1 | Revision history for Perl extension AnyEvent. |
1 | Revision history for Perl extension AnyEvent. |
2 | |
2 | |
3 | 2.0 |
3 | - preliminary and neecssarily incomplete support for IO::Async. |
4 | - INCOMPATIBLE CHANGE: |
4 | - reset SIGCHLD to DEFAULT when AnyEvent is loaded, in |
5 | poll can now either be "r" or "w" but not both. |
5 | case it was set to IGNORE, to ensure we can catch child statuses |
6 | - INCOMPATIBLE CHANGE: |
6 | even when the calling env acts stupidly. |
7 | io watcher callbacks have no arguments anymore. |
7 | |
8 | - new pure-perl model added, AnyEvent now always |
8 | 4.412 Wed Jun 24 01:35:57 CEST 2009 |
|
|
9 | - support an "untaint" attribute for AnyEvent::DNS |
|
|
10 | and set it on the default resolver. |
|
|
11 | - implement and document AnyEvent::Impl::Perl::loop. |
|
|
12 | - remove all anyevent-env variables from %ENV |
|
|
13 | when running in tainted mode. |
|
|
14 | - mention and extend the IO::Lambda benchmark. |
|
|
15 | |
|
|
16 | 4.411 Sun Jun 7 18:48:02 CEST 2009 |
|
|
17 | - do not try to use F_SETFD on windows. |
|
|
18 | |
|
|
19 | 4.41 Thu May 14 06:40:11 CEST 2009 |
|
|
20 | - work around issues in older perls (5.8.0?) when |
|
|
21 | a signal handler is deleted from the %SIG hash. |
|
|
22 | - use POSIX::_exit in child test, to avoid running |
|
|
23 | destructors. |
|
|
24 | - speed up CHLD handling by relying on SIGCHLD being |
|
|
25 | synchronously delivered, even when we roll our own |
|
|
26 | implementation. |
|
|
27 | - AnyEvent::DNS: add the "dname" resource record name |
|
|
28 | for cosmetic reasons. |
|
|
29 | |
|
|
30 | 4.4 Sun Apr 26 20:12:33 CEST 2009 |
|
|
31 | - implemented idle watchers, where applicable. |
|
|
32 | - AnyEvent->time died when Event backend was in use. |
|
|
33 | - fix a memleak in the Tk backend. |
|
|
34 | - sped up Tk timer handling. |
|
|
35 | - clip negative "after" values to 0 in AnyEvent::Impl::Event |
|
|
36 | to avoid spamming out warning messages. |
|
|
37 | - fix Qt timers without interval. |
|
|
38 | - avoid Qt zero-timeout special case, in old libqt's. |
|
|
39 | |
|
|
40 | 4.352 Mon Apr 20 16:31:11 CEST 2009 |
|
|
41 | - fix AnyEvent::Strict error messages for child watchers. |
|
|
42 | - fix/update Makefile.PL "recommends" versions. |
|
|
43 | - add AnyEvent->now_update. |
|
|
44 | |
|
|
45 | 4.351 Sat Apr 11 07:56:14 CEST 2009 |
|
|
46 | - actually make the signal pipe work under win32. |
|
|
47 | - localise $! in signal handler to avoid changing $!, although |
|
|
48 | perl itself does not seem to save/restore errno either. |
|
|
49 | - set the cloexec flag on the signal pipe (normally set by perl too). |
|
|
50 | |
|
|
51 | 4.35 Fri Mar 27 11:48:20 CET 2009 |
|
|
52 | - event models relying on AnyEvent's signal watcher emulation did |
|
|
53 | invoke the callback asynchronously, contradicting documentation |
|
|
54 | and causing signals to get lost (this includes AnyEvent's own |
|
|
55 | event loop). AnyEvent now uses the standard pipe trick to make |
|
|
56 | callback execution synchronous to the event loop. |
|
|
57 | - AnyEvent::Handle didn't free TLS context data on DESTROY |
|
|
58 | (patch by Pavel Shaydo). |
|
|
59 | - work around the ever-incompatibly-changing API of MakeMaker. |
|
|
60 | - document that changing global variables without restoring them |
|
|
61 | is a bad idea in Perl (noted by Adam Rosenstein). |
|
|
62 | - AnyEvent::Strict now barfs if ->io is passed a file. |
|
|
63 | |
|
|
64 | 4.34 Thu Feb 12 18:32:45 CET 2009 |
|
|
65 | - separately buffer TLS read data, as otherwise the read queue could |
|
|
66 | deadlock as receiving data is not expected while draining the |
|
|
67 | read queue (which cna only happen with TLS). |
|
|
68 | - raise EBADMSG error on JSON decoding errors. |
|
|
