Timeline
07/17/06:
- 17:14 Changeset [650] by
-
if we open connections outside of the accept handler, we should REALLY ...
- 17:14 Changeset [649] by
-
untab?
- 17:07 Changeset [648] by
-
r667@hope: sky | 2006-07-16 09:32:12 -0700 via chat with brad
- 17:07 Changeset [647] by
07/16/06:
- 10:38 Changeset [646] by
-
Got a bit over-excited in changeset 640.
07/15/06:
- 20:41 Changeset [645] by
-
Elaborate a bit more in the error message.
- 20:40 Changeset [644] by
-
Add another special case namespace to XMLElement.
Blech.
- 19:33 Changeset [643] by
-
Add docs and licence info to the external component stuff.
- 19:17 Changeset [642] by
-
New config block <Subdomain blah> ... </Subdomain> that, when used inside ...
- 18:53 Changeset [641] by
-
JEP-0114 (external components) support.
Seems to work with pyMSNt and ...
- 18:11 Changeset [640] by
-
Less log noise when passing stuff between vhosts.
- 14:43 Changeset [639] by
-
The start of a LiveJournal "journals" component.
This is just a toy for ...
- 11:35 Changeset [638] by
-
More work on MUC-as-component. Still doesn't really work.
I'm abandoning ...
07/14/06:
- 21:43 Changeset [637] by
-
Beginnings of a re-implementation of MUC as a component. It doesn't really ...
- 21:42 Changeset [636] by
-
Wasn't actually saving the config settings in the two Node classes.
- 21:40 Changeset [635] by
-
New Component method for sending a stanza through the vhost that owns the ...
- 21:38 Changeset [634] by
-
DJabberd::Delivery was overriding new() and breaking the configuration of ...
- 18:45 Changeset [633] by
-
Accidentally committed some test code.
- 18:41 Changeset [632] by
-
What are you still doing here? I thought I deleted you already!
- 18:40 Changeset [631] by
-
More inheritence re-jigging.
Now we have:
- Plugin
- Delivery
...
- Plugin
- 07:54 Changeset [630] by
-
log every single connection and it's incoming and outgoing xml, for now ...
- 07:27 Changeset [629] by
-
Remove excessive logging and outdated comment.
- 07:21 Changeset [628] by
-
Rejig the component class heirarchy a little bit, after reflecting on it. ...
- 06:53 Changeset [627] by
-
detab and make error more usefull
- 06:44 Changeset [626] by
-
don't display this
07/13/06:
- 20:58 Changeset [625] by
-
Completely rewritten implementation of DJabberd::Component::Easier for a ...
- 20:55 Changeset [624] by
-
A new method for building standards-compliant (I think?) error responses ...
- 20:52 Changeset [623] by
-
Make less log noise.
- 20:15 Changeset [622] by
-
Turn components back into real plugins again, except now they inherit from ...
- 19:54 Changeset [621] by
-
Components no longer get passed the $cb object and have to call ...
07/12/06:
- 20:26 Changeset [620] by
-
The start of a higher-level component base class that implements a lot of ...
- 20:25 Changeset [619] by
-
New Stanza method make_response which returns an empty stanza that is, in ...
- 13:55 Changeset [618] by
-
ridiculously pathetic counting of everything to try and figure out where ...
- 07:14 Changeset [617] by
-
changes in the monitor tool
- 03:08 Changeset [616] by
-
fix the previous refactoring to not break trusted probes
- 00:40 Changeset [615] by
-
unsubscribe
- 00:16 Changeset [614] by
-
formatting
- 00:10 Changeset [613] by
-
fix admin test
07/11/06:
- 23:51 Changeset [612] by
-
unsubscribe part 1, now we need to autorespond with an unsubscribed pack
- 23:10 Changeset [611] by
-
more refactoring to share code for people loading the roster item
- 22:58 Changeset [610] by
-
refactor
- 22:09 Changeset [609] by
-
built out roster storage plugin for LJ
- 22:05 Changeset [608] by
-
first part of the unsubscribe test
- 21:50 Changeset [607] by
-
croak earlier on bogus JID
- 18:54 Changeset [606] by
-
better reload support, w/o destroying initial memory, w/o warnings
- 18:24 Changeset [605] by
-
be able to reload the admin module at runtime
- 18:21 Changeset [604] by
-
when we leak hookchain callbacks, make gladiator show the phase they came ...
