IRC log of dig on 2012-06-07
Timestamps are in UTC.
- 01:25:20 [manu1]
- manu1 has quit (Remote host closed the connection)
- 02:09:31 [timbl]
- timbl (~timbl@c-24-62-225-11.hsd1.ma.comcast.net) has joined #dig
- 02:24:12 [manu1]
- manu1 (~chatzilla@pool-96-240-176-35.ronkva.east.verizon.net) has joined #dig
- 02:38:30 [github]
- github (~github@sh2.rs.github.com) has joined #dig
- 02:38:30 [github]
- [tabulator] timbl pushed 1 new commit to master: http://git.io/vMfUYw
- 02:38:30 [github]
- [tabulator/master] Now (a) more careful only to delete triples which are in the store, and (b) cleans up reduntant supertype triples. - Tim Berners-Lee
- 02:38:30 [github]
- github has left #dig
- 02:57:20 [rszeno]
- rszeno has quit (Quit: Leaving.)
- 07:01:55 [nunnun]
- nunnun is now known as nunnun_away
- 08:20:55 [deiu]
- deiu (~andrei@157.159.103.108) has joined #dig
- 08:20:55 [deiu]
- deiu has quit (Changing host)
- 08:20:55 [deiu]
- deiu (~andrei@unaffiliated/deiu) has joined #dig
- 08:44:15 [melvster]
- melvster (~melvin@p4FF9782A.dip.t-dialin.net) has joined #dig
- 10:10:07 [cheater_]
- cheater_ (~cheater@g229021189.adsl.alicedsl.de) has joined #dig
- 10:10:55 [cheater_]
- cheater_ has quit (Max SendQ exceeded)
- 10:12:28 [cheater__]
- cheater__ has quit (Ping timeout: 260 seconds)
- 10:12:33 [cheater]
- cheater (~cheater@g229021189.adsl.alicedsl.de) has joined #dig
- 10:13:21 [cheater]
- cheater has quit (Max SendQ exceeded)
- 10:15:31 [cheater]
- cheater (~cheater@g229021189.adsl.alicedsl.de) has joined #dig
- 10:23:30 [rszeno]
- rszeno (~rszeno@79.114.41.64) has joined #dig
- 10:56:55 [deiu]
- Anyone around with experience using virtuoso?
- 11:12:51 [RalphS]
- RalphS (Ralph@30-7-118.wireless.csail.mit.edu) has joined #dig
- 11:59:39 [timbl]
- timbl has quit (Ping timeout: 246 seconds)
- 13:00:03 [timbl]
- timbl (~timbl@31-34-205.wireless.csail.mit.edu) has joined #dig
- 13:19:48 [deiu]
- Does anyone know if there's a way of performing a SPARQL SELECT followed by an INSERT?
- 13:52:53 [Pipian_]
- Pipian_ (~pipian@18.111.100.155) has joined #dig
- 14:09:50 [Pipian_]
- Pipian_ has quit (Quit: Pipian_)
- 14:32:17 [nunnun_away]
- nunnun_away is now known as nunnun
- 15:26:01 [Pipian_]
- Pipian_ (~pipian@31-34-151.wireless.csail.mit.edu) has joined #dig
- 15:30:20 [deiu]
- deiu has quit (Ping timeout: 246 seconds)
- 15:49:32 [bblfish]
- deiu, is that not an UPDATE?
