| 1 | Revision history for Perl extension AnyEvent. |
1 | Revision history for Perl extension AnyEvent. |
|
|
2 | |
|
|
3 | - reduce memory usage and slightly speed up the pure perl backend |
|
|
4 | by only storing the file descriptor, not the file handle. |
|
|
5 | - add missing autocork method to AnyEvent::Handle (reported |
|
|
6 | by Adam Rosenstein). |
|
|
7 | - AnyEvent::DNS->resolve errornously documented an $rcode result |
|
|
8 | argument, but there isn't (spotted by Henrik Krohns). |
|
|
9 | - the naked truth about Tk - it's basically unmaintained. |
|
|
10 | |
|
|
11 | 4.32 Mon Nov 3 22:46:32 CET 2008 |
|
|
12 | - fix AnyEvent::Socket::resolve_sockaddr to properly support |
|
|
13 | unix sockets again and choose a proper default for |
|
|
14 | the socket type. |
|
|
15 | - fix call to resolve_sockaddr in tcp_connect. |
|
|
16 | |
|
|
17 | 4.31 Thu Oct 30 04:41:48 CET 2008 |
|
|
18 | - implemented AnyEvent::Handle->destroy method. |
|
|
19 | - hint about unexpected effects in TLS mode. |
|
|
20 | - speed up AnyEvent::Impl::Perl by using more arrays |
|
|
21 | instead of hashes without the slightest loss of readability :) |
|
|
22 | - work around a perl argument refcounting bug. |
|
|
23 | - with some perl patches applied, AnyEvent now seems leak-free. |
|
|
24 | |
|
|
25 | 4.3 Fri Oct 3 09:18:43 CEST 2008 |
|
|
26 | - AnyEvent will now install a no-op signal handler for SIGPIPE, |
|
|
27 | unless one has been installed already. |
|
|
28 | - warn about not loaidng AnyEvent::Impl::POE early enough |
|
|
29 | (patch by Adam Rosenstein). |
|
|
30 | - fixed a great number of bugs and corner cases in AnyEvent::Handle: |
|
|
31 | - fix a bug in where in SSL connect mode, the |
|
|
32 | client would first wait for some data by the server and |
|
|
33 | otherwise hang. (reported and analyzed in an absolutely |
|
|
34 | exemplary manner by Adam Rosenstein). |
|
|
35 | - fix a bug in where SSL EOF would not be treated |
|
|
36 | as stream EOF, putting the connection into a hung state. |
|
|
37 | (reported and analyzed in an absolutely exemplary manner by Adam |
|
|
38 | Rosenstein). |
|
|
39 | - fix a potential segfault when the TLS context |
|
|
40 | would go missing in a read callback (e.g. due to stoptls) - |
|
|
41 | Net::SSLeay of course makes no type checking whatsoever. |
|
|
42 | - AnyEvent::Handle will not stop the read watcher in TLS mode. |
|
|
43 | - AnyEvent::Handle->stoptls will send a TLS close notify |
|
|
44 | instead of simply destroying the stream now. |
|
|
45 | - fix a bug where large blocks of data written in TLS mode |
|
|
46 | would not be sent unless triggered by receives. |
|
|
47 | - on_drain will now take the tls write buffer into account. |
|
|
48 | - SSL operations have been streamlined a bit and should be faster |
|
|
49 | now, more to come, though. |
|
|
50 | - added a FAQ section, will document nontrivial issues in there. |
|
|
51 | |
|
|
52 | 4.234 Mon Sep 29 04:08:13 CEST 2008 |
|
|
53 | - fix child watcher documentation: only child exits will be reported, |
|
|
54 | no trace events. |
|
|
55 | - mention SIGPIPE in AnyEvent::Handle. |
|
|
56 | - perl backend will now use POSIX::times as monotonic clock |
|
|
57 | when available and the monotonic clock option is not. |
|
|
58 | - run even without Time::HiRes in some cases now, but do not |
|
|
59 | provide sub-second accuracy (all *supported* perl versions have |
|
|
60 | Time::HiRes). |
|
|
61 | |
|
|
62 | 4.233 Fri Aug 22 01:48:05 CEST 2008 |
|
|
63 | - fix a bug in the testsuite, causing a hang. |
|
|
64 | - clarified lots of AnyEvent::Handle constructor arguments. |
|
|
65 | |
|
|
66 | 4.232 Thu Aug 21 20:44:25 CEST 2008 |
|
|
67 | - fix a bug in AnyEvent::Handle that could cause two fatal error |
|
|
68 | callbacks on EOF, which caused AnyEvent::HTTP to signal success twice |
|
|
69 | (testcase by Leon Brocard). |
|
|
70 | - clarified on_eof/on_error documentation a bit. |
|
|
71 | |
|
|
72 | 4.231 Tue Jul 29 13:12:15 CEST 2008 |