- 15:18 Changeset [603] by
-
store the last 10 longer than 5ms requests
- 15:04 Changeset [602] by
-
use devel leak
- 14:17 Changeset [601] by
-
for fucks sake, hopefully this will catch the double accounting
- 14:16 Changeset [600] by
-
don't hold on to the vhost
- 07:19 Changeset [599] by
-
Oh, shut up already!
- 03:49 Changeset [598] by
-
was backwards
- 03:42 Changeset [597] by
-
no eliza. ;)
- 03:41 Changeset [596] by
-
better stats
- 03:26 Changeset [595] by
-
don't leak presence packets by setting them when the person's logged out ...
07/10/06:
- 19:30 Changeset [594] by
-
guard against closing closed connections more.
might be messing up stats ...
- 18:26 Changeset [593] by
-
run gladiator in child process
- 08:28 Changeset [592] by
-
send JID to bot
- 07:30 Changeset [591] by
-
fix
- 07:27 Changeset [590] by
-
put bot in same group for now
- 07:19 Changeset [589] by
-
dev tool
- 05:14 Changeset [588] by
-
slightly more usefull to say what file doesnt exist
- 03:30 Changeset [587] by
-
LiveJournal bot.
- 02:23 Changeset [586] by
-
Reverting Mart's bots 2.0, which was a gutting of Artur's bots 1.0, with ...
- 00:18 Changeset [585] by
-
presence work for bots, and just general cleanup
07/09/06:
- 21:20 Changeset [584] by
-
track gearman failures
- 20:46 Changeset [583] by
-
Some todo items relating to components.
- 20:41 Changeset [582] by
-
A delivery plugin that turns a VHost into a component. This is as opposed ...
- 20:32 Changeset [581] by
-
A delivery plugin that short-circuits delivery between two local VHosts, ...
- 20:20 Changeset [580] by
-
idiot fix to pid lookup. i blame adapting some shell crap and moving it ...
- 19:05 Changeset [579] by
-
There's not really any reason why a bot needs to save its JID, since it ...
- 18:40 Changeset [578] by
-
show details on leaked IQs/tasks
- 18:39 Changeset [577] by
-
don't annoy on connect if servers been up less than 5 minutes.
- 09:38 Changeset [576] by
-
fixing a double-return of a parser on clean close in parse_more.
don't ...
- 08:49 Changeset [575] by
-
discard the roster early, once we detect they've used it their three ...
- 08:48 Changeset [574] by
-
don't warn if can't find pid
- 08:29 Changeset [573] by
-
store initial memory vs. incremental connection memory (artur)
- 08:22 Changeset [572] by
-
uh, my bad. I'd reverted this on accident back at rev 452 or so.
- 07:56 Changeset [571] by
-
debugging junk
- 07:53 Changeset [570] by
-
guard against calling close multiple times
- 07:44 Changeset [569] by
-
notes
- 07:41 Changeset [568] by
-
fix some leaks.
- 07:19 Changeset [567] by
-
forget last broadcast presence on disconnect
- 07:18 Changeset [566] by
-
memory work, breaking links between sax handler and parser when done with ...
- 07:17 Changeset [565] by
-
remove debug
- 02:59 Changeset [564] by
-
fixing another cycle
- 02:49 Changeset [563] by
-
cycle command
- 02:49 Changeset [562] by
-
weaken upref
- 02:48 Changeset [561] by
-
weaken our up pointer
- 02:21 Changeset [560] by
-
don't let users request roster before they're authenticated
- 02:14 Changeset [559] by
-
fix CPU spin in DNS code.
this patch is a little overkill: more than ...
07/08/06:
- 18:02 Changeset [558] by
-
fix mart's whitespace
- 18:02 Changeset [557] by
-
fix potential warning.
and fix mart's whitespace
- 18:01 Changeset [556] by
-
line-bless-tracked subs
- 17:59 Changeset [555] by
-
Remove expensive debug output.
