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

Comparing AnyEvent/Changes (file contents):
Revision 1.245 by root, Tue Jun 23 23:37:32 2009 UTC vs.
Revision 1.342 by root, Tue Sep 1 16:44:58 2009 UTC

1Revision history for Perl extension AnyEvent. 1Revision history for Perl extension AnyEvent.
2
3TODO: pass values to anyevent condvar callback, make this count as recv.
4TODO: resolv.conf nameserver blabla # comment does not work
5 - update to Async::Interrupt 1.02.
6 - use same set of warningsa s common::sense 2.0.
7 - fix a potential 32 bit overflow issue due to perl having problems
8 with large hex constants in the Perl backend.
9 - do not use the slower 5.6 version fo the enc_name code in AnyEvent::DNS
10 by default.
11
125.112 Fri Aug 21 13:59:04 CEST 2009
13 - AnyEvent::Handle could cause hard crashes in Net::SSLeay when
14 using starttls with an illegal/uncreatable tls context.
15 - port to broken win32 perls.
16
175.111 Sun Aug 16 18:44:36 CEST 2009
18 - perl errornously requires sockaddr_un structures returned by the
19 kernel to have a specific length. work around this bug by padding
20 sockaddr_un structures (found to fail on OS X, as perl only uses
21 the correct code for GNU/Linux, but likely to fail on other
22 OSes as well).
23
245.11 Wed Aug 12 17:49:37 CEST 2009
25 - fixed a bug in run_cmd with input from filename.
26 - create a json coder object when none is given in write direction
27 to reduce dependency on JSON.
28
295.1 Tue Aug 11 03:17:32 CEST 2009
30 - tcp_connect now reports the correct error in $!, not an unrelated
31 one (was broken since 4.91).
32 - AE::Handle did not properly clear rtimeout/wtimeout watchers
33 when retrying a tcp_connect.
34 - new functions: AnyEvent::Util::run_cmd and close_all_fds_except.
35 - fixed wrong documentation in AE::Socket.
36
375.01 Mon Aug 10 03:16:32 CEST 2009
38 - last release broke EV child watchers :/.
39
405.0 Sun Aug 9 17:32:09 CEST 2009
41 - officially introduce and document the AE API.
42 - lowering the timeout at runtime did not have immediate effect
43 in AnyEvent::Handle.
44 - AE::Handle now has separate and independent read and write timeout
45 settings.
46 - probe for Guard module when AnyEvent::Util is loaded, not at
47 runtime (Event doesn't like eval "use" from callbacks, and it avoids
48 exporting the probe function).
49 - Impl::IOAsync's io watchers did not use the filehandle emulation
50 code to work around it's watcher limits.
51 - work around IO::Async's problems with overloaded objects.
52 - try to disable wearnings inside IO::Async, it's just too buggy.
53 - do no longer use Event by default, only EV or Perl, Event is too buggy.
54 - hackishly provide fast AE:: interface when EV or Perl is the backend.
55 - inherit VERSION from AnyEvent::VERSION in most modules.
56
574.91 Thu Aug 6 15:42:45 CEST 2009
58 - AE::Handle::starttls could get out of sync when the read buffer
59 already contains some TLS handshake.
60 - AE::Handle did not properly free the TLS session, causing bigger
61 memory leaks in Net::SSLeay than expected (reported by toaster).
62 - AE::Socket::tcp_connect will now artificially delay invoking
63 the callback to avoid returning after invoking the callback.
64 - convert many internal watcher uses to AE API.
65
664.9 Sat Aug 1 11:07:01 CEST 2009
67 - Glib silently fails when registering a timeout with
68 a negative value, so avoid that.
69 - call condvar callback immediately when it is set
70 after the condition is already signalled.
71 - check rbuf_max condition only after trying to consume
72 data.
73 - normalise signal numbers to names when using the ae's
74 signal handling, but do not document this (yet).
