| 1 | Revision history for Perl extension AnyEvent. |
1 | Revision history for Perl extension AnyEvent. |
|
|
2 | |
|
|
3 | 4.351 Sat Apr 11 07:56:14 CEST 2009 |
|
|
4 | - actually make the signal pipe work under win32. |
|
|
5 | - localise $! in signal handler to avoid changing $!, although |
|
|
6 | perl itself does not seem to save/restore errno either. |
|
|
7 | - set the cloexec flag on the signal pipe (normally set by perl too). |
|
|
8 | |
|
|
9 | 4.35 Fri Mar 27 11:48:20 CET 2009 |
|
|
10 | - event models relying on AnyEvent's signal watcher emulation did |
|
|
11 | invoke the callback asynchronously, contradicting documentation |
|
|
12 | and causing signals to get lost (this includes AnyEvent's own |
|
|
13 | event loop). AnyEvent now uses the standard pipe trick to make |
|
|
14 | callback execution synchronous to the event loop. |
|
|
15 | - AnyEvent::Handle didn't free TLS context data on DESTROY |
|
|
16 | (patch by Pavel Shaydo). |
|
|
17 | - work around the ever-incompatibly-changing API of MakeMaker. |
|
|
18 | - document that changing global variables without restoring them |
|
|
19 | is a bad idea in Perl (noted by Adam Rosenstein). |
|
|
20 | - AnyEvent::Strict now barfs if ->io is passed a file. |
|
|
21 | |
|
|
22 | 4.34 Thu Feb 12 18:32:45 CET 2009 |
|
|
23 | - separately buffer TLS read data, as otherwise the read queue could |
|
|
24 | deadlock as receiving data is not expected while draining the |
|
|
25 | read queue (which cna only happen with TLS). |
|
|
26 | - raise EBADMSG error on JSON decoding errors. |
|
|
27 | - fix some minor manpage bugs (reported by Maximilian Gaß). |
|
|
28 | - speed up select bitmask parsing quite a bit in the pure perl |
|
|
29 | backend. |
|
|
30 | - use CORE::select instead of just select in AnyEvent::Impl::Perl. |
|
|
31 | |
|
|
32 | 4.331 Tue Jan 6 21:07:25 CET 2009 |
|
|
33 | - socketpair fails on many vista machines because vista has |
|
|
34 | completely broken accept/getpeername and getsockname functions, |
|
|
35 | so we provide our own socketpair emulation that kind of works |
|
|
36 | (AnyEvent::Util::portable_pipe). |
|
|
37 | - new function: AnyEvent::Util::portable_socketpair. |
|
|
38 | - take advantage of the Guard module if it exists. |
|
|
39 | |
|
|
40 | 4.33 Fri Nov 21 02:35:40 CET 2008 |
|
|
41 | - AnyEvent::Strict did errornously flag a fileno of 0 as |
|
|
42 | illegal. |
|
|
43 | - reduce memory usage and slightly speed up the pure perl backend |
|
|
44 | by only storing the file descriptor, not the file handle. |
|
|
45 | - add missing autocork method to AnyEvent::Handle (reported |
|
|
46 | by Adam Rosenstein). |
|
|
47 | - AnyEvent::DNS->resolve errornously documented an $rcode result |
|
|
48 | argument, but there isn't (spotted by Henrik Krohns). |
|
|
49 | - the naked truth about Tk - it's basically unmaintained. |
|
|
50 | |
|
|
51 | 4.32 Mon Nov 3 22:46:32 CET 2008 |
|
|
52 | - fix AnyEvent::Socket::resolve_sockaddr to properly support |
|
|
53 | unix sockets again and choose a proper default for |
|
|
54 | the socket type. |
|
|
55 | - fix call to resolve_sockaddr in tcp_connect. |
|
|
56 | |
|
|
57 | 4.31 Thu Oct 30 04:41:48 CET 2008 |
|
|
58 | - implemented AnyEvent::Handle->destroy method. |
|
|
59 | - hint about unexpected effects in TLS mode. |
|
|
60 | - speed up AnyEvent::Impl::Perl by using more arrays |
|
|
61 | instead of hashes without the slightest loss of readability :) |
|
|
62 | - work around a perl argument refcounting bug. |
|
|
63 | - with some perl patches applied, AnyEvent now seems leak-free. |
|
|
64 | |
|
|
65 | 4.3 Fri Oct 3 09:18:43 CEST 2008 |
