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

Using timezone: Eastern Standard Time
* dwilcox joins07:32
* jcoyne joins08:17
* mikeAtUVa joins08:41
* osmandin joins08:51
* arebenji joins08:52
* chadmills leaves09:04
* whikloj joins09:07
* ajs6f joins09:18
* acoburn joins09:19
<ajs6f>That is by far the longest agenda I can remember having for a tech call.09:39
* bseeger joins
* peichman joins
* mikeAtUVa leaves09:42
* github-ff joins10:09
[fcrepo-extra-ontologies] ajs6f created FCREPO-1403 (+1 new commit): https://git.io/vrpZs
fcrepo-extra-ontologies/FCREPO-1403 654c997 A. Soroka: FCREPO-1403...
* github-ff leaves
* github-ff joins10:17
[fcrepo-extra-ontologies] ajs6f pushed 1 new commit to FCREPO-1403: https://git.io/vrpnh
fcrepo-extra-ontologies/FCREPO-1403 f253af4 A. Soroka: Corrections...
* github-ff leaves
* mikeAtUVa joins
* github-ff joins10:21
[fcrepo-extra-ontologies] ajs6f pushed 1 new commit to FCREPO-1403: https://git.io/vrpcb
fcrepo-extra-ontologies/FCREPO-1403 f792ed3 A. Soroka: Reversing...
* github-ff leaves
* github-ff joins10:51
[fcrepo-extra-ontologies] ajs6f pushed 1 new commit to FCREPO-1403: https://git.io/vrpRL
fcrepo-extra-ontologies/FCREPO-1403 d93aa08 A. Soroka: Rename objState.rdf to ObjState.rdf
* github-ff leaves
* ajs6f is here10:59
<whikloj>welcome to the hall of silence
<acoburn>*is here*11:01
<dwilcox>I'm on the call
* bbranan joins
<ajs6f>Cant break what's already broke.11:02
<bseeger>* is here *
* kat3_drx joins
* ajwagner is just calling in now11:04
* Namita joins
* kefo joins11:05
* kefo_ joins11:07
* kefo leaves11:10
* mgibney joins11:12
<ajwagner>We'd be very interested in MariaDB and Swift backed clustering but I won't have time to dive into that until end of june/july.11:14
(Failing Swift, NFS)11:15
<ajs6f>URG DROPED>
* dshalvi joins
<whikloj>sidenote: wish I was going to Dublin11:33
<ajs6f>whikloj: If you send me a very large photo of your head, I will cut it out and glue it to cardboard and carry around OR and introduce it to people as you. This is not a joke.11:36
<whikloj>ajs6f: hmmmm....would you paste to cardstock, cut out the eyes and wear it around and introduce yourself as me?11:39
<ajs6f>whikloj: I might glue it to tabs to insert in my collar and introduce you as my second head...11:41
* mikeAtUVa leaves
* escowles joins
that vocab is: http://id.loc.gov/vocabulary/preservation/eventType.html11:43
<ajs6f>escowles++ # I am getting less and less enthused about PREMIS generally.
<escowles>and i completely agree that vocab is inadequate and we need to add new types in order to use premis
ajs6f: i'm on a group working to modernize it, but it's slow going
<ajs6f>escowles: Wasn't PREMIS one of the reasons you-at-UCSD ended p constructing RDF with bags of bnodes?11:44
<escowles>ajs6f: yes, that and MADS
<escowles>stripping down that hierarchy is a big part of why it's slow going
<ajs6f>escowles: In at least two ways in my experience: hierarchy _in_ the product (assumptions about recordation, document vs. graph, xml vs. rdf) and hierarchy in _how_ the product gets built (not being able to do anything until someone else blesses it, "OMG PUBLISH OUR OWN URIS NO WAY!").11:48
Because it doesn't really work correctly.11:49
ImikeAtUVa can tell you all about that.
