| 1 | Revision history for Perl extension AnyEvent. |
1 | Revision history for Perl extension AnyEvent. |
|
|
2 | |
|
|
3 | TODO: document TCP_*** constants |
|
|
4 | TODO: catch AE::xxx for anyevent strict and anyevent debug |
|
|
5 | TODO: anyevent::trace? or debug tracing all watchers? |
|
|
6 | TODO: anyevent strict checking for $_ changes? |
|
|
7 | |
|
|
8 | TODO: redesign Impl |
|
|
9 | TODO: make ae::run mroe robust against eintr, find out which env still causes these issues Andrew Suffield |
|
|
10 | TODO: IO::Async sometimes cancels the wrong timer when you call cancel_timer. Invest extra overhead to work around this peculiar behaviour. |
|
|
11 | - INCOMPATIBLE CHANGE: AnyEvent::DNS resource records now include |
|
|
12 | the TTL as fourth element - this affects ->resolve directly, |
|
|
13 | and ->request and ->dns_unpack indirectly. Convenience functions |
|
|
14 | such as AnyEvent::DNS::{a,txt,mx,srv} are not affected. |
|
|
15 | - INCOMPATIBLE CHANGE: backend authors now should not implement |
|
|
16 | one_event or loop, but instead the AnyEvent::CondVar::_wait and _poll |
|
|
17 | methods. |
|
|
18 | - DEPRECIATION: please change |
|
|
19 | "use AnyEvent::Impl::Perl" to "use AnyEvent::Loop" and |
|
|
20 | "AnyEvent::Impl::Perl::loop" to "AnyEvent::Loop::run" |
|
|
21 | in your programs, for future compatibility. |
|
|
22 | - querying too long domain names would barf deep inside AnyEvent::DNS - since |
|
|
23 | this is hard to avoid for applications, AnyEvent:.DNS::request now checks for |
|
|
24 | for too long domain names explicitly and passes an undef to the callback |
|
|
25 | instead (reported by James Bromberger). |
|
|
26 | - new function: AnyEvent::postpone. |
|
|
27 | - IO::Async sometimes cancels the wrong timer when you call cancel_timer. |
|
|
28 | Invest extra overhead to work around this peculiar behaviour. |
|
|
29 | - split the pure perl backend into a loop module and an impl |
|
|
30 | module (AnyEvent::Loop and AnyEvent::Impl::Perl). |
|
|
31 | - some read types (regex, netstring, json, storable) could |
|
|
32 | errornously keep an AE::Handle object alive on their own |
|
|
33 | (reported by Mohammad Toossi). |
|
|
34 | - add more workarounds around child watcher bugs in IO::Async |
|
|
35 | and (newly introduced) in POE. |
|
|
36 | - use glib child watchers and try to work around its limitations, |
|
|
37 | if possible (based on a patch by Kevin Ryde). |
|
|
38 | - add support for (currently third-party) AnyEvent::Impl::FLTK. |
|
|
39 | - many read types in AnyEvent::Handle would malfunction during |
|
|
40 | transitions to TLS because of a "delete $self->{rbuf}". |
|
|
41 | (reported by Gerald Galster). |
|
|
42 | - resolve AnyEvent::Util::guard at first-use-time, not at load time, |
|
|
43 | for further speed and memory savings when not used. |
|
|
44 | - improve condvar blocking wait performance (probably only noticable |
|
|
45 | with faster backends), also make it easier for event loops that |
|
|
46 | do not support blocking waits. |
|
|
47 | - PERL_ANYEVENT_MODEL now also allows full module names, although |
|
|
48 | the usefulness of this feature is not yet clear. |
|
|
49 | |
|
|
50 | 5.34 Thu May 12 09:59:41 CEST 2011 |
|
|
51 | - AE::TLS didn't support both cert_file and key at the same time, |
|
|
52 | due to some copy&paste bug (reported and analyzed by vti). |
|
|
53 | - improve AE::Handle timeout documentation slightly. |
|
|
54 | |
|
|
55 | 5.33 Mon Feb 7 21:51:41 CET 2011 |
|
|
56 | - never appeared on CPAN, for unknown reasons. |
|
|
57 | - AnyEvent::Handle on_connect_error - do not try to destroy a handle |
|
|
58 | object when it's already gone (analyzed by Ryan Bullock). |
|
|
59 | - update idna algorithm(s) to UTS#46 version 6.0.0. |
|
|
60 | - preserve trailing dot in idn_to_ascii - testsuite says to preserve, |
|
|
61 | uts46 says it's an error, and testsuite makes more sense. |