|
|
66 | - AnyEvent will now install a no-op signal handler for SIGPIPE, |
|
|
67 | unless one has been installed already. |
|
|
68 | - warn about not loaidng AnyEvent::Impl::POE early enough |
|
|
69 | (patch by Adam Rosenstein). |
|
|
70 | - fixed a great number of bugs and corner cases in AnyEvent::Handle: |
|
|
71 | - fix a bug in where in SSL connect mode, the |
|
|
72 | client would first wait for some data by the server and |
|
|
73 | otherwise hang. (reported and analyzed in an absolutely |
|
|
74 | exemplary manner by Adam Rosenstein). |
|
|
75 | - fix a bug in where SSL EOF would not be treated |
|
|
76 | as stream EOF, putting the connection into a hung state. |
|
|
77 | (reported and analyzed in an absolutely exemplary manner by Adam |
|
|
78 | Rosenstein). |
|
|
79 | - fix a potential segfault when the TLS context |
|
|
80 | would go missing in a read callback (e.g. due to stoptls) - |
|
|
81 | Net::SSLeay of course makes no type checking whatsoever. |
|
|
82 | - AnyEvent::Handle will not stop the read watcher in TLS mode. |
|
|
83 | - AnyEvent::Handle->stoptls will send a TLS close notify |
|
|
84 | instead of simply destroying the stream now. |
|
|
85 | - fix a bug where large blocks of data written in TLS mode |
|
|
86 | would not be sent unless triggered by receives. |
|
|
87 | - on_drain will now take the tls write buffer into account. |
|
|
88 | - SSL operations have been streamlined a bit and should be faster |
|
|
89 | now, more to come, though. |
|
|
90 | - added a FAQ section, will document nontrivial issues in there. |
|
|
91 | |
|
|
92 | 4.234 Mon Sep 29 04:08:13 CEST 2008 |
|
|
93 | - fix child watcher documentation: only child exits will be reported, |
|
|
94 | no trace events. |
|
|
95 | - mention SIGPIPE in AnyEvent::Handle. |
|
|
96 | - perl backend will now use POSIX::times as monotonic clock |
|
|
97 | when available and the monotonic clock option is not. |
|
|
98 | - run even without Time::HiRes in some cases now, but do not |
|
|
99 | provide sub-second accuracy (all *supported* perl versions have |
|
|
100 | Time::HiRes). |
|
|
101 | |
|
|
102 | 4.233 Fri Aug 22 01:48:05 CEST 2008 |
|
|
103 | - fix a bug in the testsuite, causing a hang. |
|
|
104 | - clarified lots of AnyEvent::Handle constructor arguments. |
|
|
105 | |
|
|
106 | 4.232 Thu Aug 21 20:44:25 CEST 2008 |
|
|
107 | - fix a bug in AnyEvent::Handle that could cause two fatal error |
|
|
108 | callbacks on EOF, which caused AnyEvent::HTTP to signal success twice |
|
|
109 | (testcase by Leon Brocard). |
|
|
110 | - clarified on_eof/on_error documentation a bit. |
|
|
111 | |
|
|
112 | 4.231 Tue Jul 29 13:12:15 CEST 2008 |
|
|
113 | - remove some debugging code left in AnyEvent::Util::fork_call |
|
|
114 | (and no, it's impossible to implement with the broken windows |
|
|
115 | perls without resource leaks or worse). |
|
|
116 | |
|
|
117 | 4.23 Tue Jul 29 12:19:59 CEST 2008 |
|
|
118 | - document the first parameter passed to condvar callbacks to be |
|
|
119 | the callback. |
|
|
120 | - add AnyEvent::Socket::{ntoa,aton} aliases. |
|
|
121 | - optimize the AE::Handle->push_read (line) for the default |
|
|
122 | eol marker. |
|
|
123 | - optimize push_read (packstring|storable) for small packets. |
|
|
124 | - invoke on_error callback when no on_eof callback is set. |
|
|
125 | - fix a bug in push_read (storable) of unknown impact. |
|
|
126 | |
|
|
127 | 4.22 Sun Jul 20 16:34:13 CEST 2008 |
|
|
128 | - new function AnyEvent::Socket::parse_hostport. |
|
|
129 | - as the bulkheads at microsoft can't even get getprotobyname reliably |
|
|
