Summary of IRC Meeting in #apachemarmotta at Tue Feb 11 11:00:43 2014: Attendees: tkurz, aperez, jfrank, wikier
- Preface - infrastructure - ldp specs - Action: discuss if Marmotta LDP will support binary resources or not - Link: https://dvcs.w3.org/hg/ldpwg/raw-file/default/ldp.html - Link: https://dvcs.w3.org/hg/ldpwg/raw-file/default/images/ldpc1.png - aob IRC log follows: ## Preface ## [Tue Feb 11 11:00:57 2014] <wikier>: well, just let's await everybody joins the chat room [Tue Feb 11 11:01:14 2014] <wikier>: Nandana said by mail he can't do it today [Tue Feb 11 11:01:33 2014] <tkurz>: hi [Tue Feb 11 11:01:40 2014] <wikier>: and here we have jfrank, tkurz and aperez [Tue Feb 11 11:01:57 2014] <aperez>: hehehe [Tue Feb 11 11:02:12 2014] <aperez>: all unknown people [Tue Feb 11 11:02:13 2014] <aperez>: xD [Tue Feb 11 11:02:24 2014] <wikier>: xD [Tue Feb 11 11:03:21 2014] <wikier>: jfrank? [Tue Feb 11 11:03:37 2014] <wikier>: I don't expect someone else [Tue Feb 11 11:03:46 2014] <jfrank>: well then, let's go [Tue Feb 11 11:03:53 2014] <wikier>: ok [Tue Feb 11 11:03:55 2014] <wikier>: first topic ## infrastructure ## [Tue Feb 11 11:04:05 2014] <wikier>: as you saw [Tue Feb 11 11:04:11 2014] <wikier>: and jfrank is not happy with that [Tue Feb 11 11:04:15 2014] <wikier>: in the ldp branch [Tue Feb 11 11:04:22 2014] <wikier>: I created an experimental module [Tue Feb 11 11:04:26 2014] <wikier>: marmotta-ldp [Tue Feb 11 11:04:48 2014] <wikier>: my idea is to implement everything there, and keep it separated until it becomes stable [Tue Feb 11 11:04:49 2014] <jfrank>: (which does not compile btw) [Tue Feb 11 11:04:55 2014] <wikier>: both spec and implementation [Tue Feb 11 11:04:59 2014] <wikier>: does not compile, yes [Tue Feb 11 11:05:03 2014] <wikier>: sorry for that [Tue Feb 11 11:05:06 2014] <wikier>: anyway [Tue Feb 11 11:05:20 2014] <wikier>: and, in the the end, it should be merged into marmotta-core, of course [Tue Feb 11 11:05:30 2014] <wikier>: jfrank: what else do you prefer? [Tue Feb 11 11:05:33 2014] <jfrank>: so why do we have a separate branch then? [Tue Feb 11 11:06:08 2014] <wikier>: because I expect to finish the sprint, merge the branch back to develop, and ldp would be still experimental [Tue Feb 11 11:06:26 2014] <wikier>: maybe eben part of the 3.2.0 release [Tue Feb 11 11:06:33 2014] <wikier>: no idea actually [Tue Feb 11 11:07:34 2014] <wikier>: s/eben/even [Tue Feb 11 11:07:56 2014] <wikier>: ldp is planned for summer [Tue Feb 11 11:07:57 2014] <jfrank>: ldp is probably earlier done than the 3.2 release [Tue Feb 11 11:08:08 2014] <wikier>: probably [Tue Feb 11 11:08:12 2014] <wikier>: implemented [Tue Feb 11 11:08:23 2014] <wikier>: yes [Tue Feb 11 11:08:27 2014] <jfrank>: also the spec [Tue Feb 11 11:08:34 2014] <wikier>: really? [Tue Feb 11 11:08:44 2014] <wikier>: well, off-topic [Tue Feb 11 11:09:11 2014] <wikier>: I could merge the marmotta-ldp module into marmotta-core if ti makes you happier [Tue Feb 11 11:09:19 2014] <wikier>: but I'd prefer to keep it there so far [Tue Feb 11 11:09:23 2014] <jfrank>: no, lets keep it separate [Tue Feb 11 11:09:25 2014] <wikier>: other opinions? [Tue Feb 11 11:09:46 2014] <jfrank>: maybe later move the current ld-services out into a separate module [Tue Feb 11 11:10:11 2014] <wikier>: maybe [Tue Feb 11 11:10:19 2014] <wikier>: which