|
|
73 | - remove some debugging code left in AnyEvent::Util::fork_call |
|
|
74 | (and no, it's impossible to implement with the broken windows |
|
|
75 | perls without resource leaks or worse). |
|
|
76 | |
|
|
77 | 4.23 Tue Jul 29 12:19:59 CEST 2008 |
|
|
78 | - document the first parameter passed to condvar callbacks to be |
|
|
79 | the callback. |
|
|
80 | - add AnyEvent::Socket::{ntoa,aton} aliases. |
|
|
81 | - optimize the AE::Handle->push_read (line) for the default |
|
|
82 | eol marker. |
|
|
83 | - optimize push_read (packstring|storable) for small packets. |
|
|
84 | - invoke on_error callback when no on_eof callback is set. |
|
|
85 | - fix a bug in push_read (storable) of unknown impact. |
|
|
86 | |
|
|
87 | 4.22 Sun Jul 20 16:34:13 CEST 2008 |
|
|
88 | - new function AnyEvent::Socket::parse_hostport. |
|
|
89 | - as the bulkheads at microsoft can't even get getprotobyname reliably |
|
|
90 | working on their shitty fucking broken os we need to hardcode |
|
|
91 | some common protocol numbers in AnyEvent::Socket. How can |
|
|
92 | people even bother with such a pile of shit as windows. |
|
|
93 | |
|
|
94 | 4.21 Thu Jul 17 14:40:05 CEST 2008 |
|
|
95 | - INCOMPATIBLE CHANGE: AnyEvent::Util::fork_call now has a prototype |
|
|
96 | which is incompatible to previous invocation syntax. |
|
|
97 | - work around more windows perl bugs in fork_call. I cannot imagine |
|
|
98 | why anybody would bother implementing fork in such an obviously |
|
|
99 | unusable way. |
|
|
100 | - avoid calling $do_retry if the request has already finished in |
|
|
101 | AnyEvent::DNS, thus avoiding an exception (reported by Anatoly |
|
|
102 | K. Sharifulin). |
|
|
103 | |
|
|
104 | 4.2 Sat Jul 12 22:42:11 CEST 2008 |
|
|
105 | - support an "interval" argument for repeating timers. |
|
|
106 | - fix a bug in the pure perl event loop that caused timers to |
|
|
107 | be delayed under some circumstances. |
|
|
108 | - implement optional argument checking via PERL_ANYEVENT_STRICT. |
|
|
109 | - not importing anything from Event causes it to malfunction |
|
|
110 | and not use Time::HiRes, work around this bug. |
|
|
111 | - more bug workarounds against the endlessly broken Tk module, |
|
|
112 | seems to work now on 32 bit machines, but less so on 64 bit. |
|
|
113 | - minor optimisations applied to most backends. |
|
|
114 | |
|
|
115 | 4.161 Fri Jul 4 14:24:48 CEST 2008 |
|
|
116 | - fixed the t/handle/02_write.t, it read after write, with the |
|
|
117 | assumption that all data could be written before read |
|
|
118 | (reported by Jim Razmus). |
|
|
119 | |
|
|
120 | 4.160 Thu Jul 3 04:02:21 CEST 2008 |
|
|
121 | - re-registering signal watchers after unregistering one did |
|
|
122 | not work if a backend falls back on AnyEvent's default |
|
|
123 | implementation (also affected child watchers). |
|
|
124 | - new AnyEvent::Handle options: autocork and no_delay. |
|
|
125 | |
|
|
126 | 4.152 Sun Jun 22 14:15:44 CEST 2008 |
|
|
127 | - allow for 32-bit perls that implement shifts differently |
|
|
128 | on different architectures in parse_ipv4 (reported and |
|
|
129 | analysed by Keiichi DAIBA). |
|
|
130 | |
|
|
131 | 4.151 Fri Jun 6 17:34:24 CEST 2008 |
|
|
132 | - make sure specifying _only_ on_read and never pushing reads |
|
|
133 | works. |
|
|
134 | |
|
|
135 | 4.15 Fri Jun 6 13:00:46 CEST 2008 |
|
|
136 | - the pure perl backend would keep some watchers alive when more than |
|
|
137 | one watcher was registered for the same fd. |
|
|
138 | - new "packstring" and "storable" read and write types |
|
|
139 | for AnyEvent::Handle. |
|
|
140 | - allow on_eof handler to be called after on_error with EPIPE returns. |
|
|
141 | - do not immediately call on_read callback in handle constructor. |
|
|
142 | |
|
|
143 | 4.14 Thu Jun 5 20:29:31 CEST 2008 |
|
|
144 | - Fixed a bug in DNS SRV priority sorting. |
|
|
145 | - AnyEvent::Util::guard now reports runtime errors while |
|
|
