ViewVC Help
View File | Revision Log | Show Annotations | Download File
/cvs/Coro/Changes
(Generate patch)

Comparing Coro/Changes (file contents):
Revision 1.138 by root, Sun Nov 5 02:01:24 2006 UTC vs.
Revision 1.172 by root, Sun Dec 3 22:50:06 2006 UTC

1Revision history for Perl extension Coro. 1Revision history for Perl extension Coro.
2 2
3TODO: possibly convert some of the Event modules into AnyEvent users. 3TODO: transfer flags must be per state
4 4
5TODO: eg/lwp to use Coro::Select 53.01 Sun Dec 3 23:47:42 CET 2006
6TODO: Event/LWP.pm 6 - forgot to include Coro::Timer.
7 - should compile with perl 5.9.x now. 7
83.0 Sun Dec 3 22:57:25 CET 2006
9 - the "FINALLY COMPLETELY STABLE" release (coming soon:
10 the "FAMOUS LAST WORDS" release).
11 - implement a new stack sharing algorithm, which uses a stack
12 pool (size currently hardcoded).
13 - make stack sharing mandatory (it no longer uses a heuristic).
14 - eval/die no longer cause weird problems under heavy use.
15 - Coro::Event could cause livelocks if it was used but
16 no Coro::Event watchers were used.
17 - Coro::Event now uses asynccheck as prepare does not
18 check for changed watchers.
19 - Coro::Event allows multiple waiting coros and will wake up one
20 per event.
21 - Coro::Event should be cleaner and more efficient now.
22 - new utility function Coro::unblock_sub.
23 - document the sad fact that Event is no longer reentrant.
24 - putting a coroutine into the ready queue twice could under
25 some circumstances lead to stack corruption.
26 - minor incompatibility: subclassing Coro::State is supported
27 directly now without going through a _coro_state member.
28 - state/coro switching is much faster now.
29 - very minor optimisations and code/documentation cleanup.
30 - avoid problems due to compiler inlining.
31 - removed timers from Coro::Timer -> use AnyEvent instead.
32 - replaced Coro::idle coroutine by (cleaner) idle handler.
33 - updated to newest libcoro.
34 - implement enhanced support for valgrind.
35 - implement is_ready and return value for ready.
36 - removed Coro::Cont, it was a misnomer (it's generators),
37 and it was somewhat annoying to get it right. Will come back
38 if somebody asks for it :)
39 - many ->wait methods and Coro::Event could return
40 spuriously without the event having happened.
41
422.5 Tue Nov 7 12:22:33 CET 2006
43 - made Coro::Util, Coro::Select, Coro::Handle and Coro::Socket
44 use AnyEvent, moved them to Coro/.
45 - added Coro::LWP which contains all the uglyness required to
46 make LWP non-blocking.
47 - should work with perl 5.9.x now (Andreas König made me do it).
48 - fixed another bug in Coro::Select when the timeout was undef.
8 - reuse PL_start_env for all coros, saves some memory per coroutine. 49 - reuse PL_start_env for all coros, saves some memory per coroutine.
9 - manage PL_top_env differently, hopefully to avoid panic: top_env. 50 - manage PL_top_env differently, hopefully to avoid panic: top_env.
10 - timeout argument was not properly used in Coro::Socket. 51 - timeout argument was not properly used in Coro::Socket.
11 - allow limited forms of subclassing in Coro::Handle/Coro::Socket. 52 - allow limited forms of subclassing in Coro::Handle/Coro::Socket.
12 - emulate undocumented(!) functionality of IO::Socket required 53 - emulate undocumented(!) functionality of IO::Socket required
13 by LWP(!!). 54 by LWP(!!).
14 - updated eg/lwp to work with newer lwp's. 55 - updated eg/lwp to work with newer lwp's.
15 - made Coro::Util an AnyEvent user, moved it to Coro/.
16 - remove "FATAL: uncaught exception" prefix. coroutines that die 56 - remove "FATAL: uncaught exception" prefix. Coroutines that die
17 kill the whole process, just as exceptions in the main coroutine 57 kill the whole process, just as exceptions in the main "coroutine"
18 do. 58 did already.
19 59
202.1 Wed Nov 1 23:01:13 CET 2006 602.1 Wed Nov 1 23:01:13 CET 2006
21 - fix a long-standing bug in Coro::Select where select with 61 - fix a long-standing bug in Coro::Select where select with
22 zero timeout would instead change the current default filehandle. 62 zero timeout would instead change the current default filehandle.
23 - use a simpler and hopefully more robust way to clone padlists 63 - use a simpler and hopefully more robust way to clone padlists

Diff Legend

Removed lines
+ Added lines
< Changed lines
> Changed lines