some data by the server and 496 client would first wait for some data by the server and
319 otherwise hang. (reported and analyzed in an absolutely 497 otherwise hang. (reported and analysed in an absolutely
320 exemplary manner by Adam Rosenstein). 498 exemplary manner by Adam Rosenstein).
321 - fix a bug in where SSL EOF would not be treated 499 - fix a bug in where SSL EOF would not be treated
322 as stream EOF, putting the connection into a hung state. 500 as stream EOF, putting the connection into a hung state.
323 (reported and analyzed in an absolutely exemplary manner by Adam 501 (reported and analysed in an absolutely exemplary manner by Adam
324 Rosenstein). 502 Rosenstein).
325 - fix a potential segfault when the TLS context 503 - fix a potential segfault when the TLS context
326 would go missing in a read callback (e.g. due to stoptls) - 504 would go missing in a read callback (e.g. due to stoptls) -
327 Net::SSLeay of course makes no type checking whatsoever. 505 Net::SSLeay of course makes no type checking whatsoever.
328 - AnyEvent::Handle will not stop the read watcher in TLS mode. 506 - AnyEvent::Handle will not stop the read watcher in TLS mode.

Diff Legend

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