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

Using timezone: Eastern Standard Time
<pivotal-bot>Chris Beer added comment: "I think we still need to define how properties are indexed in our world. Are we happy with requiring it to be..." https://www.pivotaltracker.com/story/show/6071383800:19
Chris Beer started "Add DELETE endpoint for workspaces" https://www.pivotaltracker.com/story/show/6088733600:32
Chris Beer finished "Add DELETE endpoint for workspaces" https://www.pivotaltracker.com/story/show/60887336
* github-ff joins00:34
[fcrepo4] cbeer created workspace-delete (+1 new commit): http://git.io/ZIhtvA
fcrepo4/workspace-delete 2861320 Chris Beer: Add endpoint for deleting workspaces; fixes 60887336
* github-ff leaves
* travis-ci joins00:46
[travis-ci] futures/fcrepo4#1202 (workspace-delete - 2861320 : Chris Beer): The build failed.
[travis-ci] Change view : https://github.com/futures/fcrepo4/commit/2861320193db
[travis-ci] Build details : http://travis-ci.org/futures/fcrepo4/builds/14124510
* travis-ci leaves
<pivotal-bot>Chris Beer estimated "Fix response body from creating a node" as 0 points https://www.pivotaltracker.com/story/show/6088724600:48
Chris Beer started "Fix response body from creating a node" https://www.pivotaltracker.com/story/show/60887246
Chris Beer finished "Fix response body from creating a node" https://www.pivotaltracker.com/story/show/60887246
* github-ff joins00:49
[fcrepo4] cbeer deleted workspace-delete at 2861320: http://git.io/u8xOBw
* github-ff leaves
* kaarefc joins00:50
<pivotal-bot>Chris Beer started "Fix created workspace Location" https://www.pivotaltracker.com/story/show/6088600000:59
* travis-ci joins01:00
[travis-ci] futures/fcrepo4#1203 (http-fixes - 9e3b715 : Chris Beer): The build passed.
[travis-ci] Change view : https://github.com/futures/fcrepo4/compare/a535ee8f1b6b^...9e3b715daa1a
[travis-ci] Build details : http://travis-ci.org/futures/fcrepo4/builds/14124773
* travis-ci leaves
<pivotal-bot>Chris Beer finished "Fix created workspace Location" https://www.pivotaltracker.com/story/show/6088600001:10
* kaarefc leaves01:12
* ksclarke leaves01:20
<pivotal-bot>Chris Beer started "Support preconditions on FedoraNodes#delete" https://www.pivotaltracker.com/story/show/6088450801:28
Chris Beer finished "Support preconditions on FedoraNodes#delete" https://www.pivotaltracker.com/story/show/60884508
<bljenkins>Project fcrepo-fixity-corrupter build #445: SUCCESS in 5 min 58 sec: http://ci.fcrepo.org/jenkins/job/fcrepo-fixity-corrupter/445/01:40
* fcrepo-bot joins01:42
* github-ff joins01:47
[fcrepo4] cbeer pushed 4 new commits to http-fixes: http://git.io/S0DI0g
fcrepo4/http-fixes a19226e Chris Beer: delete fixup
fcrepo4/http-fixes 213cef7 Chris Beer: After creating a new workspace, return the path to the workspace in the location header. Fixes #60886000
fcrepo4/http-fixes b62e8cf Chris Beer: add preconditions support for DELETE and MOVE requests
* github-ff leaves
* github-ff joins01:48
[fcrepo4] cbeer force-pushed http-fixes from 439a460 to e3cd939: http://git.io/cZrUuA
fcrepo4/http-fixes 1da7638 Chris Beer: Add endpoint for deleting workspaces; fixes 60887336
fcrepo4/http-fixes b021603 Chris Beer: update response after creating a node to include the full URI to the created resource; fixes #60887246
fcrepo4/http-fixes 986b56b Chris Beer: After creating a new workspace, return the path to the workspace in the location header. Fixes #60886000
* github-ff leaves
* github-ff joins
[fcrepo4] cbeer opened pull request #156: HTTP API fixes (master...http-fixes) http://git.io/CV016g
* github-ff leaves
<pivotal-bot>Chris Beer added comment: "https://github.com/futures/fcrepo4/pull/156" https://www.pivotaltracker.com/story/show/60884508
Chris Beer added comment: "https://github.com/futures/fcrepo4/pull/156" https://www.pivotaltracker.com/story/show/60886000
Chris Beer added comment: "https://github.com/futures/fcrepo4/pull/156" https://www.pivotaltracker.com/story/show/60887246
Chris Beer added comment: "https://github.com/futures/fcrepo4/pull/156" https://www.pivotaltracker.com/story/show/60887336
* travis-ci joins01:58
[travis-ci] futures/fcrepo4#1204 (http-fixes - 439a460 : Chris Beer): The build passed.
[travis-ci] Change view : https://github.com/futures/fcrepo4/compare/9e3b715daa1a...439a460224bd
[travis-ci] Build details : http://travis-ci.org/futures/fcrepo4/builds/14125901
* travis-ci leaves
* travis-ci joins01:59
[travis-ci] futures/fcrepo4#1205 (http-fixes - e3cd939 : Chris Beer): The build passed.
[travis-ci] Change view : https://github.com/futures/fcrepo4/compare/439a460224bd...e3cd9396eace
[travis-ci] Build details : http://travis-ci.org/futures/fcrepo4/builds/14125935
* travis-ci leaves
<bljenkins>Project fcrepo-fixity-corrupter build #446: SUCCESS in 3 min 18 sec: http://ci.fcrepo.org/jenkins/job/fcrepo-fixity-corrupter/446/02:21
Project fcrepo-fedora3-federation-connector build #224: UNSTABLE in 11 min: http://ci.fcrepo.org/jenkins/job/fcrepo-fedora3-federation-connector/224/02:29
* kaarefc joins02:30
<bljenkins>Yippie, build fixed!02:31
Project fcrepo-jms-indexer-pluggable build #229: FIXED in 12 min: http://ci.fcrepo.org/jenkins/job/fcrepo-jms-indexer-pluggable/229/
* fcrepo-bot leaves02:46
<bljenkins>Project fcrepo-fixity-corrupter build #447: SUCCESS in 2 min 15 sec: http://ci.fcrepo.org/jenkins/job/fcrepo-fixity-corrupter/447/02:56
Project fcrepo-jms-indexer-pluggable build #230: UNSTABLE in 9 min 15 sec: http://ci.fcrepo.org/jenkins/job/fcrepo-jms-indexer-pluggable/230/03:03
Yippie, build fixed!
Project fcrepo-fedora3-federation-connector build #225: FIXED in 9 min 29 sec: http://ci.fcrepo.org/jenkins/job/fcrepo-fedora3-federation-connector/225/
Project fcrepo-kitchen-sink build #645: STILL FAILING in 9 min 57 sec: http://ci.fcrepo.org/jenkins/job/fcrepo-kitchen-sink/645/03:06
* kaarefc leaves04:02
* kaarefc joins05:08
* mikeAtUVa joins08:38
* ajs6f joins08:43
awoods: ping?08:54
<pivotal-bot>A. "Pholeotic" Soroka added comment: "I'm not sure when this got done, but it appears to have been finished (possibly by cbeer?) several..." https://www.pivotaltracker.com/story/show/5142663908:55
A. "Pholeotic" Soroka finished "Parameterize Dataset HTML response " https://www.pivotaltracker.com/story/show/5142663908:56
A. "Pholeotic" Soroka finished "Document the fcrepo4 "object model"" https://www.pivotaltracker.com/story/show/51571767
<ajs6f>awoods: I've got two epics in front of me for this sprint: the RDF iteration thing, and the beginnings of "self-hosted configuration". Thoughts on which to pursue?09:10
awoods: The RDF iteration thing stands behind Jersey2 migration.09:11
So work on that would first mean helping barmintor complete migration.
<awoods>Wrapping up RDF iteration makes sense (especially as we have not advertised self-hosted configuration), but I hope to make everyone's focus more clear at our sprint-planning call today.09:12
<ajs6f>awoods: That makes sense. I hope barmintor has some tasks to split off. Or that he's just plain done and I can restart RDF iteration using the new forms for injection.09:13
* kaarefc leaves09:14
* ajs6f leaves
* ajs6f joins09:19
<awoods>cbeer: Is your travis-ci account set up with the ff-tech@googlegroups.com email? If so, we should change it to point to fedora-tech@googlegroups.com09:21
* nbanks joins09:22
<ajs6f>awoods: Do you know if we can have a listener on the wiki that posts to this room, in the way that we have a listener for Travis and Github?09:24
afk bbl09:25
<awoods>ajs6f: sure
<ajs6f>I think that could be cool.
<awoods>ajs6f: we just have to create a new wiki account and have the new user "watch" the wiki space
ajs6f: would that be helpful?09:26
* kaarefc joins09:37
<ajs6f>awoods: Sure, whatever makes sense for Confluence. I'd just like to see announcements here when someone writes something.09:50
awoods: Planning meeting at 11 Eastern?09:51
<awoods>ajs6f: yes, planning @11 ET09:53
<ajs6f>awoods: Same to you.
* osmandin joins09:54
<awoods>ajs6f: Now that I think about it, Confluence will need an IRC-client-listener, such as: https://wiki.jasig.org/display/INCU/Confluence+IRC+Logger09:56
<ajs6f>awoods: Okay. Is that something we can do, or must someone at Duraspace do that?
<awoods>ajs6f: Duraspace09:57
<awoods>ajs6f: Dan Davis is the man these days managing most of the Confluence backend.
<ajs6f>awoods: Okay?09:58
<osmandin>afk (meeting 45-60 min.)
<awoods>ajs6f: I am not sure what that means09:59
<ajs6f>awoods: What means?
<ajs6f>I don't understand what you are asking about.
afk bb i 510:02
* ajs6f leaves
<awoods>ajs6f: nevermind. I said "Dan Davis is the man...", and you replied "Okay?", and I said "what?", and you said "what?"
* tecoripa joins10:03
* ermadmix joins10:06
* ajs6f joins10:07
<barmintor>you magnificent bastard
you played the long game, and you won
If I don't want to hard-code factories into the application config, I have to use CDI10:12
* barmintor shakes his fist at the sky in defeat
* ajs6f1 joins
* ajs6f leaves
* ksclarke joins
<barmintor>ajs6f1: icymi, Jersey2 wants you to use CDI or hard-code factories for hk2 dependencies10:14
* ajs6f1 leaves10:17
* kaarefc leaves10:20
* kaarefc joins10:23
* ajs6f joins10:33
CDI, baby. It's all about the CDI.10:35
<pivotal-bot>Chris Beer added "Respond to rhauch about Modeshape relicensing" https://www.pivotaltracker.com/story/show/6095679810:36
<cbeer>awoods: the notification emails are in the .travis.yml in each project.
<awoods>cbeer: perfect, thanks10:37
* ajs6f leaves10:39
* tecoripa leaves10:45
* ajs6f joins10:46
barmintor: Don't feel bad. Type-safe injection is a good thing. Enjoy the comfort and serenity of never having to track down that one mistyped package name in fifteen different XML wiring diagrams, never having to explain to someone why they need to write Java _and_ XML to make anything work, having your IDE show you a nice list of available components to fill an injection slot… type systems are one of the 20th Century's great accomplishments in informa
<barmintor>ajs6f: I'll feel better when I manage to inject a single bean
<ajs6f>JAX-RS < JavaEE, and JavaEE is built over CDI.