<escowles>ajs6f: most definitely — there is a lot of conceptual work to get everybody on the same page with normal linked data norms11:50
and those are related to both technical hierarchy and social hierarchy
<ajs6f>escowles++ # That ++ is for doing the social work.11:51
Fedora 4: We are your Linked Data social worker.
<escowles>linked data evangelist (TM)
<ajs6f>bbranan is a constructive guy.12:01
* bbranan leaves
* kat3_drx leaves12:02
<acoburn>awoods: notes are posted12:08
<awoods>thanks, acoburn.
* bseeger leaves12:11
<ajs6f>afk bbl12:21
<awoods>FYI: ModeShape 5.1 release planning: https://issues.jboss.org/projects/MODE/versions/1233005512:40
Release: 13/Jul/16
* Namita leaves12:42
* bseeger joins13:00
* bseeger leaves13:01
* bseeger joins13:07
* mikeAtUVa joins13:08
<acoburn>awoods: fyi: https://git.io/vrpFm13:16
<awoods>acoburn: is that true that the objects would end up in the binary store?
just tried it
<awoods>acoburn: hmmm13:17
<acoburn>awoods: even with the default value, if you have a really long dc:description, that could end up in the binary store
awoods: as long as it's >= 4KB in length
<whikloj>Thoughts about the MySQL and PostgreSQL documentation, it is not correct for the current release. The "ispn" has been removed from the configuration parameters, but only if you build from source.13:22
<awoods>whikloj: are you talking about the modeshape5 branch?13:25
<whikloj>awoods: nope just trying with 4.5.1
before I start building that branch
awoods: https://github.com/fcrepo4/fcrepo4/blob/fcrepo-4.5.1/fcrepo-configs/src/main/resources/config/infinispan/jdbc-postgresql/infinispan.xml
<awoods>whikloj: better? https://wiki.duraspace.org/pages/diffpagesbyversion.action?pageId=73174116&selectedPageVersions=7&selectedPageVersions=813:27
<whikloj>awoods: they are also needed for mysql, but you later removed them. So if someone is building from master then they DON'T want to have the "ispn" in there
<awoods>whikloj: Links would be helpful. I am seeing ispn in master for MySQL and PostgreSQL: https://github.com/fcrepo4/fcrepo4/blob/master/fcrepo-configs/src/main/resources/config/jdbc-mysql/repository.json#L11 and https://github.com/fcrepo4/fcrepo4/blob/master/fcrepo-configs/src/main/resources/config/jdbc-postgresql/repository.json#L1113:29
<whikloj>awoods: as am I, my eyes are playing tricks on me. Okay so adding ispn to both sets of configuration parameters would be great.13:30
<awoods>whikloj: Where are you suggesting "adding ispn to both sets of configuration parameters"??13:31
<whikloj>awoods: just like your diff showed, but the ispn is also missing from the Mysql configuration parameters13:32
awoods: I would do it, but I can't edit that page13:33
<awoods>whikloj: the "whikloj" user has write access to: https://wiki.duraspace.org/display/FEDORA4x/Configuring+JDBC+Object+Store13:35
<whikloj>awoods: oops I was viewing : https://wiki.duraspace.org/display/FEDORA451/Configuring+JDBC+Object+Store13:37
<awoods>whikloj: thanks, I have now updated: https://wiki.duraspace.org/display/FEDORA451/Configuring+JDBC+Object+Store
<mgibney>acoburn: awoods: the flip side of descriptions ending up in the binary store is that I believe when I last tested, small binary content (e.g., "files" < 4kb) can end up not being stored in the binary store ... forgive me if I'm all wrong on this point, it's been a while since I thought about it, and I'm not 100% sure if we're talking about the same thing!13:46
<acoburn>mgibney: you are correct, small binaries are, by default stored in the object store13:47
<mgibney>oh, ok. but I guess that behavior is known13:48
<awoods>mgibney: see my follow-up email. It would appear that it is possible to segregate the binaries and objects, even if the binaries are small.