69 | - fix some minor manpage bugs (reported by Maximilian Gaß). |
|
|
70 | - speed up select bitmask parsing quite a bit in the pure perl |
|
|
71 | backend. |
|
|
72 | - use CORE::select instead of just select in AnyEvent::Impl::Perl. |
|
|
73 | |
|
|
74 | 4.331 Tue Jan 6 21:07:25 CET 2009 |
|
|
75 | - socketpair fails on many vista machines because vista has |
|
|
76 | completely broken accept/getpeername and getsockname functions, |
|
|
77 | so we provide our own socketpair emulation that kind of works |
|
|
78 | (AnyEvent::Util::portable_pipe). |
|
|
79 | - new function: AnyEvent::Util::portable_socketpair. |
|
|
80 | - take advantage of the Guard module if it exists. |
|
|
81 | |
|
|
82 | 4.33 Fri Nov 21 02:35:40 CET 2008 |
|
|
83 | - AnyEvent::Strict did errornously flag a fileno of 0 as |
|
|
84 | illegal. |
|
|
85 | - reduce memory usage and slightly speed up the pure perl backend |
|
|
86 | by only storing the file descriptor, not the file handle. |
|
|
87 | - add missing autocork method to AnyEvent::Handle (reported |
|
|
88 | by Adam Rosenstein). |
|
|
89 | - AnyEvent::DNS->resolve errornously documented an $rcode result |
|
|
90 | argument, but there isn't (spotted by Henrik Krohns). |
|
|
91 | - the naked truth about Tk - it's basically unmaintained. |
|
|
92 | |
|
|
93 | 4.32 Mon Nov 3 22:46:32 CET 2008 |
|
|
94 | - fix AnyEvent::Socket::resolve_sockaddr to properly support |
|
|
95 | unix sockets again and choose a proper default for |
|
|
96 | the socket type. |
|
|
97 | - fix call to resolve_sockaddr in tcp_connect. |
|
|
98 | |
|
|
99 | 4.31 Thu Oct 30 04:41:48 CET 2008 |
|
|
100 | - implemented AnyEvent::Handle->destroy method. |
|
|
101 | - hint about unexpected effects in TLS mode. |
|
|
102 | - speed up AnyEvent::Impl::Perl by using more arrays |
|
|
103 | instead of hashes without the slightest loss of readability :) |
|
|
104 | - work around a perl argument refcounting bug. |
|
|
105 | - with some perl patches applied, AnyEvent now seems leak-free. |
|
|
106 | |
|
|
107 | 4.3 Fri Oct 3 09:18:43 CEST 2008 |
|
|
108 | - AnyEvent will now install a no-op signal handler for SIGPIPE, |
|
|
109 | unless one has been installed already. |
|
|
110 | - warn about not loaidng AnyEvent::Impl::POE early enough |
|
|
111 | (patch by Adam Rosenstein). |
|
|
112 | - fixed a great number of bugs and corner cases in AnyEvent::Handle: |
|
|
113 | - fix a bug in where in SSL connect mode, the |
|
|
114 | client would first wait for some data by the server and |
|
|
115 | otherwise hang. (reported and analyzed in an absolutely |
|
|
116 | exemplary manner by Adam Rosenstein). |
|
|
117 | - fix a bug in where SSL EOF would not be treated |
|
|
118 | as stream EOF, putting the connection into a hung state. |
|
|
119 | (reported and analyzed in an absolutely exemplary manner by Adam |
|
|
120 | Rosenstein). |
|
|
121 | - fix a potential segfault when the TLS context |
|
|
122 | would go missing in a read callback (e.g. due to stoptls) - |
|
|
123 | Net::SSLeay of course makes no type checking whatsoever. |
|
|
124 | - AnyEvent::Handle will not stop the read watcher in TLS mode. |
|
|
125 | - AnyEvent::Handle->stoptls will send a TLS close notify |
|
|
126 | instead of simply destroying the stream now. |
|
|
127 | - fix a bug where large blocks of data written in TLS mode |
|
|
128 | would not be sent unless triggered by receives. |
|
|
129 | - on_drain will now take the tls write buffer into account. |
|
|
130 | - SSL operations have been streamlined a bit and should be faster |
|
|
131 | now, more to come, though. |
|
|
132 | - added a FAQ section, will document nontrivial issues in there. |
|
|
133 | |
|
|
134 | 4.234 Mon Sep 29 04:08:13 CEST 2008 |
|
|
135 | - fix child watcher documentation: only child exits will be reported, |
