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

Comparing AnyEvent-MP/Changes (file contents):
Revision 1.61 by root, Wed Sep 9 01:47:01 2009 UTC vs.
Revision 1.152 by root, Sun Aug 28 08:13:43 2016 UTC

1Revision history for AnyEvent::MP 1Revision history for AnyEvent::MP
2 2
3TODO: global protocol fix?
4TODO: use seednodes only to broadcast?
5TODO: testsuite 3TODO: testsuite
6TODO: intro: maybe simple job pool example? 4TODO: intro: maybe simple job pool example?
7TODO: intro: mention watchdog... 5TODO: intro: mention watchdog...
8TODO: group monitoring 6TODO: maybe disbale warnings by default?
7TODO: listener-scopes (10.0.0.1:4040@vpn) and connect-scopes ("vpn,public")
8TODO: document env-variable usage
9TODO: make node objects responsible for keepalive?
9 10
10TODO: cal, cal_func 11faq: can't se anything
11TODO: fix the "trying to contatc node under wrogn address" problem. 12faq: all is asynch
122009-09-08 03:58:42 <9> trying connect to seed node 10.0.0.18:4040. 13faq: how to interface to non-perl nodes?
132009-09-08 03:58:42 <9> 10.0.0.18:4040 frank already connected, not connecting again. 14
15TODO: check gproto, nproto, on connect
16TODO: eg/*
17TODO: #TODO##d# markers
18TODO: limiting reconnecting speed when unreachable? somehow use same interval timers as for seeding and keepalive?
19TODO: multiple profiles? also some default profiles?
20TODO: export keepalive?
21TODO: $guard = con $cb->($up)
22TODO: kil $node - disconnect?
23TODO: rand is seeded by 32 bits only
24TODO: readline support
25TODO: gleeco re: AE::MP::DataConn -
262
27 - INCOMPATIBLE CHANGE: the node protocol used by this version is not
28 compatible to versions <2 - you have to upgrade all nodes.
29 - INCOMPATIBLE CHANGE: many APIs have changed, see the section
30 "PORTING FROM VERSION 1" in the AnyEvent::MP manpage for details.
31 - INCOMPATIBLE CHANGE: remove md6 auth algorithms (and use algorithms
32 based on SHA3 instead).
33 - AnyEvent::MP::Global is now optional - only seed nodes autoload it
34 and create a fully-meshed network with each other, normal nodes only
35 have to connect to seed nodes, other connections are on-demand.
36 - new "norc" and "force" options for MP::configure to ignore
37 or override the rc file.
38 - a shitload of minor and major fixes all reported and debugged
39 by Felix Ostmann.
40 - move seed code to Kernel.
41 - non-json receive framing was probably broken.
42 - fixed small typo in MP::Transport.
43 - fix (unused) provider-kv (version sometimes missing).
44 - reduce the default export list of MP::Kernel.
45 - reduced various random delays to zero, as they
46 are hopefully no longer needed with the new global
47 protocol.
48 - implement a secure mode that can suppress (some) code execution
49 requests.
50 - do not use GMP for generating alphanumeric strings even when
51 available, use a more tricky algorithm instead.
52 - anonymous node names and remote port names have been shortened
53 considerably.
54 - $AnyEvent::MP::Kernel::SRCNODE contains the node ID of the origin
55 node during message receives.
56 - aemp shell now supports "package" selection and offers an
57 $ECHO port you can send messages to.
58 - rely on perl rand() instead of using /dev/urandom when available, as
59 hopefully it is random enough.
60 - aemp shell no longer leaks one port per command.
61 - nodenames can contain %-escapes such as %n or %u.
62 - aemp now uses aemp/%n/%u as nodename.
63 - configure supports "eval", a small perl snippet to initialsie a node,
64 most useful in "aemp run eval ...".
65 - known_nodes is gone, it has little value, use all_nodes as
66 replacement.
67 - lots of small change and improvements.
68 - node_is_known has been removed.
69 - TRACE now shows the portid on SND.
70 - aggressively reduce number of external modules, to reduce
71 memory usage for nodes.
72 - rework most of the documentation.
73 - clean up MP::Kernel exports.
74
751.30 Thu Jun 30 11:30:39 CEST 2011
76 - connection errors at the right time would kill node
77 connections in a bad way (patch by Malcolm Studd).
78
791.29 Fri May 7 20:13:39 CEST 2010
80 - codename "Sadrak".
81 - fix error in callback at AnyEvent/MP/Global.pm line 339
82 (found by Sadrak).
83 - listener-less nodes had trouble sending keepalive
84 messages on write timeouts (found by Sadrak).
85 - the monitor guard could cause a memleak due
86 to autovivification if a mon was cleared after
87 the port was gone (analysed by Sadrak).
88 - do not overwrite the config file if we couldn't read it
89 for some reason (lesson demonstrated by Sadrak).
90
911.28 Thu Apr 1 21:23:54 CEST 2010
92 - accepted connections didn't correctly set up the SRCNODE
93 in some cases, leading to intra-node messages (such as monitoring)
94 to fail silently (great testcase by Sadrak).
95
961.27 Sat Mar 20 21:11:16 CET 2010
97 - avoid exceptions in the transport code if the connection
98 fails during TLS negotiation.
99 - delay new address broadcasts in seednodes, to reduce the
100 likelyhood of nodes connecting at the same time, leading to
101 message loss.
102 - improve general message passing efficiency by 8%, and special
103 case json framing for a 50% speed increase.
104 - add hints how to interface to aemp nodes from other programming
105 languages using the aemp protocol in AnyEvent::MP::Transport
