| 1 | Revision history for Perl extension AnyEvent. |
1 | Revision history for Perl extension AnyEvent. |
| 2 | |
2 | |
|
|
3 | - don't generate spurious readyness notifications when select |
|
|
4 | returns EINTR in the pure perl backend (analysed and testcase |
|
|
5 | by Bas Denissen). |
|
|
6 | |
|
|
7 | 5.251 Sat Mar 13 00:58:21 CET 2010 |
|
|
8 | - make test failed if it was run before make install on perl <= |
|
|
9 | 5.8.8, and older versions of perl make our live very difficult, |
|
|
10 | so write constants.pl during Makefile.PL time. |
|
|
11 | |
|
|
12 | 5.25 Sat Mar 13 00:23:14 CET 2010 |
|
|
13 | - fix a race condition in AnyEvent::Handle that would cause |
|
|
14 | a "bio_read: unsupported method" error (in _another_ |
|
|
15 | ssl connection) after user code threw an exception. |
|
|
16 | - added AnyEvent::Handle->destroyed method. |
|
|
17 | - speed up AnyEvent::Socket::format_address and ::format_ipv6. |
|
|
18 | - the AnyEvent::Util::fh_nonblocking function worked only |
|
|
19 | by lucky accident on win32. |
|
|
20 | - smaller and faster AnyEvent::Util::fh_nonblocking. |
|
|
21 | - when the (required!) Time::HiRes module is missing, AnyEvent |
|
|
22 | did not fall back to built-in time properly. |
|
|
23 | - do not load Fcntl at runtime, saving memory and loading time. |
|
|
24 | - precompile a number of constants and use them instead of runtime |
|
|
25 | detection and eval. |
|
|
26 | - free detection code after detection and similar memory optimisations. |
|
|
27 | - Perl backend timer interval best effort drift has been improved |
|
|
28 | (same algorithm as EV). |
|
|
29 | - update unicode idna mapping table. |
|
|
30 | |
|
|
31 | 5.24 Tue Jan 5 11:39:43 CET 2010 |
|
|
32 | - cygwin never reports errors from failed connects. 1.5 just gives |
|
|
33 | you continous readyness and EAGAIN, 1.7 is even more broken and |
|
|
34 | just hangs. work around both issues in a cygwin-specific hack. |
|
|
35 | - improve idle watcher documentation slightly. |
|
|
36 | |
|
|
37 | 5.23 Sun Dec 20 23:48:00 CET 2009 |
|
|
38 | - support IDNs in resolve_sockaddr, and therefore in tcp_connect. |
|
|
39 | - implement punycode_encode/decode, idn_nameprep, |
|
|
40 | idn_to_ascii and idn_to_unicode operations in AnyEvent::Util. |
|
|
41 | - provide $AE::VERSION. |
|
|
42 | - removed traces of "no strict 'refs'". |
|
|
43 | |
|
|
44 | 5.22 Sat Dec 5 03:51:13 CET 2009 |
|
|
45 | - downgrade-or-fail in AnyEvent::Handle::push_write, to |
|
|
46 | diagnose encoding failures earlier and more succinctly. |
|
|
47 | (this works around bugs in perl, throwing away encoding info |
|
|
48 | when passing scalar data to extensions). |
|
|
49 | - add more examples to AnyEvent::Socket manpage. |
|
|
50 | - upgrade internal warning set to the same as common::sense 2.03. |
|
|
51 | - use pack "n/a*" for pre-5.8.9 perl compatibility in AnyEvent::DNS |
|
|
52 | (John Beppu). |
|
|
53 | - AnyEvent::Socket::inet_aton now properly supports ipv6, as documented. |
|
|
54 | - add google public dns servers to fallback server set and make sure |
|
|
55 | we load-balance properly between the three sets. also add all |
|
|
56 | fallback dns servers, not just a random one, to each dns config. |
|
|
57 | |
|
|
58 | 5.21 Thu Nov 19 02:48:47 CET 2009 |
|
|
59 | - fix a problem where socket constants were called with parameters |
|
|