- 17:25 Changeset [554] by
-
groupid doesn't reference jidmap
- 17:25 Changeset [553] by
-
admin bot config
- 16:28 Changeset [552] by
-
New configuration setting ChildService which adds a JID to the list of ...
- 16:05 Changeset [551] by
-
Refactor bots to be similar in principle to components. That is, Bots no ...
- 14:34 Changeset [550] by
-
Basic support for in-process "components".
Here I'm defining a component ...
- 11:26 Changeset [549] by
-
This was rejecting id="0" as well as the intended id="" (or no id at ...
- 08:03 Changeset [548] by
-
subcommands to show all/delta/etc
- 02:58 Changeset [547] by
-
devel::gladiator mem leak command
- 02:41 Changeset [546] by
-
advertise non-SASL auth
- 02:41 Changeset [545] by
-
for artur
- 02:33 Changeset [544] by
-
bye
07/07/06:
- 17:58 Changeset [543] by
-
only run fail callback when it there is one.
- 08:18 Changeset [542] by
-
kill the proxy handler class so it becomes easier to subclass behvaiour in ...
- 08:17 Changeset [541] by
-
move common code up to superclass so we don't have to reimplement it, ...
- 08:08 Changeset [540] by
-
instead of using a coderef, provide an object that for all these intents ...
- 08:05 Changeset [539] by
-
use exml
- 08:05 Changeset [538] by
-
remove spurios comment, use exml
- 07:41 Changeset [537] by
-
give out all the info that admin port has over jabber if your username is ...
- 07:26 Changeset [536] by
-
inital bot support, just support for getting messages, test subclass with ...
- 07:25 Changeset [535] by
-
allow a jid to be a coderf, if this is the case local delivery willinvoke ...
- 07:21 Changeset [534] by
-
back out mart's resource regexp restriction.
also, if JID ctor fails, ...
- 06:25 Changeset [533] by
-
asyncify w/ gearman
- 05:33 Changeset [532] by
-
rearranging for hackability
- 05:19 Changeset [531] by
-
async roster fetching
- 04:18 Changeset [530] by
-
async-ify the LJ auth plugin, using gearman and the ...
- 04:17 Changeset [529] by
-
notes
- 02:14 Changeset [528] by
-
make all dummies appear online
- 00:31 Changeset [527] by
-
Make dummy more useful, you can now configure what users should get the ...
07/06/06:
- 16:49 Changeset [526] by
-
USR2 will give you a perl level stack trace
07/05/06:
- 16:13 Changeset [525] by
-
crashed LJ Talk.
- 07:03 Changeset [524] by
-
Licence info for MySQL authen module.
- 06:55 Changeset [523] by
-
another quirk for trillian
07/04/06:
- 19:05 Changeset [522] by
-
MySQL Authen plugin.
Submitted by Karjala, but coding style/docs slightly ...
- 18:52 Changeset [521] by
-
Allow indented comments in config.
07/03/06:
- 19:37 Changeset [520] by
-
A revised regex for JID parsing, based on JEP-0029. Doesn't support ...
07/02/06:
- 20:27 Changeset [519] by
-
prettiers
- 20:21 Changeset [518] by
-
+ # remove invalid low unicode code points which aren't allowed in ...
- 20:17 Changeset [517] by
-
list the stats too and inject them with the right unit
- 20:15 Changeset [516] by
-
change stats to counters and add a stats mode that gives you the non ...
- 20:03 Changeset [515] by
-
take options
- 19:51 Changeset [514] by
-
whitespace fix
- 19:49 Changeset [513] by
-
say per sec instead of sec to deconfuse brad
- 19:15 Changeset [512] by
-
takes admin output and dumps into ganglia
- 18:27 Changeset [511] by
-
let messages go to server. it might be a bot.
- 18:22 Changeset [510] by
-
hook for on initial presence, to send a welcome message/etc
- 17:14 Changeset [509] by
-
add a class method DJabberd->foreach_vhost since I'm sure it'll be common ...