130 | working on their shitty fucking broken os we need to hardcode |
|
|
131 | some common protocol numbers in AnyEvent::Socket. How can |
|
|
132 | people even bother with such a pile of shit as windows. |
|
|
133 | |
|
|
134 | 4.21 Thu Jul 17 14:40:05 CEST 2008 |
|
|
135 | - INCOMPATIBLE CHANGE: AnyEvent::Util::fork_call now has a prototype |
|
|
136 | which is incompatible to previous invocation syntax. |
|
|
137 | - work around more windows perl bugs in fork_call. I cannot imagine |
|
|
138 | why anybody would bother implementing fork in such an obviously |
|
|
139 | unusable way. |
|
|
140 | - avoid calling $do_retry if the request has already finished in |
|
|
141 | AnyEvent::DNS, thus avoiding an exception (reported by Anatoly |
|
|
142 | K. Sharifulin). |
|
|
143 | |
|
|
144 | 4.2 Sat Jul 12 22:42:11 CEST 2008 |
|
|
145 | - support an "interval" argument for repeating timers. |
|
|
146 | - fix a bug in the pure perl event loop that caused timers to |
|
|
147 | be delayed under some circumstances. |
|
|
148 | - implement optional argument checking via PERL_ANYEVENT_STRICT. |
|
|
149 | - not importing anything from Event causes it to malfunction |
|
|
150 | and not use Time::HiRes, work around this bug. |
|
|
151 | - more bug workarounds against the endlessly broken Tk module, |
|
|
152 | seems to work now on 32 bit machines, but less so on 64 bit. |
|
|
153 | - minor optimisations applied to most backends. |
|
|
154 | |
|
|
155 | 4.161 Fri Jul 4 14:24:48 CEST 2008 |
|
|
156 | - fixed the t/handle/02_write.t, it read after write, with the |
|
|
157 | assumption that all data could be written before read |
|
|
158 | (reported by Jim Razmus). |
|
|
159 | |
|
|
160 | 4.160 Thu Jul 3 04:02:21 CEST 2008 |
|
|
161 | - re-registering signal watchers after unregistering one did |
|
|
162 | not work if a backend falls back on AnyEvent's default |
|
|
163 | implementation (also affected child watchers). |
|
|
164 | - new AnyEvent::Handle options: autocork and no_delay. |
|
|
165 | |
|
|
166 | 4.152 Sun Jun 22 14:15:44 CEST 2008 |
|
|
167 | - allow for 32-bit perls that implement shifts differently |
|
|
168 | on different architectures in parse_ipv4 (reported and |
|
|
169 | analysed by Keiichi DAIBA). |
|
|
170 | |
|
|
171 | 4.151 Fri Jun 6 17:34:24 CEST 2008 |
|
|
172 | - make sure specifying _only_ on_read and never pushing reads |
|
|
173 | works. |
|
|
174 | |
|
|
175 | 4.15 Fri Jun 6 13:00:46 CEST 2008 |
|
|
176 | - the pure perl backend would keep some watchers alive when more than |
|
|
177 | one watcher was registered for the same fd. |
|
|
178 | - new "packstring" and "storable" read and write types |
|
|
179 | for AnyEvent::Handle. |
|
|
180 | - allow on_eof handler to be called after on_error with EPIPE returns. |
|
|
181 | - do not immediately call on_read callback in handle constructor. |
|
|
182 | |
|
|
183 | 4.14 Thu Jun 5 20:29:31 CEST 2008 |
|
|
184 | - Fixed a bug in DNS SRV priority sorting. |
|
|
185 | - AnyEvent::Util::guard now reports runtime errors while |
|
|
186 | executing the guard block as warnings. |
|
|
187 | - handle 0-byte-reads just before EOF correctly in AnyEvent::Handle. |
|
|
188 | |
|
|
189 | 4.13 Thu Jun 5 00:47:59 CEST 2008 |
|
|
190 | - AnyEvent::DNS only followed cname chains with length 2, |
|
|
191 | contrary to documentation. bumped it up to 10, thanks to |
|
|
192 | microsoft, the current broken-dns-config-king. |
|
|
193 | - AnyEvent::DNS didn't check for socket return status |
|
|
194 | but instead relied on perl not creating filehandles in that |
|
|
195 | case - too bad it gives you a bogus file handle (reported |
|
|