|
|
62 | - remove leading dots in idn_to_ascii - testsuite says to remove, |
|
|
63 | uts46 says it's an error, and testsuite makes more sense. |
|
|
64 | - fixed a bug in idn ignored character processing. |
|
|
65 | - passes the uts46 6.0.0 testsuite except for three cases which |
|
|
66 | apparently are bugs in the testsuite itself. |
|
|
67 | - passes the uts46 6.0.1 testsuite except for eleven cases which |
|
|
68 | apparently are bugs in the testsuite itself. |
|
|
69 | |
|
|
70 | 5.31 Sun Jan 23 11:49:19 CET 2011 |
|
|
71 | - work around a POE bug where POE would first poll for new events and |
|
|
72 | invoke handlers BEFORE handling the message queue. |
|
|
73 | - fix parsing of "domain" statement in resolv.conf (Johannes Plunien). |
|
|
74 | - stop_read now does just that in TLS mode. see its |
|
|
75 | documentation for caveats. |
|
|
76 | - update to common::sense 3.4 warning set (some warnings in perl impose |
|
|
77 | an arbitrary 32kb limit on regex matches, which affects |
|
|
78 | AnyEvent::Handle). |
|
|
79 | - AE::Handle had a broken rbuf_max method and missing wbuf_max method. |
|
|
80 | - new experimental AE::Handle::resettls method. |
|
|
81 | |
|
|
82 | 5.3 Fri Dec 31 05:48:13 CET 2010 |
|
|
83 | - major IOAsync patch by Paul Evans - 0.33 is required, as the major |
|
|
84 | issues with IO::Async have been solved in that version, including |
|
|
85 | the availabilty of a default loop! |
|
|
86 | - implement new wbuf_max parameter for AnyEvent::Handle. |
|
|
87 | - added Cocoa::EventLoop backend by Daisuke Murase. |
|
|
88 | - the default testsuite now honors PERL_ANYEVENT_MODEL (Paul Evans). |
|
|
89 | - clarify handle on_prepare documentation. |
|
|
90 | - set CLOEXEC flag on parent-side fd's in AnyEvent::Util::run_cmd, |
|
|
91 | to avoid leaking them to any child processes. |
|
|
92 | - try to force files into the arch-dependend lib directory, for the |
|
|
93 | half-"turly obsessive compulsive person"(s) that depend on it. |
|
|
94 | |
|
|
95 | 5.29 Sun Dec 5 10:49:21 CET 2010 |
|
|
96 | - convert EV backend to EV 4.00 API (so better upgrade EV too). |
|
|
97 | - AE::Handle: implement dynamic read_size adjustment, add |
|
|
98 | max_read_size parameter, reduce default read_size to 2048. |
|
|
99 | - add some FAQ entries for servers and/or on_eof vs. on_error. |
|
|
100 | - work around OS bugs (cygwin again) when lingering in AE::Handle. |
|
|
101 | - reduce memory usage slightly when connect was used in AE::Handle. |
|
|
102 | - wrok around more segfaults in Tk. |
|
|
103 | - document tls_ctx => undef as valid. |
|
|
104 | - detect WNOHANG value at compiletime. |
|
|
105 | - include some probably linux-only support for building AnyEvent |
|
|
106 | as part of the perl core. |
|
|
107 | - improve compatibility of t/08_idna.t to perl 5.8. |
|
|
108 | - make AnyEvent::Strict truly optional (in case somebody builds a |
|
|
109 | minimal perl...). |
|
|
110 | |
|
|
111 | 5.28 Wed Oct 13 04:14:23 CEST 2010 |
|
|
112 | - due to a glitch, AnyEvent's internal getprotobyname wasn't |
|
|
113 | actually used in tcp_server. |
|
|
114 | - implement AnyEvent::CondVar->new method, which was documented but |
|
|
115 | missing. |
|
|
116 | - support autoloading of read types in unshift_read. |
|
|
117 | - AnyEvent::Strict: do the test for invalid fh arguments differently, |
|
|
118 | resulting in more informative error messages. |
|
|
119 | - new AnyEvent::FAQ manpage. |
|
|
120 | - clarify that you can actually call push_read etc. in on_read. |
|
|
121 | - put keys and certs in /tmp, not into the current directory, |
|
|
122 | and document this in the SECURITY CONSIDERATIONS sections |
|
|
123 | of AnyEvent::TLS. |
|
|
124 | - support ";" as resolv.conf comment character in AnyEvent::DNS, as |
|
|
125 | requested by Ron Isaacson. |
|
|
126 | - document $AnyEvent::DNS::RESOLVER variable. |
