00:25:42 <melvster> melvster has quit (Remote host closed the connection)
00:27:30 <melvster> melvster (~melvster@89.176.108.70) has joined #dig
00:35:17 <cheater_1> cheater_1 (~cheater@p57AEA79B.dip0.t-ipconnect.de) has joined #dig
00:38:38 <cheater__> cheater__ has quit (Ping timeout: 256 seconds)
01:33:42 <melvster> melvster has quit (Ping timeout: 264 seconds)
02:45:17 <bblfish> bblfish (~bblfish@AAubervilliers-651-1-117-176.w86-198.abo.wanadoo.fr) has joined #dig
02:47:15 <bblfish> bblfish has quit (Remote host closed the connection)
04:33:54 <melvster> melvster (~melvster@89.176.108.70) has joined #dig
06:23:34 <bblfish> bblfish (~bblfish@AAubervilliers-651-1-117-176.w86-198.abo.wanadoo.fr) has joined #dig
06:35:12 <deiu> deiu (~andrei@92.81.103.144) has joined #dig
06:35:12 <deiu> deiu has quit (Changing host)
06:35:12 <deiu> deiu (~andrei@unaffiliated/deiu) has joined #dig
06:59:29 <bblfish> bblfish has quit (Remote host closed the connection)
07:21:22 <jmvanel> jmvanel (~jmvanel@199.0.88.79.rev.sfr.net) has joined #dig
07:24:26 <bblfish> bblfish (~bblfish@AAubervilliers-651-1-117-176.w86-198.abo.wanadoo.fr) has joined #dig
07:25:58 <deiu> deiu has quit (Ping timeout: 245 seconds)
07:32:49 <deiu> deiu (~andrei@92.81.103.144) has joined #dig
07:32:49 <deiu> deiu has quit (Changing host)
07:32:49 <deiu> deiu (~andrei@unaffiliated/deiu) has joined #dig
07:59:14 <deiu> bblfish, timbl how does one use rdflib.js to only get the RDF statements in a graph? I want to remove all the extra info that is not RDF (like protocol/browser/context info/etc.)
08:01:15 <deiu> basically to strip all header information form the graph
09:17:30 <deiu> deiu has quit (Ping timeout: 245 seconds)
09:33:26 <cheater_1> cheater_1 has quit (Ping timeout: 240 seconds)
09:34:16 <bblfish> bblfish has quit (Remote host closed the connection)
09:34:20 <cheater__> cheater__ (~cheater@p57AEA79B.dip0.t-ipconnect.de) has joined #dig
09:39:15 <bblfish> bblfish (~bblfish@AAubervilliers-651-1-117-176.w86-198.abo.wanadoo.fr) has joined #dig
09:45:04 <bblfish> bblfish has quit (Remote host closed the connection)
09:50:23 <bblfish> bblfish (~bblfish@AAubervilliers-651-1-117-176.w86-198.abo.wanadoo.fr) has joined #dig
10:34:48 <bblfish> bblfish has quit (Remote host closed the connection)
11:08:04 <bblfish> bblfish (~bblfish@AAubervilliers-651-1-117-176.w86-198.abo.wanadoo.fr) has joined #dig
11:38:31 <RalphS> RalphS (rswick@w3cvpn1.w3.org) has joined #dig
11:39:49 <DIGlogger> DIGlogger (~dig-logge@groups.csail.mit.edu) has joined #dig
11:39:49 <barjavel.freenode.net> topic is: Decentralized Information Group @ MIT http://dig.csail.mit.edu/
11:39:49 <barjavel.freenode.net> Users on #dig: DIGlogger RalphS bblfish cheater__ jmvanel melvster presbrey timbl zuzak tyteen4a03 Yudai_ sandro slvrbckt ericP mattl
12:14:45 <timbl_> timbl_ (~timbl@host86-181-11-211.range86-181.btcentralplus.com) has joined #dig
12:14:45 <timbl> timbl has quit (Read error: Connection reset by peer)
12:14:46 <timbl_> timbl_ is now known as timbl
12:42:47 <timbl> deiu,
12:42:50 <timbl> re "bblfish, timbl how does one use rdflib.js to only get the RDF statements in a graph? I want to remove all the extra info that is not RDF (like protocol/browser/context info/etc.)"
