IRC log of sparql on 2010-03-23

Timestamps are in UTC.

14:00:27 [RRSAgent]
RRSAgent has joined #sparql
14:00:27 [RRSAgent]
logging to
14:00:28 [AndyS]
zakim, ??P4 is me
14:00:28 [Zakim]
+AndyS; got it
14:00:31 [bglimm]
14:00:35 [kasei]
Zakim, aabb is me
14:00:35 [Zakim]
+kasei; got it
14:00:37 [bglimm]
scribe: bglimm
14:00:44 [AndyS]
I think ... was [IP-caller] last week.
14:00:55 [bglimm]
rrsagent, make logs world
14:01:05 [Zakim]
14:01:07 [kasei]
Zakim, mute me
14:01:07 [Zakim]
kasei should now be muted
14:01:09 [SteveH_]
Zakim, ??P14 is me
14:01:09 [Zakim]
+SteveH_; got it
14:01:42 [Zakim]
14:02:14 [Zakim]
14:02:16 [LeeF]
zakim, MIT262 has me
14:02:16 [Zakim]
+LeeF; got it
14:03:10 [bglimm]
LeeF: Main purpose today is the F2F
14:03:38 [LeeF]
zakim, who's here?
14:03:38 [Zakim]
On the phone I see +0186528aaaa, MattPerry, AndyS, kasei (muted), SteveH_, pgearon, MIT262
14:03:40 [Zakim]
MIT262 has LeeF
14:03:41 [Zakim]
On IRC I see RRSAgent, Zakim, MattPerry, AndyS, ivan, LeeF, bglimm, SteveH_, iv_an_ru, karl, sandro, pgearon, AlexPassant, kasei, trackbot
14:04:12 [pgearon]
14:04:40 [bglimm]
AndyS: Is the Oxofrd location easy to find?
14:05:13 [LeeF]
ack pgearon
14:05:19 [bglimm]
14:05:41 [bglimm]
bglimm: oeRC building on Keble road
14:05:59 [LeeF]
14:06:03 [bglimm]
LeeF: MIT location is as for F2F1.
14:07:14 [bglimm]
LeeF: bglimm and I are doing a video conference test run after the teleconf
14:07:29 [Zakim]
14:07:36 [bglimm]
... I refactored the agenda a little bit and added links to agenda topics
14:07:57 [bglimm]
... if somebody has additional links, please add them. It is all on a wiki page.
14:08:21 [Souri]
Souri has joined #sparql
14:09:35 [bglimm]
... Andy, I tried to find a link for having exists or not exists in filters and couldn't find a link to an email discussion
14:10:04 [bglimm]
AndyS: I don't think there was an email discussion. It came up in the phone calls a couple of times.
14:10:36 [bglimm]
LeeF: Thursday morning we will discuss time permitting features
14:12:00 [bglimm]
... there have been questions whether we will decide before REC/LC of the main documents whether the time permitting features will change their status or whether we keep them as time permitting till the end
14:12:30 [bglimm]
... e.g., entailment regimes or property paths could get a thumbs up from the group earlier on
14:13:18 [bglimm]
... that's something to think about for the F2F and if there are concerns let me know. We can then settle that at the F2F.
14:13:56 [bglimm]
... we gonna need people to take the lead on test cases to determine how they are run etc
14:14:30 [LeeF]
zakim, who's on the phone?
14:14:30 [Zakim]
On the phone I see +0186528aaaa, MattPerry, AndyS, kasei (muted), SteveH_, pgearon, MIT262, Souri
14:14:31 [bglimm]
... any questions or suggestion for the F2F?
14:14:32 [Zakim]
MIT262 has LeeF
14:14:56 [bglimm]
Zakim, +0186528aaaa is me
14:14:56 [Zakim]
+bglimm; got it
14:15:17 [bglimm]
AndyS: Should we have a call the Tuesday after the F2F?
14:15:30 [bglimm]
LeeF: Yes, I think that is useful.
