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

Comparing AnyEvent/Changes (file contents):
Revision 1.318 by root, Tue Aug 4 12:38:55 2009 UTC vs.
Revision 1.439 by root, Sun Jan 23 11:15:09 2011 UTC

1Revision history for Perl extension AnyEvent. 1Revision history for Perl extension AnyEvent.
2 2
3TODO: Danga::Socket backend. 3TODO: document TCP_*** constants
4TODO: delay tcp_connect 4TODO: catch AE::xxx for anyevent strict and anyevent debug
5TODO: anyevent::trace? or debug tracing all watchers?
6TODO: anyevent strict checking for $_ changes?
5 7
85.31 Sun Jan 23 11:49:19 CET 2011
9 - work around a POE bug where POE would first poll for new events and
10 invoke handlers BEFORE handling the message queue.
11 - fix parsing of "domain" statement in resolv.conf (Johannes Plunien).
12 - stop_read now does just that in TLS mode. see its
13 documentation for caveats.
14 - update to common::sense 3.4 warning set.
15 - AE::Handle had a broken rbuf_max method and missing wbuf_max method.
16 - new experimental AE::Handle::resettls method.
17
185.3 Fri Dec 31 05:48:13 CET 2010
19 - major IOAsync patch by Paul Evans - 0.33 is required, as the major
20 issues with IO::Async have been solved in that version, including
21 the availabilty of a default loop!
22 - implement new wbuf_max parameter for AnyEvent::Handle.
23 - added Cocoa::EventLoop backend by Daisuke Murase.
24 - the default testsuite now honors PERL_ANYEVENT_MODEL (Paul Evans).
25 - clarify handle on_prepare documentation.
26 - set CLOEXEC flag on parent-side fd's in AnyEvent::Util::run_cmd,
27 to avoid leaking them to any child processes.
28 - try to force files into the arch-dependend lib directory, for the
29 half-"turly obsessive compulsive person"(s) that depend on it.
30
315.29 Sun Dec 5 10:49:21 CET 2010
32 - convert EV backend to EV 4.00 API (so better upgrade EV too).
33 - AE::Handle: implement dynamic read_size adjustment, add
34 max_read_size parameter, reduce default read_size to 2048.
35 - add some FAQ entries for servers and/or on_eof vs. on_error.
36 - work around OS bugs (cygwin again) when lingering in AE::Handle.
37 - reduce memory usage slightly when connect was used in AE::Handle.
38 - wrok around more segfaults in Tk.
39 - document tls_ctx => undef as valid.
40 - detect WNOHANG value at compiletime.
41 - include some probably linux-only support for building AnyEvent
42 as part of the perl core.
43 - improve compatibility of t/08_idna.t to perl 5.8.
44 - make AnyEvent::Strict truly optional (in case somebody builds a
45 minimal perl...).
46
475.28 Wed Oct 13 04:14:23 CEST 2010
48 - due to a glitch, AnyEvent's internal getprotobyname wasn't
49 actually used in tcp_server.
50 - implement AnyEvent::CondVar->new method, which was documented but
51 missing.
52 - support autoloading of read types in unshift_read.
53 - AnyEvent::Strict: do the test for invalid fh arguments differently,
54 resulting in more informative error messages.
55 - new AnyEvent::FAQ manpage.
56 - clarify that you can actually call push_read etc. in on_read.
57 - put keys and certs in /tmp, not into the current directory,
58 and document this in the SECURITY CONSIDERATIONS sections
59 of AnyEvent::TLS.
60 - support ";" as resolv.conf comment character in AnyEvent::DNS, as
61 requested by Ron Isaacson.
62 - document $AnyEvent::DNS::RESOLVER variable.
63 - incorporated major typo patches by Abhijit Menon-Sen.
64 - AnyEvent::Handle now croaks when negative timeouts are passed.
65 - add a shitload of TCP_xxx sockopt name constants to AnyEvent::Util,
66 undocumented, but who knows when they might come in handy.
67
685.271 Tue Jun 8 12:05:46 CEST 2010
69 - backport to perl 5.8.x.