are a mess [Tue Feb 11 11:10:26 2014] <tkurz>: it is cleaner to keep it seperate for now [Tue Feb 11 11:10:32 2014] <wikier>: I'd expect a whole refactoring based on the ldp work [Tue Feb 11 11:11:25 2014] <wikier>: ok, agreement on that [Tue Feb 11 11:11:41 2014] <wikier>: any other infrastructure issue to discuss? [Tue Feb 11 11:13:11 2014] <jfrank>: not from my side [Tue Feb 11 11:13:13 2014] <wikier>: so far, I think that's it [Tue Feb 11 11:13:25 2014] <wikier>: next topic [Tue Feb 11 11:13:29 2014] <wikier>: #ldp specs ## ldp specs ## [Tue Feb 11 11:13:44 2014] <wikier>: I've been re-reading the specs [Tue Feb 11 11:14:04 2014] <wikier>: I have the same feeling that reading the wg minutes and mailing list [Tue Feb 11 11:14:18 2014] <wikier>: ldp has become something a bit strange in some parts [Tue Feb 11 11:14:31 2014] <jfrank>: quite a bit [Tue Feb 11 11:14:33 2014] <wikier>: without actually further evolving the initial use case [Tue Feb 11 11:14:42 2014] <wikier>: it'd be funny to implement some parts [Tue Feb 11 11:15:37 2014] <jfrank>: let's start with the good news... [Tue Feb 11 11:15:51 2014] <jfrank>: LDP Client should be rather straight forward [Tue Feb 11 11:15:53 2014] <tkurz>: but it is a recommendation. that gives us the freedom to let some things out [Tue Feb 11 11:16:27 2014] <tkurz>: and as soon as we will have implemented the stuff we will have created facts [Tue Feb 11 11:17:06 2014] <jfrank>: true - the minimal requirements are not so heavy [Tue Feb 11 11:17:37 2014] <jfrank>: but as soon you want to have more than that it gets messy [Tue Feb 11 11:17:40 2014] <wikier>: well, weak recommendations are always problematic [Tue Feb 11 11:18:11 2014] <wikier>: everything that let's you freedom, in the end it's because it's not well specified [Tue Feb 11 11:18:27 2014] <wikier>: but in the end the conformance tests are not either [Tue Feb 11 11:18:39 2014] <wikier>: so, yes, the core stuff should not be a bit deal at all [Tue Feb 11 11:19:38 2014] <wikier>: any suggestion where can we start? [Tue Feb 11 11:20:18 2014] <wikier>: I mean, not so easy to split for the moment [Tue Feb 11 11:20:19 2014] <wikier>: so I expected to do pair-programming with jfrank [Tue Feb 11 11:20:31 2014] <wikier>: at least to bootstrap the impl [Tue Feb 11 11:24:30 2014] <tkurz>: so let's start with the things that are clear [Tue Feb 11 11:24:35 2014] <wikier>: goof, tkurz_ is back [Tue Feb 11 11:24:41 2014] <wikier>: s/goof/good [Tue Feb 11 11:25:00 2014] <wikier>: #link https://dvcs.w3.org/hg/ldpwg/raw-file/default/ldp.html [Tue Feb 11 11:26:08 2014] <wikier>: #link https://dvcs.w3.org/hg/ldpwg/raw-file/default/images/ldpc1.png [Tue Feb 11 11:26:26 2014] <wikier>: that's the current ldp hierarchy [Tue Feb 11 11:27:28 2014] <jfrank>: looks reasonable... [Tue Feb 11 11:27:50 2014] <jfrank>: something you like to comment on that, wikier? [Tue Feb 11 11:28:26 2014] <wikier>: what? [Tue Feb 11 11:28:31 2014] <wikier>: I guess I lost something [Tue Feb 11 11:29:05 2014] <wikier>: this the current ldp hierarchy: [Tue Feb 11 11:29:06 2014] <wikier>: LDP Resource [Tue Feb 11 11:29:06 2014] <wikier>: | [Tue Feb 11 11:29:06 2014] <wikier>: |-- LDP Binary Resource [Tue Feb 11 11:29:06 2014] <wikier>: | [Tue Feb 11 11:29:06 2014] <wikier>: |-- LDP RDF Resource [Tue Feb 11 11:29:08 2014] <wikier>: | [Tue Feb 