|
|
136 | no trace events. |
|
|
137 | - mention SIGPIPE in AnyEvent::Handle. |
|
|
138 | - perl backend will now use POSIX::times as monotonic clock |
|
|
139 | when available and the monotonic clock option is not. |
|
|
140 | - run even without Time::HiRes in some cases now, but do not |
|
|
141 | provide sub-second accuracy (all *supported* perl versions have |
|
|
142 | Time::HiRes). |
|
|
143 | |
|
|
144 | 4.233 Fri Aug 22 01:48:05 CEST 2008 |
|
|
145 | - fix a bug in the testsuite, causing a hang. |
|
|
146 | - clarified lots of AnyEvent::Handle constructor arguments. |
|
|
147 | |
|
|
148 | 4.232 Thu Aug 21 20:44:25 CEST 2008 |
|
|
149 | - fix a bug in AnyEvent::Handle that could cause two fatal error |
|
|
150 | callbacks on EOF, which caused AnyEvent::HTTP to signal success twice |
|
|
151 | (testcase by Leon Brocard). |
|
|
152 | - clarified on_eof/on_error documentation a bit. |
|
|
153 | |
|
|
154 | 4.231 Tue Jul 29 13:12:15 CEST 2008 |
|
|
155 | - remove some debugging code left in AnyEvent::Util::fork_call |
|
|
156 | (and no, it's impossible to implement with the broken windows |
|
|
157 | perls without resource leaks or worse). |
|
|
158 | |
|
|
159 | 4.23 Tue Jul 29 12:19:59 CEST 2008 |
|
|
160 | - document the first parameter passed to condvar callbacks to be |
|
|
161 | the callback. |
|
|
162 | - add AnyEvent::Socket::{ntoa,aton} aliases. |
|
|
163 | - optimize the AE::Handle->push_read (line) for the default |
|
|
164 | eol marker. |
|
|
165 | - optimize push_read (packstring|storable) for small packets. |
|
|
166 | - invoke on_error callback when no on_eof callback is set. |
|
|
167 | - fix a bug in push_read (storable) of unknown impact. |
|
|
168 | |
|
|
169 | 4.22 Sun Jul 20 16:34:13 CEST 2008 |
|
|
170 | - new function AnyEvent::Socket::parse_hostport. |
|
|
171 | - as the bulkheads at microsoft can't even get getprotobyname reliably |
|
|
172 | working on their shitty fucking broken os we need to hardcode |
|
|
173 | some common protocol numbers in AnyEvent::Socket. How can |
|
|
174 | people even bother with such a pile of shit as windows. |
|
|
175 | |
|
|
176 | 4.21 Thu Jul 17 14:40:05 CEST 2008 |
|
|
177 | - INCOMPATIBLE CHANGE: AnyEvent::Util::fork_call now has a prototype |
|
|
178 | which is incompatible to previous invocation syntax. |
|
|
179 | - work around more windows perl bugs in fork_call. I cannot imagine |
|
|
180 | why anybody would bother implementing fork in such an obviously |
|
|
181 | unusable way. |
|
|
182 | - avoid calling $do_retry if the request has already finished in |
|
|
183 | AnyEvent::DNS, thus avoiding an exception (reported by Anatoly |
|
|
184 | K. Sharifulin). |
|
|
185 | |
|
|
186 | 4.2 Sat Jul 12 22:42:11 CEST 2008 |
|
|
187 | - support an "interval" argument for repeating timers. |
|
|
188 | - fix a bug in the pure perl event loop that caused timers to |
|
|
189 | be delayed under some circumstances. |
|
|
190 | - implement optional argument checking via PERL_ANYEVENT_STRICT. |
|
|
191 | - not importing anything from Event causes it to malfunction |
|
|
192 | and not use Time::HiRes, work around this bug. |
|
|
193 | - more bug workarounds against the endlessly broken Tk module, |
|
|
194 | seems to work now on 32 bit machines, but less so on 64 bit. |
|
|
195 | - minor optimisations applied to most backends. |
|
|
196 | |
|
|
197 | 4.161 Fri Jul 4 14:24:48 CEST 2008 |
|
|
198 | - fixed the t/handle/02_write.t, it read after write, with the |
|
|
199 | assumption that all data could be written before read |
|
|
200 | (reported by Jim Razmus). |
|
|
201 | |
|
|
202 | 4.160 Thu Jul 3 04:02:21 CEST 2008 |
|
|
203 | - re-registering signal watchers after unregistering one did |
|
|
204 | not work if a backend falls back on AnyEvent's default |
|
|