* nbanks leaves
<ajs6f>barmintor: I'm right behind you (I'm waiting to finish the RDF iteration stuff). If you can split off tasks, I can take 'em.
* nbanks joins10:47
* kaarefc leaves10:48
* kaarefc joins10:49
<pivotal-bot>Chris Beer estimated "FedoraContent PUT should accept checksum header" as 1 point https://www.pivotaltracker.com/story/show/6087591610:51
Chris Beer started "FedoraContent PUT should accept checksum header" https://www.pivotaltracker.com/story/show/60875916
* ajs6f leaves10:53
* ajs6f joins10:54
* tecoripa joins10:57
* ajs6f leaves10:59
<pivotal-bot>Chris Beer estimated "Remove numPids reference from fcr:identifier API" as 0 points https://www.pivotaltracker.com/story/show/6087747811:00
Chris Beer started "Remove numPids reference from fcr:identifier API" https://www.pivotaltracker.com/story/show/60877478
* github-ff joins
[fcrepo4] cbeer force-pushed http-fixes from e3cd939 to 0a3ed2f: http://git.io/cZrUuA
fcrepo4/http-fixes 4d2d959 Chris Beer: Support HTTP preconditions for DELETE and MOVE operations...
fcrepo4/http-fixes cf212a0 Chris Beer: Support checksums when modifying content at fcr:content...
fcrepo4/http-fixes 0a3ed2f Chris Beer: replace fcr:identifier?numPids parameter with 'count'...
* github-ff leaves
* ajs6f joins11:04
<pivotal-bot>Chris Beer edited "Remove numPids reference from fcr:identifier API" https://www.pivotaltracker.com/story/show/60877478
Chris Beer finished "FedoraContent PUT should accept checksum header" https://www.pivotaltracker.com/story/show/60875916
Chris Beer finished "Remove numPids reference from fcr:identifier API" https://www.pivotaltracker.com/story/show/60877478
<awoods>International Toll-Free Numbers:11:05
Participants can look up their country's toll-free number at http://www.readytalk.com/intl (Note: They'll have to enter our access code (2257295) and then enter their country
<pivotal-bot>Chris Beer added comment: "https://github.com/futures/fcrepo4/pull/156" https://www.pivotaltracker.com/story/show/6087591611:06
Chris Beer added comment: "https://github.com/futures/fcrepo4/pull/156" https://www.pivotaltracker.com/story/show/60877478
<awoods>All: https://wiki.duraspace.org/display/FF/Sprints+B8+and+B9+Planning
<pivotal-bot>Mike Durbin unstarted "Update modeshape to support a quick path to determine if a node has a child with a given name." https://www.pivotaltracker.com/story/show/6059511011:07
Mike Durbin edited "Update modeshape to support a quick path to determine if a node has a child with a given name." https://www.pivotaltracker.com/story/show/60595110
<awoods>barmintor: standup?11:08
<pivotal-bot>Mike Durbin edited "Ensure that the implications of versioning are compatible with policy enforcement" https://www.pivotaltracker.com/story/show/60769426
Chris Beer added comment: "Updated REST API page." https://www.pivotaltracker.com/story/show/6087591611:09
Chris Beer added comment: "Updated REST API page." https://www.pivotaltracker.com/story/show/60877478
<barmintor>on the line
<pivotal-bot>Chris Beer added comment: "Updated REST API page." https://www.pivotaltracker.com/story/show/60887336
* ermadmix_ joins11:11
<pivotal-bot>Ye Cao started "Solr query recipes" https://www.pivotaltracker.com/story/show/5745338811:15
Andrew Woods accepted "Improve position and styling of content replacement form in html view." https://www.pivotaltracker.com/story/show/60895822
* travis-ci joins11:17
[travis-ci] futures/fcrepo4#1207 (http-fixes - 0a3ed2f : Chris Beer): The build passed.
[travis-ci] Change view : https://github.com/futures/fcrepo4/compare/e3cd9396eace...0a3ed2f89840
[travis-ci] Build details : http://travis-ci.org/futures/fcrepo4/builds/14146117
* travis-ci leaves
gregjansen: I can barely hear you. Can you get closer / speak up?11:22
<bljenkins>Project fcrepo-fixity-corrupter build #448: SUCCESS in 1 min 29 sec: http://ci.fcrepo.org/jenkins/job/fcrepo-fixity-corrupter/448/
* kaarefc leaves11:23
<bljenkins>Project fcrepo-kitchen-sink build #646: STILL FAILING in 4 min 8 sec: http://ci.fcrepo.org/jenkins/job/fcrepo-kitchen-sink/646/11:27
* ermadmix leaves11:31
<pivotal-bot>Chris Beer added "Update administrative search to support rdf:type translations (or, at least, make sure it works)" https://www.pivotaltracker.com/story/show/6096250611:34
Scott Prater added "Write recipe for converting sample Fedora 3 content models to Fedora 4" https://www.pivotaltracker.com/story/show/6096348211:41
<mikeAtUVa>I've got to leave the call early, but with regards to versioning, unless someone else wanted to pick up the work, I intend to complete the two tickets in the backlog this sprint and would consider them to be a reasonable stopping point for Versioning in this release.11:44
<awoods>mikeAtUVa: thanks, we can chat later11:45
<osmandin>afp..brb in 10 mins11:52
<awoods>tecoripa: https://wiki.duraspace.org/display/FF/Design+-+Customizable+Search+Index12:06
<pivotal-bot>Scott Prater added "Acceptance test: external search" https://www.pivotaltracker.com/story/show/6096629412:07
<pivotal-bot>Chris Beer added comment: "https://issues.jboss.org/browse/MODE-2098" https://www.pivotaltracker.com/story/show/6080602412:10
Chris Beer accepted "Propose a Modeshape Property API e.g. Property#getNodes() for retrieving multi-valued referenced nodes" https://www.pivotaltracker.com/story/show/60806024
<ajs6f>barmintor: injecting yet?
<barmintor>ajs6f: nope12:14
<pivotal-bot>Scott Prater added "Acceptance test: versioning" https://www.pivotaltracker.com/story/show/60967090
<ajs6f>I like to inject brine into large cuts of meat.
barmintor: Where is it blowing up?
<pivotal-bot>Scott Prater edited "Acceptance test: versioning" https://www.pivotaltracker.com/story/show/6096709012:15
* ermadmix joins
<pivotal-bot>Chris Beer edited "Respond to rhauch about Modeshape relicensing" https://www.pivotaltracker.com/story/show/6095679812:26
<barmintor>osmandin: mute12:32
<ajs6f>tecorpa wins!12:33
<osmandin>barmintor: was mute12:34
<barmintor>osmandin: sorry, thought that was your line.
<ajs6f>Scott forever!12:38
<pivotal-bot>Chris Beer started "Simple Administrative Search" https://www.pivotaltracker.com/story/show/6071379612:45
<barmintor>I don't think benchmarks tell you where bottlenecks are- but they tell you when you can stop fiddling with things.12:48
<pivotal-bot>Scott Prater added "Template for Module API documentation" https://www.pivotaltracker.com/story/show/6097059012:49
<ajs6f>Benchmarks over many different platforms may tell you a little about bottlenecks, but that's not an efficient way to do that.
* ermadmix leaves12:53
* mikeAtUVa leaves
* ermadmix joins12:58
<ajs6f>awoods: If you are concerned about performance, we need to spend lots of time turning all the code into map-reduce algorithms and figuring our a way to run Hadoop on GPUs.13:05
That's the practical approach.
<osmandin>premature optimization root of all evil13:07
<ajs6f>It's not premature. We should have been optimizing about a decade ago.
<osmandin>i think some things can be optimized when you're writing code or designing your system... but some stuff has to wait until you have hard data13:13
<ajs6f>We had hard data for Fedora3's poor performance for many years. We just never had the time or motivations to do anything about it.13:14
<cbeer>ajs6f: and i think we discovered earlier that fcrepo3 performance is really bad on updates, not ingest.13:15
<ajs6f>cbeer: aye. And I'm still suspicious about all the XML processing that goes on for updates...
<osmandin>afk (lunch)13:16
<escowles>cbeer ajs6f : so we should probably come up with a better recipe that what we have now: e.g., load 10k objects, then update 1k of them, do some searches, etc.
<ajs6f>Do we need a cross-hatched profile form? Repo form vs. workflow form?13:17
<awoods>barmintor: You said you wrote an email about tasks?
<ajs6f>I can imagine a couple of workfow profiles that could be matched with repo profiles to generate a performance test.
<awoods>barmintor: Where did that get sent? I am not seeing it.
<barmintor>where the what got the who?
<pivotal-bot>Scott Prater added "Create sample Performance Testing Profile (PTP) for common use case" https://www.pivotaltracker.com/story/show/6097389213:19
A. "Pholeotic" Soroka started "Create sample Performance Testing Profile (PTP) for common use case" https://www.pivotaltracker.com/story/show/60973892
Scott Prater added "Create use case for common preservation repository profile" https://www.pivotaltracker.com/story/show/6097424213:22
Scott Prater added comment: "@ajs6f : I'll provide the use case: https://www.pivotaltracker.com/story/show/60974242" https://www.pivotaltracker.com/story/show/6097389213:23
Eric James added "Review strategy and scope of implementing solr index on repo" https://www.pivotaltracker.com/story/show/60974368
Eric James started "Review strategy and scope of implementing solr index on repo" https://www.pivotaltracker.com/story/show/60974368
Eric James added "support for dynamic directories within filesystem connector" https://www.pivotaltracker.com/story/show/6097465213:25
Eric James added "benchmarking filesize limits for REST API file ingest" https://www.pivotaltracker.com/story/show/6097476613:26
Eric James edited "benchmarking filesize limits for REST API file ingest" https://www.pivotaltracker.com/story/show/6097476613:27
Scott Prater added comment: "@ajs6f : I think a crosshatched profile is a good idea. Nouns and adjectives, then verbs: how many objec..." https://www.pivotaltracker.com/story/show/60974242
* tecoripa leaves
<pivotal-bot>A. "Pholeotic" Soroka added comment: "I think it's two kinds of profiles. A test profile comprises a repo profile and a workflow profile..." https://www.pivotaltracker.com/story/show/6097424213:29
<pivotal-bot>A. "Pholeotic" Soroka added comment: "https://wiki.duraspace.org/display/FF/Test+-+Repository+Profiles+for+Performance+Testing" https://www.pivotaltracker.com/story/show/60973892
* ermadmix_ leaves13:33
<pivotal-bot>Esme Cowles added comment: "@ajs6f: so a repo profile is: server hardware, network connectivity between ingest client and server, repo c..." https://www.pivotaltracker.com/story/show/6097424213:45
Esme Cowles started "Make property values in HTML UI editable" https://www.pivotaltracker.com/story/show/6078128013:46
Esme Cowles added comment: "I think there are two issues here, plus one additional feature to add before this is ready: ""13:50
1. I had some F..." https://www.pivotaltracker.com/story/show/60781280
A. "Pholeotic" Soroka added comment: "No, I was thinking that a repo profile is along the lines shown here: ""
https://wiki.duraspace.org/..." https://www.pivotaltracker.com/story/show/60974242
* ermadmix leaves13:52
<pivotal-bot>Esme Cowles added comment: "I'm going to update the Repository Profiles for Performance Testing page into three sections: ""14:03
1. Repository ..." https://www.pivotaltracker.com/story/show/60974242
A. "Pholeotic" Soroka added comment: "I'd really like to keep "number of objects, kind of objects" separate from "do these things agains..." https://www.pivotaltracker.com/story/show/6097424214:04
Osman Din started "Test Basic AuthZ" https://www.pivotaltracker.com/story/show/6061567014:08
Esme Cowles added comment: "I'm assuming that we are starting from an empty repo at the beginning of the test. So the test procedure wo..." https://www.pivotaltracker.com/story/show/6097424214:09
A. "Pholeotic" Soroka added comment: "I'm _not_ assuming you start from an empty repo. I think that's not a good assumption to make. How..." https://www.pivotaltracker.com/story/show/6097424214:11
<osmandin>Greg, does sometime after 3 pm work for you today for our discussion?14:17
I mean at or after 3 pm14:18
<pivotal-bot>Chris Beer finished "Simple Administrative Search" https://www.pivotaltracker.com/story/show/60713796
* github-ff joins
[fcrepo4] cbeer force-pushed sparql from 993bb63 to b4d4496: http://git.io/Rl-hQw
fcrepo4/sparql 76204e2 Chris Beer: define all the Dublin Core elements as pre-defined properties, and use the dc/elements/1.1 namespace instead of dcterms.