- 17:04 Changeset [508] by
-
list all connected users
- 17:01 Changeset [507] by
-
keep track of all active servers so we can traverse them for stats
- 16:45 Changeset [506] by
-
fix whitespace
- 03:25 Changeset [505] by
-
stanza processing latency tracking
- 01:45 Changeset [504] by
-
track auth success/failures
- 01:44 Changeset [503] by
-
connection listing
- 01:28 Changeset [502] by
-
how did that work before?
- 01:22 Changeset [501] by
-
account for trailing whitespace
- 01:19 Changeset [500] by
-
be more paranoid
- 01:13 Changeset [499] by
-
keep track of how many stanzas go to local delivery and how many go to s2s ...
- 01:07 Changeset [498] by
-
some more stats. probably badly named keys. artur/I will be fixing I'm ...
- 01:01 Changeset [497] by
-
keep track of modifications
- 00:51 Changeset [496] by
-
make the tests pass after Brads change
- 00:44 Changeset [495] by
-
more admin commands
- 00:28 Changeset [494] by
-
did that already
- 00:28 Changeset [493] by
-
roster caching, as well as roster load coalescing: don't kick off 3 async ...
- 00:15 Changeset [492] by
-
You can now set adminport, to which you can telnet in and get a list of ...
07/01/06:
- 23:55 Changeset [491] by
-
centralize roster loading to remove 3 big sections of duplicated code, and ...
- 16:51 Changeset [490] by
-
Fix the RosterGet fallback implementation.
Was using DJabberd::Roster ...
- 16:47 Changeset [489] by
-
Prefer a local log config (in ./etc) over a global one (in /etc). Also, ...
- 08:40 Changeset [488] by
-
test both InMemoryOnly and SQLite, you can control the SQLite with the ...
- 08:34 Changeset [487] by
-
globals aren't safe.
what if you had multiple vhosts, like in a test.
- 08:33 Changeset [486] by
-
updates
- 08:31 Changeset [485] by
-
added InMemoryOnly plugin we can use for tests
- 08:28 Changeset [484] by
-
guard against failures creating connection objects. like SSL ...
- 03:20 Changeset [483] by
-
support for declaring IPs of health check monitor machines (BIG-IP, ...
- 03:12 Changeset [482] by
-
fixed
- 03:12 Changeset [481] by
-
don't take down entire server on DNS lookup failure
- 03:10 Changeset [480] by
-
in some cases a dns timeout kills the server
- 02:11 Changeset [479] by
-
update
06/30/06:
- 21:02 Changeset [478] by
-
let IQ results/errors be delivered to unavailable connection
- 20:54 Changeset [477] by
-
todo
- 20:54 Changeset [476] by
-
some clients do stuff before being available, we should respond to their ...
- 20:54 Changeset [475] by
-
don't put yourself on your roster. makes iChat loop asking for vcard ...
- 20:54 Changeset [474] by
-
proper vcard format
- 19:45 Changeset [473] by
-
use avatar endpoint, which resizes to 96x96 for iChat
- 08:03 Changeset [472] by
-
downbless into a separate class where I can put accessors
- 07:56 Changeset [471] by
-
cleanup
- 07:50 Changeset [470] by
-
not required
- 07:47 Changeset [469] by
-
Implement JEP-0153 (Avatars), but kinda turned on its head: instead this ...
- 07:28 Changeset [468] by
-
FIXME for the IQ handler code
- 07:19 Changeset [467] by
-
move all the SQLite code to a subclass
- 05:07 Changeset [466] by
-
example for vcards
- 05:06 Changeset [465] by
-
found a bug
- 05:06 Changeset [464] by
-
detect right domain
- 02:54 Changeset [463] by
-
configurable SSL. no more magic.
- 02:06 Changeset [462] by
-
more cases we don't have to load rosteritems
- 01:54 Changeset [461] by
-
was falling through
- 01:49 Changeset [460] by
-
bug fixes to previous 'dont_load_rosteritem' patch
- 01:22 Changeset [459] by
-
way to flag presence packets as internally generated/trusted to avoid some ...