205 | implementation (also affected child watchers). |
|
|
206 | - new AnyEvent::Handle options: autocork and no_delay. |
|
|
207 | |
|
|
208 | 4.152 Sun Jun 22 14:15:44 CEST 2008 |
|
|
209 | - allow for 32-bit perls that implement shifts differently |
|
|
210 | on different architectures in parse_ipv4 (reported and |
|
|
211 | analysed by Keiichi DAIBA). |
|
|
212 | |
|
|
213 | 4.151 Fri Jun 6 17:34:24 CEST 2008 |
|
|
214 | - make sure specifying _only_ on_read and never pushing reads |
9 | works. |
215 | works. |
|
|
216 | |
|
|
217 | 4.15 Fri Jun 6 13:00:46 CEST 2008 |
|
|
218 | - the pure perl backend would keep some watchers alive when more than |
|
|
219 | one watcher was registered for the same fd. |
|
|
220 | - new "packstring" and "storable" read and write types |
|
|
221 | for AnyEvent::Handle. |
|
|
222 | - allow on_eof handler to be called after on_error with EPIPE returns. |
|
|
223 | - do not immediately call on_read callback in handle constructor. |
|
|
224 | |
|
|
225 | 4.14 Thu Jun 5 20:29:31 CEST 2008 |
|
|
226 | - Fixed a bug in DNS SRV priority sorting. |
|
|
227 | - AnyEvent::Util::guard now reports runtime errors while |
|
|
228 | executing the guard block as warnings. |
|
|
229 | - handle 0-byte-reads just before EOF correctly in AnyEvent::Handle. |
|
|
230 | |
|
|
231 | 4.13 Thu Jun 5 00:47:59 CEST 2008 |
|
|
232 | - AnyEvent::DNS only followed cname chains with length 2, |
|
|
233 | contrary to documentation. bumped it up to 10, thanks to |
|
|
234 | microsoft, the current broken-dns-config-king. |
|
|
235 | - AnyEvent::DNS didn't check for socket return status |
|
|
236 | but instead relied on perl not creating filehandles in that |
|
|
237 | case - too bad it gives you a bogus file handle (reported |
|
|
238 | and analysed by Vladimir Timofeev). |
|
|
239 | - fix queue management logic in AnyEvent::Handle: |
|
|
240 | when on_read was registered, the queue was empty |
|
|
241 | and no progress could be made AnyEvent::Handle would |
|
|
242 | enter an endless loop. |
|
|
243 | - correctly start reading again when the handle became |
|
|
244 | busy again after idling. |
|
|
245 | - correctly treat tls shut-downs as EOF condition |
|
|
246 | (for the time being, would be nice to have a callback for that). |
|
|
247 | - correctly call eof callback under all conditions |
|
|
248 | (hopefully). |
|
|
249 | - the timeout callback did not expect that $self can go |
|
|
250 | away any time. |
|
|
251 | |
|
|
252 | 4.12 Tue Jun 3 10:58:04 CEST 2008 |
|
|
253 | - include AnyEvent::Intro, a tutorial for anyevent, |
|
|
254 | anyevent::socket and anyevent::handle. |
|
|
255 | - allow more options in on_error. |
|
|
256 | |
|
|
257 | 4.11 Fri May 30 23:42:25 CEST 2008 |
|
|
258 | - INCOMPATIBLE CHANGE: replace ptr by real PTR lookup, provide |
|
|
259 | reverse_lookup and reverse_verify to replace it, support |
|
|
260 | v4mapped and v4compat addresses. |
|
|
261 | - provide more documentation for the resolver class. |
|
|
262 | - really replace longest run of :0: by :: in format_address, |
|
|
263 | also properly convert :: and ::1 again. |
|
|
264 | - support NAPTR record name and decode it. |
|
|
265 | - implement random weight sampling for SRV records, as per |
|
|
266 | rfc 2782. |
|
|
267 | - correctly abort on srv-record targets of ".". |
|
|
268 | - added AnyEvent::DNS::wait_for_slot. |
|
|
269 | - in the unlikely event of a virtual circuit connection |
|
|
270 | being invalidated by a delayed udp reply, AnyEvent::DNS |
|
|
271 | could die. this has been fixed. |
|
|
272 | - plug a probable memleak in the DNS vc code. |
|
|
273 | - use configured timeout also for VC request phase. |
|
|
274 | - implement timeout and max_outstanding methods |
|
|
275 | for resolver class. |
|
|
276 | - update version numbers in all modules. |