fcrepo4/sparql 13cdace Chris Beer: Implement basic SPARQL-based administrative search
fcrepo4/sparql 5990834 Chris Beer: wire in a sparql search form.
* github-ff leaves
<pivotal-bot>A. "Pholeotic" Soroka added comment: "https://wiki.duraspace.org/display/FF/Test+-+Workflow+Profiles+for+Performance+Testing" https://www.pivotaltracker.com/story/show/6097424214:19
Chris Beer added comment: "Addressed code-review comments." https://www.pivotaltracker.com/story/show/60713796
Esme Cowles added comment: "I certainly don't expect repos to start fresh every morning. But if the contents of the repository impact t..." https://www.pivotaltracker.com/story/show/6097424214:21
* gregjansen joins14:23
<pivotal-bot>A. "Pholeotic" Soroka added comment: "I don't think we disagree about what to specify. I'm saying that we ought to specify the two thing..." https://www.pivotaltracker.com/story/show/60974242
A. "Pholeotic" Soroka added comment: "https://wiki.duraspace.org/display/FF/Test+-+Platform+Profiles+for+Performance+Testing" https://www.pivotaltracker.com/story/show/6097424214:29
* travis-ci joins14:31
[travis-ci] futures/fcrepo4#1209 (sparql - b4d4496 : Chris Beer): The build passed.
[travis-ci] Change view : https://github.com/futures/fcrepo4/compare/993bb634d808...b4d449685647
[travis-ci] Build details : http://travis-ci.org/futures/fcrepo4/builds/14156570
* travis-ci leaves
<pivotal-bot>Esme Cowles added comment: "I agree with all of your points about keeping the initial repo setup separate. In fact, it might make sense..." https://www.pivotaltracker.com/story/show/60974242
Gregory Jansen added "Prepare and test UNC content models for F4" https://www.pivotaltracker.com/story/show/6098251814:34
Gregory Jansen edited "Prepare and test UNC content models for F4" https://www.pivotaltracker.com/story/show/60982518
A. "Pholeotic" Soroka added comment: "I agree with that entirely. Right now, I'm just setting up "human-readable" templates so that at l..." https://www.pivotaltracker.com/story/show/60974242
A. "Uriculoneous" Soroka added comment: "https://wiki.duraspace.org/display/FF/Test+-+Performance+Test+Profiles" https://www.pivotaltracker.com/story/show/6097389214:35
<bljenkins>Project fcrepo-fixity-corrupter build #449: SUCCESS in 1 min 30 sec: http://ci.fcrepo.org/jenkins/job/fcrepo-fixity-corrupter/449/14:42
<pivotal-bot>Gregory Jansen added "Run benchmark tests for single Fedora node on server hardware." https://www.pivotaltracker.com/story/show/6098381614:46
Gregory Jansen added "Run benchmark tests of F4 4/8 node cluster at UNC" https://www.pivotaltracker.com/story/show/6098396814:48
<bljenkins>Yippie, build fixed!14:49
Project fcrepo-jms-indexer-pluggable build #232: FIXED in 8 min 11 sec: http://ci.fcrepo.org/jenkins/job/fcrepo-jms-indexer-pluggable/232/
Project fcrepo-kitchen-sink build #647: STILL FAILING in 7 min 51 sec: http://ci.fcrepo.org/jenkins/job/fcrepo-kitchen-sink/647/14:50
* nbanks leaves14:58
* nbanks joins15:00
* ermadmix joins15:01
<osmandin>gregjansen: do you have time to chat today about auth (as discussed on phone)
<ajs6f>barmintor: thoughts? still no good?15:03
<barmintor>ajs6f: Trying to figure out how to get the ContainerWrapper to work15:04
ajs6f: let me push a branch
<ajs6f>k. just shout when/whereever
Still on branch jeresey2?
<gregjansen>osmandin: yeah sure. I need a few minutes, but we could talk at 3:30 EST.. hangout?15:05
<ajs6f>afk bbs15:06
<osmandin>gregjansen: great..see you at hangout at 330
* ajs6f leaves
* github-ff joins
[fcrepo4] barmintor created cdi (+1 new commit): http://git.io/F6CjdA
fcrepo4/cdi 61e29f5 Benjamin Armintor: start trying to get Weld to work with ContainerWrapper
* github-ff leaves
* travis-ci joins15:08
[travis-ci] futures/fcrepo4#1211 (cdi - 61e29f5 : Benjamin Armintor): The build has errored.
[travis-ci] Change view : https://github.com/futures/fcrepo4/commit/61e29f5e0c9d
[travis-ci] Build details : http://travis-ci.org/futures/fcrepo4/builds/14158979
* travis-ci leaves
* ajs6f joins15:11
* nbanks leaves15:12
<barmintor>ajs6f http://git.io/F6CjdA15:15
<ajs6f>barmintor: that's the cdi-ification bracnh?15:20
<osmandin>does this mean getting rid of spring eventually?
<ajs6f>it'll take me a while to pull it (fluttery network). Is it building (if not, where can I step in)?15:21
osmandin: No, but it raises the question. There is a CDI-Spring bridge.
osmandin: But we'd like to keep things simple, of course.
<osmandin>ajs6f: ok thx..i was trying openwebeans w/ spring15:22
<barmintor>osmandin: I honestly can't tell.
ajs6f: We'll just use the Spring-CDI bridge behind the HK2-CDI bridge!15:23
<ajs6f>osmandin: Any opnions?
cool. owb is said to be very fast, faster than Weld.
barmintor: Are you building up i-test gear from scratch? Did you look at Pax Exam?
* mikeAtUVa joins15:24
<barmintor>ajs6f: I'm trying to synthesize our IT gear with the examples I've found online15:25
IDK what pax exam is
<ajs6f>"A Bridge Too Far"
barmintor: I think we definitely need to look at Pax Exam. Or something else like it, but that's the framework with which I am familiar.
<barmintor>I totally belive you.
<bljenkins>Project fcrepo-fixity-corrupter build #450: SUCCESS in 1 min 8 sec: http://ci.fcrepo.org/jenkins/job/fcrepo-fixity-corrupter/450/15:27
<ajs6f>I'd give you a link, but our wifi here is so awful that I am literally restricted to text.15:28
You can find it quick enough. It's an Ops4j joint.
It's an integration test framework that supports OSGi, CDI and other suchlike containers.
Think of it as the CDI equiv of the Spring test framework we've been using.15:29
And it has a cute mascot.15:30
barmintor: Have you been working entirely at the HTTP layer? IOW, are you trying to accept Spring beans from the kernel and use them in the HTTP layer?
<barmintor>ajs6f: Step 1: An IT spins up, and fails because it can't get Spring beans (in progress)15:31
<bljenkins>Project fcrepo-kitchen-sink build #648: STILL FAILING in 4 min 12 sec: http://ci.fcrepo.org/jenkins/job/fcrepo-kitchen-sink/648/
* ermadmix leaves15:32
<ajs6f>(via that bridge or by some other means)
barmintor: Two possibilities: keep kernel as spring (what you are doing) in which case we will either need a bridge or we will have to write our own simple bridge. Or, redo kernel in CDI and make other modules accept CDI beans.
The reason we didn't try that months ago (when I started experimenting with CDI) is because we got all Spring-tastic on the JMS stuff in fcrepo-jms.15:33
* ermadmix joins
<ajs6f>And I didn't want to try to rewrite that.
<barmintor>ajs6f: You misapprehend what I am doing- I merely want to get to a point where I have a test whose success I can work towards.
ajs6f: I think keeping Spring around under CDI is a last resort.
<ajs6f>barmintor: I think we have lots of good tests. What is the nature of the new one you want you introduce?15:34
<barmintor>One of the existing ones, running with CDI-y stuff
* github-ff joins15:35
[fcrepo4] awoods pushed 1 new commit to master: http://git.io/NpG6Jw
fcrepo4/master ecbf380 Andrew Woods: Update Travis-CI email address
* github-ff leaves
<ajs6f>barmintor: Agreed, we do not want to ride two horses.
As my grandfather said, with one tuchus, you cannot dance at two weddings.
Right, That's what I was meant by "having two possiblities".
In order to get a test working, we can either supply it with the Spring beans it now expects, or convert it to expect CDI beans.
<barmintor>Right. If I take the Spring out of the web layer, everything falls apart15:36
<ajs6f>barmintor: Do you want to talk (hangout) about this real quick? I feel like we're talking past each other.
<barmintor>I have a meeting in 20 minutes
<ajs6f>barmintor: Right. I'm looking at your commit (above) and seeing that you seem to be rebuilding some of the functionality of the Spring stuff in the web layer. I'm saying that Pax Exam does that.
* barmintor is trying to pull Pax Exam into that branch right now15:37
<ajs6f>barmintor: okay. Maybe later?
(Spring stuff for tests, that is.)
Rock on, you rockin' rocker.
PE _should_ be able to manage your bean container and your HTTP container for you. I've used it to do that, but that was in the spring and I don't know wether we kept that branch around. It was one of the "experimental CDI" branches I cut for cbeer to examine.15:39
<awoods>barmintor/ajs6f: I would like to have a serious talk about how far we go with the Jersey2 work before we say "enough is enough". This has taken on a life of its own, and the driving use-cases are not clearly high-priority (yet).
<barmintor>awoods: tomorrow morning?15:40
<awoods>barmintor: after stand-up? that would work.