- 15:49:55 [bblfish]
- I have not used this heavily, so I really don't know
- 17:18:30 [deiu]
- deiu (~andrei@2a01:e35:8b67:4160:50b1:3b7c:6b82:3ea6) has joined #dig
- 17:18:41 [deiu]
- deiu has quit (Changing host)
- 17:18:41 [deiu]
- deiu (~andrei@unaffiliated/deiu) has joined #dig
- 17:24:10 [bblfish]
- betehess: does the following improvement to RDFReaderFactory make sense? https://gist.github.com/2890174
- 17:24:27 [bblfish]
- (perhaps it should be called apply() )
- 17:35:20 [betehess]
- ok with apply
- 17:35:29 [betehess]
- maybe it should be a partial function
- 17:35:52 [betehess]
- I believe this is why I didn't provide the function in the first place
- 17:36:08 [betehess]
- we shouldn't lie in our type :-)
- 17:36:18 [bblfish]
- ah yes
- 17:36:31 [bblfish]
- It's tricky with those return types
- 17:36:34 [betehess]
- or if you prefer to make total, we could return an Option
- 17:36:48 [betehess]
- or even provide 2 flavours
- 17:36:59 [betehess]
- one total function, and one partial one
- 17:37:13 [betehess]
- hrmmm
- 17:37:21 [bblfish]
- yes, the option one looks good
- 17:37:30 [bblfish]
- with apply
- 17:37:41 [bblfish]
- def apply(ser: RDFSerialization): Option[RDFReader[Rdf,_]]
- 17:37:43 [betehess]
- now that I think about it, it should be ok to provide the implementation for all the RDFSerialization, for both Jena and Sesame
- 17:37:55 [betehess]
- so we could go directly with apply as a total function
- 17:38:00 [betehess]
- which would be ideal
- 17:38:22 [bblfish]
- ah yes, I see you mean that those that don't have a serialisation just return None for those
- 17:38:40 [betehess]
- and because this is feasible with no big effort, you can go directly with apply, and we'll adapt later the other implementations
- 17:38:53 [betehess]
- so apply with direct type (no Option)
- 17:39:04 [bblfish]
- so you mean that def RDFXMLReader: Option[RDFReader[Rdf, RDFXML]]
- 17:39:09 [betehess]
- no no
- 17:39:12 [bblfish]
- ah ok
- 17:39:25 [betehess]
- just s/ser/apply/ in your gist :-)
- 17:39:37 [betehess]
- ah, can you please put the return type explicitly as well?
- 17:39:46 [bblfish]
- yes
- 17:39:53 [betehess]
- I don't want Scala to infer an intersection type :-)
- 17:40:28 [bblfish]
- https://gist.github.com/0a0b0bffce0c9ec1217f
- 17:40:34 [betehess]
- inference and subtyping put together have some weird results if you don't pay attention...
- 17:40:43 [bblfish]
- but with the others filled in I suppose
- 17:40:56 [betehess]
- but...
- 17:41:00 [betehess]
- this _ is weird
- 17:41:04 [betehess]
- hooo
- 17:41:05 [betehess]
- I remember
- 17:41:15 [betehess]
- we would need something like dependent types
- 17:41:25 [betehess]
- that was my issue :-)
- 17:41:34 [bblfish]
- ah I thought it might be
- 17:41:40 [betehess]
- there is another way though
- 17:41:41 [bblfish]
- I wonder if we're putting types in too many places
- 17:41:46 [betehess]
- let me give you a gist
- 17:41:53 [betehess]
- (never)
- 17:43:09 [betehess]
- overloading should be enough https://gist.github.com/0c956028ea5d083d87c3#comments
- 17:44:17 [betehess]
- don't know if you'll have problems with erasure there...
- 17:44:52 [betehess]
- worst case, the current thing is not that bad :-)
- 17:45:11 [bblfish]
- It's just that I want to have actors that can write to the hard drive in some format
- 17:45:19 [betehess]
- yes ?
- 17:45:35 [bblfish]
- and so I want to be able to get the right writer ( and reader ) given a format I don't know ahead of time
- 17:45:37 [betehess]
- we need to make sure that the framework supports the format
- 17:45:47 [betehess]
- those are different issues
- 17:45:58 [bblfish]
- so I could write a match sequence of course
- 17:46:00 [betehess]
- you have two different pieces
- 17:46:11 [betehess]
- String => Option[RDFSerialization]
- 17:46:23 [betehess]
- where you may not know how to map it
- 17:46:29 [bblfish]
- yes, I had supposed that was alrady sorted
- 17:46:33 [betehess]
- no
- 17:46:41 [betehess]
- these are very different things
- 17:46:59 [bblfish]
- I just wanted that apply( ) method you just wrote :-)
- 17:47:31 [bblfish]
- I think erasure is only with generics, but I'll see...