14:15:35 [Zakim]
14:15:37 [pgearon]
Thanks Lee
14:15:40 [Zakim]
14:15:41 [bglimm]
14:15:42 [Zakim]
14:15:44 [Zakim]
14:15:48 [Zakim]
14:15:50 [Zakim]
14:16:00 [Zakim]
14:16:13 [bglimm]
LeeF, what do I need to do for the (short) minutes?
14:16:34 [bglimm]
rrsagent, create minutes
14:16:34 [RRSAgent]
I have made the request to generate bglimm
14:16:54 [bglimm]
rrsagent, make logs world
14:17:08 [kasei]
LeeF, thanks for sending reassurances to Nicholas on the comments list. although the draft response is shorter than your reassurance :)
14:17:50 [bglimm]
chair: LeeF
14:18:08 [kasei]
I saw you approved the RV-2 draft, but not NH-1. Could you, or someone else approve that (which just says we're going to tackle the result format issue)?
14:18:14 [bglimm]
Zakim, who is here?
14:18:14 [Zakim]
On the phone I see bglimm
14:18:15 [Zakim]
On IRC I see RRSAgent, Zakim, AndyS, ivan, LeeF, bglimm, SteveH_, iv_an_ru, karl, sandro, pgearon, AlexPassant, kasei, trackbot
14:18:40 [kasei]
14:19:24 [kasei]
Did you read over the latest response from Rob Vesse? Wondering if the followup questions are for you or SteveH.
14:21:16 [LeeF]
bglimm, the minutes magic stuff is at
14:21:23 [AndyS]
Only PP comment could be for me - aggs is Steve. And the PP comment is more a remark - he's emailed Steve and I off list as well and this feels like a discussion not a formal comment. This design works-ish and does not work.
14:21:55 [bglimm]
how can I get the IRC log into a wiki page?
14:22:13 [AndyS]
Lee/chair - feel free to rule on whether the PP part needs a response. He's not actually asking for a change - just telling us what he's done.
14:22:34 [kasei]
OK. Do you consider the RV-2 thread to be resolved, then?
14:22:53 [bglimm]
ah, I see
14:23:04 [LeeF]
AndyS, I don't think it needs a response
14:24:01 [bglimm]
member:rrsagent, create minutes
14:24:11 [bglimm]
member:rrsagent, create minutes
14:24:12 [AndyS]
For my part yes. For aggs - still a couple of dangling comments I'm afraid - values and DISTINCT (important detail) and SAMPLE/DISTINCT
14:24:36 [Zakim]
14:24:38 [Zakim]
SW_(SPARQL)10:00AM has ended
14:24:39 [Zakim]
Attendees were MattPerry, +1.310.729.aabb, AndyS, kasei, SteveH_, pgearon, LeeF, Souri, bglimm
14:24:43 [kasei]
LeeF, I've got a few things for the SD F2F agenda. Can I hand them off to you or just update the wiki?
14:25:38 [LeeF]
go ahead and update the wiki, please
14:25:42 [LeeF]
i appreciate it
14:37:32 [bglimm]
Andy, are you still here?
14:37:41 [bglimm]
Here's the street view for the entrance:
14:37:42 [bglimm],-1.258517&sspn=0.007849,0.01929&ie=UTF8&hq=&hnear=7+Keble+Rd,+Oxford+OX1+3,+United+Kingdom&ll=51.759487,-1.258535&spn=0.007849,0.01929&z=16&layer=c&cbll=51.759521,-1.258425&panoid=qkHCzOBCL0OZKoBFavyWQw&cbp=12,342.25,,0,5
14:51:39 [LeeF]
bglimm, our video conf IP address is
14:51:48 [LeeF]
which is
14:52:12 [LeeF]
and if it's useful, we also have an ISDN number of: 6172580349
14:52:13 [bglimm]
ok, I'll head over to the video conference room now
14:52:16 [ericP]
ericP has joined #sparql
14:52:17 [LeeF]
wait that's not right
14:52:22 [bglimm]
yes, I have all that from Sandro
14:52:25 [LeeF]
ISDN is: 6172580340
14:52:25 [ericP]
get your act together, man
14:52:29 [LeeF]
14:52:40 [bglimm]
I somewhere have our data as well
14:54:04 [AxelPolleres]
AxelPolleres has joined #sparql
14:54:14 [AxelPolleres]
14:54:35 [AxelPolleres]
... got caught in the time-shift trap.