11 11:29:10 2014] <wikier>: |-- LDP Container [Tue Feb 11 11:29:12 2014] <wikier>: | [Tue Feb 11 11:29:14 2014] <wikier>: |-- LDP Basic Container [Tue Feb 11 11:29:16 2014] <wikier>: |-- LDP Direct Container [Tue Feb 11 11:29:18 2014] <wikier>: |-- LDP Indirect Container [Tue Feb 11 11:29:40 2014] <wikier>: the first question for me is: are we going to implement binary resources? i.e., content [Tue Feb 11 11:29:48 2014] <tkurz>: yes we should [Tue Feb 11 11:29:51 2014] <jfrank>: why not? [Tue Feb 11 11:30:08 2014] <tkurz>: we already had this even in LMF [Tue Feb 11 11:30:12 2014] <wikier>: we were discussing to remove the content support from lmf in marmotta [Tue Feb 11 11:30:26 2014] <tkurz>: no we should not [Tue Feb 11 11:30:26 2014] <wikier>: this just turns the way back again [Tue Feb 11 11:31:14 2014] <jfrank>: i dont see a problem here [Tue Feb 11 11:31:39 2014] <tkurz>: I think it is quite useful to have it for e.g. implementing a semantic cms on marmotta [Tue Feb 11 11:32:09 2014] <wikier>: then we are back in the old lmf idea? [Tue Feb 11 11:32:58 2014] <jfrank>: not exactly [Tue Feb 11 11:33:22 2014] <jfrank>: in lmf, every resource could be binary and rdf at the same time [Tue Feb 11 11:33:36 2014] <jfrank>: in ldp, a resource is either binary OR rdf [Tue Feb 11 11:34:01 2014] <jfrank>: this is a much clearer separation [Tue Feb 11 11:34:54 2014] <tkurz>: exactly [Tue Feb 11 11:35:09 2014] <tkurz>: and this is very useful [Tue Feb 11 11:35:47 2014] <tkurz>: but I think for the moment it is a nice to have [Tue Feb 11 11:36:27 2014] <tkurz>: we should keep the idea on the webservice level but maybe not implement it in the backend for the moment [Tue Feb 11 11:37:34 2014] <wikier>: I'd prefer to discuss this in the mailing list with a broader team [Tue Feb 11 11:37:47 2014] <tkurz>: okay for me [Tue Feb 11 11:37:59 2014] <wikier>: because I see Marmotta just at the "LDP RDF Resource" level [Tue Feb 11 11:38:06 2014] <wikier>: but that's just a personal opinion [Tue Feb 11 11:38:06 2014] <jfrank>: why? [Tue Feb 11 11:38:13 2014] <wikier>: you may be right [Tue Feb 11 11:38:32 2014] <wikier>: I don't want to convert Marmotta in a Semantic CMS [Tue Feb 11 11:38:51 2014] <jfrank>: but into a LDP implementation? [Tue Feb 11 11:39:03 2014] <wikier>: just because some use cases from LDP (bug tracker) include managing binary data [Tue Feb 11 11:39:40 2014] <wikier>: 6.4.4 [Tue Feb 11 11:39:51 2014] <wikier>: LDP servers may accept an HTTP POST of non-RDF representations (LDP-BRs) for creation of any kind of resource, [Tue Feb 11 11:40:02 2014] <wikier>: may, not must [Tue Feb 11 11:40:20 2014] <jfrank>: sure, but it would be an easy one... [Tue Feb 11 11:40:23 2014] <wikier>: which it's okay regarding section 3 [Tue Feb 11 11:40:42 2014] <jfrank>: but ok, let's move this discussion to the ML [Tue Feb 11 11:40:43 2014] <wikier>: ok, let's vote it in the mailing list [Tue Feb 11 11:41:08 2014] <wikier>: #action discuss if Marmotta LDP will support binary resources or not [Tue Feb 11 11:41:12 2014] <wikier>: ok [Tue Feb 11 11:42:26 2014] <wikier>: how can we approach the initial implementation? [Tue Feb 11 11:42:33 2014] <aperez>: wikier I will read you later [Tue Feb 11 11:42:40 2014] <aperez>: Ainga wants to record a video about the ECIR [Tue Feb 11 11:42:43 2014] <aperez>: .... [Tue Feb 11 