| 1 | Revision history for libev, a high-performance and full-featured event loop. |
1 | Revision history for libev, a high-performance and full-featured event loop. |
| 2 | |
2 | |
| 3 | TODO: ev_time_freeze, ev_time_thaw |
|
|
| 4 | TODO: ev_walk |
3 | TODO: ev_walk |
| 5 | TODO: ev_stop_all |
4 | TODO: remain |
| 6 | TODO: fix signal handling(?) under win32 |
5 | - ev_unloop and ev_loop wrongly used a global variable to exit loops, |
| 7 | 3.54 |
6 | instead of using a per-loop variable (bug caught by accident...). |
|
|
7 | - Denis F. Latypoff corrected many typos in example code snippets. |
|
|
8 | - the ev_set_io_collect_interval interpretation has changed. |
|
|
9 | - honor autoconf detection of EV_USE_CLOCK_SYSCALL, also double- |
|
|
10 | check that the syscall number is available before trying to |
|
|
11 | use it (reported by ry@tinyclouds). |
|
|
12 | - use GetSystemTimeAsFileTime instead of _timeb on windows, for |
|
|
13 | slightly higher accuracy. |
|
|
14 | - properly declare ev_loop_verify even when !EV_MULTIPLICITY. |
|
|
15 | - implement ev_loop_depth. |
|
|
16 | |
|
|
17 | 3.6 Tue Apr 28 02:49:30 CEST 2009 |
| 8 | - multiple timers becoming ready within an event loop iteration |
18 | - multiple timers becoming ready within an event loop iteration |
| 9 | will be invoked in the "correct" order now. |
19 | will be invoked in the "correct" order now. |
| 10 | - do not leave the event loop early just because we have no active |
20 | - do not leave the event loop early just because we have no active |
| 11 | watchers, fixing a problem when embedding a kqueue loop |
21 | watchers, fixing a problem when embedding a kqueue loop |
| 12 | that has active kernel events but no registered watchers |
22 | that has active kernel events but no registered watchers |
| 13 | (reported by blacksand blacksand). |
23 | (reported by blacksand blacksand). |
| 14 | - correctly zero the idx values for arrays, so destroying and |
24 | - correctly zero the idx values for arrays, so destroying and |
| 15 | reinitialising the default loop actually works (patch by |
25 | reinitialising the default loop actually works (patch by |
| 16 | Malek Hadj-Ali). |
26 | Malek Hadj-Ali). |
|
|
27 | - implement ev_suspend and ev_resume. |
| 17 | - new EV_CUSTOM revents flag for use by applications. |
28 | - new EV_CUSTOM revents flag for use by applications. |
|
|
29 | - add documentation section about priorites. |
|
|
30 | - add a glossary to the dcoumentation. |
|
|
31 | - extend the ev_fork description slightly. |
|
|
32 | - optimize a jump out of call_pending. |
| 18 | |
33 | |
| 19 | 3.53 Sun Feb 15 02:38:20 CET 2009 |
34 | 3.53 Sun Feb 15 02:38:20 CET 2009 |
| 20 | - fix a bug in event pipe creation on win32 that would cause a |
35 | - fix a bug in event pipe creation on win32 that would cause a |
| 21 | failed assertion on event loop creation (patch by Malek Hadj-Ali). |
36 | failed assertion on event loop creation (patch by Malek Hadj-Ali). |
| 22 | - probe for CLOCK_REALTIME support at runtime as well and fall |
37 | - probe for CLOCK_REALTIME support at runtime as well and fall |