12:43:05 <bblfish> hi
12:43:33 <timbl> sts = kb.statementsMatching(undefined, undefined, undefined, kb.sym(''htttp://ti…"));
12:43:34 <bblfish> I was going to say, it seems to me that rdflib.js is doing things wrong by putting all the metadata into the graph
12:43:50 <timbl> Not into the graph.
12:43:53 <timbl> Into the store.
12:44:07 <bblfish> ah, it makes the difference? 
12:44:19 <bblfish> ok then. 
12:44:22 <timbl> the store has many graphs.
12:44:27 <timbl> It is a quad store.
12:44:46 <bblfish> yes, I would have thought it did. I find it difficult to use. 
12:45:05 <bblfish> I think perhaps it's because the N3 serialiser does not serialise the different graphs
12:45:09 <bblfish> ?
12:45:47 <bblfish> Somehow when I last looked I never get it to show me what is in each graph.
12:46:22 <timbl> The serializer serializes   the set of statements you give it as a merged graph,  There is a separametr dump store option to serialize it all as N3 shoring the separate stores.
12:46:30 <timbl> shoeing the separate groans i mean
12:46:40 <timbl> I hate spel autocorrect
12:47:45 <bblfish> ah great. I'll look for that.
12:48:09 <timbl> It is the same as cwm, by default you query and dump the noon graph.
12:48:17 <timbl> s/noon/union/
12:48:31 <bblfish> btw, timbl have you seen https://github.com/antoniogarrote/rdfstore-js ?
12:49:23 <bblfish> Not sure what happened to antoniogarrote - he seems to have dissapeared from the map in the last month. But he seems to have take rdflib and gone quite far with it in other directions
12:49:59 <timbl> It has full sparql 1.0 
12:50:29 <timbl> It has some neat things
12:50:41 <timbl> Does it also does everything that rdflib.js does?
12:50:57 <bblfish> I am not sure how good it is on the linked data side
12:51:13 <bblfish> that's kind of what I am trying to figure out.
12:51:28 <bblfish> Adrian Gschwend from Switzerland was thinking of starting a mailing list for rdf-js 
12:51:47 <bblfish> because we seem to have efforts going on here and there, and its not clear that the groups are communicting
12:52:21 <timbl> rdf-js in general?
12:52:24 <bblfish> yes
12:52:47 <timbl> community group i wonder, instead of ddfining an api defining code
12:53:24 <bblfish> yes a community group could be one thing. I can ping you and him about it ktk@netlabs.org
12:53:50 <bblfish> for example the thing is that we need a model-view-controller framework to make life easier for JS devs
12:54:22 <bblfish> antoniogarrotte wrote one https://github.com/antoniogarrote/semantic-ko
12:54:49 <bblfish> ( the examples don't work just out of the box on his site - but if you look at the issue DB you'll see I posted the fix there )
12:55:48 <bblfish> ( github is restricting JS access to their site )
12:56:15 <bblfish> so there is no group one can go to currently where people can point one to the right JS lib.
12:57:23 <bblfish> or where specialist JS devs can point us to what is missing. 
13:00:02 <bblfish> I think he was thinking of putting the group somewhere else than community groups at w3c, as it is a bit ( more complicted than usual ) to subscribe there.
13:00:22 <bblfish> but perhaps if enough people here turn out there, then that would be good enough 
13:00:24 <bblfish> :-)
13:05:13 <bblfish> At http://stample.co/ we have a lot of frontend code written with backbone.js, so I need to see how I can hack that to make it work with a graph store.
13:09:29 <bblfish> DIGlogger, pointer?