|
|
127 | - incorporated major typo patches by Abhijit Menon-Sen. |
|
|
128 | - AnyEvent::Handle now croaks when negative timeouts are passed. |
|
|
129 | - add a shitload of TCP_xxx sockopt name constants to AnyEvent::Util, |
|
|
130 | undocumented, but who knows when they might come in handy. |
|
|
131 | |
|
|
132 | 5.271 Tue Jun 8 12:05:46 CEST 2010 |
|
|
133 | - backport to perl 5.8.x. |
|
|
134 | |
|
|
135 | 5.27 Sun Jun 6 12:12:05 CEST 2010 |
|
|
136 | - postpone differently in AnyEvent::Socket now, as |
|
|
137 | when not, canceling the connection attempt might fail |
|
|
138 | (found by Felix Antonius Wilhelm Ostmann). |
|
|
139 | - explicitly check for non-stream sockets in AE::Handle, too many |
|
|
140 | clueless people fell into the trap of this somehow working. |
|
|
141 | - simplified and reworked the "OTHER MODULES" section. |
|
|
142 | - better/more condvar examples. |
|
|
143 | |
|
|
144 | 5.261 Wed Apr 28 16:13:36 CEST 2010 |
|
|
145 | - AF_INET6 was not properly used from Socket6 during configuration |
|
|
146 | time (found by Andy Grundman). |
|
|
147 | |
|
|
148 | 5.26 Mon Apr 12 04:49:35 CEST 2010 |
|
|
149 | - don't generate spurious readyness notifications when select |
|
|
150 | returns EINTR in the pure perl backend (analysed and testcase |
|
|
151 | by Bas Denissen). |
|
|
152 | - give same examples in SYNOPSIS sections of AnyEvent and AE. |
|
|
153 | - provider faster implementations for AE::time/now/now_update for |
|
|
154 | EV and Perl backends. |
|
|
155 | - provide faster AE::cv and AE::time implementations for all |
|
|
156 | backends. |
|
|
157 | - no longer support lower/mixed-case signal names (this was |
|
|
158 | never documented nor universally supported). |
|
|
159 | - some more memory ยต-optimisations, and somewhat less messy |
|
|
160 | code to cope with both AnyEvent and AE APIs. |
|
|
161 | |
|
|
162 | 5.251 Sat Mar 13 00:58:21 CET 2010 |
|
|
163 | - make test failed if it was run before make install on perl <= |
|
|
164 | 5.8.8, and older versions of perl make our live very difficult, |
|
|
165 | so write constants.pl during Makefile.PL time. |
|
|
166 | |
|
|
167 | 5.25 Sat Mar 13 00:23:14 CET 2010 |
|
|
168 | - fix a race condition in AnyEvent::Handle that would cause |
|
|
169 | a "bio_read: unsupported method" error (in _another_ |
|
|
170 | ssl connection) after user code threw an exception. |
|
|
171 | - added AnyEvent::Handle->destroyed method. |
|
|
172 | - speed up AnyEvent::Socket::format_address and ::format_ipv6. |
|
|
173 | - the AnyEvent::Util::fh_nonblocking function worked only |
|
|
174 | by lucky accident on win32. |
|
|
175 | - smaller and faster AnyEvent::Util::fh_nonblocking. |
|
|
176 | - when the (required!) Time::HiRes module is missing, AnyEvent |
|
|
177 | did not fall back to built-in time properly. |
|
|
178 | - do not load Fcntl at runtime, saving memory and loading time. |
|
|
179 | - precompile a number of constants and use them instead of runtime |
|
|
180 | detection and eval. |
|
|
181 | - free detection code after detection and similar memory optimisations. |
|
|
182 | - Perl backend timer interval best effort drift has been improved |
|
|
183 | (same algorithm as EV). |
|
|
184 | - update unicode idna mapping table. |
|
|
185 | |
|
|
186 | 5.24 Tue Jan 5 11:39:43 CET 2010 |
|
|
187 | - cygwin never reports errors from failed connects. 1.5 just gives |
|
|
188 | you continous readyness and EAGAIN, 1.7 is even more broken and |
|
|
189 | just hangs. work around both issues in a cygwin-specific hack. |
|
|
190 | - improve idle watcher documentation slightly. |
|
|
191 | |
|
|
192 | 5.23 Sun Dec 20 23:48:00 CET 2009 |
|
|
193 | - support IDNs in resolve_sockaddr, and therefore in tcp_connect. |
|
|
194 | - implement punycode_encode/decode, idn_nameprep, |
|
|
195 | idn_to_ascii and idn_to_unicode operations in AnyEvent::Util. |