70
715.27 Sun Jun 6 12:12:05 CEST 2010
72 - postpone differently in AnyEvent::Socket now, as
73 when not, canceling the connection attempt might fail
74 (found by Felix Antonius Wilhelm Ostmann).
75 - explicitly check for non-stream sockets in AE::Handle, too many
76 clueless people fell into the trap of this somehow working.
77 - simplified and reworked the "OTHER MODULES" section.
78 - better/more condvar examples.
79
805.261 Wed Apr 28 16:13:36 CEST 2010
81 - AF_INET6 was not properly used from Socket6 during configuration
82 time (found by Andy Grundman).
83
845.26 Mon Apr 12 04:49:35 CEST 2010
85 - don't generate spurious readyness notifications when select
86 returns EINTR in the pure perl backend (analysed and testcase
87 by Bas Denissen).
88 - give same examples in SYNOPSIS sections of AnyEvent and AE.
89 - provider faster implementations for AE::time/now/now_update for
90 EV and Perl backends.
91 - provide faster AE::cv and AE::time implementations for all
92 backends.
93 - no longer support lower/mixed-case signal names (this was
94 never documented nor universally supported).
95 - some more memory ยต-optimisations, and somewhat less messy
96 code to cope with both AnyEvent and AE APIs.
97
985.251 Sat Mar 13 00:58:21 CET 2010
99 - make test failed if it was run before make install on perl <=
100 5.8.8, and older versions of perl make our live very difficult,
101 so write constants.pl during Makefile.PL time.
102
1035.25 Sat Mar 13 00:23:14 CET 2010
104 - fix a race condition in AnyEvent::Handle that would cause
105 a "bio_read: unsupported method" error (in _another_
106 ssl connection) after user code threw an exception.
107 - added AnyEvent::Handle->destroyed method.
108 - speed up AnyEvent::Socket::format_address and ::format_ipv6.
109 - the AnyEvent::Util::fh_nonblocking function worked only
110 by lucky accident on win32.
111 - smaller and faster AnyEvent::Util::fh_nonblocking.
112 - when the (required!) Time::HiRes module is missing, AnyEvent
113 did not fall back to built-in time properly.
114 - do not load Fcntl at runtime, saving memory and loading time.
115 - precompile a number of constants and use them instead of runtime
116 detection and eval.
117 - free detection code after detection and similar memory optimisations.
118 - Perl backend timer interval best effort drift has been improved
119 (same algorithm as EV).
120 - update unicode idna mapping table.
121
1225.24 Tue Jan 5 11:39:43 CET 2010
123 - cygwin never reports errors from failed connects. 1.5 just gives
124 you continous readyness and EAGAIN, 1.7 is even more broken and
125 just hangs. work around both issues in a cygwin-specific hack.
126 - improve idle watcher documentation slightly.
127
1285.23 Sun Dec 20 23:48:00 CET 2009
129 - support IDNs in resolve_sockaddr, and therefore in tcp_connect.
130 - implement punycode_encode/decode, idn_nameprep,
131 idn_to_ascii and idn_to_unicode operations in AnyEvent::Util.
132 - provide $AE::VERSION.
133 - removed traces of "no strict 'refs'".
134
1355.22 Sat Dec 5 03:51:13 CET 2009
136 - downgrade-or-fail in AnyEvent::Handle::push_write, to
137 diagnose encoding failures earlier and more succinctly.
138 (this works around bugs in perl, throwing away encoding info
139 when passing scalar data to extensions).
140 - add more examples to AnyEvent::Socket manpage.
141 - upgrade internal warning set to the same as common::sense 2.03.
142 - use pack "n/a*" for pre-5.8.9 perl compatibility in AnyEvent::DNS
143 (John Beppu).
144 - AnyEvent::Socket::inet_aton now properly supports ipv6, as documented.
145 - add google public dns servers to fallback server set and make sure
146 we load-balance properly between the three sets. also add all
147 fallback dns servers, not just a random one, to each dns config.
148
1495.21 Thu Nov 19 02:48:47 CET 2009
150 - fix a problem where socket constants were called with parameters
151 (spotted by David Friedland).
152 - fork_call never use'd POSIX (reported by Daisuke Maki).
153 - improve perl 5.6 compatibility further (but it still won't work
154 unless you rip out everything but the core).