- 01:14 Changeset [458] by
-
more notes
- 01:03 Changeset [457] by
-
notes
06/29/06:
- 23:13 Changeset [456] by
-
livejournal roster/auth modules
- 22:25 Changeset [455] by
-
update LJ roster plugin, at least the core of it
- 20:55 Changeset [454] by
-
Implement basic Service Discovery in the MUC Plugin, plus make it less ...
- 19:37 Changeset [453] by
-
$self->{vhost} isn't set yet.
- 19:15 Changeset [452] by
-
-- lots of connection/stream/xmlparser reuse bug fixes
-- boot people ...
- 19:04 Changeset [451] by
-
trillian's jabber plugin is so fucking terrible.
- 02:04 Changeset [450] by
-
implement configurable parser sharing, fix some TODOs, declare types of ...
- 01:47 Changeset [449] by
-
comment cleanup
- 01:30 Changeset [448] by
-
rename 'jabberhandler' to 'saxhandler'
06/28/06:
- 22:55 Changeset [447] by
-
less/more useful logging
- 22:47 Changeset [446] by
-
remove debugging. :P
- 22:47 Changeset [445] by
-
declare fields
- 22:45 Changeset [444] by
-
DNS timeouts
- 22:14 Changeset [443] by
-
notes
- 20:27 Changeset [442] by
-
An attempt at allowing MUC rooms to be configured in the conf file. This ...
- 19:48 Changeset [441] by
-
New config option to log the server's PID to some file.
- 18:33 Changeset [440] by
-
don't think we actually want that weak. we properly break any possible ...
- 18:09 Changeset [439] by
-
more debugging
- 17:58 Changeset [438] by
-
notes
- 17:58 Changeset [437] by
-
notes
- 17:31 Changeset [436] by
-
let's play with the world
- 03:37 Changeset [435] by
-
bug as reported by Ask: we were rejecting incoming stanzas with a from= ...
06/27/06:
- 22:23 Changeset [434] by
-
some Jabber clients request roster first, then send initial presence. in ...
- 19:12 Changeset [433] by
-
work with unix domain sockets
- 19:11 Changeset [432] by
-
make test harness work with unix domain sockets
- 19:10 Changeset [431] by
-
support running on a unix domain socket
- 17:36 Changeset [430] by
-
client-only mode
- 17:25 Changeset [429] by
-
better debugging test
- 17:22 Changeset [428] by
-
still parse once byte at a time, but read bigger chunks
- 06:18 Changeset [427] by
-
tiny little memory gain
- 06:04 Changeset [426] by
-
remove now-misleading comment
- 06:02 Changeset [425] by
-
don't store the barejid string in the jidmap. there are other APIs for ...
- 05:53 Changeset [424] by
-
accidentally committed while debugging
- 05:52 Changeset [423] by
-
fix stream closing bug with new parser-on-demand changes
- 04:34 Changeset [422] by
-
don't rely on auto-vivify, and delete field, instead of assign to {}, to ...
- 04:34 Changeset [421] by
-
remove debuggin
- 04:33 Changeset [420] by
-
fix mem leak via closure
- 01:50 Changeset [419] by
-
connection load script
- 01:30 Changeset [418] by
-
fix memory leak: we were never unregistering the mapping from JID -> ...
- 01:29 Changeset [417] by
-
forgot this in last commit
- 01:29 Changeset [416] by
-
don't hold onto XML parsers. create one as needed.
06/26/06:
- 21:47 Changeset [415] by
-
memory footprint:
-- don't store the (massive) StreamStart object in the ...
- 21:45 Changeset [414] by
-
smaller memory footprint
- 21:07 Changeset [413] by
-
remove duplicate/unused information/method.
- 21:06 Changeset [412] by
-
mark these as not working
- 20:54 Changeset [411] by
-
fixes XMLParser memory leaks, working around Perl 5.8.4 bug (fixed in Perl ...
- 06:58 Changeset [410] by
-
- rename ds_client to ds_conn (remnant from pre-s2s support, when ...
- 06:56 Changeset [409] by
-
add accessor to get default declared namespace
06/22/06:
- 22:55 Changeset [408] by
-
r4060@colinux-kuiki: hachi | 2006-06-22 18:21:03 -0400 Serverout queue ...
