Log of the #duraspace-ff channel on chat.freenode.net

Using timezone: Eastern Standard Time
* escowles leaves01:37
* awoods joins01:52
* dwilcox joins07:47
* mikeAtUVa joins08:31
* acoburn joins08:39
* ksclarke joins08:54
* ksclarke leaves09:21
* ksclarke joins
* osmandin joins09:38
* whikloj joins09:44
* ajs6f1 joins10:48
* ajs6f leaves
<f4jenkins>Project fcrepo4-T2 build #228: SUCCESS in 6 min 24 sec: http://jenkins.fcrepo.org/job/fcrepo4-T2/228/10:54
escowles: NonRDFSources now include rdf:type=ldp:NonRDFSource triple
* github-ff joins10:57
[fcrepo-camel-toolbox] awoods opened pull request #20: Consolidate web modules (master...fcrepo-1523) http://git.io/vJhrB
* github-ff leaves
* ajs6f1 leaves11:00
<acoburn>awoods: I'll be a bit late to the mtg11:01
* escowles joins11:03
* ajs6f joins11:05
* doylejo joins11:10
* dhlamb joins11:27
whikloj: I'd be willing to spend some time, but have zero Java 8 knowledge11:36
awoods ^^^
<mikeAtUVa>can we sell this to people as "java 8 training"?
<whikloj>awoods: I could test, possibly fix but not sure.
* doylejo leaves11:44
* osmandin leaves11:47
<ruebot>awoods, mikeAtUVa: https://jira.duraspace.org/browse/FCREPO-1524 let me know if that's good. if so, i'll start working on it this afternoon.12:05
<mikeAtUVa>ruebot: I'll take a look now...12:06
<dhlamb>acoburn: ping12:12
* dwilcox leaves12:15
<acoburn>dhlamb: pong12:20
<dhlamb>acoburn: i've run into an interesting situation with fcrepo-camel and transactions12:21
acoburn: it's pretty edge, and i may have a solution. if i made a ticket, could you review it?
<acoburn>dhlamb: of course
<dhlamb>acoburn: both the ticket, and work i put into it
<acoburn>dhlamb: absolutely12:22
dhlamb: now I'm curious
<dhlamb>acoburn: cool. it's not playing well with hibernate. i'm trying to dodge the global transaction bullet for the time being so i can continue work. but it seems to be making transactions even when i don't set up the component to do so, simply by merit of declaring the route as transacted.12:23
acoburn: but of course, eventually, global transactions will come into play. i (islandora) will be dealing with jms, jdbc, and fedora at the same time pretty soon.12:24
<acoburn>whenever you declare a route as transacted it will invoke the transaction machinery
and that will apply throughout a route (across direct: endpoints)
<dhlamb>acoburn: i see at one spot you're automaking the transaction manager if it doesn't exist when requesting the tx template
<acoburn>dhlamb: yes, that will create the default tx manager12:25
<dhlamb>acoburn: and then checking to see if the route is transacted later. was thinking you could return null if manager isn't set up in config and then check for transacted route AND non-null template
acoburn: seem reasonable?12:26
* dwilcox joins12:27
<acoburn>dhlamb: I'll have to look at that code again, as I recall, there was some reason why the tx code had to be implemented that way
dhlamb: but I'm certainly open to alternate implementations12:28
<dhlamb>acoburn: yeah, i don't doubt that. it's certainly gonna try to push you into best-practice of using transactions. hibernate certainly does it.
acoburn: well, problem is that it doesn't seem to be committing the transaction all the time when it autogenerates the manager/template.12:29
acoburn: so my route only executes once. after that, it just gets hung with a never committed fedora transaction.
<acoburn>dhlamb: do you have code to point to?12:30
<dhlamb>acoburn: i can push, but it's pretty rooted in context12:31
<acoburn>dhlamb: a gist would be fine
<dhlamb>acoburn: ok, moment12:32
<acoburn>dhlamb: or I can wait until you push your branch to a public repo
<dhlamb>acoburn: spring config -> https://github.com/Islandora-Labs/islandora/blob/services/camel/services/src/main/webapp/WEB-INF/applicationContext.xml#L8-L6012:35
acoburn: and the route -> https://github.com/Islandora-Labs/islandora/blob/services/camel/services/src/main/java/ca/islandora/services/routes/CollectionEndpoint.java
<acoburn>dhlamb: I see, you have multiple tx managers12:36
<dhlamb>acoburn: well, one's commented out. i have the whole JTA / JOTM rabbit hole to go down eventually12:37
<acoburn>dhlamb: right, but is the idea that you are transacting some event flow with hibernate and then nesting a fedora transaction in that?12:38
<dhlamb>acoburn: ideally, but things collide. right now i'm trying not to use a fedora transaction and just let hibernate be the only one.
<acoburn>dhlamb: or is it that you do /not/ want the fcrepo interaction to be transacted
<dhlamb>acoburn: ^ that, but only to plug the hole so i can keep working on things12:39
acoburn: while i get global transactions sorted out
<acoburn>dhlamb: transactions don't cross seda: endpoints
<dhlamb>acoburn: sneaky sneaky12:40
acoburn: didn't think about ducking out of the thread
<acoburn>dhlamb: so you could, for instance, do any necessary pre-processing sans transaction
dhlamb: and then initiate the transacted() part on the other side of seda:12:41
<dhlamb>acoburn: let me try and mash some things around to see what happens12:42
acoburn: and if i can get it going that way, then i'll have room to breathe and maybe we can start a larger conversation about how to get all these transactions to play nicely together globally
<acoburn>dhlamb: that sounds good, I recall seeing some mechanism for deciding which transaction applies to a given component.12:45
acoburn: yup, that did it. just pushed the fedora call to it's own seda route.12:53
acoburn: sidestepping FTW
<acoburn>dhlamb: nice, I'm glad that worked!