155 - prefer Net::DNS::Resolver over ipconfig, if installed, on win32.
156 uses 10MB of RAM, but doesn't open a console window. *sigh*.
157
1585.202 Wed Oct 14 22:35:44 CEST 2009
159 - AnyEvent::DNS would unexpectedly clobber $_ under windows
160 (analysed by Matthias Waldorf).
161 - AnyEvent::Handle::run_cmd can now pass the PID of the
162 newly-created process, which is much less useful than it might
163 sound (based on patch by Yann Kerherve).
164
1655.201 Tue Sep 29 12:09:25 CEST 2009
166 - AnyEvent:Handle::on_starttls/on_stoptls methods were broken
167 (reported by Torsten Foertsch).
168 - common::sense 2.0 could cause tcp_server to throw an exception
169 (analysed by elmex).
170
1715.2 Mon Sep 14 07:04:49 CEST 2009
172 - INCOMPATIBLE CHANGE: do no longer support register_read_type
173 and register_write_type in AnyEvent::Handle, instead support
174 package names (the facility was mostly abused).
175 - implement "packagename-as-read/write type" support in
176 AnyEvent::Handle.
177 - AnyEvent::Handle: new options "keepalive" and "oobinline".
178 - oobinline set by default to avoid security issues.
179 - the pure-perl event loop backend wrongly detected times() fallback
180 support (spotted by Pavel Boldin).
181
1825.12 Tue Sep 1 20:26:50 CEST 2009
183 - be more lenient when parsing resolv.conf files, as some people
184 use hashmarks as comment indicator inside directives (reported
185 by Michael S. Fischer).
186 - use same set of warnings as common::sense 2.0.
187 - fix a potential 32 bit overflow issue due to perl having problems
188 with large hex constants in the Perl backend.
189 - do not use the slower 5.6 version fo the enc_name code in AnyEvent::DNS
190 by default.
191 - fix some prototypes.
192 - (optionally) require Async::Interrupt 1.02.
193 - replace opendns fallbacks by verizon/level3 public dns servers.
194
1955.112 Fri Aug 21 13:59:04 CEST 2009
196 - AnyEvent::Handle could cause hard crashes in Net::SSLeay when
197 using starttls with an illegal/uncreatable tls context.
198 - port to broken win32 perls.
199
2005.111 Sun Aug 16 18:44:36 CEST 2009
201 - perl errornously requires sockaddr_un structures returned by the
202 kernel to have a specific length. work around this bug by padding
203 sockaddr_un structures (found to fail on OS X, as perl only uses
204 the correct code for GNU/Linux, but likely to fail on other
205 OSes as well).
206
2075.11 Wed Aug 12 17:49:37 CEST 2009
208 - fixed a bug in run_cmd with input from filename.
209 - create a json coder object when none is given in write direction
210 to reduce dependency on JSON.
211
2125.1 Tue Aug 11 03:17:32 CEST 2009
213 - tcp_connect now reports the correct error in $!, not an unrelated
214 one (was broken since 4.91).
215 - AE::Handle did not properly clear rtimeout/wtimeout watchers
216 when retrying a tcp_connect.
217 - new functions: AnyEvent::Util::run_cmd and close_all_fds_except.
218 - fixed wrong documentation in AE::Socket.
219
2205.01 Mon Aug 10 03:16:32 CEST 2009
221 - last release broke EV child watchers :/.
222
2235.0 Sun Aug 9 17:32:09 CEST 2009
224 - officially introduce and document the AE API.
225 - lowering the timeout at runtime did not have immediate effect
226 in AnyEvent::Handle.
227 - AE::Handle now has separate and independent read and write timeout
228 settings.
229 - probe for Guard module when AnyEvent::Util is loaded, not at
230 runtime (Event doesn't like eval "use" from callbacks, and it avoids
231 exporting the probe function).
232 - Impl::IOAsync's io watchers did not use the filehandle emulation
233 code to work around it's watcher limits.
234 - work around IO::Async's problems with overloaded objects.
235 - try to disable wearnings inside IO::Async, it's just too buggy.
236 - do no longer use Event by default, only EV or Perl, Event is too buggy.