60 | (spotted by David Friedland). |
|
|
61 | - fork_call never use'd POSIX (reported by Daisuke Maki). |
|
|
62 | - improve perl 5.6 compatibility further (but it still won't work |
|
|
63 | unless you rip out everything but the core). |
|
|
64 | - prefer Net::DNS::Resolver over ipconfig, if installed, on win32. |
|
|
65 | uses 10MB of RAM, but doesn't open a console window. *sigh*. |
|
|
66 | |
|
|
67 | 5.202 Wed Oct 14 22:35:44 CEST 2009 |
|
|
68 | - AnyEvent::DNS would unexpectedly clobber $_ under windows |
|
|
69 | (analysed by Matthias Waldorf). |
|
|
70 | - AnyEvent::Handle::run_cmd can now pass the PID of the |
|
|
71 | newly-created process, which is much less useful than it might |
|
|
72 | sound (based on patch by Yann Kerherve). |
|
|
73 | |
|
|
74 | 5.201 Tue Sep 29 12:09:25 CEST 2009 |
|
|
75 | - AnyEvent:Handle::on_starttls/on_stoptls methods were broken |
|
|
76 | (reported by Torsten Foertsch). |
|
|
77 | - common::sense 2.0 could cause tcp_server to throw an exception |
|
|
78 | (analysed by elmex). |
|
|
79 | |
|
|
80 | 5.2 Mon Sep 14 07:04:49 CEST 2009 |
|
|
81 | - INCOMPATIBLE CHANGE: do no longer support register_read_type |
|
|
82 | and register_write_type in AnyEvent::Handle, instead support |
|
|
83 | package names (the facility was mostly abused). |
|
|
84 | - implement "packagename-as-read/write type" support in |
|
|
85 | AnyEvent::Handle. |
|
|
86 | - AnyEvent::Handle: new options "keepalive" and "oobinline". |
|
|
87 | - oobinline set by default to avoid security issues. |
|
|
88 | - the pure-perl event loop backend wrongly detected times() fallback |
|
|
89 | support (spotted by Pavel Boldin). |
|
|
90 | |
|
|
91 | 5.12 Tue Sep 1 20:26:50 CEST 2009 |
|
|
92 | - be more lenient when parsing resolv.conf files, as some people |
|
|
93 | use hashmarks as comment indicator inside directives (reported |
|
|
94 | by Michael S. Fischer). |
|
|
95 | - use same set of warnings as common::sense 2.0. |
|
|
96 | - fix a potential 32 bit overflow issue due to perl having problems |
|
|
97 | with large hex constants in the Perl backend. |
|
|
98 | - do not use the slower 5.6 version fo the enc_name code in AnyEvent::DNS |
|
|
99 | by default. |
|
|
100 | - fix some prototypes. |
|
|
101 | - (optionally) require Async::Interrupt 1.02. |
|
|
102 | - replace opendns fallbacks by verizon/level3 public dns servers. |
|
|
103 | |
|
|
104 | 5.112 Fri Aug 21 13:59:04 CEST 2009 |
|
|
105 | - AnyEvent::Handle could cause hard crashes in Net::SSLeay when |
|
|
106 | using starttls with an illegal/uncreatable tls context. |
|
|
107 | - port to broken win32 perls. |
|
|
108 | |
|
|
109 | 5.111 Sun Aug 16 18:44:36 CEST 2009 |
|
|
110 | - perl errornously requires sockaddr_un structures returned by the |
|
|
111 | kernel to have a specific length. work around this bug by padding |
|
|
112 | sockaddr_un structures (found to fail on OS X, as perl only uses |
|
|
113 | the correct code for GNU/Linux, but likely to fail on other |
|
|
114 | OSes as well). |
|
|
115 | |
|
|
116 | 5.11 Wed Aug 12 17:49:37 CEST 2009 |
|
|
117 | - fixed a bug in run_cmd with input from filename. |
|
|
118 | - create a json coder object when none is given in write direction |
|
|
119 | to reduce dependency on JSON. |
|
|
120 | |
|
|
121 | 5.1 Tue Aug 11 03:17:32 CEST 2009 |
|
|
122 | - tcp_connect now reports the correct error in $!, not an unrelated |
|