<ajs6f>awoods: I feel like we backed into this without assembling the right away team. barmintor is only dealing with CDI because he has too, whereas I spent severl weeks in the spring actually doing a CDI reimpl of the kernel and HTTP. I agree that the use cases aren't there right now to justify a massive reallocation of effort. But part of our conversation should be about the future direction for our plumbing and how we resource it.15:41
<awoods>ajs6f: agreed15:42
<ajs6f>awoods/barmintor: the new thing that cbeer and now barmintor have turned up is that one of our most important libraries is moving out from under us towards the CDI space. No matter what, we have to decide what to do about that. We knew that CDI was interesting on some level, and we know that there are other choices for JAX-RS (and other choices to support HTTP API, for that matter). We know that because we tried most of 'em.15:45
We don't want to get into the same box we were in for years with Fedora3: an obsolete web services library that made our lives painful.15:46
But we don't want to let one library (no matter how important) drive our component selection.15:47
<barmintor>it's also true that writing further and further away from the intended container with bridge libraries is not exactly going to be a boon to performance
<ajs6f>Good gracious me, no.
* travis-ci joins15:48
[travis-ci] futures/fcrepo4#1212 (master - ecbf380 : Andrew Woods): The build passed.
[travis-ci] Change view : https://github.com/futures/fcrepo4/compare/97ae37e48c81...ecbf3800e7c8
[travis-ci] Build details : http://travis-ci.org/futures/fcrepo4/builds/14160358
* travis-ci leaves
<ajs6f>We chose Spring because we had some experience with it and were able to move quick through alpha using it. This kind of concern is one of the reasons I wanted to revisit that decision and make a more conscious, informed choice about wiring and containers.
<awoods>We need to design with flexibility in mind. I do not think that means we need to always be bleeding-edge.15:49
<barmintor>I agree, though it is worth noting that JAXRS2/CDI/Jersey2 stuff is I think 2-3 years old now
<ajs6f>I don't want to choose for bleeding edge. I want to choose for the centers of strong communities, and I don't think that pushes towards or away Spring or CDI.15:50
<barmintor>my experience with Fedora 3 -> Spring leads me to think that changing to CDI will accompany a major version rev15:51
<ajs6f>CDI is at the heart of JavaEE. It's not experimental stuff, although as barmintor points out, it's not as perfectly documented as we would like, either. I would say that MODE probably has a smaller community around it than JavaEE. {grin}
<barmintor>whether that's 3->4 or 4->5 is a separate question
<ajs6f>5? Seriously?15:52
Unless you are thinking of a very different rate of release than we had for the last decade.
<bljenkins>Project fcrepo-fixity-corrupter build #451: SUCCESS in 1 min 2 sec: http://ci.fcrepo.org/jenkins/job/fcrepo-fixity-corrupter/451/15:53
<ajs6f>(Although I don't disagree that the choice of DI and C framework is a major, major choice.)
<barmintor>I think it was a structural mistake not to release Fedora 4 on Spring in 2010, so yes to different rate :)
<ajs6f>Mm. That's a fair point.
<barmintor>I just think about what retrofitting Spring ont FCFG was like (and it was not pleasant, despite the functional analogs), and look at CDI v. Spring… it's just not a mid-version change15:55
<bljenkins>Project fcrepo-kitchen-sink build #649: STILL FAILING in 2 min 59 sec: http://ci.fcrepo.org/jenkins/job/fcrepo-kitchen-sink/649/15:56
<ajs6f>I'm not denying that this is a major decision (that's why I've been bleating about it for months) but there are some crucial differences between using a tugboat to haul forward the the giant bloated carcass of Fedora3 in the harbor of Spring and refactoring a recently-minted codebase that has never even been beta-released.15:57
This wouldn't be a mid-version change, either, because we've never released Fedora 4.0.15:58
<barmintor>right, right: I'm just saying 4.0 is decision time for CDI or Spring
<ajs6f>Okay, but are you saying that if we don't release with CDI (assuming we end up wanting to) we should wait for another couple of years, or are you saying that whenever we do release with CDI (assuming we want to) we should signify it with an appropriate version bump?16:00
<awoods>ajs6f/barmintor: I think there are two decisions to be made: CDI vs. Spring, and now vs. after Christmas.
<ajs6f>I am a little worried, too, that if we decide to go to CDI and release with Spring anyway, we'll cause some confusion in an already-unruly community of devs.
awoods: After Christmas but before March?16:01
<awoods>ajs6f: sure, potentially (timing is part of the discussion)
<barmintor>ajs6f: I'm saying the change will end up feeling like a major rev regardless
<awoods>ajs6f/barmintor: Given the resources that have been put onto sprints 8&9, it will be a significant failure if we do not put out a user-facing release sometime during sprint 9.16:02
<ajs6f>awoods: After C. but before M. sounds quite reasonable to me. But waiting until well into the spring makes me a little anxious. I just don't see why we'll be more ready to tackle plumbing the longer we wait.
awoods; Agreed, and I'm not pressing to do this _now_.
Especially because barmintor is being beaten about the head and neck by it.16:03
<awoods>ajs6f: and I am not pressing to not do it.
<ajs6f>awoods: But I would like to commit to choosing a framework _and_ at least starting work on the reimpl _before_ March.
<awoods>ajs6f/barmintor: I think the plan is to discuss this tomorrow morning after stand-up.
<ajs6f>awoods: I hate to say it, but I may not be here for that. Unexpected medical appointment.16:04
<awoods>ajs6f: I think it is safe to say that your position is known.
<awoods>ajs6f: there we have it
<ajs6f>This is why we should use ML. Our reimpl could just be an elegant proof that a reimpl exists.
afk bbs16:06
* ajs6f leaves
* ajs6f joins16:09
<cbeer>ajs6f: can you weigh on in https://www.pivotaltracker.com/story/show/60883814, https://www.pivotaltracker.com/story/show/60885748 and https://www.pivotaltracker.com/story/show/60885870?16:13
<pivotal-bot>feature: Normalize responses for Nodes with a trailing slash (unscheduled) / owner:
<cbeer>oh, and https://www.pivotaltracker.com/story/show/6088719016:14
<pivotal-bot>feature: Could/should fcr:export/fcr:import and fcr:backup/fcr:restore get grouped under a single endpoint? (unscheduled) / owner:
A. "Uriculoneous" Soroka added comment: "I agree with this, as long as we are talking about URLs that end in paths. I think that URLs th..." https://www.pivotaltracker.com/story/show/60883814
A. "Uriculoneous" Soroka added comment: "200. We have created an identifier, but we have not created a resource. It's perfectly possible..." https://www.pivotaltracker.com/story/show/6088574816:15
Chris Beer added comment: "+1. I think it's only paths that exhibit weird responses anyway." https://www.pivotaltracker.com/story/show/60883814
Chris Beer deleted "Should POST fcr:identifier send back status 200 or 201?" https://www.pivotaltracker.com/story/show/6088574816:16
Chris Beer deleted "If fcr:identifier returns only a single ID, should we send it in the Location header too?" https://www.pivotaltracker.com/story/show/60885756
<pivotal-bot>A. "Uriculoneous" Soroka added comment: "A plain 500 with a very good explanatory message?" https://www.pivotaltracker.com/story/show/6088587016:18
A. "Uriculoneous" Soroka added comment: "_Maybe_ a 417 Expectation Failed? " https://www.pivotaltracker.com/story/show/6088587016:19
A. "Uriculoneous" Soroka added comment: "It seems like we are mixing verbs into our URLs. " https://www.pivotaltracker.com/story/show/60887190
Chris Beer added comment: "They've been mixed. Suggestions for unmixing them?" https://www.pivotaltracker.com/story/show/6088719016:20
A. "Uriculoneous" Soroka added comment: "How about a header on messages to/from the ordinary resource endpoint indicating that rather th..." https://www.pivotaltracker.com/story/show/6088719016:21
A. "Uriculoneous" Soroka added comment: "And for multi-import/export (aka backup/restore)... I dunno. This is where HTTP isn't fun. We'r..." https://www.pivotaltracker.com/story/show/6088719016:22
Chris Beer added comment: "conneg gone crazy? I guess we could do this with a catch-all GET on FedoraNodes.. Sorting it out for POSTs c..." https://www.pivotaltracker.com/story/show/60887190
A. "Uriculoneous" Soroka added comment: "Is "backup" supposed to be able to export only a portion of the repo, or only the _whole thing_..." https://www.pivotaltracker.com/story/show/6088719016:24
Chris Beer added comment: "Whole thing. It's a JCR concept we're just passing through." https://www.pivotaltracker.com/story/show/6088719016:25
A. "Uriculoneous" Soroka added comment: "Then frankly, madam, I don't give a damn. No one in their right mind is going to serialize thei..." https://www.pivotaltracker.com/story/show/6088719016:27
<ajs6f>afk bbs16:38
* ajs6f leaves
* ajs6f joins16:41
* ajs6f leaves16:58
* ajs6f joins
* osmandin leaves17:00
* tecoripa joins17:01
* ajs6f leaves17:02
* github-ff joins17:06
[fcrepo4] cbeer force-pushed http-fixes from 0a3ed2f to 9d7cee9: http://git.io/cZrUuA
fcrepo4/http-fixes 5c2cda7 Chris Beer: After creating a new workspace, return the path to the workspace in the location header. Fixes #60886000
fcrepo4/http-fixes e99f818 Chris Beer: Support HTTP preconditions for DELETE and MOVE operations...
fcrepo4/http-fixes f6df776 Chris Beer: Support checksums when modifying content at fcr:content...
* github-ff leaves
* nbanks joins17:18
* gregjansen leaves
<pivotal-bot>Andrew Woods added comment: "Sounds good. I will await the new and improved version." https://www.pivotaltracker.com/story/show/6078128017:21
* travis-ci joins17:22
[travis-ci] futures/fcrepo4#1213 (http-fixes - 9d7cee9 : Chris Beer): The build passed.
[travis-ci] Change view : https://github.com/futures/fcrepo4/compare/0a3ed2f89840...9d7cee91ad94
[travis-ci] Build details : http://travis-ci.org/futures/fcrepo4/builds/14165276
* travis-ci leaves
* nbanks leaves17:26
* github-ff joins17:30
[fcrepo4] awoods pushed 1 new commit to master: http://git.io/9LzVFg
fcrepo4/master 7564c9b Andrew Woods: Merge pull request #147 from futures/sparql...
* github-ff leaves
<pivotal-bot>Andrew Woods delivered "Simple Administrative Search" https://www.pivotaltracker.com/story/show/60713796
Andrew Woods edited "Triage GitHub-Issues" https://www.pivotaltracker.com/story/show/6045879617:34
<bljenkins>Project fcrepo-fixity-corrupter build #452: SUCCESS in 2 min 33 sec: http://ci.fcrepo.org/jenkins/job/fcrepo-fixity-corrupter/452/17:37
Project fcrepo-fedora3-federation-connector build #230: FAILURE in 6 min 18 sec: http://ci.fcrepo.org/jenkins/job/fcrepo-fedora3-federation-connector/230/17:41
Project fcrepo-jms-indexer-pluggable build #235: UNSTABLE in 10 min: http://ci.fcrepo.org/jenkins/job/fcrepo-jms-indexer-pluggable/235/17:45
* travis-ci joins17:48
[travis-ci] futures/fcrepo4#1215 (master - 7564c9b : Andrew Woods): The build passed.