- 09:03 Changeset [407] by
-
intracluster stanza delivery. no delivery plugin yet.
- 08:47 Changeset [406] by
-
processing of clustermessages
- 08:38 Changeset [405] by
-
sending ClusterMessage storables around
- 08:32 Changeset [404] by
-
r4056@colinux-kuiki: hachi | 2006-06-22 04:32:53 -0400 Here is the base ...
- 08:27 Changeset [403] by
-
start of intradomain protocol packet parsing
- 08:25 Changeset [402] by
-
r4052@colinux-kuiki: hachi | 2006-06-22 04:25:28 -0400 Add the base ...
- 08:15 Changeset [401] by
-
start of Connection::ClusterIn
- 05:43 Changeset [400] by
-
very rough start on cluster stuff.
commit early, commit often. hacking ...
- 05:41 Changeset [399] by
-
add cluster test
- 05:41 Changeset [398] by
-
style
- 05:28 Changeset [397] by
-
add start of function for inter-cluster testing
- 03:22 Changeset [396] by
-
count and loglevel
- 03:21 Changeset [395] by
-
don't allow two connections with same full JID. (same resource)
- 03:11 Changeset [394] by
-
fix count
- 03:10 Changeset [393] by
-
make test pass
- 03:05 Changeset [392] by
-
added failing test for dupe resource booting
- 03:04 Changeset [391] by
-
added end-sream
- 03:02 Changeset [390] by
-
more notes
- 02:28 Changeset [389] by
-
new method and notes
- 01:37 Changeset [388] by
-
memory usage notes
06/21/06:
- 21:10 Changeset [387] by
-
r4034@colinux-kuiki: hachi | 2006-06-21 17:11:06 -0400 Benchmarking ...
- 20:40 Changeset [386] by
-
found a bug
- 20:39 Changeset [385] by
-
todo update
- 20:04 Changeset [384] by
-
Fixes runaway memory/CPU leak/recursion/crash:
# set unavailable ...
- 17:59 Changeset [383] by
-
when you get your roster, you want it to re-generated subscribe requests ...
- 07:59 Changeset [382] by
-
allow testing all with MUC
- 05:29 Changeset [381] by
-
I missread the spec, we shouldn't send any broadcast to them except the ...
- 05:04 Changeset [380] by
-
two more race conditions, all tests now pass on my box
- 05:02 Changeset [379] by
-
race condition on wich packet got there first
- 04:58 Changeset [378] by
-
yawn
- 04:56 Changeset [377] by
-
not needed anymore, not doing subdomains in vhosts but rather as different ...
- 04:55 Changeset [376] by
-
make this test use a higher port so it doesnt try to use my server
- 04:44 Changeset [375] by
-
cached the sqlite connection so it gets shared between all instances so ...
- 02:38 Changeset [374] by
-
everything but vcard passes again. vcard passes but has some scary ...
- 01:03 Changeset [373] by
-
fixing tests
06/20/06:
- 23:55 Changeset [372] by
-
more test fixups
- 23:45 Changeset [371] by
-
test fixes
- 23:43 Changeset [370] by
-
cluster notes
- 23:42 Changeset [369] by
-
test fixes/additions (forgotten commit)/notes
- 23:41 Changeset [368] by
-
quell warning
- 23:31 Changeset [367] by
-
fix up htdigest test
- 23:17 Changeset [366] by
-
fix up this test so it passes again.
it was broken by the handling of not ...
- 22:56 Changeset [365] by
-
more robust shutdown of test servers
- 18:22 Changeset [364] by
-
don't hit the slow LDAP so hard
- 17:36 Changeset [363] by
-
untested fix for MUC destroying presence probes. (my guess)
- 06:09 Changeset [362] by
-
questions/notes, including possible bug we've been hunting
- 04:32 Changeset [361] by
-
some notes
06/19/06:
- 23:03 Changeset [360] by
-
r3994@colinux-kuiki: hachi | 2006-06-19 19:03:34 -0400 Start adding ...