|
|
277 | |
|
|
278 | 4.1 Thu May 29 05:45:40 CEST 2008 |
|
|
279 | - INCOMPATIBLE CHANGE: renamed xxx_ip functions to xxx_address |
|
|
280 | which mirrors their purpose better (old names still available |
|
|
281 | for a while). Also moved AnyEvent::DNS::addr to |
|
|
282 | AnyEvent::Socket::resolve_sockaddr. |
|
|
283 | - implement AnyEvent->time and AnyEvent->now. |
|
|
284 | - fix IPv6 support in pack_sockaddr. |
|
|
285 | - officially un-experimentalise all newly introduced code. |
|
|
286 | - support unix domain sockets everywhere by specifying |
|
|
287 | a host of "unix/" and the pathname as service. |
|
|
288 | - implement an activity timeout in AnyEvent::Handle. |
|
|
289 | - added a regex read type to AnyEvent::Handle. |
|
|
290 | - added a json read/write type to AnyEvent::Handle. |
|
|
291 | - always croak in on_error in AnyEvent::Handle. |
|
|
292 | - document how to subclass AnyEvent::Handle. |
|
|
293 | - implement AnyEvent::Util::fork_call. |
|
|
294 | - add support for IPv6 nameservers and nameserver statements. |
|
|
295 | - work around _yet_ _another_ windows perl bug in where empty |
|
|
296 | select masks cause errors in the pure perl backend (this |
|
|
297 | workaround cannot be implemented for the other event loops, |
|
|
298 | but EV also works around this). |
|
|
299 | - supply AnyEvent::Socket::address_type and make good use of it. |
|
|
300 | - clarify and fix the Handle documentation w.r.t $self vs. $handle |
|
|
301 | vs. callback arguments. |
|
|
302 | - add some recommends to the META.yml. |
|
|
303 | |
|
|
304 | 4.05 Mon May 26 19:44:06 CEST 2008 |
|
|
305 | - some platforms ignore the rfc and prepend an extra sa_len member to |
|
|
306 | the sockaddr structure, cater for those. also use sockaddr_family, |
|
|
307 | didn't know it was there, but comes in handy for the workaround. |
|
|
308 | - undo the import WIN32 hack. |
|
|
309 | |
|
|
310 | 4.04 Mon May 26 08:03:31 CEST 2008 |
|
|
311 | - try to work around yet another windows bug: failed connects are reported |
|
|
312 | as if it were out-of-band data. windows users: you suck. Only |
|
|
313 | EV, Glib, Event and the pure perl backend can handle this condition. |
|
|
314 | - optimize write algorithm in AnyEvent::Handle. |
|
|
315 | - properly parse PERL_ANYEVENT_PROTOCOLS and use the info correctly. |
|
|
316 | - tcp_connect now properly iterates through all targets. |
|
|
317 | - check wether IPv6 sockets can actually be created, otherwise, disable ipv6 |
|
|
318 | support entirely. also hardcode some AF_INET6 constants for the |
|
|
319 | sake of perl 5.8. |
|
|
320 | - vastly improved nameserver/suffix detection algorithm for windows |
|
|
321 | (still a hack...) |
|
|
322 | - try to use a fallback dns server if no dns servers could be found. |
|
|
323 | - splatter around some local $SIG{__DIE__}'s to work around bad |
|
|
324 | die handlers. |
|
|
325 | |
|
|
326 | 4.03 Sun May 25 05:05:57 CEST 2008 |
|
|
327 | - work around perl bugs on windows, where perl returns undocumented |
|
|
328 | error codes for sysread, syswrite etc. |
|
|
329 | - fix AnyEvent::Handle::unshift_read to really unshift and not push. |
|
|
330 | - AnyEvent::Handle could sometimes stumble over 0-byte writes. |
|
|
331 | - fix unpack_sockaddr, now works for AF_INET6 sockaddr's. |
|
|
332 | - updated documentation. |
|
|
333 | - lots of workarounds for perl 5.8 quirks. |
|
|
334 | - implement a more flexible type system for both reads and writes |
|
|
335 | in AnyEvent::Handle. |
|
|
336 | - support IPv6 in tcp_server. |
|
|
337 | - pass local hostname and port to the prepare callback in tcp_server. |
|
|
338 | - make types extendable in AnyEvent::Handle. |
|
|
339 | - croak, not die, in AnyEvent::Handle; |
|
|
340 | |
|
|