<dhlamb>acoburn: i'm pushing myself to finish one of our islandora services in entirety so i can document and then have others attempt to create their own. once i get that done, i'll look into getting all those transactions working together harmoniously.12:54
acoburn: would be a HUGE win for us if we can get jms, jdbc, and fedora rolling back together in case of exception12:55
acoburn: right now we just drop messages and shrug, lol12:56
acoburn: embarassing, really
<acoburn>dhlamb: what about a redelivery policy? (retry on exception)
dhlamb: though rolling back would be what you want in that case12:57
<dhlamb>acoburn: i mean w/r/t our current stack. not what i'm working on now. but yes, redeliver X number of times and if fail is continuous, deliver to dead letter channel is what i envision.
* awoods leaves13:04
* awoods joins13:15
* awoods leaves13:37
* awoods joins13:50
* acoburn1 joins13:55
* acoburn leaves13:57
* acoburn1 leaves14:33
* acoburn joins
* dwilcox leaves14:41
* github-ff joins14:44
[fcrepo-camel-toolbox] acoburn opened pull request #21: added maven configuration for feature.xml generation (master...fcrepo-1479) http://git.io/vUeEK
* github-ff leaves
* dwilcox joins14:47
* yinlin joins15:05
* awead_away leaves15:07
* awead joins15:12
* awead leaves15:16
* awead joins15:17
* dwilcox leaves15:22
* awead leaves15:40
* awead joins15:41
* osmandin joins15:55
* dhlamb leaves16:08
* ajs6f leaves16:09
* awead leaves16:14
* ajs6f joins
<acoburn>awoods: ping16:17
<awoods>on a cll16:18
<ajs6f>That's a Welsh call.
<acoburn>I thought that was spelled cwl16:19
Who knew?
* osmandin leaves16:28
* awead joins
<acoburn>escowles: ping16:46
* yinlin leaves16:52
* mikeAtUVa leaves17:04
* awoods leaves17:06
* ajs6f leaves17:08
* ajs6f joins17:11
* ajs6f leaves
* github-ff joins17:19
[fcrepo-camel-toolbox] acoburn pushed 1 new commit to master: http://git.io/vUvK9
fcrepo-camel-toolbox/master 3b012c7 Andrew Woods: Make configuration properties consistent...
* github-ff leaves
* awoods joins
* github-ff joins17:20
[fcrepo-camel-toolbox] acoburn closed pull request #20: Consolidate web modules (master...fcrepo-1523) http://git.io/vJhrB
* github-ff leaves
* travis-ci joins17:24
fcrepo4-labs/fcrepo-camel-toolbox#63 (master - 3b012c7 : Andrew Woods): The build passed.
Change view : https://github.com/fcrepo4-labs/fcrepo-camel-toolbox/compare/d142b434fff9...3b012c7465f8
Build details : http://travis-ci.org/fcrepo4-labs/fcrepo-camel-toolbox/builds/61684307
* travis-ci leaves
* github-ff joins17:43
[fcrepo-camel-toolbox] acoburn opened pull request #22: update readme files with new build information (master...fix-readme) http://git.io/vUvSy
* github-ff leaves
* acoburn leaves17:47
* awead leaves17:48
* awead joins17:53
* whikloj leaves17:55
* github-ff joins18:07
[fcrepo-camel-toolbox] awoods pushed 2 new commits to master: http://git.io/vUvNP
fcrepo-camel-toolbox/master 1d689fe Aaron Coburn: update readme files with new build information
fcrepo-camel-toolbox/master f62d187 Andrew Woods: Merge pull request #22 from acoburn/fix-readme...
* github-ff leaves
<f4jenkins>Project fcrepo-camel-toolbox build #75: UNSTABLE in 4 min 26 sec: http://jenkins.fcrepo.org/job/fcrepo-camel-toolbox/75/18:11
Aaron Coburn: update readme files with new build information
* travis-ci joins18:12
fcrepo4-labs/fcrepo-camel-toolbox#65 (master - f62d187 : Andrew Woods): The build passed.
Change view : https://github.com/fcrepo4-labs/fcrepo-camel-toolbox/compare/3b012c7465f8...f62d187bd3c5
Build details : http://travis-ci.org/fcrepo4-labs/fcrepo-camel-toolbox/builds/61690005
* travis-ci leaves
* github-ff joins18:15
[fcrepo-camel-toolbox] awoods pushed 1 new commit to master: http://git.io/vUvpA
fcrepo-camel-toolbox/master ae614a4 Andrew Woods: Update README.md
* github-ff leaves
* ksclarke leaves
* travis-ci joins18:18
fcrepo4-labs/fcrepo-camel-toolbox#66 (master - ae614a4 : Andrew Woods): The build passed.18:19
Change view : https://github.com/fcrepo4-labs/fcrepo-camel-toolbox/compare/f62d187bd3c5...ae614a4ad881
Build details : http://travis-ci.org/fcrepo4-labs/fcrepo-camel-toolbox/builds/61690818
* travis-ci leaves
<f4jenkins>Yippee, build fixed!
Project fcrepo-camel-toolbox build #76: FIXED in 4 min 24 sec: http://jenkins.fcrepo.org/job/fcrepo-camel-toolbox/76/
awoods: Update README.md
* ksclarke joins18:41
* awead leaves19:22
* awead joins20:01
* dhlamb joins20:19
* dhlamb leaves21:40

Generated by Sualtam