[travis-ci] Change view : https://github.com/futures/fcrepo4/compare/ecbf3800e7c8...7564c9b5fa3c
[travis-ci] Build details : http://travis-ci.org/futures/fcrepo4/builds/14166579
* travis-ci leaves
<bljenkins>Project fcrepo-fixity-corrupter build #453: SUCCESS in 2 min 37 sec: http://ci.fcrepo.org/jenkins/job/fcrepo-fixity-corrupter/453/18:11
Yippie, build fixed!18:19
Project fcrepo-fedora3-federation-connector build #231: FIXED in 10 min: http://ci.fcrepo.org/jenkins/job/fcrepo-fedora3-federation-connector/231/
Yippie, build fixed!
Project fcrepo-jms-indexer-pluggable build #236: FIXED in 10 min: http://ci.fcrepo.org/jenkins/job/fcrepo-jms-indexer-pluggable/236/
<cbeer>awoods: if you haven't seen it, https://github.com/futures/fcrepo4/pull/156 is ready to go18:20
<pivotal-bot>Chris Beer accepted "Simple Administrative Search" https://www.pivotaltracker.com/story/show/60713796
<awoods>cbeer: I am reviewing it now
<bljenkins>Project fcrepo-kitchen-sink build #650: STILL FAILING in 10 min: http://ci.fcrepo.org/jenkins/job/fcrepo-kitchen-sink/650/18:21
<awoods>cbeer: you got a lot of pivotal mileage out of that one PR.18:23
<cbeer>just inflating my numbers.18:24
if i were smarter, i'd have made some of those tickets have more points.
especially the 0 point ones.
* ksclarke leaves18:25
* ermadmix leaves18:41
<pivotal-bot>Andrew Woods added "DELETE Non-Existent Workspace" https://www.pivotaltracker.com/story/show/6100283818:43
* ermadmix joins
<pivotal-bot>Andrew Woods edited "FedoraContent PUT should accept checksum header" https://www.pivotaltracker.com/story/show/60875916
Andrew Woods edited "Remove numPids reference from fcr:identifier API" https://www.pivotaltracker.com/story/show/60877478
* escowles leaves18:44
* github-ff joins
[fcrepo4] barmintor pushed 1 new commit to cdi: http://git.io/jT2Dmg
fcrepo4/cdi 995e305 Benjamin Armintor: forget Weld- Let's try Pax Exam\!
* github-ff leaves
<pivotal-bot>Andrew Woods added comment: "Pending one comment relating to removal of Versioning logic from FedoraNodes.18:45
https://github.com/futures/fc..." https://www.pivotaltracker.com/story/show/60887246
* travis-ci joins18:46
[travis-ci] futures/fcrepo4#1216 (cdi - 995e305 : Benjamin Armintor): The build has errored.
[travis-ci] Change view : https://github.com/futures/fcrepo4/compare/61e29f5e0c9d...995e3050d16a
[travis-ci] Build details : http://travis-ci.org/futures/fcrepo4/builds/14170350
* travis-ci leaves
<pivotal-bot>Andrew Woods delivered "Parameterize Dataset HTML response " https://www.pivotaltracker.com/story/show/5142663918:47
Andrew Woods accepted "Parameterize Dataset HTML response " https://www.pivotaltracker.com/story/show/51426639
Andrew Woods estimated "Document the fcrepo4 "object model"" as 1 point https://www.pivotaltracker.com/story/show/5157176718:57
Andrew Woods delivered "Document the fcrepo4 "object model"" https://www.pivotaltracker.com/story/show/51571767
Andrew Woods accepted "Document the fcrepo4 "object model"" https://www.pivotaltracker.com/story/show/51571767
* github-ff joins18:58
[fcrepo4] cbeer force-pushed http-fixes from 9d7cee9 to f3dd9d0: http://git.io/cZrUuA
fcrepo4/http-fixes 8a4f7cb Chris Beer: update response after creating a node to include the full URI to the...
fcrepo4/http-fixes 30ef4d5 Chris Beer: After creating a new workspace, return the path to the workspace in the location header. Fixes #60886000
fcrepo4/http-fixes d097e17 Chris Beer: Support HTTP preconditions for DELETE and MOVE operations...
* github-ff leaves
<bljenkins>Project fcrepo-fixity-corrupter build #454: SUCCESS in 1 min 24 sec: http://ci.fcrepo.org/jenkins/job/fcrepo-fixity-corrupter/454/19:06
<pivotal-bot>Andrew Woods added "Default Storage Parameter Directory: tmp to cwd" https://www.pivotaltracker.com/story/show/61004336
Andrew Woods edited "Default Storage Parameter Directory: tmp to cwd" https://www.pivotaltracker.com/story/show/6100433619:07
Andrew Woods edited "Single system property to specify directory for all fcrepo content" https://www.pivotaltracker.com/story/show/60788492
Andrew Woods estimated "Single system property to specify directory for all fcrepo content" as 1 point https://www.pivotaltracker.com/story/show/60788492
Andrew Woods edited "Single system property to specify directory for all fcrepo content" https://www.pivotaltracker.com/story/show/60788492
Andrew Woods edited "Prepare and test UNC content models for F4" https://www.pivotaltracker.com/story/show/6098251819:08
<bljenkins>Project fcrepo-fedora3-federation-connector build #232: UNSTABLE in 6 min 33 sec: http://ci.fcrepo.org/jenkins/job/fcrepo-fedora3-federation-connector/232/19:11
* travis-ci joins19:13
[travis-ci] futures/fcrepo4#1217 (http-fixes - f3dd9d0 : Chris Beer): The build passed.
[travis-ci] Change view : https://github.com/futures/fcrepo4/compare/9d7cee91ad94...f3dd9d011fd3
[travis-ci] Build details : http://travis-ci.org/futures/fcrepo4/builds/14171159
* travis-ci leaves
* ermadmix leaves19:14
<pivotal-bot>Andrew Woods edited "Use Maven parallelism in CI" https://www.pivotaltracker.com/story/show/5960319019:20
Andrew Woods edited "Improve/update Maven descriptive metadata" https://www.pivotaltracker.com/story/show/57310572
Andrew Woods delivered "Add DELETE endpoint for workspaces" https://www.pivotaltracker.com/story/show/6088733619:22
Andrew Woods delivered "Fix response body from creating a node" https://www.pivotaltracker.com/story/show/60887246
Andrew Woods delivered "Fix created workspace Location" https://www.pivotaltracker.com/story/show/60886000
Andrew Woods delivered "Support preconditions on FedoraNodes#delete" https://www.pivotaltracker.com/story/show/60884508
Andrew Woods delivered "FedoraContent PUT should accept checksum header" https://www.pivotaltracker.com/story/show/60875916
Andrew Woods delivered "Remove numPids reference from fcr:identifier API" https://www.pivotaltracker.com/story/show/60877478
Chris Beer accepted "Add DELETE endpoint for workspaces" https://www.pivotaltracker.com/story/show/6088733619:24
Chris Beer accepted "Fix response body from creating a node" https://www.pivotaltracker.com/story/show/60887246
Chris Beer accepted "Fix created workspace Location" https://www.pivotaltracker.com/story/show/60886000
Chris Beer accepted "Support preconditions on FedoraNodes#delete" https://www.pivotaltracker.com/story/show/60884508
Chris Beer accepted "FedoraContent PUT should accept checksum header" https://www.pivotaltracker.com/story/show/60875916
Chris Beer accepted "Remove numPids reference from fcr:identifier API" https://www.pivotaltracker.com/story/show/60877478
* github-ff joins19:25
[fcrepo4] awoods pushed 1 new commit to master: http://git.io/EnySHQ
fcrepo4/master e5ba8d6 Andrew Woods: Merge pull request #156 from futures/http-fixes...
* github-ff leaves
* tecoripa leaves19:26
<pivotal-bot>Andrew Woods added "Collect Stakeholder Configuration Management Needs" https://www.pivotaltracker.com/story/show/6100533219:28
<bljenkins>Project fcrepo-fixity-corrupter build #455: SUCCESS in 1 min 49 sec: http://ci.fcrepo.org/jenkins/job/fcrepo-fixity-corrupter/455/19:31
<pivotal-bot>Chris Beer started "DELETE Non-Existent Workspace" https://www.pivotaltracker.com/story/show/6100283819:33
Chris Beer finished "DELETE Non-Existent Workspace" https://www.pivotaltracker.com/story/show/61002838
* github-ff joins19:36
[fcrepo4] cbeer created http-fixes (+1 new commit): http://git.io/5UBXHA
fcrepo4/http-fixes 3f841c7 Chris Beer: Return a 404 Not Found when deleting a workspace that doesn't exist.
* github-ff leaves
<pivotal-bot>Chris Beer added comment: "https://github.com/futures/fcrepo4/pull/157" https://www.pivotaltracker.com/story/show/61002838
* github-ff joins
[fcrepo4] cbeer opened pull request #157: Return a 404 Not Found when deleting a workspace that doesn't exist. (master...http-fixes) http://git.io/HGOLJQ
* github-ff leaves
* travis-ci joins19:38
[travis-ci] futures/fcrepo4#1219 (master - e5ba8d6 : Andrew Woods): The build passed.
[travis-ci] Change view : https://github.com/futures/fcrepo4/compare/7564c9b5fa3c...e5ba8d65eecf
[travis-ci] Build details : http://travis-ci.org/futures/fcrepo4/builds/14172343
* travis-ci leaves
<bljenkins>Project fcrepo-jms-indexer-pluggable build #238: UNSTABLE in 8 min 41 sec: http://ci.fcrepo.org/jenkins/job/fcrepo-jms-indexer-pluggable/238/
* travis-ci joins19:48
[travis-ci] futures/fcrepo4#1220 (http-fixes - 3f841c7 : Chris Beer): The build passed.
[travis-ci] Change view : https://github.com/futures/fcrepo4/commit/3f841c782b81
[travis-ci] Build details : http://travis-ci.org/futures/fcrepo4/builds/14172629
* travis-ci leaves
* ermadmix joins19:52
<bljenkins>Project fcrepo-fixity-corrupter build #456: SUCCESS in 1 min 5 sec: http://ci.fcrepo.org/jenkins/job/fcrepo-fixity-corrupter/456/19:54
Yippie, build fixed!19:59
Project fcrepo-jms-indexer-pluggable build #239: FIXED in 6 min 53 sec: http://ci.fcrepo.org/jenkins/job/fcrepo-jms-indexer-pluggable/239/
<pivotal-bot>Andrew Woods delivered "DELETE Non-Existent Workspace" https://www.pivotaltracker.com/story/show/6100283820:14
* github-ff joins
[fcrepo4] awoods pushed 1 new commit to master: http://git.io/Of3zlg
fcrepo4/master 9eaac0f Andrew Woods: Merge pull request #157 from futures/http-fixes...
* github-ff leaves
<bljenkins>Project fcrepo-fixity-corrupter build #457: SUCCESS in 1 min 53 sec: http://ci.fcrepo.org/jenkins/job/fcrepo-fixity-corrupter/457/20:19
* ermadmix leaves20:23
<pivotal-bot>Andrew Woods added "Fix Checksum Mismatch Message" https://www.pivotaltracker.com/story/show/6100798820:24
Andrew Woods edited "Fix Checksum Mismatch Message" https://www.pivotaltracker.com/story/show/6100798820:25
<bljenkins>Project fcrepo-jms-indexer-pluggable build #240: UNSTABLE in 7 min 43 sec: http://ci.fcrepo.org/jenkins/job/fcrepo-jms-indexer-pluggable/240/
* travis-ci joins20:26
[travis-ci] futures/fcrepo4#1222 (master - 9eaac0f : Andrew Woods): The build passed.
