Log of the #fcrepo channel on chat.freenode.net

Using timezone: Eastern Standard Time
* manez joins04:41
* manez leaves04:46
* manez joins05:41
* manez leaves05:46
* manez joins06:42
* manez leaves06:47
* manez joins07:43
* manez leaves07:47
* manez joins08:00
* coblej joins08:02
* thomz leaves08:10
* manez leaves08:30
* manez joins08:34
* coblej leaves08:41
* coblej joins08:45
* dwilcox joins08:46
* youn joins09:05
* thomz joins09:06
* thomz leaves09:07
* dhlamb joins09:14
* ajs6f joins09:20
* bseeger joins09:26
* peichman joins09:58
<ajs6f>awoods: Any objection to closing https://jira.duraspace.org/browse/FCREPO-127510:11
?
<ruebot>awoods: i should know this, but i think it changed, what's the right way to get a count of resources in 4.7.0-RC? (doing testing)
<ajs6f>ruebot: Use your fingers _and_ your toes. It took me a while to get the hang of it.
* ajs6f leaves10:13
* ajs6f joins
<awoods>ajs6f: Is the idea in closing that ticket that we are just going to do Memento and care less about what Marmotta is doing?10:14
ruebot: solr index?10:15
<ruebot>awoods: i was afraid of that.
<ajs6f>awoods: More or less. The idea is that we are in an independent conv with HvS and crew, and what Marmotta has done is not as interesting as what we are going to do directly with the Memento team.
<ruebot>awoods: doing this "Migration of Existing Repositories" on the testing.10:16
<awoods>ajs6f: So, close as "Won't Fix" with your above comment?
<ajs6f>awoods: Sure.10:17
<awoods>ajs6f++
ruebot++
* acoburn joins10:18
* ajs6f leaves10:30
* ajs6f joins10:38
* ajs6f leaves10:39
* ajs6f joins10:40
<awoods>Team: I am going to go dark shortly to get organized for upcoming travel.
* coblej leaves10:48
* whikloj joins10:55
* coblej joins11:03
* osmandin joins11:04
<ajs6f>I wonder if awoods means this kind of dark : https://pitofultimatedarkshadows.files.wordpress.com/2014/10/olga-pit640x480.jpg11:20
<osmandin>ajs6f: The weather is already kind of gloomy and dark around Boston.11:29
<ajs6f>osmandin: It's not very cheerful down here either. We've really had a nasty wet week.
* bseeger leaves11:34
<whikloj>Do not avoooid the hounds of hell. Do not avooooid the beasts of brimstone. Do not avoooid the... puppies of purgatory.11:45
* coblej leaves11:58
* ajs6f leaves12:03
* osmandin leaves12:05
* ajs6f joins12:06
* anarchivist leaves12:17
* anarchivist joins12:18
* ajs6f leaves12:19
* ajs6f joins12:25
* coblej joins12:27
* ajs6f leaves12:30
* dbernstein_ joins12:35
* dbernstein leaves
* acoburn leaves
* manez leaves12:38
* bseeger joins
* ajs6f joins12:59
* peichman leaves13:02
* peichman joins13:03
* tolloid joins
* tolloid leaves13:05
* tolloid joins13:07
* coblej leaves13:24
* acoburn joins13:29
* acoburn leaves
* github-ff joins
[fcrepo4-vagrant] barmintor pushed 2 new commits to hyconn2016: https://git.io/vPL09
fcrepo4-vagrant/hyconn2016 d4085df Benjamin Armintor: fix varnish config and install nodejs for rails
fcrepo4-vagrant/hyconn2016 77d03f1 Benjamin Armintor: update solr schema
* github-ff leaves
* ajs6f leaves13:35
* ajs6f joins13:37
* bseeger leaves13:39
* bseeger joins
* acoburn joins13:44
* tolloid leaves13:52
<acoburn>ajs6f: I finished removing the jcr.Session from the kernel-api. All unit/integration tests pass
ajs6f: rather than submitting the entire change set, though, I'm thinking it would be easier if I start with just the new interfaces
ajs6f: thoughts?
ajs6f: the whole thing is a monster: 83 files changed, 1595 insertions(+), 1790 deletions(-)13:54
* jgpawletko leaves13:56
* coblej joins
<ajs6f>acoburn: Are there no natural boundaries?14:00
acoburn: On another note: https://www.infoq.com/articles/java9-osgi-future-modularity
<acoburn>ajs6f: not really
<ajs6f>acoburn: Well, yeah, if you can even start by showing the new types, I guess that's a place to start.14:01
* tolloid joins
* tolloid leaves14:02
* peichman leaves14:03
* peichman joins14:05
<acoburn>ajs6f: https://github.com/fcrepo4/fcrepo4/pull/110714:15
* github-ff joins
[fcrepo4] acoburn opened pull request #1107: Add core types to be used to replace the use of jcr.Session in the kernel API (master...fcrepo-1868) https://git.io/vPLVh
* github-ff leaves
* coblej leaves14:20
<ajs6f>acoburn: I'm saying I think the session thing is simpler.14:27
acoburn: It doesn't know about transaction boundaries. That's client side.