341 | 4.0 Sat May 24 19:58:08 CEST 2008 |
|
|
342 | - added (experimental) AnyEvent::DNS module. |
|
|
343 | - added (experimental) AnyEvent::Socket module. |
|
|
344 | - added (experiemntal) support for TLS/SSL in AnyEvent::Handle. |
|
|
345 | - make condvar->end call ->send as documented. |
|
|
346 | - make condvar constructor support the documented cb => argument. |
|
|
347 | - overload condvars so you can use them instead of code references. |
|
|
348 | - support more of the POSIX forms instead of a pure dotted quad for |
|
|
349 | verbatim ip addresses. |
|
|
350 | - AnyEvent::Socket::inet_aton will use AnyEvent::DNS now and not fork anymore. |
|
|
351 | - try very hard not to freeze on broken windows perls. |
|
|
352 | - make on_eof optional in AnyEvent::Handle. |
|
|
353 | - support service names in tcp_connect and tcp_server. |
|
|
354 | - work around netbsd/openbsd bug causing unix sockets not to |
|
|
355 | become writable until empty. |
|
|
356 | - only make a dummy read for error detection when getpeername |
|
|
357 | returns ENOTCONN. |
|
|
358 | - defend AnyEvent::Util::inet_aton against Coro::LWP's brutal |
|
|
359 | override. |
|
|
360 | - new AnyEvent::Guard::cancel. |
|
|
361 | |
|
|
362 | 3.5 Sat May 17 23:17:14 CEST 2008 |
|
|
363 | - gracefully deal with systems not defining CLOCK_MONOTONIC |
|
|
364 | (as opposed to not supporting the clock itself, which worked in 3.41). |
|
|
365 | - added AnyEvent::Util::tcp_connect and AnyEvent::Util::tcp_server |
|
|
366 | helpers, to replace the ill-designed AnyEvent::Socket class. |
|
|
367 | - added some examples (eg/connect, eg/handle). |
|
|
368 | - added AnyEvent::Util::guard. |
|
|
369 | - AnyEvent::Util now exports most of its functions. |
|
|
370 | - fixed a bug in *_read_chunk, where the callback was called with $self as |
|
|
371 | first and second argument. |
|
|
372 | - simplified and fixed regex handling in AnyEvent::Handle::*_read_line. |
|
|
373 | - fix manpage condvars examples to use new syntax. |
|
|
374 | - considerably improve inet_aton with EV::ADNS backend to support |
|
|
375 | cname chains (common dns configuration bug, e.g. with www.google.de) |
|
|
376 | and aaaa records if no a record could be found. |
|
|
377 | |
|
|
378 | 3.41 Sun May 11 19:53:13 CEST 2008 |
|
|
379 | - work around a bug in perl 5.8.8's local. Reported by Yi Ma Mao. |
|
|
380 | |
|
|
381 | 3.4 Sun May 11 00:29:25 CEST 2008 |
|
|
382 | - complete re-implementation of condvars with a hopefully much more useful |
|
|
383 | API (backward compatible functions exist but are undocumented). |
|
|
384 | - AnyEvent::Handle fully rewritten with a hopefully more powerful API |
|
|
385 | (still subject to change). |
|
|
386 | - take advantage of CLOCK_MONOTONIC, if available, in AnyEvent::Impl::Perl. |
|
|
387 | - provide AnyEvent::post_detect and @AnyEvent::post_detect, which |
|
|
388 | allows module authors to avoid forcing event loop detection. |
|
|
389 | (used by Coro::AIO, Coro::BDB, Coro::AnyEvent for example). |
|
|
390 | - remove coro backends: Coro now provides generic support for AnyEvent |
|
|
391 | via Coro::AnyEvent. |
|
|
392 | |
|
|
393 | 3.3 Mon Apr 28 09:51:06 CEST 2008 |
|
|
394 | - added AnyEvent::Handle, AnyEvent::Socket and AnyEvent::Util modules. |
|
|
395 | - fix a bug in the pure perl backend that kept watchers alive |
|
|
396 | when multiple watchers were registered for the same fh/poll combo. |
|
|
397 | - add a benchmark section showing AnyEvent overhead and comparing |
|
|
398 | the different event loops with each other. |
|
|
399 | - prefer pure perl over tk when autoprobing, as it's about as fast, |
|
|
400 | but doesn't crash with many watchers. |
|
|
401 | - declare Qt support non-experimental. |
|
|
402 | - clarify the confusing section about the file descriptor being |