[travis-ci] Change view : https://github.com/futures/fcrepo4/compare/e5ba8d65eecf...9eaac0f929ae
[travis-ci] Build details : http://travis-ci.org/futures/fcrepo4/builds/14173670
* travis-ci leaves
<pivotal-bot>Andrew Woods accepted "DELETE Non-Existent Workspace" https://www.pivotaltracker.com/story/show/6100283820:34
* ermadmix joins
<bljenkins>Project fcrepo-fixity-corrupter build #458: SUCCESS in 2 min 38 sec: http://ci.fcrepo.org/jenkins/job/fcrepo-fixity-corrupter/458/20:51
Project fcrepo-kitchen-sink build #651: STILL FAILING in 8 min 4 sec: http://ci.fcrepo.org/jenkins/job/fcrepo-kitchen-sink/651/20:59
Yippie, build fixed!21:00
Project fcrepo-fedora3-federation-connector build #236: FIXED in 11 min: http://ci.fcrepo.org/jenkins/job/fcrepo-fedora3-federation-connector/236/
* ermadmix leaves21:06
* ksclarke joins21:13
* ermadmix joins21:38
* ermadmix leaves22:09
* ermadmix joins22:57
* ermadmix leaves23:29
* ermadmix joins23:34
* ermadmix leaves00:05
* ermadmix joins00:20
* ermadmix leaves00:51
* ermadmix joins00:57
* kaarefc joins01:18
* ermadmix leaves01:29
* ksclarke leaves02:00
* ermadmix joins02:06
* kaarefc leaves02:14
* ermadmix leaves02:38
* ermadmix joins02:54
* kaarefc joins03:22
* ermadmix leaves03:25
* ermadmix joins03:47
* ermadmix leaves04:18
* ermadmix joins05:03
* ermadmix leaves05:34
* ermadmix joins06:09
* fasseg joins06:15
<pivotal-bot>Frank Asseg added comment: "Not sure what is meant by this ticket, will talk to Chris/Andrew later" https://www.pivotaltracker.com/story/show/6095679806:16
* ermadmix leaves06:58
* ermadmix joins07:39
* ermadmix leaves08:10
* ermadmix joins08:12
* github-ff joins08:39
[fcrepo-aws-puppet] fasseg deleted slim-setup at 4add23d: http://git.io/3UEIYw
* github-ff leaves
* ajs6f joins
<pivotal-bot>A. "Uriculoneous" Soroka added comment: "@awoods : Do we actually have an authoritative list of contributors anywhere?" https://www.pivotaltracker.com/story/show/5731057208:40
A. "Uriculoneous" Soroka added comment: "@awoods : _How_ does it fail?" https://www.pivotaltracker.com/story/show/5960319008:41
* ermadmix leaves08:44
* ermadmix joins08:47
* kaarefc leaves09:01
awoods: ping?09:09
<ajs6f>Given the discussion betwen you and barmintor and I yesterday, I'm going to switch back to the RDF iteration thing, working from current master (Jersey1). Right?09:10
<awoods>ajs6f: yes. Thank you.
<ajs6f>I'm going to be wanting some intermediate merges again… further bulletins as events warrant.09:11
ajs6f: I also had a question about rendering items in the UI...
ajs6f: can the UI now support large numbers of items?
Here's the situation:09:12
the kernel can supply iterators of triples which would allow cheap presentation of big RDF....
but only a fraction of the HTTP endpoints use that capability.
The most important step is
to switch from the current serialization machinery (which expects a Jena Dataset)
to a new class (as yet unwritten) which will handle streams.09:13
a MessageBodyWriter<RdfStream>
<awoods>ajs6f: that makes sense
<ajs6f>There's an issue to be worked out which both barmintor and I have seen:
You can't logout the session in scope until the stream has completed (because it is lazy) or the stream will explode.09:14
My first solution was to do that "asynch" logout callback.
But this won't work with Jersey1/JAX-RS1 serialization. So it will have to be refined.
And that's my first task.
We'll know it's done when our JAX-RS methods can reliably feature an RdfStream return type.09:15
(Which will can serialized appropriately in the same pattern as is now used for Datasets.)
In fact, I will make that my first intermediate merge, so that other people can use it / opine about it.09:16
<awoods>ajs6f: I am not sure I follow the "will have to be refined" bit... but more importantly, how do you see getting around the logout issue and its impact on disrupting iteration?
<ajs6f>That's what I have to figure out.
* escowles joins09:17
<awoods>ajs6f: seems like an offset/paging mechanism is in order... and leveraging what Mode offers in that regard.
<ajs6f>The serialization machinery will have to somehow signal exhaustion of the stream to trigger a callback..
I don't see how that would help.
afk bbl09:18
* ajs6f leaves
<fasseg>awoods: can you elaborate a bit what is the relicensing issue with modeshape is? Can't we package modeshape anymore under an ASL? https://www.pivotaltracker.com/s/projects/684825/stories/6095679809:23
Or shall I wait for chris to come on?
* ajs6f joins09:25
* kaarefc joins09:26
<awoods>fasseg: ModeShape's relicensing is a good thing09:27
<ajs6f>I already gave rhauch my okay.09:28
<awoods>fasseg: They were GPL before, which was problematic for us.
<fasseg>I see
<ajs6f>Come on, fasseg. _Everybody's_ doing it.
You don't want to be left out, do you?09:29
<pivotal-bot>Andrew Woods added comment: "see attached "compile.out"" https://www.pivotaltracker.com/story/show/59603190
<fasseg>So rhauch just wants my OK? I feel important ;)09:31
what's the new license then? asl2?
aha asl just found the discussion on their list
"So far reaction has been nothing but positive, so we're going to proceed with relicensing ModeShape under the ASL 2.0, starting with the first release of ModeShape 4.0. Thanks everyone!"09:33
I don't have the slightest idea what to do....modeshape is obviuosly going ahead with relicensing, and the ticket says I should "discuss the issues" with randall....09:34
what issue? the issue im having with asl over gpl? im all against copyleft so yeah
no isue form my side...09:35
<awoods>fasseg: Are you back on clustering?09:36
<fasseg>"!("/(/§!)?!?="§=)`?=§" Puppet sh*t09:37
err yes.
it seems almost impossible to get an oracle jdk installed
<awoods>fasseg: If AWS is not working for you, should you transition?
fasseg: is the oracle jdk necessary?09:38
<fasseg>lemme try the oracle jvm, and I will just do it by hand, and If I don't run into the same pathNotFoundException, I'd create results with 3Nodes/3 threads and 7nodes/7threads instead of 3/5 and 7/1109:39
It should not be...but it would not be the first time for me to find an incompability between openjdk and the oraclejvm
<awoods>I wonder if other cluster testers (cbeer/gregjansen/osmandin) also see the the pathNotFoundException when using more threads than nodes.09:40
<fasseg>I dont see them at home
neither on SCC09:41
<fasseg>upping the IOps or increasing the replication timeout in the infinispan.xml didn't help either
<pivotal-bot>Andrew Woods edited "Review strategy and scope of implementing solr index on repo" https://www.pivotaltracker.com/story/show/6097436809:42
<awoods>good to know
<fasseg>and on e of the differences was that I was using oracle's jvm on the other tests, I guess I could run the tests on SCC with the openjdk JVM and check if the timeouts occur
* kaarefc leaves09:43
<awoods>fasseg: I just wonder if it is worth working out the details of AWS issues if we have several other clusters available for testing.
<fasseg>sure Id be happy to go on another cluster if AWS doesn't work out...09:45
just give me today then to create some n/n nodes/threads results on AWS09:46
and then we at least have a baseline for AWS09:47
<awoods>fasseg: sounds good
fasseg: If you are not aware, we are focused on putting out a release during sprint-9 which will cover the following capabilities: https://wiki.duraspace.org/display/FF/Fedora+4.0+Key+Capabilities09:48
* osmandin joins
<awoods>fasseg: obviously your clustering/performance work plays right into that plan, but we need to make sure you are still able to make progress.
* osmandin leaves09:49
<fasseg>Yeah we talked about this last week, and I took on the clustering feature
* tecoripa joins09:52
* osmandin joins
* ksclarke joins09:54
<osmandin>w.r.t clustering and pathnotfound exception. . . i saw it at times . . . irrespective of # of threads being used09:59
<fasseg>osmandin: I get these consistently every object ingest on AWS, so not a single object gets ingested when I use more threads than nodes....10:03
<awoods>All: it looks like google-hangout will not be sufficient for our daily stand-ups. Shall we just do ReadyTalk?10:04
<fasseg>osmandin: you did get these sporadically, right? so finetuning the replication timeout parameter would likely help...
<osmandin>fasseg: yes, sporadically
frankly, i'm not too comfortable10:05
<fasseg>awoods: If we have to, that means using skype for me :)
<osmandin>i mean frankly speaking
seeing these stack traces10:06
<awoods>fasseg: There is an international number for ReadyTalk, but the issue for you is just being able to call a phone number from your computer?10:07
<osmandin>fasseg: as an aside, does the benchtool check to see that all nodes are up and running?
<awoods>osmandin: I doubt it
<fasseg>awoods: yeah but holding quality is even worse and I dont have a headset for the phone ;)
osmandin: Nah, I just used it to fire against a load balancer10:08
but looking at the logs the same time on the vms
* kaarefc joins10:09
<fasseg>osmandin: but one could probalby use the responses from the httpcomponents to easily list the nodes that were used in the test10:10
<osmandin>fasseg: you mean by grabbing the num. of nodes listed on the page..(or is there a direct http call to get the number of nodes)?10:13
<fasseg>nah just add every IP from the HttpResponse Object to a Unique List and spit out the List at the end10:14
<pivotal-bot>Andrew Woods added comment: "https://wiki.duraspace.org/display/FF/Project+Team ("Technical Team")" https://www.pivotaltracker.com/story/show/5731057210:15
<fasseg>osmandin: You know what im saying?10:16
<osmandin>fasseg: no, i need to look at the code
<fasseg>osmandin: if you can wait for a bit I can update the code...
<osmandin>fasseg: sure10:17
* kaarefc leaves10:19
<pivotal-bot>Andrew Woods added comment: "Shall we have a special topic on "property indexing"? At the least, it would be good to make clear what the..." https://www.pivotaltracker.com/story/show/6071383810:29
* osmandin leaves10:30
* osmandin joins
<fasseg>osmandin: Can you try if the nodelist works for you?10:34
I just pushed a change10:35
It basically takes the Location header from the HTTP response and add the host and port to a HashSet...10:36
<osmandin>fasseg: sure, i can test in a bit..thx10:37
<escowles>fasseg: i've seen the pathNotFoundException occasionally (once or twice when doing 25k objects, for example)10:38
fasseg: that was with 1 node, 1 thread
<fasseg>escowles: hmm ok I dont think I have ever seen the exception on a single node, but I didn't ingest 25k objects10:39
<escowles>i've done 10 or so 25k ingest runs with fcrepo4, and i've seen a handful of them, and always after 15-20k objects already ingested10:40
* gregjansen joins10:44
<awoods>gregjansen: is the google-hangout limit now 15?10:51
<cbeer>fasseg: rhauch sent you an email a couple weeks ago trying to get your blessing for the LGPL / ASL relicensing10:55
said he never got a reply10:56
<fasseg>cbeer: Ah I see, lemme check my mails...