|
|
196 | - provide $AE::VERSION. |
|
|
197 | - removed traces of "no strict 'refs'". |
|
|
198 | |
|
|
199 | 5.22 Sat Dec 5 03:51:13 CET 2009 |
|
|
200 | - downgrade-or-fail in AnyEvent::Handle::push_write, to |
|
|
201 | diagnose encoding failures earlier and more succinctly. |
|
|
202 | (this works around bugs in perl, throwing away encoding info |
|
|
203 | when passing scalar data to extensions). |
|
|
204 | - add more examples to AnyEvent::Socket manpage. |
|
|
205 | - upgrade internal warning set to the same as common::sense 2.03. |
|
|
206 | - use pack "n/a*" for pre-5.8.9 perl compatibility in AnyEvent::DNS |
|
|
207 | (John Beppu). |
|
|
208 | - AnyEvent::Socket::inet_aton now properly supports ipv6, as documented. |
|
|
209 | - add google public dns servers to fallback server set and make sure |
|
|
210 | we load-balance properly between the three sets. also add all |
|
|
211 | fallback dns servers, not just a random one, to each dns config. |
|
|
212 | |
|
|
213 | 5.21 Thu Nov 19 02:48:47 CET 2009 |
|
|
214 | - fix a problem where socket constants were called with parameters |
|
|
215 | (spotted by David Friedland). |
|
|
216 | - fork_call never use'd POSIX (reported by Daisuke Maki). |
|
|
217 | - improve perl 5.6 compatibility further (but it still won't work |
|
|
218 | unless you rip out everything but the core). |
|
|
219 | - prefer Net::DNS::Resolver over ipconfig, if installed, on win32. |
|
|
220 | uses 10MB of RAM, but doesn't open a console window. *sigh*. |
|
|
221 | |
|
|
222 | 5.202 Wed Oct 14 22:35:44 CEST 2009 |
|
|
223 | - AnyEvent::DNS would unexpectedly clobber $_ under windows |
|
|
224 | (analysed by Matthias Waldorf). |
|
|
225 | - AnyEvent::Handle::run_cmd can now pass the PID of the |
|
|
226 | newly-created process, which is much less useful than it might |
|
|
227 | sound (based on patch by Yann Kerherve). |
|
|
228 | |
|
|
229 | 5.201 Tue Sep 29 12:09:25 CEST 2009 |
|
|
230 | - AnyEvent:Handle::on_starttls/on_stoptls methods were broken |
|
|
231 | (reported by Torsten Foertsch). |
|
|
232 | - common::sense 2.0 could cause tcp_server to throw an exception |
|
|
233 | (analysed by elmex). |
|
|
234 | |
|
|
235 | 5.2 Mon Sep 14 07:04:49 CEST 2009 |
|
|
236 | - INCOMPATIBLE CHANGE: do no longer support register_read_type |
|
|
237 | and register_write_type in AnyEvent::Handle, instead support |
|
|
238 | package names (the facility was mostly abused). |
|
|
239 | - implement "packagename-as-read/write type" support in |
|
|
240 | AnyEvent::Handle. |
|
|
241 | - AnyEvent::Handle: new options "keepalive" and "oobinline". |
|
|
242 | - oobinline set by default to avoid security issues. |
|
|
243 | - the pure-perl event loop backend wrongly detected times() fallback |
|
|
244 | support (spotted by Pavel Boldin). |
|
|
245 | |
|
|
246 | 5.12 Tue Sep 1 20:26:50 CEST 2009 |
|
|
247 | - be more lenient when parsing resolv.conf files, as some people |
|
|
248 | use hashmarks as comment indicator inside directives (reported |
|
|
249 | by Michael S. Fischer). |
|
|
250 | - use same set of warnings as common::sense 2.0. |
|
|
251 | - fix a potential 32 bit overflow issue due to perl having problems |
|
|
252 | with large hex constants in the Perl backend. |
|
|
253 | - do not use the slower 5.6 version fo the enc_name code in AnyEvent::DNS |
|
|
254 | by default. |
|
|
255 | - fix some prototypes. |
|
|
256 | - (optionally) require Async::Interrupt 1.02. |
|
|
257 | - replace opendns fallbacks by verizon/level3 public dns servers. |
|
|
258 | |
|
|
259 | 5.112 Fri Aug 21 13:59:04 CEST 2009 |
|
|
260 | - AnyEvent::Handle could cause hard crashes in Net::SSLeay when |
|
|
261 | using starttls with an illegal/uncreatable tls context. |
|
|
262 | - port to broken win32 perls. |
|
|
263 | |
|
|
264 | 5.111 Sun Aug 16 18:44:36 CEST 2009 |
|
|
265 | - perl errornously requires sockaddr_un structures returned by the |