13:09:29 <bblfish> See http://dig.csail.mit.edu/irc/dig/2013-09-16#T13-09-29
13:14:50 <deiu> deiu (~andrei@unaffiliated/deiu) has joined #dig
13:15:12 <deiu> bblfish, ping
13:17:31 <bblfish> hi, it looks like timbl answered your question above deiu. Look here: ttp://dig.csail.mit.edu/irc/dig/2013-09-16#T13-09-29
13:17:40 <bblfish> sorry http://dig.csail.mit.edu/irc/dig/2013-09-16#T13-09-29
13:19:22 <bblfish> deiu: I took your code from rww.io and got Amelie from stample to put it on our clone of rdflib https://github.com/stample/rdflib.js/tree/master/example/ldp
13:19:49 <bblfish> the idea is to turn that into an rdflib example that would be fully dynamic
13:20:03 <timbl_> timbl_ (~timbl@host86-181-11-211.range86-181.btcentralplus.com) has joined #dig
13:20:41 <bblfish> a fully dynamic js based LDP/file explorer exaample
13:20:47 <deiu> I see
13:20:50 <deiu> that would be nice, yes
13:21:01 <bblfish> Amelie is going to work on it a little bit to get the hang of RDF.
13:21:08 <bblfish> It's mostly just taking your code.
13:21:09 <timbl> timbl has quit (Ping timeout: 248 seconds)
13:21:09 <timbl_> timbl_ is now known as timbl
13:21:31 <bblfish> Then I can see if rww.io's code transformed to rdflib can work on RWW-Play without change
13:22:00 <bblfish> so that would be a way of testing RWW-Play https://github.com/stample/rww-play
13:22:50 <bblfish> the fun thing is that one could have links from one LDPC on rww.io to another LDPC on another server.
13:23:02 <bblfish> and the JS should be able to just follow along
13:24:13 <deiu> yes, that's what I'm looking for myself (distributed LDPC)
13:24:27 <deiu> and thanks timbl for the example
13:25:34 <bblfish> mind you looking at the ontologies used for the file system they seem ripe for some improvements. For example rww.io gives the size of directories in bytes which is a bit odd I think.
13:26:00 <bblfish> as far as virtual resources giving the size in bytes is going to be a bit difficult for RDF as different representations could have different sizes
13:26:14 <bblfish> perhaps a size in triples would be more useful.
13:26:52 <bblfish> ( but the same would be true even for images, as with content Neg the sizes of resources will be different )
13:27:22 <deiu> well, rww.io provides an accurate representation of the filesystem
13:27:49 <deiu> there are no virtual resources hosted on it
13:28:09 <bblfish> yes, but you're not meant to know or care as a client what the database behind it is
13:28:23 <deiu> you don't :)
13:28:32 <deiu> you just use REST to manage the resources
13:29:34 <bblfish> Yes, so say you have 3 different representations. .rdf .n3 .ttl and you patch one, don't you want all the others to be updated too?
13:29:41 <bblfish> or do we have to patch all three?
13:30:01 <bergi> bergi (~bergi@static.132.160.46.78.clients.your-server.de) has joined #dig
13:30:41 <bblfish> anyway, the point is that specifying bytes on content negotiated resources is a good quick thing to do, but longer term it's a bit naïve
13:31:56 <deiu> no, you don't have three representations on rww.io
13:32:10 <bblfish> yes, but I do on rwwPlay
13:32:14 <deiu> ah
13:32:27 <bblfish> in fact I wrote it so that it can return any representation
13:32:36 <deiu> yes, that's preferred 
13:32:37 <betehess> betehess (~betehess@2001:470:8b2d:804:6152:e6d2:d728:d478) has joined #dig
13:33:07 <bblfish> well, it's good we're implementing and discussing these things. This is how we can work out what the issues are :-)
13:33:07 <deiu> BTW, you could create the document with a specific extension on the fly
13:33:20 <deiu> after conneg
13:33:24 <bblfish> which document?
13:33:42 <deiu> doc.ttl / doc.n3 / etc
13:33:53 <bblfish> we do create them on the fly
13:33:59 <bblfish> we only store .ttl
13:34:09 <deiu> then why do you need to update three documents?
13:34:30 <deiu> <bblfish> Yes, so say you have 3 different representations. .rdf .n3 .ttl and you patch one, don't you want all the others to be updated too?
13:34:33 <bblfish> we don't. But it makes no sense to say that </2013/card> size 1023 .
13:35:01 <deiu> it makes for rww.io, since it's basically a filesystem
13:35:10 <bblfish> so is rwwPlay
13:35:17 <bblfish> it's built on a file system
13:35:28 <deiu> for example, it helps with quota management
13:35:53 <bblfish> well it's just a modelling problem deiu
13:36:21 <deiu> do you have a use case where you need to know the number of triples?