75 - pure perl signal emulation did not properly set nonblocking
76 mode on the signal pipe on win32 (this is unlikely to have caused
77 any issues).
78 - new module: AnyEvent::Debug.
79 - AnyEvent::Strict now checks that a signal was specified
80 by name (not name or number).
81 - reduce memory footprint in typical cases by ~50kb by
82 only compiling signal code when necessary.
83 - add AnyEvent::Handle->rbuf_max.
84 - grab the AE namespace for future new API, implement stubs for
85 said future/alternative API.
86 - new function: AnyEvent::Socket::format_hostport.
87
884.881 Tue Jul 28 12:51:53 CEST 2009
89 - work around a bug in local in pre-5.10 perls, causing
90 AnyEvent::Handle to recurse when it should not
91 (analyzed by elmex).
92
934.88 Tue Jul 28 04:04:37 CEST 2009
94 - re-bless the handle into a dummy package after calling
95 AnyEvent::Handle::destroy, so the user does not need to check
96 for errors after every push_write etc.
97 - do not attempt to run t/02_signals.t on obviously broken
98 platforms.
99
1004.87 Sun Jul 26 02:06:16 CEST 2009
101 - do not attempt to linger when there is no longer a valid fh.
102 - enforce tls mode to be either accept or connect, do not simply
103 segfault in Net::SSLeay.
104 - AnyEvent::Handle can now call tcp_connect itself (new parameters
105 connect, on_prepare, on_connect and on_connect_error). Updated
106 tutorial accordingly.
107 - add AnyEvent::Impl::Irssi backend.
108
1094.86 Mon Jul 20 23:52:29 CEST 2009
110 - since the verbose warning is not enough, explicitly document
111 that versions before 1.33 of Net::SSLeay are not secure.
112 - work around signal handling races in Event and (...) Event::Lib.
113 - try to align signal-race timer to full-second boundaries.
114 - work around Tk not liking negative timeouts.
115 - don't complain of different grades of environmental unfriendlyness
116 in IO::Async.
117
1184.85 Sat Jul 18 06:16:14 CEST 2009
119 - nail the signal race problem in perl once and for all
120 (see $AnyEvent::MAX_SIGNAL_LATENCY).
121 - take advantage of Async::Interrupt if it is available.
122 - load Time::HiRes and Guard modules on demand only.
123 - add optional/recommended modules section to AnyEvent
124 documentation.
125 - reduce memory usage considerably (and reduce startup penalty)
126 by not using "strict", "warnings" and "overload" modules.
127 - work around buggy windows/openbsd perls and provide EBADMSG
128 and EPROTO ourselves when missing.
129 - improve perl 5.6 compatibility of the core event loop.
130 - made Net::SSLeay version 1.33 a soft requirement.
131
1324.83 Fri Jul 17 16:56:26 CEST 2009
133 - implement AnyEvent::Socket::getprotobyname.
134 - AnyEvent::CondVar's will now detect recursive blocking
135 waits and will croak, as too many people fall into
136 this trap.
137 - AnyEvent::Handle will now call ->destroy on itself after
138 executing the on_error callback, instead of doing some
139 half-baked internal shutdown, for fatal errors.
140 - clarify on_eof behaviour w.r.t. the read queue and
141 on_read callbacks.
142 - ignore some possible spurious wake-ups in tcp_connect.
143
1444.82 Sat Jul 11 00:34:55 CEST 2009
145 - POE and Event backends didn't accept some callable objects as
146 callbacks.
147 - use Config module instead of POSIX module to detect signal names
148 in AnyEvent::Strict and AnyEvent::Impl::EventLib, as the POSIX
149 module doesn't even have all POSIX signals :/.
150 - use more workarounds around the many refcnt/corruption bugs in
151 Event::Lib.
152 - work around a race condition in perl's select, causing t/03_child.t
153 to rarely fail.