<osmandin>is the meeting at readytalk today?11:00
<awoods>866-740-1260, participant code: 2257295
but true
<awoods>better solution?
<barmintor>ajs6f: what time do you need to leave today?
* tecoripa1 joins11:01
dialing in now
<ajs6f>barmintor: 5ish or so. Why?11:03
<fasseg>cbeer: I answered Rhauch I haven't seen his mail in my gmail account :(
<ajs6f>awoods: Google hangout.
<barmintor>ajs6f: I thought you needed to leave early for a medical thing
<pivotal-bot>Frank Asseg added comment: "Responded to his email" https://www.pivotaltracker.com/story/show/60956798
<ajs6f>barmintor: Oh, right— sorry. No, I was able to switch that.
<pivotal-bot>Frank Asseg accepted "Respond to rhauch about Modeshape relicensing" https://www.pivotaltracker.com/story/show/60956798
<cbeer>awoods: https://support.google.com/a/answer/2677443?hl=en
<barmintor>awoods: I am on the call (muted)
Why not?11:05
<cbeer>there's some indication google apps hangouts allow 15 people
<fasseg>Video calls can have a maximum of 10 participants. from https://support.google.com/plus/answer/1216376?hl=en
or is there an unannounced beta of sorts11:06
<cbeer>fasseg: https://support.google.com/a/answer/2677443?hl=en
they announced it all over last year
but they don't advertise it well, apparently.
<osmandin>jetty trace: https://gist.github.com/osmandin/7547603
not sure what's not getting cleaned up11:07
<fasseg>and normally they're so good at changing *evrything* everyday ;)
<osmandin>ERROR 10:57:06.967 (RepositoryNodeTypeManager) Node types were read from the system content, and appear to be inconsistent or invalid: repo
<pivotal-bot>Mike Durbin started "Add versioning documentation/walkthrough to the wiki" https://www.pivotaltracker.com/story/show/6079875811:12
* github-ff joins11:13
[fcrepo4] ajs6f created RDFIterationForevva (+1 new commit): http://git.io/3cF1fQ
fcrepo4/RDFIterationForevva 5c8fa63 ajs6f: Minor change to JcrRdfTools.getTreeTriples()
* github-ff leaves
<pivotal-bot>A. "Mutchkin" Soroka added comment: "Looks like a failure in which one test is outrunning another between modules. Given that RSS is hea..." https://www.pivotaltracker.com/story/show/5960319011:24
A. "Mutchkin" Soroka added comment: "@awoods : can you open a ticket for this with the stacktrace?" https://www.pivotaltracker.com/story/show/5960319011:25
* tecoripa1 leaves
<awoods>barmintor: are you on the line?
* travis-ci joins11:28
[travis-ci] futures/fcrepo4#1223 (RDFIterationForevva - 5c8fa63 : ajs6f): The build passed.
[travis-ci] Change view : https://github.com/futures/fcrepo4/commit/5c8fa6347b32
[travis-ci] Build details : http://travis-ci.org/futures/fcrepo4/builds/14204939
* travis-ci leaves
<gregjansen>one thing that we need for large files is probably a "Setup Profile" that captures what we mean by large...11:35
<bljenkins>Project fcrepo-fixity-corrupter build #459: SUCCESS in 1 min 29 sec: http://ci.fcrepo.org/jenkins/job/fcrepo-fixity-corrupter/459/
<barmintor>awoods: I'm just hanging out
<gregjansen>barmintor: we stayed on ready talk11:36
<gregjansen>I know
<barmintor>"your conference will begin when the chairperson arrives, please wait."11:37
<fasseg>osmandin: for me only the load-balancer gets printed out in the node-list, ill have to check if I can get the IP of the machines behind the lb somehow...11:38
<gregjansen>barmintor: awoods had to run to a meeting,
barmintor: so we only had a brief chat
<osmandin>fasseg: ok thx.. our nodes need to be rebooted, so i didnt get a chance to test yet11:39
<bljenkins>Project fcrepo-kitchen-sink build #652: STILL FAILING in 5 min 10 sec: http://ci.fcrepo.org/jenkins/job/fcrepo-kitchen-sink/652/11:40
<gregjansen>barmintor: awoods has meeting lined up until 2pm EST, then he can probably hangout again
* kaarefc joins11:43
<osmandin>fasseg: wouldn't it be simpler to just this info from fedora itself?11:44
it should be a close reflection of the actual configuration11:45
if it's not, then well . . .
<pivotal-bot>Chris Beer added "MODE-2098" https://www.pivotaltracker.com/story/show/6105106011:55
<fasseg>osmandin: hmm it's not that easy, since then we will have an additional GET per ingest in order to disocer the node the GET came from, which does not necessarily be the same node :/12:10
<osmandin>fasseg: i was thinking about the case when somebody is about to run the benchtool but one of the nodes is down. so grabbing fcrepo:clusterNodeView (or whatever) might be helpful (perhaps even doing a few random checks in a separate thread)12:19
<fasseg>Oh so you mean just one GET at the beginning and maybe one at the end ?12:20
than parse out the node number and print it out to the user?
<osmandin>yeah something like that... just a nice to have12:21
<fasseg>like "fcrepo was running with N nodes at the start, and was running on M nodes after the benchmark"?
<osmandin>i remember generating a graph or two and only realizing 80% of the objects were making it due to a node firewall issue12:22
<cbeer>http://www.jgroups.org/manual/html/protlist.html#d0e4161 => bind_addr special values after the table12:44
<pivotal-bot>Chris Beer added "Run fcrepo3 benchmarks on Stanford server" https://www.pivotaltracker.com/story/show/6105760813:05
Chris Beer estimated "Run fcrepo3 benchmarks on Stanford server" as 3 points https://www.pivotaltracker.com/story/show/61057608
Chris Beer started "Run fcrepo3 benchmarks on Stanford server" https://www.pivotaltracker.com/story/show/61057608
Chris Beer added "Establish cluster on Stanford nodes" https://www.pivotaltracker.com/story/show/6105769613:06
Chris Beer estimated "Establish cluster on Stanford nodes" as 2 points https://www.pivotaltracker.com/story/show/61057696
Chris Beer started "Establish cluster on Stanford nodes" https://www.pivotaltracker.com/story/show/61057696
Chris Beer finished "Establish cluster on Stanford nodes" https://www.pivotaltracker.com/story/show/61057696
* ermadmix leaves13:07
<pivotal-bot>Chris Beer added comment: "Using jgroups config: https://gist.github.com/cbeer/fd3997e40fe014eab071" https://www.pivotaltracker.com/story/show/61057696
<ajs6f>Is anyone else having intermitten i-test failures with the JMS component?13:11
<awoods>ajs6f: yes
I thought osmandin or someone fixed that?
<osmandin>parsing issue....sometimes msgs are truncated
ajs6f: I fixed a related issue13:12
<ajs6f>Right, right.
Do we know why/how msgs get truncated?
Does the machine cough or giggle at the wrong moment?
<osmandin>ajs6f i didn't look..only looked at the output (i.e. the message)13:13
Well, I'm not going to fool with it now.
It works at least one out of two tries, and that's better than most of my ode.
<osmandin>ok, i can take a look again.. and if i find out the cause, create a ticket
<osmandin>for myself13:14
<ajs6f>If you have time, great! thx
* github-ff joins13:23
[fcrepo4] ajs6f pushed 1 new commit to RDFIterationForevva: http://git.io/8SMYqA
fcrepo4/RDFIterationForevva 037d879 ajs6f: Introducing RdfStreamProvider...
* github-ff leaves
<ajs6f>awoods: time for an intermediate merge? I have the streaming rdf serializer ^^^
Now it's time to use it.
<awoods>ajs6f: would later this evening work for you?
<ajs6f>No. If you don't merge it now, I will throw it away. Well, all right. Just because it's you, awoods. I'll wait.13:24
<awoods>ajs6f: you're the best.
<ajs6f>Working on Fedora makes everyone happy.
<awoods>ajs6f: now that you mention it, that is probably true... F4 brings out the best in people.13:25
<ajs6f>Actually, cbeer has seen the older way of doing this (from a week ago). maybe cbeer can merge it?
F4 just makes people so happy they aren't working on F3 that they become better people.
<awoods>if cbeer is willing to review and merge... that works for me.
Anyway, let me make a PR. That has to happen no matter what.13:26
<pivotal-bot>A. "Mutchkin" Soroka finished "Alter outward-facing contracts in fcrepo-kernel to expose new iterator-based machinery for RDF." https://www.pivotaltracker.com/story/show/5924323613:27
* github-ff joins13:28
[fcrepo4] ajs6f opened pull request #158: Rdf iteration forevva (master...RDFIterationForevva) http://git.io/6z7j9g
* github-ff leaves
<pivotal-bot>A. "Mutchkin" Soroka added comment: "https://github.com/futures/fcrepo4/pull/158" https://www.pivotaltracker.com/story/show/59244014
<osmandin>afk (meeting)13:30
* ermadmix joins13:33
* travis-ci joins13:36
[travis-ci] futures/fcrepo4#1224 (RDFIterationForevva - 037d879 : ajs6f): The build passed.
[travis-ci] Change view : https://github.com/futures/fcrepo4/compare/5c8fa6347b32...037d8798c09a
[travis-ci] Build details : http://travis-ci.org/futures/fcrepo4/builds/14211221
* travis-ci leaves
<fasseg>osmandin: I added the suggest feature to the bench-tool, and it will now tell you how many nodes are in the cluster, it will also check after the test if the cluster size is the same as before..13:40
<bljenkins>Project fcrepo-fixity-corrupter build #460: SUCCESS in 1 min 12 sec: http://ci.fcrepo.org/jenkins/job/fcrepo-fixity-corrupter/460/13:42
<cbeer>ajs6f: very clever, and much nicer than before13:43
shipped it.13:44
* github-ff joins13:45
[fcrepo4] cbeer pushed 1 new commit to master: http://git.io/3F-3vg
fcrepo4/master 760b842 Chris Beer: Merge pull request #158 from futures/RDFIterationForevva...
* github-ff leaves
<ajs6f>cbeer, you make me blush.
Just started looking at HttpTripleUtil and UriAwareResourceModelFactory. When we said we will do everything in RDF, we weren't kidding.13:47
<cbeer>heading to the office13:49
<bljenkins>Project fcrepo-kitchen-sink build #653: STILL FAILING in 7 min 45 sec: http://ci.fcrepo.org/jenkins/job/fcrepo-kitchen-sink/653/13:50
Yippie, build fixed!13:51
Project fcrepo-jms-indexer-pluggable build #243: FIXED in 9 min 53 sec: http://ci.fcrepo.org/jenkins/job/fcrepo-jms-indexer-pluggable/243/
* travis-ci joins13:58
[travis-ci] futures/fcrepo4#1226 (master - 760b842 : Chris Beer): The build passed.
[travis-ci] Change view : https://github.com/futures/fcrepo4/compare/9eaac0f929ae...760b84277b0c
[travis-ci] Build details : http://travis-ci.org/futures/fcrepo4/builds/14212210
* travis-ci leaves
<awoods>barmintor/gregjansen/ajs6f/others: google-hangout?14:02
<awoods>jersey2 etc
<ajs6f>I thought we decided that?