<acoburn>mgibney: meaning, a file < 4KB would be stored in leveldb or your jdbc database
<awoods>acoburn: ...unless you set minimumStringSize
<acoburn>mgibney: yes this is known behavior
awoods: correct
<mgibney>ok, thanks
ah, sorry; I was only looking at the IRC; I now see the group post to which you were referring13:51
* escowles leaves13:57
* bseeger leaves14:00
* bseeger joins14:05
<acoburn>awoods: I'll try attending the OR meeting remotely (it will be quite early though)14:12
<ajs6f>acoburn: That means you don't have to wear pants.14:18
<acoburn>ajs6f: or comb my hair
* mgibney leaves14:23
<whikloj>or comb your pants14:27
<ajs6f>There is an Apache project named Flink. It's like they just don't even care anymore.14:34
* bseeger leaves14:57
* github-ff joins15:01
[fcrepo-extra-ontologies] awoods opened pull request #4: Add FQ URIs and change versioning string (FCREPO-1403...pull/3) https://git.io/vrhmJ
* github-ff leaves
<ajs6f>dwilcox: "Functionality to support for standard import/export" — is that really a leadership group conversation? At the very least, I would want to supply some context to it.15:05
<dwilcox>ajs6f: It's a Leadership conversation only in terms of getting resources to pursue the work, not detailing the nature of the work itself15:06
<ajs6f>dqilcox: Okay, that makes more sense. In that case, are you not presupposing a fairly specific plan for that work?15:07
dwilcox: ^^^ # got your nick wrong15:08
quick show of hands— anyone coming into Dublin by the Saturday before OR (11 June)?15:10
<dwilcox>ajs6f: There are a few different approaches we could take in terms of implementation. I think at this point we're just looking for a general commitment to put resources toward this as a goal, and we'll figure out the details in due course.15:11
<ajs6f>dwilcox: Okay. You know what you're doing. Just to be clear— you're not expecting to have a commitment from the tech folks for design or impl before then, right? Because there's a whole conversation there that hasn't happened.15:13
<dwilcox>ajs6f: No, I have no expectation that such a design would be available before the LG meeting. We'll work that out after
<ajs6f>acoburn: This "standard import/export" thing— you could actually use the bidirectional 3store connection to do some of that… if it was crafty enough.15:14
dwilcox: Okay. I
dwilcox: Okay.
<acoburn>ajs6f: in a meeting
* manez joins15:37
* bseeger joins16:12
* manez leaves16:21
<ajs6f>dwilcox: As long as you are asking the LG for resources, why not add a line item to fund the printing of album covers?16:24
* arebenji leaves16:31
* ajs6f leaves16:44
* dwilcox leaves16:57
* peichman leaves
* bseeger leaves17:08
* acoburn leaves17:15
* github-ff joins17:22
[fcrepo4] awoods closed pull request #1041: Remove more modeshape dependencies from the HTTP layers (master...fcrepo-1694.2) https://git.io/vraZy
* github-ff leaves
* travis-ci joins17:42
fcrepo4/fcrepo4#4499 (master - a8799ee : Aaron Coburn): The build passed.
Change view : https://github.com/fcrepo4/fcrepo4/compare/ded76af2258f...a8799ee4e74c
Build details : https://travis-ci.org/fcrepo4/fcrepo4/builds/134868148
* travis-ci leaves
* dwilcox joins17:51
* whikloj leaves17:59
* jcoyne leaves18:07
* dwilcox leaves18:18
* dwilcox joins18:20
* dwilcox leaves18:22
* kefo leaves19:02
* dwilcox joins20:31
* dwilcox leaves21:08
* dwilcox joins21:11
* dwilcox leaves21:18
* github-ff joins21:26
[fcrepo4] acoburn opened pull request #1048: Remove modeshape's ServletCredentials from the HTTP layers (master...fcrepo-1694.6) https://git.io/vrjeS
* github-ff leaves
* dwilcox joins22:14
* dwilcox leaves22:42