13:36:44 <bblfish> <> a ldp:Container; ldp:created <m1> . <m1> reperesentation <m1.ttl> . <m1.ttl> mime "text/turtle"; size 1023 .
13:36:45 <deiu> BTW, rww.io returns the number of triples it stores after a POST/PUT
13:36:50 <deiu> ah
13:37:18 <bblfish> so then you can specify the types of each representation.
13:37:25 <bblfish> and the sizes if you wish.
13:38:08 <deiu> wouldn't it add a nasty overhead if you open each resource document to read the number of triples?
13:38:11 <bblfish> though that does not make much for quota management, because those three representations are not using more space on disk.
13:39:19 <bblfish> well you could just put the size of the ttl doc if you want.
13:39:33 <deiu> that's what I currently do
13:40:39 <bblfish> The size relation is not defined as "the size of the turtle version of the file"
13:41:00 <bblfish> so for you it works because you are only serving ttl files
13:41:23 <bblfish> on the other hand your hash urls are going to be <card.ttl#me> 
13:41:44 <bblfish> which means your identifiers are going to be tied to turtle
13:42:15 <deiu> no no
13:42:24 <bblfish> ah?
13:42:24 <deiu> I don't use extensions for RDF resources
13:42:44 <deiu> if you go to rww.io, you'll see that RDF documents are listed as "RDF"
13:43:00 <deiu> whereas doc.ttl will be listed as "text/turtle"
13:43:12 <bblfish> rww.io does not stop you though from doing this http://stample.rww.io/2013/
13:43:30 <deiu> of course not
13:43:40 <bblfish> let me see
13:43:45 <deiu> but if you want to have a generic RDF document, you should not add .ttl
13:43:54 <deiu> then you just have to do conneg 
13:45:35 <bblfish> $ cwm http://stample.rww.io/2013/
13:45:35 <bblfish> …
13:45:35 <bblfish> <test>     a <http://www.w3.org/2000/01/rdf-schema#Resource>;
13:45:36 <bblfish>          stat:mtime 1379339055;
13:45:38 <bblfish>          stat:size 114 .
13:45:59 <bblfish> so you are saying that <test> has size 114 .
13:46:07 <deiu> yes, the resource
13:46:10 <bblfish> but if I get the RDF/XML version
13:47:25 <deiu> it should say the same
13:48:30 <bblfish> $ curl -i -H "Accept: application/rdf+xml" http://stample.rww.io/2013/test
13:48:30 <bblfish> HTTP/1.1 200 OK
13:48:30 <bblfish> Date: Mon, 16 Sep 2013 13:48:07 GMT
13:48:31 <bblfish> Server: Apache/2.2.22 (Ubuntu)
13:48:33 <bblfish> X-Powered-By: PHP/5.3.10-1ubuntu3.8
13:48:35 <bblfish> Set-Cookie: SID=klv3iud3f70irvjb8im6t1i305; expires=Sat, 15-Sep-2018 13:48:07 GMT; path=/; domain=.rww.io
13:48:37 <bblfish> Expires: Thu, 19 Nov 1981 08:52:00 GMT
13:48:39 <bblfish> Cache-Control: no-store, no-cache, must-revalidate, post-check=0, pre-check=0
13:48:41 <bblfish> Pragma: no-cache
13:48:43 <bblfish> Set-Cookie: showMetaFiles=1; expires=Wed, 16-Oct-2013 13:48:07 GMT
13:48:45 <bblfish> User: dns:86.198.32.176
13:48:47 <bblfish> Filename: /www/rww.io/data/stample.rww.io/2013/test
13:48:49 <bblfish> Link: <http://stample.rww.io/2013/.acl.test>; rel=acl
13:48:51 <bblfish> MS-Author-Via: DAV, SPARQL
13:48:53 <bblfish> ETag: "1379339055-e2cecaf480021cf333dc080130e46e2a"
13:48:55 <bblfish> Updates-Via: ws://stample.rww.io:81
13:48:57 <bblfish> Triples: 1
13:49:00 <bblfish> Content-Length: 189
13:49:01 <bblfish> Content-Type: application/rdf+xml
13:49:03 <bblfish> there it says 189
13:49:05 <bblfish> ...
