1 | Revision history for Perl extension AnyEvent. |
1 | Revision history for Perl extension AnyEvent. |
2 | |
2 | |
3 | TODO: parse ipv6 mapped addresses transparently |
3 | 4.9 |
|
|
4 | - Glib silently fails when registering a timeout with |
|
|
5 | a negative value, so avoid that. |
|
|
6 | - normalise signal numbers to names when using the ae's |
|
|
7 | signal handling, but do not document this (yet). |
|
|
8 | - pure perl signal emulation did not properly set nonblocking |
|
|
9 | mode on the signal pipe on win32 (this is unlikely to have caused |
|
|
10 | any issues). |
|
|
11 | - new module: AnyEvent::Debug. |
|
|
12 | - AnyEvent::Strict now checks that a signal was specified |
|
|
13 | by name (not name or number). |
|
|
14 | - reduce memory footprint in typical cases by ~50kb by |
|
|
15 | only compiling signal code when necessary. |
|
|
16 | - add AnyEvent::Handle->rbuf_max. |
|
|
17 | - grab the AE namespace for future new API, implement stubs for |
|
|
18 | said future/alternative API. |
|
|
19 | |
|
|
20 | 4.881 Tue Jul 28 12:51:53 CEST 2009 |
|
|
21 | - work around a bug in local in pre-5.10 perls, causing |
|
|
22 | AnyEvent::Handle to recurse when it should not |
|
|
23 | (analyzed by elmex). |
|
|
24 | |
|
|
25 | 4.88 Tue Jul 28 04:04:37 CEST 2009 |
|
|
26 | - re-bless the handle into a dummy package after calling |
|
|
27 | AnyEvent::Handle::destroy, so the user does not need to check |
|
|
28 | for errors after every push_write etc. |
|
|
29 | - do not attempt to run t/02_signals.t on obviously broken |
|
|
30 | platforms. |
|
|
31 | |
|
|
32 | 4.87 Sun Jul 26 02:06:16 CEST 2009 |
|
|
33 | - do not attempt to linger when there is no longer a valid fh. |
|
|
34 | - enforce tls mode to be either accept or connect, do not simply |
|
|
35 | segfault in Net::SSLeay. |
|
|
36 | - AnyEvent::Handle can now call tcp_connect itself (new parameters |
|
|
37 | connect, on_prepare, on_connect and on_connect_error). Updated |
|
|
38 | tutorial accordingly. |
|
|
39 | - add AnyEvent::Impl::Irssi backend. |
|
|
40 | |
|
|
41 | 4.86 Mon Jul 20 23:52:29 CEST 2009 |
|
|
42 | - since the verbose warning is not enough, explicitly document |
|
|
43 | that versions before 1.33 of Net::SSLeay are not secure. |
|
|
44 | - work around signal handling races in Event and (...) Event::Lib. |
|
|
45 | - try to align signal-race timer to full-second boundaries. |
|
|
46 | - work around Tk not liking negative timeouts. |
|
|
47 | - don't complain of different grades of environmental unfriendlyness |
|
|
48 | in IO::Async. |
|
|
49 | |
|
|
50 | 4.85 Sat Jul 18 06:16:14 CEST 2009 |
|
|
51 | - nail the signal race problem in perl once and for all |
|
|
52 | (see $AnyEvent::MAX_SIGNAL_LATENCY). |
|
|
53 | - take advantage of Async::Interrupt if it is available. |
|
|
54 | - load Time::HiRes and Guard modules on demand only. |
|
|
55 | - add optional/recommended modules section to AnyEvent |
|
|
56 | documentation. |
|
|
57 | - reduce memory usage considerably (and reduce startup penalty) |
|
|
58 | by not using "strict", "warnings" and "overload" modules. |
|
|
59 | - work around buggy windows/openbsd perls and provide EBADMSG |
|
|
60 | and EPROTO ourselves when missing. |
|
|
61 | - improve perl 5.6 compatibility of the core event loop. |
|
|
62 | - made Net::SSLeay version 1.33 a soft requirement. |
|
|
63 | |
|
|
64 | 4.83 Fri Jul 17 16:56:26 CEST 2009 |
|
|
65 | - implement AnyEvent::Socket::getprotobyname. |
|
|
66 | - AnyEvent::CondVar's will now detect recursive blocking |
|
|
67 | waits and will croak, as too many people fall into |
|
|
68 | this trap. |
|
|
69 | - AnyEvent::Handle will now call ->destroy on itself after |
|
|
70 | executing the on_error callback, instead of doing some |
|
|
71 | half-baked internal shutdown, for fatal errors. |
|
|
72 | - clarify on_eof behaviour w.r.t. the read queue and |
|
|
73 | on_read callbacks. |
|
|
74 | - ignore some possible spurious wake-ups in tcp_connect. |
|
|
75 | |
|
|
76 | 4.82 Sat Jul 11 00:34:55 CEST 2009 |
|
|
77 | - POE and Event backends didn't accept some callable objects as |
|
|
78 | callbacks. |
|