|
|
403 | kept alive. |
|
|
404 | - document the race between loading of an event module and |
|
|
405 | child processes exiting. |
|
|
406 | - support POE as "backend" (with some caveats, POE is not generic |
|
|
407 | enough, and darn slow). |
|
|
408 | - support Wx and Prima through POE. |
|
|
409 | - optimise perl backend to use 20% less memory and take advantage |
|
|
410 | of typical timeout behaviour. It can now compete with |
|
|
411 | select/poll-based C event loops in most cases (it is usually |
|
|
412 | faster than Event and Glib :). |
|
|
413 | - roughly cut EV memory use in half and increase its speed by 30%, |
|
|
414 | by removing undocumented functionality. Did something similar |
|
|
415 | to other event loops where possible. |
|
|
416 | |
|
|
417 | 3.2 Thu Apr 24 10:10:40 CEST 2008 |
|
|
418 | - do not die when anyevent watchers are destroyed while running callbacks |
|
|
419 | in the pureperl backend (could only happen when two watchers are registered |
|
|
420 | for the same fh/poll combo). |
|
|
421 | - support autoloading for child watchers, was broken in all versions. |
|
|
422 | - implement PERL_ANYEVENT_MODEL env variable. |
|
|
423 | - (experimental) implement interface to the (very crashy/buggy) Event::Lib module. |
|
|
424 | - (experimental) implement interface to the Qt module (cannot be autoprobed). |
|
|
425 | - this release is IO:AnyEvent-proof. |
|
|
426 | |
|
|
427 | 3.12 Tue Apr 22 07:11:46 CEST 2008 |
|
|
428 | - reinstate AUTHOR section that got lost somehow. |
|
|
429 | - do not hang in the testsuite with badly broken perls (activestate, |
|
|
430 | strawberry...), but instead diagnose the problem and continue. |
|
|
431 | - use INT instead of CHLD in an attempt to support broken windows perls |
|
|
432 | better (this decreases the test quality, unfortunately). |
|
|
433 | - do not send the signal to the process group |
|
|
434 | (no problem for CHLD, bad for INT :). |
|
|
435 | |
|
|
436 | 3.11 Sat Apr 19 06:57:31 CEST 2008 |
|
|
437 | - major documentation rework. |
|
|
438 | - document the fact that child watchers only watch for zombies. |
|
|
439 | - fix the child watcher example. |
|
|
440 | |
|
|
441 | 3.1 Wed Apr 16 17:09:01 CEST 2008 |
|
|
442 | - work around recurring bugs in Tk by dup'ing filehandles, the |
|
|
443 | only method with good success chances on Tk (the bugs apparently |
|
|
444 | don't get fixed anytime soon). |
|
|
445 | - lift the restriction of only one watcher per fh direction |
|
|
446 | (as the Tk bug workaround also lifts it and only Tk imposed |
|
|
447 | such strong limits). |
|
|
448 | - changed probe order to prefer coro adaptors. |
|
|
449 | - explain why recursion into the event loop is not supported |
|
|
450 | unless the backend supports it (only Coro::EV does without |
|
|
451 | any restrictions...). |
|
|
452 | - add simple manpages for all backend modules. |
|
|
453 | |
|
|
454 | 3.0 Mon Apr 7 21:30:23 CEST 2008 |
|
|
455 | - Coro::Signal changed semantics, roll our own, also cleaning |
|
|
456 | up the Coro implementation in general. |
|
|
457 | - rename Coro backend to CoroEvent. |
|
|
458 | - add some decision helping paragraph to the manpage that should |
|
|
459 | help people to decide whether AnyEvent is the right thing for them. |
|
|
460 | |
|
|
461 | 2.9 Mon Jan 28 13:31:54 CET 2008 |
|
|
462 | - update for EV 3.0 API changes. |
|
|
463 | |
|
|
464 | 2.8 Sun Nov 25 15:06:03 CET 2007 |
|
|
465 | - waitpid can validly return 0. accept this fact of life |
|
|
466 | instead of reporting it to any watchers. |
|
|
467 | |
|
|
468 | 2.7 Fri Nov 23 11:41:14 CET 2007 |
|
|
469 | - force use of AnyEvent::Impl::Perl in testsuite, there is |
|
|
470 | too much breakage outside AnyEvent. |
|
|
471 | - deliver signals synchronously in AnyEvent::Impl::Perl |