14:54:41 [LeeF]
AxelPolleres, :-D
14:54:46 [AxelPolleres]
14:54:52 [AxelPolleres]
well well well...
14:55:18 [AxelPolleres]
wanted to add:
14:55:31 [bglimm]
LeeF: I can't find our IP, but I head to the room now and ask. I'll take my laptop with me, so we can talk over IRC.
14:55:38 [bglimm_]
bglimm_ has joined #sparql
14:55:41 [LeeF]
14:55:46 [LeeF]
you can also just call us, which is fine as well
14:55:51 [AxelPolleres]
it would make a lot of sense to remind people to pick up ANY open actions they can, before the F2F (especially testing)
14:56:06 [LeeF]
AxelPolleres, i'd suggest you send mail
14:56:14 [AxelPolleres]
will do
14:56:15 [AxelPolleres]
14:56:18 [AxelPolleres]
14:56:39 [AxelPolleres]
did you have a chance to get ahead there?
14:58:10 [AxelPolleres]
I saw at least discussions ongoing on the mail thread today (
14:58:34 [LeeF]
no, i haven't yet
14:58:42 [LeeF]
i'm mildly hopeful i'll do something today or tomorrow
14:58:45 [LeeF]
but can't guarantee it
14:58:48 [LeeF]
i know it's a major action
14:58:56 [AxelPolleres]
Andy, Paul, did you have a chance to sync on ?
15:00:18 [pgearon]
15:01:17 [LeeF]
pgearon, did the WITH keyword come from existing implementations, or was it picked specifically for the current work?
15:01:42 [pgearon]
I think it was picked up specifically for this. It was one of the proposals from F2F2
15:01:43 [LeeF]
trying to figure out how to make a datasets proposal for update that makes clear that WITH affects default graph for insert/delete while FROM does default graph (merge) for WHERE
15:01:50 [LeeF]
15:02:55 [bglimm]
15:03:09 [bglimm]
are you recieving the call?
15:03:35 [LeeF]
it says connected
15:03:37 [LeeF]
but i only see myself :)
15:03:51 [bglimm]
Here it seems you didn't answer
15:03:59 [bglimm]
Can you try and call us?
15:04:03 [LeeF]
what number?
15:04:09 [bglimm]
15:04:18 [bglimm]
That's our IP
15:04:27 [LeeF]
15:04:43 [bglimm]
I hear you
15:04:52 [LeeF]
there's nobody there!
15:04:59 [LeeF]
now there is :)
15:05:42 [Zakim]
SW_(SPARQL)10:00AM has now started
15:05:49 [Zakim]
15:05:54 [Zakim]
15:05:55 [Zakim]
SW_(SPARQL)10:00AM has ended
15:05:55 [Zakim]
Attendees were [IPcaller]
15:06:25 [ivan]
ivan has joined #sparql
15:06:49 [bglimm]
It seems you have to call us for some reason
15:07:19 [LeeF]
OK, no problem
15:07:33 [LeeF]
i've written down the IP
15:07:43 [bglimm]
Can I try and call you again to see whether that is also working?
15:07:50 [LeeF]
15:07:53 [LeeF]
i'll be ready this time
15:08:12 [bglimm]
15:08:36 [bglimm]
it just says dialing and is making a dialing tone
15:08:42 [LeeF]
...i clicked "answer" and it said "connected to OERC" but then nothing else
15:08:53 [LeeF]
try one last time? i'll try a different button
15:08:58 [bglimm]
Here it says not conneted :-(
15:09:01 [bglimm]
15:09:10 [bglimm]
15:09:17 [LeeF]
no differnet here
15:09:20 [bglimm]
15:09:21 [LeeF]
let me check the instructions again
15:09:38 [bglimm]
maybe some firewall problem
15:10:24 [LeeF]
could be... i don't see anything in the instructions that would explain what's going wrong
15:10:34 [LeeF]
as long as it works with us calling you though, should be OK
15:10:45 [bglimm]
could you check your call rate?