11:42:58 2014] <aperez>: to interview the zaizi presenters.... [Tue Feb 11 11:43:06 2014] <wikier>: aperez: have fun [Tue Feb 11 11:43:06 2014] <aperez>: hahahhaha [Tue Feb 11 11:43:25 2014] <wikier>: ok [Tue Feb 11 11:43:29 2014] <aperez>: hahaha [Tue Feb 11 11:44:01 2014] <wikier>: ok, back to the business [Tue Feb 11 11:44:04 2014] <wikier>: how can we approach the initial implementation? [Tue Feb 11 11:44:35 2014] <jfrank>: for me there are two questions to start with: [Tue Feb 11 11:44:41 2014] <wikier>: as soon as we have a jax-rs draft, it's would be easier to split into more concrete tasks [Tue Feb 11 11:44:47 2014] <tkurz>: the first thing is to create service stubs [Tue Feb 11 11:44:54 2014] <jfrank>: does the sesame-turtle rio support relative uris? [Tue Feb 11 11:45:06 2014] <wikier>: the parser? [Tue Feb 11 11:45:09 2014] <wikier>: yes [Tue Feb 11 11:45:15 2014] <jfrank>: parser and serializer [Tue Feb 11 11:45:26 2014] <wikier>: parser yes [Tue Feb 11 11:45:44 2014] <wikier>: serializaer I'm not sure, but it doesn't really matter, right? [Tue Feb 11 11:46:11 2014] <jfrank>: parser is more importatn [Tue Feb 11 11:46:25 2014] <wikier>: parser support is, for sure [Tue Feb 11 11:50:13 2014] <wikier>: http://openrdf.callimachus.net/sesame/2.7/apidocs/org/openrdf/rio/Rio.html#parse(java.io.InputStream, java.lang.String, org.openrdf.rio.RDFFormat, org.openrdf.model.Resource...) [Tue Feb 11 11:50:43 2014] <wikier>: Model parse(InputStream in, String baseURI, RDFFormat dataFormat, Resource... contexts) [Tue Feb 11 11:51:48 2014] <wikier>: ok, so the create service stubs (jax-rs implementation of ldp) is the first tasks we have to work on [Tue Feb 11 11:51:57 2014] <wikier>: and then, the other will come more naturally [Tue Feb 11 11:52:05 2014] <wikier>: volunteer/s for that? [Tue Feb 11 11:52:13 2014] <wikier>: I could not start before tomorrow [Tue Feb 11 11:52:35 2014] <wikier>: anyone else? [Tue Feb 11 11:52:39 2014] <jfrank>: i'll take a look [Tue Feb 11 11:52:55 2014] <jfrank>: won't be much, but anyways [Tue Feb 11 11:53:04 2014] <wikier>: exactly [Tue Feb 11 11:53:20 2014] <wikier>: I'm creating the issue in jira [Tue Feb 11 11:53:26 2014] <wikier>: I'll assign to you [Tue Feb 11 11:53:34 2014] <jfrank>: +1 [Tue Feb 11 11:53:38 2014] <wikier>: great [Tue Feb 11 11:57:30 2014] <wikier>: MARMOTTA-440 [Tue Feb 11 11:57:37 2014] <wikier>: here we go [Tue Feb 11 12:03:30 2014] <wikier>: and then let's see [Tue Feb 11 12:03:37 2014] <wikier>: I thing is better to go step by step [Tue Feb 11 12:03:40 2014] <wikier>: and run later [Tue Feb 11 12:04:00 2014] <wikier>: what do you think, guys? [Tue Feb 11 12:05:50 2014] <jfrank>: ok [Tue Feb 11 12:06:01 2014] <wikier>: ok ## aob ## [Tue Feb 11 12:06:20 2014] <wikier>: aob? [Tue Feb 11 12:06:34 2014] <wikier>: nothing from my side [Tue Feb 11 12:06:51 2014] <jfrank>: nothing here too [Tue Feb 11 12:07:01 2014] <wikier>: we'll continue by mail, having other meeting when necessary, I think [Tue Feb 11 12:07:07 2014] <jfrank>: +1 [Tue Feb 11 12:07:34 2014] <wikier>: good [Tue Feb 11 12:07:39 2014] <jfrank>: do you start the binary discussion? [Tue Feb 11 12:07:50 2014] <wikier>: I'm writing the mail, yes [Tue Feb 11 12:08:25 2014] <wikier>: let's see [Tue Feb 11 12:08:35 2014] <wikier>: ASFBot: meeting ends Meeting ended at Tue Feb 11 12:08:35 2014