|
|
472 | or any other event loop that relies on AnyEvents child |
|
|
473 | watcher emulation. *Could* help with hanging testsuite |
|
|
474 | (except when Event or EV are installed). |
|
|
475 | |
|
|
476 | 2.6 Fri Nov 9 20:36:35 CET 2007 |
|
|
477 | - fix bug in testsuite. |
|
|
478 | - move EV adaptor modules to AnyEvent. |
|
|
479 | - add Coro+EV adaptor module. |
|
|
480 | |
|
|
481 | 2.55 Tue Nov 6 17:41:32 CET 2007 |
|
|
482 | - add EV to the list of supported event models. |
|
|
483 | - do not auto-reset pid watchers, pass pid and status to them. |
|
|
484 | - allow a pid of zero to watch for all children in child watchers. |
|
|
485 | |
|
|
486 | 2.54 Wed Jul 18 17:36:23 CEST 2007 |
|
|
487 | - work around a perl bug that results in |
|
|
488 | BEGIN not safe after errors--compilation aborted without |
|
|
489 | any discernible reason or error message by once |
|
|
490 | again not use'ing strict. |
|
|
491 | |
|
|
492 | 2.53 Sun Jul 8 10:51:53 CEST 2007 |
|
|
493 | - make Glib interface work again (spotted by elmex). |
|
|
494 | |
|
|
495 | 2.52 Wed Mar 7 18:36:16 CET 2007 |
|
|
496 | - child watchers stopped working permanently when |
|
|
497 | all current child watchers were destroyed. |
|
|
498 | |
|
|
499 | 2.51 Mon Dec 11 21:33:24 CET 2006 |
|
|
500 | - work around bugs in perl where eval "require Module" returns |
|
|
501 | true even if the module couldn't be loaded :/. |
|
|
502 | |
|
|
503 | 2.5 Mon Dec 11 02:15:28 CET 2006 |
|
|
504 | - avoid $AUTOLOAD because many perls corrupt it. |
|
|
505 | - AnyEvent::detect forces autodetection. |
|
|
506 | - implement signal watchers (experimental) |
|
|
507 | (Tk does not support async signals (<= 804.027 at least), |
|
|
508 | so they might get delayed indefinitely on Tk). |
|
|
509 | - implement child watchers (experimental). |
|
|
510 | - moved default condvar implementation into base module, simplifying |
|
|
511 | most implementation modules. |
|
|
512 | |
|
|
513 | 2.1 Fri Nov 24 15:50:48 CET 2006 |
|
|
514 | - better docs. |
|
|
515 | - simple testsuite. |
|
|
516 | - added AnyEvent->one_event method for special purposes. |
|
|
517 | |
|
|
518 | 2.0 Wed Nov 1 02:21:30 CET 2006 |
|
|
519 | - INCOMPATIBLE CHANGE: poll can now either be "r" or "w" but not both. |
|
|
520 | - INCOMPATIBLE CHANGE: io watcher callbacks have no arguments anymore. |
|
|
521 | - EXPERIMENTAL: |
|
|
522 | new pure-perl model added, AnyEvent now always finds a suitable |
|
|
523 | event model. |
10 | - improved documentation. |
524 | - improved documentation. |
11 | - reduced cpu overhead. |
525 | - reduced cpu overhead. |
12 | |
526 | |
13 | 1.02 Fri Jan 13 14:15:40 CET 2006 |
527 | 1.02 Fri Jan 13 14:15:40 CET 2006 |
14 | - add COPYING to clarify license. |
528 | - add COPYING to clarify license. |
… | |
… | |
36 | - fixed documentation. |
550 | - fixed documentation. |
37 | - clarified some corner cases. |
551 | - clarified some corner cases. |
38 | - vastly improved documentation with a long example. |
552 | - vastly improved documentation with a long example. |
39 | |
553 | |
40 | 0.2 Thu Dec 1 22:18:49 CET 2005 |
554 | 0.2 Thu Dec 1 22:18:49 CET 2005 |
41 | - check wether a valid method was called, to avoid endless loops. |
555 | - check whether a valid method was called, to avoid endless loops. |
42 | - move condvars into their own class (cleaner, safer). |
556 | - move condvars into their own class (cleaner, safer). |
43 | - don't die when some modules are not available. |
557 | - don't die when some modules are not available. |
44 | - Tk is too broken to allow for words. Will work once Tk works. |
558 | - Tk is too broken to allow for words. Will work once Tk works. |
45 | |
559 | |
46 | 0.1 Thu Dec 1 19:44:57 CET 2005 |
560 | 0.1 Thu Dec 1 19:44:57 CET 2005 |