- 17:47:32 [betehess]
- this is a mistake in jena and sesame api designs
- 17:47:32 [betehess]
- especially in our case, as the format is coming from the outside
- 17:47:33 [betehess]
- ok, but that's not a ReaderFactory
- 17:47:59 [betehess]
- well, you won't gain that much
- 17:48:05 [betehess]
- you want the String thing
- 17:48:13 [bblfish]
- ReaderFactory(Turtle)
- 17:48:20 [betehess]
- it should be part of your implementation
- 17:48:50 [betehess]
- you can offer somewhere a Map[String, RDFSerialization] that your application supports
- 17:48:51 [bblfish]
- I think I need both.
- 17:49:11 [bblfish]
- There are an infinnite ways to map strings to things (mime-types, Jena strings, ...)
- 17:49:29 [betehess]
- you need all of them?
- 17:49:35 [betehess]
- mime-type should be enough
- 17:49:40 [betehess]
- as it's coming from the outside
- 17:49:45 [bblfish]
- no, I'll provide Map[String, RDFSerialization] myself
- 17:49:52 [betehess]
- for the rest, just use the right type
- 17:49:53 [betehess]
- yes
- 17:50:07 [bblfish]
- then I can use the apply() method above
- 17:50:16 [bblfish]
- to get the correct writer and reader
- 17:50:31 [betehess]
- in this case, you'll need *one* apply
- 17:50:38 [betehess]
- which does the pattern matching
- 17:50:44 [betehess]
- so you'll need the total function
- 17:51:12 [betehess]
- and I guess you'll need some dependent method type here
- 17:51:45 [betehess]
- worst case, you do something nasty (but ok) with an asInstanceOf
- 17:52:19 [bblfish]
- is it really improtant to have the type of the serialisation in the object?
- 17:52:23 [betehess]
- the apply method should be just written in term of the other functions
- 17:52:31 [betehess]
- you won't do much without it
- 17:52:52 [betehess]
- you'll never be able to compose things together
- 17:53:01 [betehess]
- Scala won't even let you
- 17:54:56 [bblfish]
- a full function may be difficult, because some frameworks may not have all the serialisations. They may not have RDFXMLAbbrev for example
- 17:55:57 [bblfish]
- RDFXMLAbbrev of course could be a subclass of RDFXML
- 17:58:57 [bblfish]
- in fact I think one could argue that serialisations of RDF should not be a sealed class
- 17:59:15 [bblfish]
- it will always be possible to have new serialisations appear
- 17:59:37 [bblfish]
- and one should not make it difficult for people to produce new serialisation plugins
- 18:00:48 [bblfish]
- Json-ld for example is one serilisation. and pretty much all XML could be read into rdf xml using GRDDL
- 18:02:11 [bblfish]
- It's a bit the same as with akka actors not having strongly typed messages
- 19:13:40 [timbl]
- tabulator.org seems to have changed its ssh key ad no long has mine
- 19:14:03 [timbl]
- I wonder whether presbrey has upgraded it or something
- 19:14:19 [Pipian_]
- Pipian_ has quit (Ping timeout: 245 seconds)
- 19:15:27 [manu-db]
- manu-db has quit (Remote host closed the connection)
- 19:15:29 [Pipian_]
- Pipian_ (~pipian@31-34-151.wireless.csail.mit.edu) has joined #dig
- 19:22:55 [deiu]
- deiu has quit (Ping timeout: 264 seconds)
- 19:23:43 [Pipian__]
- Pipian__ (~pipian@30-9-35.wireless.csail.mit.edu) has joined #dig
- 19:25:44 [Pipian_]
- Pipian_ has quit (Ping timeout: 252 seconds)
- 20:05:09 [manu-db]
- manu-db (~msporny@digitalbazaar.com) has joined #dig
- 20:21:23 [melvster1]
- melvster1 (~melvin@p4FF96533.dip.t-dialin.net) has joined #dig
- 20:22:02 [RalphS]
- RalphS has quit ()
- 20:22:24 [melvster]
- melvster has quit (Ping timeout: 246 seconds)
- 20:33:34 [bblfish]
- betehess: just created a file using PUT and akka
- 20:38:11 [bblfish]
- and was able to read it again with a GET :-)
- 20:39:11 [mattl]
- last.fm, linkedin and more all recommending people change their passwords.