196 | and analysed by Vladimir Timofeev). |
|
|
197 | - fix queue management logic in AnyEvent::Handle: |
|
|
198 | when on_read was registered, the queue was empty |
|
|
199 | and no progress could be made AnyEvent::Handle would |
|
|
200 | enter an endless loop. |
|
|
201 | - correctly start reading again when the handle became |
|
|
202 | busy again after idling. |
|
|
203 | - correctly treat tls shut-downs as EOF condition |
|
|
204 | (for the time being, would be nice to have a callback for that). |
|
|
205 | - correctly call eof callback under all conditions |
|
|
206 | (hopefully). |
|
|
207 | - the timeout callback did not expect that $self can go |
|
|
208 | away any time. |
|
|
209 | |
|
|
210 | 4.12 Tue Jun 3 10:58:04 CEST 2008 |
|
|
211 | - include AnyEvent::Intro, a tutorial for anyevent, |
|
|
212 | anyevent::socket and anyevent::handle. |
|
|
213 | - allow more options in on_error. |
|
|
214 | |
|
|
215 | 4.11 Fri May 30 23:42:25 CEST 2008 |
|
|
216 | - INCOMPATIBLE CHANGE: replace ptr by real PTR lookup, provide |
|
|
217 | reverse_lookup and reverse_verify to replace it, support |
|
|
218 | v4mapped and v4compat addresses. |
|
|
219 | - provide more documentation for the resolver class. |
|
|
220 | - really replace longest run of :0: by :: in format_address, |
|
|
221 | also properly convert :: and ::1 again. |
|
|
222 | - support NAPTR record name and decode it. |
|
|
223 | - implement random weight sampling for SRV records, as per |
|
|
224 | rfc 2782. |
|
|
225 | - correctly abort on srv-record targets of ".". |
|
|
226 | - added AnyEvent::DNS::wait_for_slot. |
|
|
227 | - in the unlikely event of a virtual circuit connection |
|
|
228 | being invalidated by a delayed udp reply, AnyEvent::DNS |
|
|
229 | could die. this has been fixed. |
|
|
230 | - plug a probable memleak in the DNS vc code. |
|
|
231 | - use configured timeout also for VC request phase. |
|
|
232 | - implement timeout and max_outstanding methods |
|
|
233 | for resolver class. |
|
|
234 | - update version numbers in all modules. |
|
|
235 | |
|
|
236 | 4.1 Thu May 29 05:45:40 CEST 2008 |
|
|
237 | - INCOMPATIBLE CHANGE: renamed xxx_ip functions to xxx_address |
|
|
238 | which mirrors their purpose better (old names still available |
|
|
239 | for a while). Also moved AnyEvent::DNS::addr to |
|
|
240 | AnyEvent::Socket::resolve_sockaddr. |
|
|
241 | - implement AnyEvent->time and AnyEvent->now. |
|
|
242 | - fix IPv6 support in pack_sockaddr. |
|
|
243 | - officially un-experimentalise all newly introduced code. |
|
|
244 | - support unix domain sockets everywhere by specifying |
|
|
245 | a host of "unix/" and the pathname as service. |
|
|
246 | - implement an activity timeout in AnyEvent::Handle. |
|
|
247 | - added a regex read type to AnyEvent::Handle. |
|
|
248 | - added a json read/write type to AnyEvent::Handle. |
|
|
249 | - always croak in on_error in AnyEvent::Handle. |
|
|
250 | - document how to subclass AnyEvent::Handle. |
|
|
251 | - implement AnyEvent::Util::fork_call. |
|
|
252 | - add support for IPv6 nameservers and nameserver statements. |
|
|
253 | - work around _yet_ _another_ windows perl bug in where empty |
|
|
254 | select masks cause errors in the pure perl backend (this |
|
|
255 | workaround cannot be implemented for the other event loops, |
|
|
256 | but EV also works around this). |
|
|
257 | - supply AnyEvent::Socket::address_type and make good use of it. |
|
|
258 | - clarify and fix the Handle documentation w.r.t $self vs. $handle |
|
|
259 | vs. callback arguments. |
|
|
260 | - add some recommends to the META.yml. |
|
|
261 | |
|
|
