| 1 | Revision history for AnyEvent::MP |
1 | Revision history for AnyEvent::MP |
| 2 | |
2 | |
| 3 | TODO: testsuite |
3 | TODO: testsuite |
| 4 | TODO: intro: maybe simple job pool example? |
4 | TODO: intro: maybe simple job pool example? |
| 5 | TODO: intro: mention watchdog... |
5 | TODO: intro: mention watchdog... |
|
|
6 | TODO: maybe disbale warnings by default? |
|
|
7 | TODO: networkdomains |
|
|
8 | TODO: group proxy nodes - store group membership for other nodes |
|
|
9 | TODO: switch to AnyEvent::Log |
| 6 | |
10 | |
|
|
11 | faq: can't se anything |
|
|
12 | faq: all is asynch |
|
|
13 | faq: how to interface to non-perl nodes? |
|
|
14 | |
|
|
15 | TODO: AnyEvent::MP gets grp_reg, grp_get |
|
|
16 | TODO: AnyEvent::MP keeps list of all "global" nodes |
|
|
17 | TODO: Global loaded on demand only, adds grp database and node address database |
|
|
18 | TODO: transport connects use any global node to find addresses (maybe itself) |
|
|
19 | TODO: not really |
|
|
20 | TODO: nuke global_req/g_reply |
|
|
21 | TODO: use AE::log |
|
|
22 | TODO: maybe check g_proto version? |
|
|
23 | TODO: connect instantly in node->connect |
|
|
24 | |
|
|
25 | 2 |
|
|
26 | - INCOMPATIBLE CHANGE: the node protocol used by this version is not |
|
|
27 | compatible to versions <2 - you have to upgrade all node |
|
|
28 | - new "norc" and "force" options for MP::configure to ignore |
|
|
29 | or override the rc file. |
|
|
30 | - move seed code to Kernel. |
|
|
31 | - fixed small typo in MP::Transport. |
|
|
32 | - fix (unused) provider-kv (version sometimes missing). |
|
|
33 | - reduce the default export list of MP::Kernel. |
|
|
34 | - reduced various random delays to zero, as they |
|
|
35 | are hopefully no longer needed with the new global |
|
|
36 | protocol. |
|
|
37 | |
|
|
38 | 1.30 Thu Jun 30 11:30:39 CEST 2011 |
|
|
39 | - connection errors at the right time would kill node |
|
|
40 | connections in a bad way (patch by Malcolm Studd). |
|
|
41 | |
|
|
42 | 1.29 Fri May 7 20:13:39 CEST 2010 |
|
|
43 | - codename "Sadrak". |
|
|
44 | - fix error in callback at AnyEvent/MP/Global.pm line 339 |
|
|
45 | (found by Sadrak). |
|
|
46 | - listener-less nodes had trouble sending keepalive |
|
|
47 | messages on write timeouts (found by Sadrak). |
|
|
48 | - the monitor guard could cause a memleak due |
|
|
49 | to autovivification if a mon was cleared after |
|
|
50 | the port was gone (analysed by Sadrak). |
|
|
51 | - do not overwrite the config file if we couldn't read it |
|
|
52 | for some reason (lesson demonstrated by Sadrak). |
|
|
53 | |
|
|
54 | 1.28 Thu Apr 1 21:23:54 CEST 2010 |
|
|
55 | - accepted connections didn't correctly set up the SRCNODE |
|
|
56 | in some cases, leading to intra-node messages (such as monitoring) |
|
|
57 | to fail silently (great testcase by Sadrak). |
|
|
58 | |
|
|
59 | 1.27 Sat Mar 20 21:11:16 CET 2010 |
|
|
60 | - avoid exceptions in the transport code if the connection |
|
|
61 | fails during TLS negotiation. |
|
|
62 | - delay new address broadcasts in seednodes, to reduce the |
|
|
63 | likelyhood of nodes connecting at the same time, leading to |
|
|
64 | message loss. |
|
|
65 | - improve general message passing efficiency by 8%, and special |
|
|
66 | case json framing for a 50% speed increase. |
| 7 | - add hints how to interface to aemp nodes from other programming |
67 | - add hints how to interface to aemp nodes from other programming |
| 8 | languages using the aemp protocol in AnyEvent::MP::Transport |
68 | languages using the aemp protocol in AnyEvent::MP::Transport |
| 9 | (it can really be quite simple). |
69 | (it can really be quite simple). |
|
|
70 | - handle unsupported remote framings more gracefully. |
| 10 | |
71 | |
| 11 | 1.26 Wed Dec 30 16:48:37 CET 2009 |
72 | 1.26 Wed Dec 30 16:48:37 CET 2009 |
| 12 | - INCOMPATIBLE CHANGE: the high-level node protocol changed, |
73 | - INCOMPATIBLE CHANGE: the high-level node protocol changed, |
| 13 | all nodes need to upgrade. |
74 | all nodes need to upgrade. |
| 14 | - remote kil's did not work (reported by Eugene Grigoriev). |
75 | - remote kil's did not work (reported by Eugene Grigoriev). |