<acoburn>ajs6f: not sure I follow14:28
ajs6f: you mean the FedoraSession::hasState method?14:29
<ajs6f>acoburn: Yeah, I don't think it belongs there.
<acoburn>ajs6f: maybe….14:30
<ajs6f>acoburn: If you are opening more than one FedoraSession (so that you have to keep track of where each one is) then _you_ should do that keeping-track-of.
<acoburn>ajs6f: the SessionProvider class in fcrepo-http-common needs to know if a given session is part of a "batch operation"
<ajs6f>acoburn: Then _it_ should keep track of that.14:31
<acoburn>ajs6f: that should be possible
ajs6f: (of course it's possible; I mean possible w/o tremendous work)14:32
<ajs6f>acoburn: Another way to put this: there should be no situation in which more than one session is participating in a batch. There is just a session. It may get committed at the end of an HTTP request, it may get committed after twelve. It doesn't know or care.
It only knows about getting commmitted.
Fedora 4: Get committed.
(to an asylum)
<acoburn>ajs6f: I'll re-work the impl w/o the hasState method14:33
<ajs6f>acoburn: Now aren't you sorry you asked?
<acoburn>:-)
ajs6f: any other comments?14:34
<ajs6f>acoburn: I think you look very nice today.
<acoburn>ajs6f: why thank you, sir
ajs6f: there may be value in knowing if a session "is alive"; WDYT about a method like `boolean isAlive()`14:37
ajs6f: the alternative is to just catch exceptions
<ajs6f>acoburn: Maybe… maybe a session is always alive. Maybe you can call commit as much as you want?14:38
<acoburn>ajs6f: the whole point of the hasState method was to avoid an endless proliferation of isX() tests on the Session object14:39
ajs6f: I'm fine with not including the isAlive method
<ajs6f>acoburn what is IsX()?
acoburn: You mean isActive isDead isOnVacation?14:40
<acoburn>ajs6f: isAlive(), isCommitted(), isNew(), isHungry()
<ajs6f>acoburn: If there is only one state, we need none of that. I think we could do two methods:
acoburn: commit(), which flushes changes according to the usual semantics and begins a new scope, and abort(), which throws away changes and begins a new scope.14:41
acoburn: and you can call each as many times as you like.
acoburn: begin is implicit.14:42
<acoburn>ajs6f: ok, the issue with that is this &$^@* HttpResourceConverter class in fcrepo-http-common
<ajs6f>acoburn: Eh?
acoburn: Didn't barmintor fix that?/14:43
<acoburn>ajs6f: no
ajs6f: https://github.com/fcrepo4/fcrepo4/blob/master/fcrepo-http-commons/src/main/java/org/fcrepo/http/commons/api/rdf/HttpResourceConverter.java
<ajs6f>acoburn: I thought he made that work over IDs only?
<acoburn>ajs6f: it's completely entangled with modeshape
ajs6f: you mean in his recent effort? that was never merged14:44
<ajs6f>acoburn: Yes, recent effort… )*&*&*()&)**&)& IT WAS NEVER MERGED WHY NOT?!
<acoburn>ajs6f: I have no idea
<ajs6f>acoburn: Anyone? Anyone? Bueller? Bueller?
acoburn: I know you're not going to want to hear this, but we should get that in first. It will make all that JCR in that class (and in several others) go away.14:45
<acoburn>ajs6f: a week of work … down the drain
<ajs6f>acoburn: Oh, jeeez. Wait, maybe there's a way to salvage this.14:46
<acoburn>ajs6f: and I'll be on other projects starting next week
<ajs6f>acoburn: Is there no way to rebase? Did you look at that and it is awful?
acoburn: I'm sorry. I thought barmintor's stuff got merged.
<acoburn>ajs6f: someone will need to rebase over the jena upgrade
<ajs6f>acoburn: Yes.
barmintor: ping
poop poop poop14:48
<whikloj>ajs6f/acoburn: I thought barmintor's stuff wasn't complete?14:50
<ajs6f>whikloj: I don't know.
whikloj: Do you know if there was a ticket?14:51
<acoburn>ajs6f: my understanding was that dbernstein was going to pick up that work and continue it
<ajs6f>acoburn:whikloj: I reviewed it and other people did too.
<whikloj>ajs6f: To confirm we are talking about the converter refactor right?
<ajs6f>whikloj: Yes
<acoburn>ajs6f: this was the whole point of my email last week asking about that very PR
<ajs6f>acoburn: I think you are talking about something else.
<whikloj>ajs6f: Yes there was a ticket, I'll see if I can find it
<ajs6f>acoburn: (Than the converter refactor)
<acoburn>ajs6f: I don't think so14:52
<ajs6f>acoburn: It's the converter refactor that will hit HttpResourceThingy
I want to know why that wasn't merged.
<acoburn>ajs6f: the converter refactor will hit everything
<ajs6f>acoburn: Well, a hell of a lot, yes.