154
1554.81 Thu Jul 9 10:30:30 CEST 2009
156 - AnyEvent::Handle didn't properly diagnose write errors
157 (it expected -1 from syswrite, how lame... :).
158 - support file descriptors in addition to file handles
159 in AnyEvent->io.
160 - new env variables: PERL_ANYEVENT_RESOLV_CONF,
161 PERL_ANYEVENT_MAX_OUTSTANDING_DNS, PERL_ANYEVENT_CA_FILE
162 and PERL_ANYEVENT_CA_PATH.
163 - provide a sensible synopsis section for AnyEvent::TLS.
164 - add a "supported backends" section to the manpage.
165 - added simple io watcher test to testsuite, using a
166 portable_socketpair.
167 - tried to improve the stability of the Event::Lib backend,
168 YMMV.
169
1704.8 Mon Jul 6 23:45:16 CEST 2009
171 - AnyEvent::DNS did not properly follow CNAME records with
172 uppercase targets.
173 - AnyEvent::DNS would errornously return AAAA records
174 with v4 mapped addresses (a faulty record) as ipv4 addresses,
175 causing AnyEvent::Socket to throw an exception.
176 - added new module AnyEvent::TLS for easier SSL/TLS context
177 creation, with many options including hostname verification,
178 secure default configuration, lots of documentation and,
179 predefined diffie-hellman keys for perfect forward security
180 and much more. get it while it's still fresh!
181 - use AnyEvent::TLS in AnyEvent::Handle for context management.
182 - load AnyEvent::Handle only on demand in AnyEvent::DNS,
183 so AnyEvent::Socket users have smaller memory footprint
184 in the common case.
185 - add AnyEvent::Handle->push_shutdown method.
186 - add an additional error message parameter to AnyEvent::Handle's
187 on_error callback (for TLS, $! is still available).
188 - add AnyEvent::Handle on_starttls/on_stoptls callbacks.
189 - make AnyEvent::Handle more robust against early conenction
190 failures (during new), and return C<undef> in that case
191 from the constructor.
192 - AnyEvent::Handle will now try to load only JSON::XS first,
193 then fall back to JSON.
194 - format_ipv4/format_ipv6 are now exported by default, for symmetry,
195 and because it was documented that way.
196
1974.451 Fri Jul 3 00:28:58 CEST 2009
198 - do not clear rbuf when shutting down an AnyEvent::Handle
199 object - doing so breaks AnyEvent::HTTP.
200
2014.45 Mon Jun 29 22:59:26 CEST 2009
202 - a write error could cause AnyEvent::Handle to create
203 an I/O watcher with an undefined $fh.
204 - special-case mapped ipv4 addresses in both
205 AnyEvent::Socket::format_address and parse_address,
206 to treat them just like ipv4 addresses.
207 - updated and overhauled the AnyEvent::Intro doc.
208 - implement AnyEvent::Socket::format_ipv4/ipv6.
209 - slightly speed up portable_pipe/socketpair.
210 - expand condvar begin/end documentation.
211
2124.42 Fri Jun 26 08:32:18 CEST 2009
213 - preliminary and neccesarily incomplete support for IO::Async.
214 - reset SIGCHLD to DEFAULT when AnyEvent is loaded, in
215 case it was set to IGNORE, to ensure we can catch child statuses
216 even when the calling env acts stupidly.
217 - updated benchmarks with IO::Async, which performs very well.
2 218
34.412 Wed Jun 24 01:35:57 CEST 2009 2194.412 Wed Jun 24 01:35:57 CEST 2009
4 - support an "untaint" attribute for AnyEvent::DNS 220 - support an "untaint" attribute for AnyEvent::DNS
5 and set it on the default resolver. 221 and set it on the default resolver.
6 - implement and document AnyEvent::Impl::Perl::loop. 222 - implement and document AnyEvent::Impl::Perl::loop.

Diff Legend

Removed lines
+ Added lines
< Changed lines
> Changed lines