262 | 4.05 Mon May 26 19:44:06 CEST 2008 |
|
|
263 | - some platforms ignore the rfc and prepend an extra sa_len member to |
|
|
264 | the sockaddr structure, cater for those. also use sockaddr_family, |
|
|
265 | didn't know it was there, but comes in handy for the workaround. |
|
|
266 | - undo the import WIN32 hack. |
|
|
267 | |
|
|
268 | 4.04 Mon May 26 08:03:31 CEST 2008 |
|
|
269 | - try to work around yet another windows bug: failed connects are reported |
|
|
270 | as if it were out-of-band data. windows users: you suck. Only |
|
|
271 | EV, Glib, Event and the pure perl backend can handle this condition. |
|
|
272 | - optimize write algorithm in AnyEvent::Handle. |
|
|
273 | - properly parse PERL_ANYEVENT_PROTOCOLS and use the info correctly. |
|
|
274 | - tcp_connect now properly iterates through all targets. |
|
|
275 | - check wether IPv6 sockets can actually be created, otherwise, disable ipv6 |
|
|
276 | support entirely. also hardcode some AF_INET6 constants for the |
|
|
277 | sake of perl 5.8. |
|
|
278 | - vastly improved nameserver/suffix detection algorithm for windows |
|
|
279 | (still a hack...) |
|
|
280 | - try to use a fallback dns server if no dns servers could be found. |
|
|
281 | - splatter around some local $SIG{__DIE__}'s to work around bad |
|
|
282 | die handlers. |
|
|
283 | |
|
|
284 | 4.03 Sun May 25 05:05:57 CEST 2008 |
|
|
285 | - work around perl bugs on windows, where perl returns undocumented |
|
|
286 | error codes for sysread, syswrite etc. |
|
|
287 | - fix AnyEvent::Handle::unshift_read to really unshift and not push. |
|
|
288 | - AnyEvent::Handle could sometimes stumble over 0-byte writes. |
|
|
289 | - fix unpack_sockaddr, now works for AF_INET6 sockaddr's. |
|
|
290 | - updated documentation. |
|
|
291 | - lots of workarounds for perl 5.8 quirks. |
|
|
292 | - implement a more flexible type system for both reads and writes |
|
|
293 | in AnyEvent::Handle. |
|
|
294 | - support IPv6 in tcp_server. |
|
|
295 | - pass local hostname and port to the prepare callback in tcp_server. |
|
|
296 | - make types extendable in AnyEvent::Handle. |
|
|
297 | - croak, not die, in AnyEvent::Handle; |
|
|
298 | |
|
|
299 | 4.0 Sat May 24 19:58:08 CEST 2008 |
|
|
300 | - added (experimental) AnyEvent::DNS module. |
|
|
301 | - added (experimental) AnyEvent::Socket module. |
|
|
302 | - added (experiemntal) support for TLS/SSL in AnyEvent::Handle. |
|
|
303 | - make condvar->end call ->send as documented. |
|
|
304 | - make condvar constructor support the documented cb => argument. |
|
|
305 | - overload condvars so you can use them instead of code references. |
|
|
306 | - support more of the POSIX forms instead of a pure dotted quad for |
|
|
307 | verbatim ip addresses. |
|
|
308 | - AnyEvent::Socket::inet_aton will use AnyEvent::DNS now and not fork anymore. |
|
|
309 | - try very hard not to freeze on broken windows perls. |
|
|
310 | - make on_eof optional in AnyEvent::Handle. |
|
|
311 | - support service names in tcp_connect and tcp_server. |
|
|
312 | - work around netbsd/openbsd bug causing unix sockets not to |
|
|
313 | become writable until empty. |
|
|
314 | - only make a dummy read for error detection when getpeername |
|
|
315 | returns ENOTCONN. |
|
|
316 | - defend AnyEvent::Util::inet_aton against Coro::LWP's brutal |
|
|
317 | override. |
|
|
318 | - new AnyEvent::Guard::cancel. |
|
|
319 | |
|
|
320 | 3.5 Sat May 17 23:17:14 CEST 2008 |
|
|
321 | - gracefully deal with systems not defining CLOCK_MONOTONIC |
|
|
322 | (as opposed to not supporting the clock itself, which worked in 3.41). |
|
|