106 (it can really be quite simple).
107 - handle unsupported remote framings more gracefully.
108
1091.26 Wed Dec 30 16:48:37 CET 2009
110 - INCOMPATIBLE CHANGE: the high-level node protocol changed,
111 all nodes need to upgrade.
112 - remote kil's did not work (reported by Eugene Grigoriev).
113 - documented monitor_timeout, connect_interval, auth_offer,
114 auth_accept, framing_format, autocork and nodelay options
115 and make them configurable via bin/aemp.
116 - some doc updates.
117 - remote kil's were not synchronous to message sends which affected
118 nobody, as remote kils were broken.
119 - remote mon/unmonitor requests were not synchronous to message
120 sends either sometimes, although it likely didnt matter.
121
1221.24 Wed Dec 9 14:59:52 CET 2009
123 - codename "the foul egg".
124 - increased "unmonitored local port died" message level to 2, so
125 it is displayed by default now.
126 - generate random seed at configure time, not module loading time,
127 so fork&configure works as expected.
128 - decreased "seed"-related message priority in Global from 0 (d'oh)
129 to 9, as they are not signifcant.
130 - new experimental module AnyEvent::MP::DataConn to establish
131 additional data connections between nodes.
132 - bin/aemp now loads AnyEvent::Watchdog::Util before trying
133 to call restart.
134 - bin/aemp setcert didn't properly handle the filename argument.
135 - removed empty and unused timeout= parameter from protocol greeting.
136
1371.23 Fri Nov 6 18:46:26 CET 2009
138 - listener-less nodes were misinformed by their masters about
139 the location of new nodes, and therefore didn't connect
140 properly in all cases.
141 - messages send during node-up processing could sometimes get lost,
142 which would hamper global's ability to mesh the network.
143 - fixed AnyEvent::MP::Global::grp_reg to return a guard also
144 in scalar context.
145 - fixed AnyEvent::MP::Kernel::mon_nodes to return a guard also
146 in scalar context.
147 - try to improve error reporting when automatically loading
148 a function - being unable to load a module will now
149 stop the process and report the error.
150
1511.22 Sat Oct 17 03:41:47 CEST 2009
152 - fix two rcv-bugs: after adding a tagged rcv, the default
153 callback would not be used or changed anymore (reported
154 by Felix Antonius Wilhelm Ostmann, with apologies for
155 taking so long to releae the fix).
156 - added Coro::MP introduction to AnyEvent::MP::Intro.
157 - add peval function.
158
1591.21 Mon Oct 5 21:58:36 CEST 2009
160 - always delay kil messages.
161 - remote monitor requests were never cleaned up (found by
162 Sadrak and analysed by elmex).
163 - move $VERSION to AnyEvent::MP and AnyEvent::MP only.
164
1651.2 Thu Oct 1 22:59:35 CEST 2009
166 - fix a rather big bug where message processing could be
167 inhibited completely.
168 - grp_mon could not return a guard due to a typo.
169 - implement primitive aemp shell.
170 - do not use bound address as listen length (reported by
171 Felix Antonius Wilhelm Ostman).
172 - no longer export an undefined "reg" function from AE::MP.
173 - the mon_guard return value no longer keeps an additional
174 reference to the passed refs.
175
1761.1 Fri Sep 11 04:34:03 CEST 2009
177 - bumped the transport protocol to version 1 - sorry, but there
178 were too many bugfixes/changes.
179 - new function AnyEvent::MP::cal for simple rpc.
14 - renamed AnyEvent::MP::Global functions to grp_reg|get|mon. 180 - renamed AnyEvent::MP::Global functions to grp_reg|get|mon.
181 - implemented monitoring for Global groups.
15 - removed all userspace time-outs from the transport, instead 182 - removed all userspace time-outs from the transport, instead
16 rely on tcp retransmit timeouts and tcp keepalive. 183 rely on tcp retransmit timeouts and tcp keepalive.
17 - spawn now delays spawns on the local node artificially, so 184 - spawn now delays spawns on the local node artificially, so
18 it can return before invoking the init function. 185 it can return before invoking the init function.
19 - break endless recursion between ping-pong ports on the local 186 - break endless recursion between ping-pong ports on the local
… …
27 - rely on perl srand'ing better than we could. 194 - rely on perl srand'ing better than we could.
28 - try to backport to 5.8.2, unfortunately, recursive dependencies 195 - try to backport to 5.8.2, unfortunately, recursive dependencies
29 rely on Module::Build, which fails at 5.8.7, so YMMV. 196 rely on Module::Build, which fails at 5.8.7, so YMMV.
30 - use a vastly different reseeding strategy. 197 - use a vastly different reseeding strategy.
31 - use seed nodes to notify nodes about new nodes. 198 - use seed nodes to notify nodes about new nodes.
199 - in the case where a ndoe learns new addresses after it tried
200 connecting, prefer the new addrseses over the old ones.
32 201
331.0 Wed Sep 2 20:40:43 CEST 2009 2021.0 Wed Sep 2 20:40:43 CEST 2009
34 - randomise more connection times, but reduce delay to 0.2s (seeds) 203 - randomise more connection times, but reduce delay to 0.2s (seeds)
35 and 0.05s (normal connects). 204 and 0.05s (normal connects).
36 - upgrade to and require AnyEvent::Watchdog 1.0. 205 - upgrade to and require AnyEvent::Watchdog 1.0.

Diff Legend

– Removed lines
+ Added lines
< Changed lines
> Changed lines