15:10:52 [bglimm]
maybe put it on auto?
15:11:01 [LeeF]
let me see if I can find it
15:11:05 [bglimm]
1920K is our call rate
15:11:23 [bglimm]
and it is supposed to negotiate, but maybe that does not work properly
15:11:45 [AndyS]
AxelPolleres, pgearon -- current test frame works for update - need to define the update properties but action-result is OK isn't it?
15:12:08 [LeeF]
trying to find the right settings
15:12:10 [bglimm]
we'll try again with lower call rate
15:12:17 [LeeF]
i have "max upstream rate" 1152
15:12:28 [bglimm]
we were using 1900 something
15:12:36 [bglimm]
now we try with 1024
15:12:54 [bglimm]
didn't work again
15:12:54 [AxelPolleres]
Andy, will check.
15:13:27 [bglimm]
Can you call us again and if that works, we can just do it like that for the F2F?
15:13:57 [pgearon]
AndyS: I've just been concerned about defining the data to be updated. Though it may be fine to use the current method of defining data, and make all modifications on the default graph. I believe that this involves an implicit LOAD
15:13:58 [LeeF]
yup, will try in one sec
15:14:29 [pgearon]
is that what you were thinking of?
15:14:35 [bglimm]
15:14:42 [LeeF]
no idea why it doesn't work the other way, but this should do OK
15:15:06 [bglimm]
I'll disconnect now
15:15:21 [LeeF]
ok, thanks for the test
15:15:26 [bglimm]
Ok, I think we'll manage on Thursday
15:15:33 [bglimm]
You just need to call us then
15:16:02 [AndyS]
pgearon - can define the initial graph store like did for query?
15:17:19 [pgearon]
I suppose that's how QUERY worked isn't it? I should have thought of it that way. Instead, I used default-graph-uri= in all my tests to point to the data for each specific test
15:17:59 [pgearon]
that was easier than doing a LOAD, then the query, then a DROP. But for Update tests that sequence will be required
15:18:45 [pgearon]
but since QUERY was defined that way, then I suppose it works for UPDATE too. OK, I'm good with that
15:20:11 [bglimm]
bglimm has joined #sparql
15:23:55 [AndyS]
pgearon - how it's done might vary across systems - the declaration needs to say "starting with X,Y do ACTION, the result should be Z"
15:24:45 [AndyS]
In memory - i just release the dataset and GC does the rest - in a DB I use an in-memory version mostly + a little to check the persistence layer.
15:31:29 [AndyS]
Try to avoid dropping SQL tables - waaaaaayyyyyy too slow - so I clear those. In TDB I do delete DBs in testing cos it's quickish.
15:32:19 [AndyS]
For SPARQL tests of query (and update) I try to do most testing of the processors in-memory for speed -- then I run them more frequently. I sort of groan when a SQL-backed tests need running.
16:09:18 [SteveH_]
heh, it takes ages to run the SPARQL query tests in 4store, it creates a new disk-bacdked store for every test - dumb I guess
16:12:23 [AndyS]
Slow it maybe but correct it is. Last thing I want is to debug the test framework.
16:18:44 [SteveH_]
16:19:14 [SteveH_]
...though, it's an evil perl script, with a sprinkling of python, so correctness is hard to guarantee!
16:19:37 [LeeF]
I bet it was correct until the python got in there.
16:19:44 [SteveH_]
16:19:54 [kasei]
16:20:37 [Zakim]
Zakim has left #sparql
17:36:44 [AxelPolleres]
AxelPolleres has left #sparql
18:03:24 [ivan]
ivan has joined #sparql
18:39:10 [LeeF]
LeeF has joined #sparql
20:09:28 [ivan]
ivan has joined #sparql