IRC log of dig on 2007-05-30

Timestamps are in UTC.

00:31:10 [lkagal]
lkagal (n=lkagal1@79.68.171.66.subscriber.vzavenue.net) has joined #dig
01:26:46 [DanC_lap]
DanC_lap has quit (Read error: 110 (Connection timed out))
01:47:16 [DanC_lap]
DanC_lap (n=connolly@206.135.195.2) has joined #dig
02:36:43 [sandro]
sandro has quit (kornbluth.freenode.net irc.freenode.net)
02:36:43 [DanC]
DanC has quit (kornbluth.freenode.net irc.freenode.net)
02:36:43 [lkagal]
lkagal has quit (kornbluth.freenode.net irc.freenode.net)
02:36:43 [eikeon]
eikeon has quit (kornbluth.freenode.net irc.freenode.net)
02:37:42 [das05r]
das05r has quit (Read error: 104 (Connection reset by peer))
02:46:05 [DanC_lap]
DanC_lap has quit (Read error: 110 (Connection timed out))
02:57:22 [irc.freenode.net]
Disconnected from irc.freenode.net (Connection reset by peer)
11:43:02 [DIGlogger]
DIGlogger (n=dig-logg@pink-panther.csail.mit.edu) has joined #dig
11:43:02 [kubrick.freenode.net]
topic is: DIG (http://dig.csail.mit.edu/ ) meets Mondays at 2:30pm (except 28 May, memorial day)
11:43:02 [kubrick.freenode.net]
Users on #dig: DIGlogger RalphS das05r lkagal sandro eikeon DanC
12:55:48 [lkagal]
lkagal has left #dig
12:55:55 [lkagal]
lkagal (n=lkagal1@79.68.171.66.subscriber.vzavenue.net) has joined #dig
13:09:48 [presbrey]
presbrey (n=Joe@31-34-215.wireless.csail.mit.edu) has joined #dig
13:11:16 [timbl]
timbl (n=timbl@206.135.195.3) has joined #dig
13:15:49 [jambo]
jambo (i=jambo@30-7-68.wireless.csail.mit.edu) has joined #dig
13:23:04 [dli01]
dli01 (i=dli@31-35-67.wireless.csail.mit.edu) has joined #dig
13:26:31 [timbl]
RalphS?
13:27:03 [timbl]
I wonder whether you could confirm the UROP CSAIL accounts .. my cert is protected by a password I don't know
13:57:29 [DanC_lap]
DanC_lap (n=connolly@206.135.195.2) has joined #dig
13:57:48 [lkagal]
lkagal has quit (Remote closed the connection)
14:02:12 [kennyluck]
kennyluck (n=kennyluc@30-7-218.wireless.csail.mit.edu) has joined #dig
14:07:29 [lkagal]
lkagal (n=lkagal1@79.68.171.66.subscriber.vzavenue.net) has joined #dig
14:42:29 [timbl]
timbl has quit (Read error: 113 (No route to host))
14:44:54 [DanC_lap]
DanC_lap has quit (Read error: 113 (No route to host))
15:09:04 [RalphS]
RalphS has quit (Read error: 110 (Connection timed out))
15:16:06 [timbl]
timbl (n=timbl@207.47.10.130.static.nextweb.net) has joined #dig
15:20:50 [lkagal]
lkagal has quit ()
15:57:45 [lkagal]
lkagal (n=lkagal1@30-5-196.wireless.csail.mit.edu) has joined #dig
15:59:56 [DanC_lap]
DanC_lap (n=connolly@207.47.10.130.static.nextweb.net) has joined #dig
16:20:12 [timbl]
timbl has quit ()
16:24:22 [timbl]
timbl (n=timbl@207.47.10.130.static.nextweb.net) has joined #dig
16:28:57 [RalphS]
RalphS (n=swick@30-6-188.wireless.csail.mit.edu) has joined #dig
17:43:22 [presbrey]
presbrey has left #dig
18:04:59 [lkagal]
lkagal has quit ()
18:07:09 [lkagal]
lkagal (n=lkagal1@30-5-196.wireless.csail.mit.edu) has joined #dig
19:55:15 [DanC_lap]
DanC_lap has quit (Read error: 110 (Connection timed out))
19:57:07 [timbl]
timbl has quit (Read error: 110 (Connection timed out))
20:22:26 [RalphS]
RalphS has quit ("bye for today")
20:44:22 [DanC_lap]
DanC_lap (n=connolly@207.47.10.130.static.nextweb.net) has joined #dig
20:45:16 [timbl]
timbl (n=timbl@207.47.10.130.static.nextweb.net) has joined #dig
20:54:40 [jambo]
timbl, i'm having some really bad problems with seeAlsos, expanding nodes, and browsing in general on the live version of tabulator
20:55:01 [timbl]
Oh. Sorry!
20:55:10 [timbl]
303s should work for th first time.
20:55:21 [timbl]
But it may be that I didn't test seeAlso
20:55:54 [jambo]
some nodes aren't expanding at all :(
20:56:43 [timbl]
There is one crucial place where the calll back fures out whether the new event is relevant.
20:56:47 [timbl]
Functio n is relevant()
20:57:38 [jambo]
okay
20:57:47 [timbl]
var relevant = function() { // Is the loading of this URI relevam to the display of subject?
20:59:12 [timbl]
I guess it should also query the store to find out whether <uri> seeAlso <rd>
20:59:37 [timbl]
I feel that is a pain as it happens every time the callback fires, even when it is for anothre URI
21:00:20 [timbl]
Do you have a test case?