323 | - added AnyEvent::Util::tcp_connect and AnyEvent::Util::tcp_server |
|
|
324 | helpers, to replace the ill-designed AnyEvent::Socket class. |
|
|
325 | - added some examples (eg/connect, eg/handle). |
|
|
326 | - added AnyEvent::Util::guard. |
|
|
327 | - AnyEvent::Util now exports most of its functions. |
|
|
328 | - fixed a bug in *_read_chunk, where the callback was called with $self as |
|
|
329 | first and second argument. |
|
|
330 | - simplified and fixed regex handling in AnyEvent::Handle::*_read_line. |
|
|
331 | - fix manpage condvars examples to use new syntax. |
|
|
332 | - considerably improve inet_aton with EV::ADNS backend to support |
|
|
333 | cname chains (common dns configuration bug, e.g. with www.google.de) |
|
|
334 | and aaaa records if no a record could be found. |
|
|
335 | |
|
|
336 | 3.41 Sun May 11 19:53:13 CEST 2008 |
|
|
337 | - work around a bug in perl 5.8.8's local. Reported by Yi Ma Mao. |
|
|
338 | |
|
|
339 | 3.4 Sun May 11 00:29:25 CEST 2008 |
|
|
340 | - complete re-implementation of condvars with a hopefully much more useful |
|
|
341 | API (backward compatible functions exist but are undocumented). |
|
|
342 | - AnyEvent::Handle fully rewritten with a hopefully more powerful API |
|
|
343 | (still subject to change). |
|
|
344 | - take advantage of CLOCK_MONOTONIC, if available, in AnyEvent::Impl::Perl. |
|
|
345 | - provide AnyEvent::post_detect and @AnyEvent::post_detect, which |
|
|
346 | allows module authors to avoid forcing event loop detection. |
|
|
347 | (used by Coro::AIO, Coro::BDB, Coro::AnyEvent for example). |
|
|
348 | - remove coro backends: Coro now provides generic support for AnyEvent |
|
|
349 | via Coro::AnyEvent. |
|
|
350 | |
|
|
351 | 3.3 Mon Apr 28 09:51:06 CEST 2008 |
|
|
352 | - added AnyEvent::Handle, AnyEvent::Socket and AnyEvent::Util modules. |
|
|
353 | - fix a bug in the pure perl backend that kept watchers alive |
|
|
354 | when multiple watchers were registered for the same fh/poll combo. |
|
|
355 | - add a benchmark section showing AnyEvent overhead and comparing |
|
|
356 | the different event loops with each other. |
|
|
357 | - prefer pure perl over tk when autoprobing, as it's about as fast, |
|
|
358 | but doesn't crash with many watchers. |
|
|
359 | - declare Qt support non-experimental. |
|
|
360 | - clarify the confusing section about the file descriptor being |
|
|
361 | kept alive. |
|
|
362 | - document the race between loading of an event module and |
|
|
363 | child processes exiting. |
|
|
364 | - support POE as "backend" (with some caveats, POE is not generic |
|
|
365 | enough, and darn slow). |
|
|
366 | - support Wx and Prima through POE. |
|
|
367 | - optimise perl backend to use 20% less memory and take advantage |
|
|
368 | of typical timeout behaviour. It can now compete with |
|
|
369 | select/poll-based C event loops in most cases (it is usually |
|
|
370 | faster than Event and Glib :). |
|
|
371 | - roughly cut EV memory use in half and increase its speed by 30%, |
|
|
372 | by removing undocumented functionality. Did something similar |
|
|
373 | to other event loops where possible. |
| 2 | |
374 | |
| 3 | 3.2 Thu Apr 24 10:10:40 CEST 2008 |
375 | 3.2 Thu Apr 24 10:10:40 CEST 2008 |
| 4 | - do not die when anyevent watchers are destroyed while running callbacks |
376 | - do not die when anyevent watchers are destroyed while running callbacks |
| 5 | in the pureperl backend (could only happen when two watchers are registered |
377 | in the pureperl backend (could only happen when two watchers are registered |
| 6 | for the same fh/poll combo). |
378 | for the same fh/poll combo). |