<ajwagner>This one isn't it: https://jira.duraspace.org/browse/FCREPO-2105
<acoburn>ajwagner: I believe so14:53
<ajwagner>and this one https://jira.duraspace.org/browse/FCREPO-2111
<ajs6f>acoburn: We have several massive PRs colliding like elephants that have eaten too many fermented palm fruits.
<acoburn>ajs6f: so pick one
<ajs6f>It's this: https://jira.duraspace.org/browse/FCREPO-2111
<whikloj>ajs6f/acoburn: I found this one, but I thought there was another.https://jira.duraspace.org/browse/FCREPO-211114:54
ajs6f++
<ajs6f>acoburn: Not that I pick that one, but that that's the one that matters from barmintor.
acoburn ;The other one isn't relvant here and I don't care about it now.
https://github.com/fcrepo4/fcrepo4/pull/1086
<acoburn>ajs6f: ok, what I want to know is whether I should spend any more time on the jcr.Session thing, b/c I have other things to do as well14:55
<ajs6f>acoburn: I am sorry, acoburn. Hm. Can you make it happen if we ignore barmintor's work for now?
<acoburn>ajs6f: if this isn't going to go forward, then I'll do something else
<ajs6f>acoburn: You are here and he is not.
<acoburn>ajs6f: I can make it happen, but I don't want it to be a wasted effort14:56
<ajs6f>acoburn: I don't think it is. Let's think about the scenario: you move this through. Then barmintor or someone wearing one of his snazzy hipster hats comes along and tries to move his PR forward.14:57
acoburn: Obvsiouly there will be a huge amount of rebasing. But's that already true.
acoburn: Can you think of any way in which the actual purposes of the two pieces of work are at odds?14:58
* coblej joins
<ajs6f>all: I need to switch VPNs, be back in a mo14:59
* ajs6f leaves
* ajs6f joins15:00
back
<barmintor>acoburn to go first IMO
<ajs6f>DONE
acoburn: Fling yourself into the breach.15:01
acoburn: We will back you to the hilt.
* coblej leaves15:02
* osmandin joins15:04
<acoburn>ajs6f: ok, I'll keep moving this forward then
<ajs6f>acoburn: Damn the torpedoes, full speed ahead!15:06
* youn leaves15:20
* dbernstein leaves15:22
* osmandin leaves15:59
* coblej joins
* bseeger leaves16:00
* bseeger joins16:03
* coblej leaves16:04
<peichman>has anyone else noticed the fcrepo-indexing-* routes indexing their /audit collections when doing a reindexing using fcrepo-reindexing REST service, despite config that says to ignore /audit?16:05
I am running version 4.6.0 of the fcrepo-camel-toolbox
I should clarify, I have audit.collection specified in the org.fcrepo.camel.indexing.{solr,triplestore}.cfg config files, but not in the reindexing config file. Do I need it there as well?16:07
acoburn: ping ^^16:08
<acoburn>peichman: it is actually indexing those resources in the triplestore/solr? or is it just logging that it is processing the resource?16:09
<peichman>acoburn: they end up in Solr and the triplestore
acoburn: (which is obviously less than ideal!)
<acoburn>peichman: right16:10
peichman: let me look into that
<peichman>acoburn: thanks
* coblej joins16:22
* github-ff joins
[fcrepo4-vagrant] barmintor pushed 1 new commit to hyconn2016: https://git.io/vPLFG
fcrepo4-vagrant/hyconn2016 a4c1fa4 Benjamin Armintor: fix solr libs and configs for text extraction
* github-ff leaves
* bseeger leaves16:27
* dwilcox leaves16:55
* ajs6f leaves17:00
* coblej leaves17:08
* peichman leaves17:16
* acoburn leaves17:26
* coblej joins17:29
* dwilcox joins17:36
* coblej leaves17:39
* coblej joins17:42
* github-ff joins17:49
[fcrepo4-vagrant] barmintor pushed 1 new commit to hyconn2016: https://git.io/vPtUc
fcrepo4-vagrant/hyconn2016 fe19075 Benjamin Armintor: include lucene analysis deps in solr config
* github-ff leaves
* whikloj leaves17:50
* coblej leaves17:52
* dbernstein joins
* mjgiarlo leaves18:01
* mjgiarlo joins18:14
* dbernstein leaves18:15
* dbernstein joins18:17
* manez joins18:30
* manez leaves19:10
* coblej joins19:23
* mjgiarlo leaves19:29
* coblej leaves19:34
* mjgiarlo joins19:37
* coblej joins20:18
* coblej leaves20:20
* peichman joins21:22
* mjgiarlo leaves21:38
* peichman leaves
* peichman joins21:45
* mjgiarlo joins21:49
* peichman leaves22:03
* github-ff joins22:18
[fcrepo4-vagrant] barmintor tagged hyconn2016 at hyconn2016: https://git.io/vPt4C
* github-ff leaves
* dwilcox leaves22:41
* peichman joins22:45
* f4jenkins joins23:12
* dhlamb leaves23:47