13:49:07 <bblfish> There it says Content-Length: 189
13:49:27 <deiu> ah..yes
13:49:44 <bblfish> But well you just mean: this resource is taking up so much space on my disk
13:49:51 <deiu> nope
13:50:23 <deiu> content-length just says that the data you GET (based on conneg) is 189
13:50:33 <bblfish> I know that
13:50:55 <bblfish> so your stat:size is saying: this resource takes so much space on our disk
13:50:55 <deiu> if you want to store the data as rdf+xml, it will then take 189 too
13:51:11 <deiu> yes, I agree there's a missing piece of info
13:51:25 <bblfish> well I can kind of see the point. 
13:51:49 <deiu> maybe I can somehow add a triple saying that a given resource is serialized as turtle
13:52:20 <bblfish> well then you'd have the same problem
13:52:25 <bblfish> because if you do 
13:52:35 <bblfish> <card> mime "text/turtle" .
13:52:46 <bblfish> then presumably you may want to also tell the user 
13:52:52 <bblfish> <card> mime "application/rdf+xml"
13:53:01 <bblfish> what you really want to do then is 
13:53:40 <bblfish> <card> representation [ mime "text/turtle"; size 114 ], [ mime "rdf/xml" ] .
13:53:55 <bblfish> ( wihtout for example giving the size of the rdf/xml cause you don't care .
13:54:11 <deiu> well, then why do it at all?
13:54:22 <bblfish> to let the user know you serve it up?
13:54:32 <deiu> the user can do a HEAD or OPTIONS 
13:54:39 <bblfish> then don't do it.
13:54:54 <deiu> :)
13:54:58 <bblfish> I am speaking about modelling the size
13:55:18 <bblfish> if you model the size on <card> then it means something liek: this is taking so much space on my disk
13:55:36 <deiu> yes, I need to think about it
13:55:40 <bblfish> if you do it on a representation, then it means: this representation si so and so big
13:55:41 <deiu> but you're right
13:55:58 <bblfish> s/mean/can mean/
13:56:34 <deiu> I need to run now, got a train to catch
13:56:41 <deiu> I'll think about it
13:56:52 <bblfish> ok. Me too. 
13:56:58 <deiu> I might just remove it for now, until we can think of a better way to model it
13:57:17 <deiu> talk to you soon!
13:57:40 <bblfish> anyay, if size means the resource has that size, then perhaps it's makes more sense than I thought. 
13:57:51 <bblfish> I suppose for people in DB land this would be difficult to calculate.
13:58:01 <bblfish> which is where counting in triples may be better.
13:59:17 <bblfish> Same with people using rdf stores. I suppose we'll end up supporting both sizes in bytes, and in triples.
14:03:26 <deiu> deiu has quit (Ping timeout: 240 seconds)
14:10:08 <bblfish> timbl: Adrian at Netlabs sent us an answer. He seems to be willing to put the energy into this. Adrian works with Bart Van Leeuwen the firefighter who uses RDF ( and is based in holland )
14:10:51 <bblfish> presbrey: would you be interested in a JS-RDF community group?
14:38:27 <deiu> deiu (~andrei@unaffiliated/deiu) has joined #dig
14:43:12 <deiu> deiu has quit (Quit: Leaving)
14:50:52 <timbl> timbl has quit (Ping timeout: 264 seconds)
15:02:12 <timbl> timbl (~timbl@host86-181-11-211.range86-181.btcentralplus.com) has joined #dig
15:24:36 <RalphS> RalphS has quit ()
15:49:47 <RalphS> RalphS (rswick@w3cvpn3.w3.org) has joined #dig
18:01:56 <timbl> timbl has quit (Quit: timbl)
18:40:20 <timbl> timbl (~timbl@host86-181-11-211.range86-181.btcentralplus.com) has joined #dig
20:20:27 <RalphS> RalphS has quit ()
20:39:41 <timbl> timbl has quit (Quit: timbl)
21:00:28 <jorash> jorash (A@unaffiliated/jorash) has joined #dig
21:57:46 <bblfish> bblfish has quit (Remote host closed the connection)
22:34:04 <jmvanel> jmvanel has quit (Ping timeout: 264 seconds)