|
|
266 | kernel to have a specific length. work around this bug by padding |
|
|
267 | sockaddr_un structures (found to fail on OS X, as perl only uses |
|
|
268 | the correct code for GNU/Linux, but likely to fail on other |
|
|
269 | OSes as well). |
| 2 | |
270 | |
| 3 | 5.11 Wed Aug 12 17:49:37 CEST 2009 |
271 | 5.11 Wed Aug 12 17:49:37 CEST 2009 |
| 4 | - fixed a bug in run_cmd with input from filename. |
272 | - fixed a bug in run_cmd with input from filename. |
| 5 | - create a json coder object when none is given in write direction |
273 | - create a json coder object when none is given in write direction |
| 6 | to reduce dependency on JSON. |
274 | to reduce dependency on JSON. |
| … | |
… | |
| 65 | - new function: AnyEvent::Socket::format_hostport. |
333 | - new function: AnyEvent::Socket::format_hostport. |
| 66 | |
334 | |
| 67 | 4.881 Tue Jul 28 12:51:53 CEST 2009 |
335 | 4.881 Tue Jul 28 12:51:53 CEST 2009 |
| 68 | - work around a bug in local in pre-5.10 perls, causing |
336 | - work around a bug in local in pre-5.10 perls, causing |
| 69 | AnyEvent::Handle to recurse when it should not |
337 | AnyEvent::Handle to recurse when it should not |
| 70 | (analyzed by elmex). |
338 | (analysed by elmex). |
| 71 | |
339 | |
| 72 | 4.88 Tue Jul 28 04:04:37 CEST 2009 |
340 | 4.88 Tue Jul 28 04:04:37 CEST 2009 |
| 73 | - re-bless the handle into a dummy package after calling |
341 | - re-bless the handle into a dummy package after calling |
| 74 | AnyEvent::Handle::destroy, so the user does not need to check |
342 | AnyEvent::Handle::destroy, so the user does not need to check |
| 75 | for errors after every push_write etc. |
343 | for errors after every push_write etc. |
| … | |
… | |
| 297 | 4.3 Fri Oct 3 09:18:43 CEST 2008 |
565 | 4.3 Fri Oct 3 09:18:43 CEST 2008 |
| 298 | - AnyEvent will now install a no-op signal handler for SIGPIPE, |
566 | - AnyEvent will now install a no-op signal handler for SIGPIPE, |
| 299 | unless one has been installed already. |
567 | unless one has been installed already. |
| 300 | - warn about not loaidng AnyEvent::Impl::POE early enough |
568 | - warn about not loaidng AnyEvent::Impl::POE early enough |
| 301 | (patch by Adam Rosenstein). |
569 | (patch by Adam Rosenstein). |
| 302 | - fixed a great number of bugs and corner cases in AnyEvent::Handle: |
570 | - fixed a great number of bugs and corner cases in AnyEvent::Handle. |
| 303 | - fix a bug in where in SSL connect mode, the |
571 | - fix a bug in where in SSL connect mode, the |
| 304 | client would first wait for some data by the server and |
572 | client would first wait for some data by the server and |
| 305 | otherwise hang. (reported and analyzed in an absolutely |
573 | otherwise hang. (reported and analysed in an absolutely |
| 306 | exemplary manner by Adam Rosenstein). |
574 | exemplary manner by Adam Rosenstein). |
| 307 | - fix a bug in where SSL EOF would not be treated |
575 | - fix a bug in where SSL EOF would not be treated |
| 308 | as stream EOF, putting the connection into a hung state. |
576 | as stream EOF, putting the connection into a hung state. |
| 309 | (reported and analyzed in an absolutely exemplary manner by Adam |
577 | (reported and analysed in an absolutely exemplary manner by Adam |
| 310 | Rosenstein). |
578 | Rosenstein). |
| 311 | - fix a potential segfault when the TLS context |
579 | - fix a potential segfault when the TLS context |
| 312 | would go missing in a read callback (e.g. due to stoptls) - |
580 | would go missing in a read callback (e.g. due to stoptls) - |
| 313 | Net::SSLeay of course makes no type checking whatsoever. |
581 | Net::SSLeay of course makes no type checking whatsoever. |
| 314 | - AnyEvent::Handle will not stop the read watcher in TLS mode. |
582 | - AnyEvent::Handle will not stop the read watcher in TLS mode. |
