… | |
… | |
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 | |
6 | |
|
|
7 | faq: can't se anything |
|
|
8 | faq: all is asynch |
|
|
9 | faq: how to interface to non-perl nodes? |
|
|
10 | |
|
|
11 | TODO: maybe disbale warnings by default? |
|
|
12 | TODO: lies sadrak mails |
|
|
13 | |
|
|
14 | - listener-less nodes had trouble sending keepalive |
|
|
15 | messages on write timeouts (found by Sadrak). |
|
|
16 | - the monitor guard could cause a memleak due |
|
|
17 | to autovivification if a mon was cleared after |
|
|
18 | the port was gone (analysed by Sadrak). |
|
|
19 | |
|
|
20 | 1.28 Thu Apr 1 21:23:54 CEST 2010 |
|
|
21 | - accepted conenctions didn't corretcly set up the SRCNODE |
|
|
22 | in some cases, leading to intra-node messages (such as monitoring) |
|
|
23 | to fail silently (great testcase by Sadrak). |
|
|
24 | |
|
|
25 | 1.27 Sat Mar 20 21:11:16 CET 2010 |
|
|
26 | - avoid exceptions in the transport code if the connection |
|
|
27 | fails during TLS negotiation. |
|
|
28 | - delay new address broadcasts in seednodes, to reduce the |
|
|
29 | likelyhood of nodes connecting at the same time, leading to |
|
|
30 | message loss. |
|
|
31 | - improve general message passing efficiency by 8%, and special |
|
|
32 | case json framing for a 50% speed increase. |
|
|
33 | - add hints how to interface to aemp nodes from other programming |
|
|
34 | languages using the aemp protocol in AnyEvent::MP::Transport |
|
|
35 | (it can really be quite simple). |
|
|
36 | - handle unsupported remote framings more gracefully. |
|
|
37 | |
|
|
38 | 1.26 Wed Dec 30 16:48:37 CET 2009 |
7 | - INCOMPATIBLE CHANGE: the high-level node protocol changed, |
39 | - INCOMPATIBLE CHANGE: the high-level node protocol changed, |
8 | all nodes need to upgrade. |
40 | all nodes need to upgrade. |
9 | - remote kil's did not work (reported by Eugene Grigoriev). |
41 | - remote kil's did not work (reported by Eugene Grigoriev). |
10 | - documented monitor_timeout, connect_interval, auth_offer, |
42 | - documented monitor_timeout, connect_interval, auth_offer, |
11 | auth_accept, framing_format, autocork and nodelay options |
43 | auth_accept, framing_format, autocork and nodelay options |
12 | and make them configurable via bin/aemp. |
44 | and make them configurable via bin/aemp. |
13 | - some doc updates. |
45 | - some doc updates. |
14 | - remote kil's were not synchronous to message sends which affected |
46 | - remote kil's were not synchronous to message sends which affected |
15 | nobody, as remote kils were broken. |
47 | nobody, as remote kils were broken. |
|
|
48 | - remote mon/unmonitor requests were not synchronous to message |
|
|
49 | sends either sometimes, although it likely didnt matter. |
16 | |
50 | |
17 | 1.24 Wed Dec 9 14:59:52 CET 2009 |
51 | 1.24 Wed Dec 9 14:59:52 CET 2009 |
18 | - codename "the foul egg". |
52 | - codename "the foul egg". |
19 | - increased "unmonitored local port died" message level to 2, so |
53 | - increased "unmonitored local port died" message level to 2, so |
20 | it is displayed by default now. |
54 | it is displayed by default now. |