- 20:45:41 [melvster1]
- wow
- 20:46:04 [mattl]
- heh, i justed a quote from you, melvster1 -- working on my resume.
- 20:46:13 [melvster1]
- you did? :)
- 20:47:02 [mattl]
- trying to put in some stuff about GNU social and you did a write up for the final report of the FSW incubator
- 20:47:09 [melvster1]
- ahhh
- 20:47:41 [melvster1]
- what's the status of GNU social now?
- 20:48:02 [mattl]
- we're expecting to merge with StatusNet
- 20:48:10 [mattl]
- and continue under the GNU social name
- 20:48:19 [melvster1]
- https://my-profile.eu/ is turning into a really nice standards oriented social net
- 20:48:37 [melvster1]
- deiu, who's normally here during the daytime, is the developer
- 20:48:41 [mattl]
- ah nice
- 20:49:03 [melvster1]
- actually he was doing the status.net/identi.ca integration yesterday :)
- 20:49:08 [mattl]
- aha
- 20:49:31 [bblfish]
- ah deiu is working on that?
- 20:49:31 [mattl]
- i need to look more at this
- 20:49:42 [melvster1]
- sure, hi bblfish
- 20:49:52 [bblfish]
- betehess: so here is the code for the aktor https://github.com/bblfish/Play20/blob/693109a2192356bec5cc9591e9eb2e83ddaab274/framework/src/webid/src/main/scala/org/w3/readwriteweb/play/ResourceManager.scala
- 20:50:14 [bblfish]
- I suppose he's doing some webid work there?
- 20:50:25 [melvster1]
- webid is supported yes
- 20:50:31 [bblfish]
- on status net?
- 20:50:41 [melvster1]
- no ... on my profile
- 20:51:01 [melvster1]
- csarven tried to get webid into status net ... but it was probably a bit ahead of the curve at that point
- 20:51:33 [bblfish]
- ah yes. my-profile certainly shows the good side of it
- 20:51:43 [melvster1]
- webid of course becomes more important after all these password leaks
- 20:51:51 [bblfish]
- yes. duh
- 20:52:14 [melvster1]
- in the next days my profile will have a friend based wall ... so it will be very much like facebook in functionality but user centric
- 20:52:40 [bblfish]
- ah, he's gaining a huge advantage. I have to work hard to catch up
- 20:52:41 [bblfish]
- ;-)
- 20:52:57 [melvster1]
- friendly competition :)
- 20:53:07 [bblfish]
- indeed
- 20:53:12 [melvster1]
- mattl: have you seen that ostatus is a bit stalled right now?
- 20:54:31 [bblfish]
- betehess: if you look at line 97 you'll see I implemented the find(serialisation) on ReaderFactory. Things compile ok in fact with [Rdf,_] type
- 20:54:56 [timbl]
- timbl has quit (Quit: timbl)
- 21:09:48 [bblfish]
- betehess: added patch to banana, but not completely stuck on it https://github.com/w3c/banana-rdf/commit/8024443ec1c80069c2cf86e07d8e6323db59ed0b
- 21:49:11 [Pipian__]
- Pipian__ has quit (Quit: Pipian__)
- 23:36:43 [melvster1]
- melvster1 has quit (Remote host closed the connection)
- 23:45:30 [bblfish_]
- bblfish_ (~bblfish@AAubervilliers-651-1-325-4.w83-200.abo.wanadoo.fr) has joined #dig
- 23:45:42 [bblfish]
- bblfish has quit (Ping timeout: 252 seconds)
- 23:53:57 [bblfish_]
- bblfish_ has quit (Ping timeout: 248 seconds)