|
123 | one (was broken since 4.91). |
|
|
124 | - AE::Handle did not properly clear rtimeout/wtimeout watchers |
|
|
125 | when retrying a tcp_connect. |
|
|
126 | - new functions: AnyEvent::Util::run_cmd and close_all_fds_except. |
| 3 | - fixed wrong documentation in AE::Socket. |
127 | - fixed wrong documentation in AE::Socket. |
| 4 | |
128 | |
| 5 | 5.01 Mon Aug 10 03:16:32 CEST 2009 |
129 | 5.01 Mon Aug 10 03:16:32 CEST 2009 |
| 6 | - last release broke EV child watchers :/. |
130 | - last release broke EV child watchers :/. |
| 7 | |
131 | |
| … | |
… | |
| 54 | - new function: AnyEvent::Socket::format_hostport. |
178 | - new function: AnyEvent::Socket::format_hostport. |
| 55 | |
179 | |
| 56 | 4.881 Tue Jul 28 12:51:53 CEST 2009 |
180 | 4.881 Tue Jul 28 12:51:53 CEST 2009 |
| 57 | - work around a bug in local in pre-5.10 perls, causing |
181 | - work around a bug in local in pre-5.10 perls, causing |
| 58 | AnyEvent::Handle to recurse when it should not |
182 | AnyEvent::Handle to recurse when it should not |
| 59 | (analyzed by elmex). |
183 | (analysed by elmex). |
| 60 | |
184 | |
| 61 | 4.88 Tue Jul 28 04:04:37 CEST 2009 |
185 | 4.88 Tue Jul 28 04:04:37 CEST 2009 |
| 62 | - re-bless the handle into a dummy package after calling |
186 | - re-bless the handle into a dummy package after calling |
| 63 | AnyEvent::Handle::destroy, so the user does not need to check |
187 | AnyEvent::Handle::destroy, so the user does not need to check |
| 64 | for errors after every push_write etc. |
188 | for errors after every push_write etc. |
| … | |
… | |
| 289 | - warn about not loaidng AnyEvent::Impl::POE early enough |
413 | - warn about not loaidng AnyEvent::Impl::POE early enough |
| 290 | (patch by Adam Rosenstein). |
414 | (patch by Adam Rosenstein). |
| 291 | - fixed a great number of bugs and corner cases in AnyEvent::Handle: |
415 | - fixed a great number of bugs and corner cases in AnyEvent::Handle: |
| 292 | - fix a bug in where in SSL connect mode, the |
416 | - fix a bug in where in SSL connect mode, the |
| 293 | client would first wait for some data by the server and |
417 | client would first wait for some data by the server and |
| 294 | otherwise hang. (reported and analyzed in an absolutely |
418 | otherwise hang. (reported and analysed in an absolutely |
| 295 | exemplary manner by Adam Rosenstein). |
419 | exemplary manner by Adam Rosenstein). |
| 296 | - fix a bug in where SSL EOF would not be treated |
420 | - fix a bug in where SSL EOF would not be treated |
| 297 | as stream EOF, putting the connection into a hung state. |
421 | as stream EOF, putting the connection into a hung state. |
| 298 | (reported and analyzed in an absolutely exemplary manner by Adam |
422 | (reported and analysed in an absolutely exemplary manner by Adam |
| 299 | Rosenstein). |
423 | Rosenstein). |
| 300 | - fix a potential segfault when the TLS context |
424 | - fix a potential segfault when the TLS context |
| 301 | would go missing in a read callback (e.g. due to stoptls) - |
425 | would go missing in a read callback (e.g. due to stoptls) - |
| 302 | Net::SSLeay of course makes no type checking whatsoever. |
426 | Net::SSLeay of course makes no type checking whatsoever. |
| 303 | - AnyEvent::Handle will not stop the read watcher in TLS mode. |
427 | - AnyEvent::Handle will not stop the read watcher in TLS mode. |