Whatevs. omw
* ajs6f leaves14:04
<barmintor>ajs6f to get on the call kkthx
<pivotal-bot>Mike Durbin started "Review and possibly revise or extend the fcrepo-http-api endpoints relating to versions to ensure they are useful f..." https://www.pivotaltracker.com/story/show/60769514
* ajs6f joins14:10
<bljenkins>Project fcrepo-fixity-corrupter build #461: SUCCESS in 1 min 42 sec: http://ci.fcrepo.org/jenkins/job/fcrepo-fixity-corrupter/461/14:11
<ajs6f>I'm just getting a "Please wait…" from the hangout.
all I want is to work with Adam!
<ajs6f>Oh, there it goes. barmintor, you flatterer!
<bljenkins>Project fcrepo-fedora3-federation-connector build #239: UNSTABLE in 6 min 34 sec: http://ci.fcrepo.org/jenkins/job/fcrepo-fedora3-federation-connector/239/
Project fcrepo-jms-indexer-pluggable build #244: UNSTABLE in 7 min 30 sec: http://ci.fcrepo.org/jenkins/job/fcrepo-jms-indexer-pluggable/244/14:17
<ajs6f>barmintor: It should already be happening, I think.14:34
(exportable bundles, if you meant that in the OSGi sense).
<barmintor>ajs6f: I don't think it is
We can fix that.
We did put the bundle plugin in there.
Maybe the config got bent out of shape.
<barmintor>For one: When I put in activators and tried to build, I got tons of library resolution conflicts
<ajs6f>Bundle activators? You want to use those?14:35
<barmintor>I'm just following instrucitons
<ajs6f>Not saying we can't, just thinking that there are other ways to do that kind of thing.
Instructions from ?
<ajs6f>When did OSGi come into the Jersey 2 thing?
<barmintor>CDI, OSGi, log in my eye14:36
We only *need* CDI
<ajs6f>"Thou hypocrite, first cast out the beam out of thine own eye..."
We only need CDI now, because we're only interested in wiring now, right? But it would be nice to have a module framework in play. In case I hadn't mentioned that.14:37
<barmintor>I can't believe I'm typing this, ut +1
<ajs6f>Pax Exam should be able to stat a Weld or OWB container w/o any OSGi jazz.14:38
<ajs6f>+1 to modularization?
<bljenkins>Project fcrepo-fixity-corrupter build #462: SUCCESS in 1 min 27 sec: http://ci.fcrepo.org/jenkins/job/fcrepo-fixity-corrupter/462/
<ajs6f>I think I'm going to faint.
<bljenkins>Yippie, build fixed!14:44
Project fcrepo-fedora3-federation-connector build #240: FIXED in 7 min 13 sec: http://ci.fcrepo.org/jenkins/job/fcrepo-fedora3-federation-connector/240/
Yippie, build fixed!14:45
Project fcrepo-jms-indexer-pluggable build #245: FIXED in 8 min 0 sec: http://ci.fcrepo.org/jenkins/job/fcrepo-jms-indexer-pluggable/245/
Project fcrepo-kitchen-sink build #654: STILL FAILING in 7 min 40 sec: http://ci.fcrepo.org/jenkins/job/fcrepo-kitchen-sink/654/14:46
<ajs6f>awoods: Did we ever establish the key capabilities around… audit?14:50
all: question: if the http api is asked for a serialization we don't support, should we throw a 406 or should we just return some kind of RDF, like rdf/xml?15:10
<escowles>ajs6f: we can do both, right? return 406 with data in the response body15:14
Fedora 4: The best of all possible worlds.
* osmandin leaves
* osmandin joins15:19
<cbeer>ajs6f: 406 and screw them.15:24
with a response body "PATCHES WELCOME"15:25
<ajs6f>I'm still trying to figure out whether a MessageBodyWriter can actually set the response code. That's pretty late in the chain.
<cbeer>at least it doesn't give the appearance of working, if you have a lazy RDF parser that doesn't check the response coe.
ajs6f: do we need to pre-validate?
<ajs6f>What about robust-itudiness?
prevalidate the mimetype? I hope not. That's ugly in a controller.15:26
<cbeer>ajs6f: we are robust. they sent something nonsense, we told them so.
<ajs6f>I'm going to let our two chefs fight it out with kitchen knives.
<cbeer>fail fast, yada yada.
<escowles>i don't have a strong opinion for sending data in addition to the 406 -- i think we can assume it's not a browser (which would have html in the accept header), so a nicer version of "PATCHES WELCOME" wouldn't be that bad15:28
<ajs6f>Okay, a 406 with "Are you looking at me? Are. You. Looking. At. Me? Because there's nobody else here…"
awoods: ping?15:40
* ajs6f leaves15:53
* ajs6f joins15:55
<tecoripa>general question about test platform profiles: are they supposed to be abstract, or specific to a given instance of a platform?15:59
<ajs6f>The template is the abstraction. A profile for a given platform should be specific to a platform (e.g. Wisconsin's cluster)
<tecoripa>I thought the former ("a platform with two CPUs, two nodes in a cluster, etc."), but the child pages I see refer to specific platforms.16:00
but I stand corrected.
This page contains a lot of the same info: https://wiki.duraspace.org/display/FF/Test+Clusters
<ajs6f>Then we should REFACTOR!16:02
<escowles>tecoripa: yes, i think it will make sense to move that into the new templates and consolidate the test results in general
<tecoripa>escowles: ok, I'll create a ticket and do that16:03
<pivotal-bot>Scott Prater started "Create use case for common preservation repository profile" https://www.pivotaltracker.com/story/show/6097424216:43
Scott Prater added comment: "Test - Repository Profile: Clustered16:44
https://wiki.duraspace.org/display/FF/Test+-+Repository+Profile%3A+Clu..." https://www.pivotaltracker.com/story/show/60974242
<pivotal-bot>Scott Prater added "Refactor Test Clusters wiki page into Test Platform Profiles" https://www.pivotaltracker.com/story/show/6107756016:50
* github-ff joins16:57
[fcrepo4] barmintor pushed 1 new commit to cdi: http://git.io/Cbbrww
fcrepo4/cdi 1bf1da7 Benjamin Armintor: wrapping up progress on modularization and Pax-Exam
* github-ff leaves
<ajs6f>barmintor: did you have any luck?16:58
<barmintor>^ that's just stashing changes for later
ajs6f: sort of
<ajs6f>In the Fedora context, "sort of" can be a kind f triumph.
<barmintor>ajs6f: that was a fascinating email thread between you and MODE, btw16:59
<ajs6f>"fascinating". I'm going to take that as a complement.
<barmintor>it was!17:00
<ajs6f>Even better!
It's always fun to watch a man (me) who seems to be committed to non-Euclidean principles of software architecture try to convince sane people of his views.17:01
I do think we'll get some good action on the configuration front.
I think rhauch sees where we're coming from there.
It would be nice to be able to treat MODE repos as a little more "managable".17:02
<barmintor>ajs6f: one of the things I noticed (rhauch mentioned this as well) is that Spring's take on bean config is more straightforward (setters and getters), whereas CDI wants config producers, basically
<ajs6f>I hadn't thought about some of the issues he raised with self-hosted config and clustering.
<barmintor>yeah, that's a different issue
<ajs6f>barmintor: I think CDI pushes you to let the framework key off of types as purely as possible.17:03
<barmintor>but the reflection thing they do now is just not sufficient for plugin or extension, IMO
<ajs6f>setters/getters pollute that with names.
IMO too, but I think we need to be very concrete with them, and I think they have the right to expect that.17:04
We need to give them specific instances of pain and suffering.
_I_ certainly wouldn't listen to a bedraggled, fiflthy and possibly lunatic person like me leering and mumbling about "component lifecycles" and "type-safe injection". I'd want to see some specific examples of where what MODE does now isn't allowing people to extend and expand.17:06
And I think we can show them that in the federation SPI.
{sigh} I know I pushed to layer HTML over the endpoints, but making it all work in the iterated world is painful.17:07
<barmintor>escowles: are your numbers for performance coming from ff-jmeter-madness?
<ajs6f>Out for the day. See y'all later.17:09
* ajs6f leaves
<escowles>barmintor: no, i'm using benchtool17:10
* jonathangee leaves17:13
<barmintor>I could have sworn I fixed this failsafe.argLine bug in the root pom17:21
hey, we have two different versions of the surefire plugin configured in the root pom17:24
ah, one's the report plugin17:26
seems a little suspicious that they have different versions, but w/e17:27
<bljenkins>Project fcrepo-fixity-corrupter build #463: SUCCESS in 2 min 8 sec: http://ci.fcrepo.org/jenkins/job/fcrepo-fixity-corrupter/463/
* osmandin leaves17:28
* kaarefc leaves17:29
<bljenkins>Project fcrepo-jms-indexer-pluggable build #246: UNSTABLE in 9 min 53 sec: http://ci.fcrepo.org/jenkins/job/fcrepo-jms-indexer-pluggable/246/17:35
Project fcrepo-kitchen-sink build #655: STILL FAILING in 8 min 11 sec: http://ci.fcrepo.org/jenkins/job/fcrepo-kitchen-sink/655/
* tecoripa leaves17:44
* tecoripa joins17:45
* tecoripa leaves
* gregjansen leaves18:03
* ksclarke leaves18:07
<pivotal-bot>Andrew Woods added "Multi-threaded build failure" https://www.pivotaltracker.com/story/show/6108478618:14
Andrew Woods edited "MODE-2098" https://www.pivotaltracker.com/story/show/6105106018:23
* ermadmix leaves18:30
<pivotal-bot>Andrew Woods accepted "Establish cluster on Stanford nodes" https://www.pivotaltracker.com/story/show/6105769618:58
* jonathangee joins19:00
<pivotal-bot>Andrew Woods added comment: "There has been no new activity on this ticket since Nov 1st. Are you suggesting that the third incremental ..." https://www.pivotaltracker.com/story/show/59243236
Andrew Woods edited "Investigate issues encountered on the AWS cluster instances" https://www.pivotaltracker.com/story/show/5956958019:03
* fasseg leaves19:06
<pivotal-bot>Andrew Woods added "Federate over large files" https://www.pivotaltracker.com/story/show/6108774019:07
Andrew Woods edited "Federate over large files" https://www.pivotaltracker.com/story/show/61087740
Andrew Woods edited "Federate over large files" https://www.pivotaltracker.com/story/show/6108774019:08
Andrew Woods added "Add properties to federated nodes" https://www.pivotaltracker.com/story/show/6108785619:09
Andrew Woods edited "Add properties to federated nodes" https://www.pivotaltracker.com/story/show/61087856
Andrew Woods edited "fcr:fixity bug using the FileSystemConnector" https://www.pivotaltracker.com/story/show/5600291619:10
Andrew Woods edited "fcr:fixity bug using the FileSystemConnector" https://www.pivotaltracker.com/story/show/56002916
Andrew Woods added "Serving large files" https://www.pivotaltracker.com/story/show/6108799619:13
Andrew Woods edited "Serving large files" https://www.pivotaltracker.com/story/show/61087996

Generated by Sualtam