|
79 | - use Config module instead of POSIX module to detect signal names |
|
|
80 | in AnyEvent::Strict and AnyEvent::Impl::EventLib, as the POSIX |
|
|
81 | module doesn't even have all POSIX signals :/. |
|
|
82 | - use more workarounds around the many refcnt/corruption bugs in |
|
|
83 | Event::Lib. |
|
|
84 | - work around a race condition in perl's select, causing t/03_child.t |
|
|
85 | to rarely fail. |
|
|
86 | |
|
|
87 | 4.81 Thu Jul 9 10:30:30 CEST 2009 |
|
|
88 | - AnyEvent::Handle didn't properly diagnose write errors |
|
|
89 | (it expected -1 from syswrite, how lame... :). |
|
|
90 | - support file descriptors in addition to file handles |
|
|
91 | in AnyEvent->io. |
|
|
92 | - new env variables: PERL_ANYEVENT_RESOLV_CONF, |
|
|
93 | PERL_ANYEVENT_MAX_OUTSTANDING_DNS, PERL_ANYEVENT_CA_FILE |
|
|
94 | and PERL_ANYEVENT_CA_PATH. |
|
|
95 | - provide a sensible synopsis section for AnyEvent::TLS. |
|
|
96 | - add a "supported backends" section to the manpage. |
|
|
97 | - added simple io watcher test to testsuite, using a |
|
|
98 | portable_socketpair. |
|
|
99 | - tried to improve the stability of the Event::Lib backend, |
|
|
100 | YMMV. |
|
|
101 | |
|
|
102 | 4.8 Mon Jul 6 23:45:16 CEST 2009 |
|
|
103 | - AnyEvent::DNS did not properly follow CNAME records with |
|
|
104 | uppercase targets. |
|
|
105 | - AnyEvent::DNS would errornously return AAAA records |
|
|
106 | with v4 mapped addresses (a faulty record) as ipv4 addresses, |
|
|
107 | causing AnyEvent::Socket to throw an exception. |
|
|
108 | - added new module AnyEvent::TLS for easier SSL/TLS context |
|
|
109 | creation, with many options including hostname verification, |
|
|
110 | secure default configuration, lots of documentation and, |
|
|
111 | predefined diffie-hellman keys for perfect forward security |
|
|
112 | and much more. get it while it's still fresh! |
|
|
113 | - use AnyEvent::TLS in AnyEvent::Handle for context management. |
|
|
114 | - load AnyEvent::Handle only on demand in AnyEvent::DNS, |
|
|
115 | so AnyEvent::Socket users have smaller memory footprint |
|
|
116 | in the common case. |
|
|
117 | - add AnyEvent::Handle->push_shutdown method. |
|
|
118 | - add an additional error message parameter to AnyEvent::Handle's |
|
|
119 | on_error callback (for TLS, $! is still available). |
|
|
120 | - add AnyEvent::Handle on_starttls/on_stoptls callbacks. |
|
|
121 | - make AnyEvent::Handle more robust against early conenction |
|
|
122 | failures (during new), and return C<undef> in that case |
|
|
123 | from the constructor. |
|
|
124 | - AnyEvent::Handle will now try to load only JSON::XS first, |
|
|
125 | then fall back to JSON. |
|
|
126 | - format_ipv4/format_ipv6 are now exported by default, for symmetry, |
|
|
127 | and because it was documented that way. |
|
|
128 | |
|
|
129 | 4.451 Fri Jul 3 00:28:58 CEST 2009 |
|
|
130 | - do not clear rbuf when shutting down an AnyEvent::Handle |
|
|
131 | object - doing so breaks AnyEvent::HTTP. |
|
|
132 | |
|
|
133 | 4.45 Mon Jun 29 22:59:26 CEST 2009 |
|
|
134 | - a write error could cause AnyEvent::Handle to create |
|
|
135 | an I/O watcher with an undefined $fh. |
|
|
136 | - special-case mapped ipv4 addresses in both |
|
|
137 | AnyEvent::Socket::format_address and parse_address, |
|
|
138 | to treat them just like ipv4 addresses. |
|
|
139 | - updated and overhauled the AnyEvent::Intro doc. |
|
|
140 | - implement AnyEvent::Socket::format_ipv4/ipv6. |
4 | - slightly speed up portable_pipe/socketpair. |
141 | - slightly speed up portable_pipe/socketpair. |
|
|
142 | - expand condvar begin/end documentation. |
5 | |
143 | |
6 | 4.42 Fri Jun 26 08:32:18 CEST 2009 |
144 | 4.42 Fri Jun 26 08:32:18 CEST 2009 |
7 | - preliminary and neccesarily incomplete support for IO::Async. |
145 | - preliminary and neccesarily incomplete support for IO::Async. |
8 | - reset SIGCHLD to DEFAULT when AnyEvent is loaded, in |
146 | - reset SIGCHLD to DEFAULT when AnyEvent is loaded, in |
9 | case it was set to IGNORE, to ensure we can catch child statuses |
147 | case it was set to IGNORE, to ensure we can catch child statuses |