21:01:28 [jambo]
checking one with kenny right now
21:02:40 [jambo]
with your foaf, it seems impossible to expand the node for #i http://dig.csail.mit.edu/2005/ajar/ajaw/tab.html?uri=http://www.w3.org/People/Berners-Lee/card
21:03:37 [jambo]
When you open Mentions->Male->E.g. timbl , the outline actually collapses back to its starting point
21:04:27 [jambo]
after what seems like the amount of time it would take it to expand
21:05:09 [timbl]
There are funny things which I have seen happen when something is open in two places on the utliner.
21:05:28 [timbl]
test case -- what starting URI for Mentions-> ...
21:05:29 [timbl]
?
21:05:41 [jambo]
http://dig.csail.mit.edu/2005/ajar/ajaw/tab.html?uri=http://www.w3.org/People/Berners-Lee/card
21:06:47 [jambo]
i would contrast it with the stable version, but unfortunately dbpedia takes over and makes it explode
21:07:48 [timbl]
Yes, I had to put a hack in to stop that
21:08:42 [timbl]
Ok. I think that what is happening is that when you open up 'tim berners-lee'
21:09:13 [timbl]
in the eg link is that it fires a callback on http://www.w3.org/People/Berners-Lee/card which causes the top level to repaint.
21:09:40 [timbl]
Which deletes anyhting opened out.
21:11:42 [timbl]
The repaint may be a feature
21:11:51 [timbl]
But the closeure of nesting isn't
21:12:32 [timbl]
Bt why *is* the callback for http://www.w3.org/People/Berners-Lee/card fired again, when it has already ben read?
21:44:16 [jambo_]
jambo_ (i=jambo@SENIOR-FIVE-SIXTY-SEVEN.MIT.EDU) has joined #dig
21:44:40 [jambo_]
timbl, im going to try and find the issue
21:52:29 [timbl]
you mean find the bug?
21:53:07 [jambo_]
yeah, I figured out the weird cases where it collapses itself, so now its just the random ones where it decides not to expand
21:53:41 [timbl]
Did yo fix the former?
21:53:50 [timbl]
the weird cases?
21:53:58 [dli01]
dli01 has quit (Read error: 110 (Connection timed out))
21:54:26 [jambo_]
no fix yet, I've just isolated the common thing between them
21:54:34 [timbl]
ok
21:54:42 [timbl]
I think at
21:54:42 [timbl]
if (!force && typeof this.requested[docuri]!="undefined") {
21:54:42 [timbl]
log.debug("We already have "+docuri+". Skipping.")
21:54:42 [timbl]
this.fireCallbacks('done',args)
21:54:42 [timbl]
return null
21:54:43 [timbl]
}
21:54:46 [timbl]
in sources
21:54:56 [timbl]
it may be firing the repaint o fthe outline
21:55:00 [jambo_]
okay
21:55:05 [timbl]
at the high level.
21:55:11 [jambo_]
right
21:55:44 [timbl]
It may be unnecessary to fir request and then done when the document has already been loaded.
21:57:40 [timbl]
Keep the fire of request till after that test, and not fire done in teher.
21:57:59 [timbl]
(The number of dones must exactly match th number of requests in the end)
21:58:13 [jambo_]
so just have it return, instead?
21:58:20 [timbl]
Could try that.
21:58:45 [timbl]
Other things might break... but I think i added code to do an inital outline_expand whether or not there are any callbaqcsk
21:59:31 [timbl]
Yes: render() // inital open, or else full if re-open
21:59:31 [timbl]
tabulate.js line 1513
22:00:49 [lkagal]
lkagal has quit ()
22:01:48 [timbl]
i think in the longer trem it would be better (a) to keep an index of callback routines per event type per URI, and (b) when a redirect happnes, move all callbacks from the uri A to the uri B, or put a callback for B to call A (might be better)
22:03:10 [timbl]
Oh, by the way, there is a bug with photos in bubbles in the map that they come out huge
22:03:49 [jambo]
jambo has quit (Read error: 110 (Connection timed out))
22:04:08 [kennyluck]
kennyluck has left #dig
22:05:08 [jambo_]
jambo_ is now known as jambo
22:11:01 [jambo]
Im aware, ill make that next on my plate
22:11:44 [jambo]
it turns out that the outline is actually expanding and then collapsing very very quickly
22:14:09 [timbl]
yes, it it that the high-level outline gets a repaint event
22:14:16 [timbl]
The outermost
22:14:37 [timbl]
because it has subscribed to events for http:/.../card
22:15:01 [jambo]
oh, i see now
22:15:08 [jambo]
yeah, that explains the other cases too
22:15:32 [timbl]
Ths only happens when the same URI (module #fragment) occurs at more than one level.
22:16:46 [timbl]
The TAG takes a break
22:18:32 [jambo]
So then how to get around it taking up events that shouldnt belong to it is the problem, huh?
22:25:33 [timbl]
Well, one would be to say, it can expand but it has to preserve its expanded children. That would be heavy, but would alow update wehn new data becomes availbel.
22:26:16 [timbl]
A second way would be to make its basic registration for the event be a one-off: make sure the callback is deleted.
22:26:27 [timbl]
Currenty callbacks are dleted of they return false, IIRC
22:26:34 [timbl]
That might have been messed up
22:36:24 [timbl]
You could have a look at the state of the callback lists
22:40:10 [jambo]
Thats what I am doing right now
22:40:29 [jambo]
But I am going to get some dinner first :)
23:14:16 [lkagal]
lkagal (n=lkagal1@79.68.171.66.subscriber.vzavenue.net) has joined #dig
23:15:44 [lkagal]
lkagal has quit (Client Quit)