146 | executing the guard block as warnings. |
|
|
147 | - handle 0-byte-reads just before EOF correctly in AnyEvent::Handle. |
|
|
148 | |
|
|
149 | 4.13 Thu Jun 5 00:47:59 CEST 2008 |
|
|
150 | - AnyEvent::DNS only followed cname chains with length 2, |
|
|
151 | contrary to documentation. bumped it up to 10, thanks to |
|
|
152 | microsoft, the current broken-dns-config-king. |
|
|
153 | - AnyEvent::DNS didn't check for socket return status |
|
|
154 | but instead relied on perl not creating filehandles in that |
|
|
155 | case - too bad it gives you a bogus file handle (reported |
|
|
156 | and analysed by Vladimir Timofeev). |
|
|
157 | - fix queue management logic in AnyEvent::Handle: |
|
|
158 | when on_read was registered, the queue was empty |
|
|
159 | and no progress could be made AnyEvent::Handle would |
|
|
160 | enter an endless loop. |
|
|
161 | - correctly start reading again when the handle became |
|
|
162 | busy again after idling. |
|
|
163 | - correctly treat tls shut-downs as EOF condition |
|
|
164 | (for the time being, would be nice to have a callback for that). |
|
|
165 | - correctly call eof callback under all conditions |
|
|
166 | (hopefully). |
|
|
167 | - the timeout callback did not expect that $self can go |
|
|
168 | away any time. |
| 2 | |
169 | |
| 3 | 4.12 Tue Jun 3 10:58:04 CEST 2008 |
170 | 4.12 Tue Jun 3 10:58:04 CEST 2008 |
| 4 | - include AnyEvent::Intro, a tutorial for anyevent, |
171 | - include AnyEvent::Intro, a tutorial for anyevent, |
| 5 | anyevent::socket and anyevent::handle. |
172 | anyevent::socket and anyevent::handle. |
| 6 | - allow more options in on_error. |
173 | - allow more options in on_error. |
| … | |
… | |
| 104 | - support service names in tcp_connect and tcp_server. |
271 | - support service names in tcp_connect and tcp_server. |
| 105 | - work around netbsd/openbsd bug causing unix sockets not to |
272 | - work around netbsd/openbsd bug causing unix sockets not to |
| 106 | become writable until empty. |
273 | become writable until empty. |
| 107 | - only make a dummy read for error detection when getpeername |
274 | - only make a dummy read for error detection when getpeername |
| 108 | returns ENOTCONN. |
275 | returns ENOTCONN. |
| 109 | - defend AnyEvent::Util::inet_aton against Coro::LWP brutal |
276 | - defend AnyEvent::Util::inet_aton against Coro::LWP's brutal |
| 110 | override. |
277 | override. |
| 111 | - new AnyEvent::Guard::cancel. |
278 | - new AnyEvent::Guard::cancel. |
| 112 | |
279 | |
| 113 | 3.5 Sat May 17 23:17:14 CEST 2008 |
280 | 3.5 Sat May 17 23:17:14 CEST 2008 |
| 114 | - gracefully deal with systems not defining CLOCK_MONOTONIC |
281 | - gracefully deal with systems not defining CLOCK_MONOTONIC |
| … | |
… | |
| 128 | |
295 | |
| 129 | 3.41 Sun May 11 19:53:13 CEST 2008 |
296 | 3.41 Sun May 11 19:53:13 CEST 2008 |
| 130 | - work around a bug in perl 5.8.8's local. Reported by Yi Ma Mao. |
297 | - work around a bug in perl 5.8.8's local. Reported by Yi Ma Mao. |
| 131 | |
298 | |
| 132 | 3.4 Sun May 11 00:29:25 CEST 2008 |
299 | 3.4 Sun May 11 00:29:25 CEST 2008 |
| 133 | - complete re-implementation of condvars with a hopefully much mroe useful |
300 | - complete re-implementation of condvars with a hopefully much more useful |
| 134 | API (backward compatible functions exist but are undocumented). |
301 | API (backward compatible functions exist but are undocumented). |
| 135 | - AnyEvent::Handle fully rewritten with a hopefully more powerful API |
302 | - AnyEvent::Handle fully rewritten with a hopefully more powerful API |
| 136 | (still subject to change). |
303 | (still subject to change). |
| 137 | - take advantage of CLOCK_MONOTONIC, if available, in AnyEvent::Impl::Perl. |
304 | - take advantage of CLOCK_MONOTONIC, if available, in AnyEvent::Impl::Perl. |
| 138 | - provide AnyEvent::post_detect and @AnyEvent::post_detect, which |
305 | - provide AnyEvent::post_detect and @AnyEvent::post_detect, which |