237 - hackishly provide fast AE:: interface when EV or Perl is the backend.
238 - inherit VERSION from AnyEvent::VERSION in most modules.
239
2404.91 Thu Aug 6 15:42:45 CEST 2009
241 - AE::Handle::starttls could get out of sync when the read buffer
242 already contains some TLS handshake.
6 - AE::Handle did not properly free the TLS session, causing bigger 243 - AE::Handle did not properly free the TLS session, causing bigger
7 memory leaks in Net::SSLeay than expected (reported by toaster). 244 memory leaks in Net::SSLeay than expected (reported by toaster).
245 - AE::Socket::tcp_connect will now artificially delay invoking
246 the callback to avoid returning after invoking the callback.
247 - convert many internal watcher uses to AE API.
8 248
94.9 Sat Aug 1 11:07:01 CEST 2009 2494.9 Sat Aug 1 11:07:01 CEST 2009
10 - Glib silently fails when registering a timeout with 250 - Glib silently fails when registering a timeout with
11 a negative value, so avoid that. 251 a negative value, so avoid that.
12 - call condvar callback immediately when it is set 252 - call condvar callback immediately when it is set
… …
29 - new function: AnyEvent::Socket::format_hostport. 269 - new function: AnyEvent::Socket::format_hostport.
30 270
314.881 Tue Jul 28 12:51:53 CEST 2009 2714.881 Tue Jul 28 12:51:53 CEST 2009
32 - work around a bug in local in pre-5.10 perls, causing 272 - work around a bug in local in pre-5.10 perls, causing
33 AnyEvent::Handle to recurse when it should not 273 AnyEvent::Handle to recurse when it should not
34 (analyzed by elmex). 274 (analysed by elmex).
35 275
364.88 Tue Jul 28 04:04:37 CEST 2009 2764.88 Tue Jul 28 04:04:37 CEST 2009
37 - re-bless the handle into a dummy package after calling 277 - re-bless the handle into a dummy package after calling
38 AnyEvent::Handle::destroy, so the user does not need to check 278 AnyEvent::Handle::destroy, so the user does not need to check
39 for errors after every push_write etc. 279 for errors after every push_write etc.
… …
2614.3 Fri Oct 3 09:18:43 CEST 2008 5014.3 Fri Oct 3 09:18:43 CEST 2008
262 - AnyEvent will now install a no-op signal handler for SIGPIPE, 502 - AnyEvent will now install a no-op signal handler for SIGPIPE,
263 unless one has been installed already. 503 unless one has been installed already.
264 - warn about not loaidng AnyEvent::Impl::POE early enough 504 - warn about not loaidng AnyEvent::Impl::POE early enough
265 (patch by Adam Rosenstein). 505 (patch by Adam Rosenstein).
266 - fixed a great number of bugs and corner cases in AnyEvent::Handle: 506 - fixed a great number of bugs and corner cases in AnyEvent::Handle.
267 - fix a bug in where in SSL connect mode, the 507 - fix a bug in where in SSL connect mode, the
268 client would first wait for some data by the server and 508 client would first wait for some data by the server and
269 otherwise hang. (reported and analyzed in an absolutely 509 otherwise hang. (reported and analysed in an absolutely
270 exemplary manner by Adam Rosenstein). 510 exemplary manner by Adam Rosenstein).
271 - fix a bug in where SSL EOF would not be treated 511 - fix a bug in where SSL EOF would not be treated
272 as stream EOF, putting the connection into a hung state. 512 as stream EOF, putting the connection into a hung state.
273 (reported and analyzed in an absolutely exemplary manner by Adam 513 (reported and analysed in an absolutely exemplary manner by Adam
274 Rosenstein). 514 Rosenstein).
275 - fix a potential segfault when the TLS context 515 - fix a potential segfault when the TLS context
276 would go missing in a read callback (e.g. due to stoptls) - 516 would go missing in a read callback (e.g. due to stoptls) -
277 Net::SSLeay of course makes no type checking whatsoever. 517 Net::SSLeay of course makes no type checking whatsoever.
278 - AnyEvent::Handle will not stop the read watcher in TLS mode. 518 - AnyEvent::Handle will not stop the read watcher in TLS mode.

Diff Legend

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