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

Using timezone: Eastern Standard Time
* fcrepo-bot joins07:13
* github-ff joins08:09
[fcrepo4] sprater opened pull request #218: Fix assertEquals() messages to indicate error. (master...fcrepo-62392884) http://git.io/A7i0Qw
* github-ff leaves
<pivotal-bot_>Scott Prater added comment: "Pull request: https://github.com/futures/fcrepo4/pull/218" https://www.pivotaltracker.com/story/show/62392884
Scott Prater finished "Fix assert messages on authz IT tests" https://www.pivotaltracker.com/story/show/62392884
* fasseg joins08:12
* fcrepo-bot leaves08:17
<pivotal-bot_>Frank Asseg added comment: "I already did some tests with a preliminary 3.7.0 Version, and it seems to me that Modeshape has decreased i..." https://www.pivotaltracker.com/story/show/6341588408:21
<fasseg>awoods: Do you want me to run the benchmarks again to verify that the performance didn't decrease as bad as I fear with Modeshape 3.7.0 or shall I get straight to creating a PR for fcrepo4 on Mode 3.7?08:23
and a happy new year btw ;)08:24
<pivotal-bot_>A. "Horbulaco" Soroka added comment: "Isn't this exactly what GraphSubjects is for? Doing translation between internal and external iden..." https://www.pivotaltracker.com/story/show/6366919408:46
* ajs6f joins08:55
* fasseg_ joins09:00
* fasseg leaves
<awoods>fasseg: You expect a performance hit with mode3.7?09:10
* github-ff joins09:12
[fcrepo4] awoods closed pull request #218: Fix assertEquals() messages to indicate error. (master...fcrepo-62392884) http://git.io/A7i0Qw
* github-ff leaves
<pivotal-bot_>Andrew Woods delivered "Fix assert messages on authz IT tests" https://www.pivotaltracker.com/story/show/62392884
* ajs6f leaves09:14
<fasseg_>awoods: yes from my tests with the version I built from master last week (not the released 3.7.0) I could see a perfomance impact. It was about 15-20% slower on large files (~1GB)
<awoods>fasseg: Was the leveldb OOM issued resolved?09:15
<fasseg>awoods: I hope it's better with the released version since there were a couple of committs in between my tests and the release
awoods: yes
awoods: I could run ingests with a heap of *only* 500mb09:16
* ajs6f joins
<awoods>fasseg: Please go ahead and run performance tests to make sure things are not horrible before we commit to 3.7.
<pivotal-bot_>Frank Asseg added comment: "Will do!" https://www.pivotaltracker.com/story/show/6196991009:17
Frank Asseg added comment: "I will first check the impact on performance before upgrading" https://www.pivotaltracker.com/story/show/6341588409:18
* mikeAtUVa joins09:19
* travis-ci joins09:30
[travis-ci] futures/fcrepo4#1482 (master - 8ca495f : Andrew Woods): The build passed.
[travis-ci] Change view : https://github.com/futures/fcrepo4/compare/79f2b356ec64...8ca495f71fbf
[travis-ci] Build details : http://travis-ci.org/futures/fcrepo4/builds/16868206
* travis-ci leaves
<pivotal-bot_>Andrew Woods accepted "Fix assert messages on authz IT tests" https://www.pivotaltracker.com/story/show/62392884
Andrew Woods added comment: "@frankasseg, please create a ticket for the performance-checking work." https://www.pivotaltracker.com/story/show/6341588409:40
Frank Asseg added "Check the performance impact of Modeshape 3.7.0" https://www.pivotaltracker.com/story/show/6371730209:42
Frank Asseg added comment: "added this ticket: https://www.pivotaltracker.com/story/show/63717302" https://www.pivotaltracker.com/story/show/63415884
<fasseg>I seem to be unable to set the repository.json config path using the property fcrepo.modeshape.configuration anymore:09:49
I try starting with CATALINA_OPTS="-Dfcrepo.modeshape.configuration=file:///home/ruckus/dev/tomcat7/repository.json" bin/catalina.sh run
and fedora complains that it cannot find home/ruckus/dev/tomcat7/repository.json (relative path) did the handling there change over the last 2 weeks?09:50
<pivotal-bot_>Andrew Woods added comment: "There are some basic ways of leveraging Tomcat and Apache to achieve some (if not all) of the issue raised ..." https://www.pivotaltracker.com/story/show/6366919409:52
<awoods>fasseg: have you tried: CATALINA_OPTS="-Dfcrepo.modeshape.configuration=/home/ruckus/dev/tomcat7/repository.json09:55
<fasseg>awoods: same error: java.io.FileNotFoundException: ServletContext resource [/home/ruckus/dev/tomcat7/repository.json] cannot be resolved to URL because it does not exist
while file /home/ruckus/dev/tomcat7/repository.json givees me "/home/ruckus/dev/tomcat7/repository.json: ASCII text" so it does exist09:56
<awoods>fasseg: does it work with Mode3.6?09:58
<fasseg>nah it's from a fresh fcrepo install from master
so I'm currntly trying with mode 3.6 because I wanted to accept this one: https://www.pivotaltracker.com/story/show/6196991009:59
<pivotal-bot_>bug: Adding a node to a federated read-only file system is enabled in the frontend and creates a NPE when used (delivered) / owner: Chris Beer
<awoods>fasseg: I am not seeing the change that would effect the repository.json loading behavior.10:01
<fasseg>I couldn't see anything either....10:02
Hmm maybe I still have the wrong mode version in my local repository10:03
<awoods>fasseg: You may want to use git-bisect to find out when it stopped working.
<fasseg>I'll have to read up on that...10:04
<awoods>fasseg: http://git-scm.com/book/en/Git-Tools-Debugging-with-Git#Binary-Search10:05
<pivotal-bot_>Frank Asseg added comment: "@cbeer The NPE is gone, but now i'm getting this exception when trying to add a Object/Datastream: ""10:13
java.lan..." https://www.pivotaltracker.com/story/show/61969910
<barmintor>coffee before argument, afk10:54
<barmintor>awoods: line?
* scossu joins
<barmintor>nt:folder allows any nt:hierarchyNode subtype as a child. If we introduced a new subtype of nt:hierarchyNode (fedora:resource), you woul dbe able to put them under nt:folder objects11:51
awoods: I have to drop12:01
we missed you ajs6f.12:50
<pivotal-bot_>Andrew Woods added comment: "Note, 3.7.1 is out." https://www.pivotaltracker.com/story/show/6341588413:44
<ajs6f>Sorry, unexpected meeting.13:48
Did you get the content model stuff sorted?13:49
<awoods>ajs6f: Not exactly13:51
<ajs6f>Cool. But almost, right?13:53
* ajs6f leaves13:55
* ajs6f joins14:03
* ajs6f leaves14:41
* ajs6f joins14:43
* ajs6f leaves14:58
* ajs6f joins15:02
* scossu leaves15:05
* ajs6f leaves15:30
* ajs6f joins15:38
* fasseg leaves15:52
<awoods>cbeer: Will you be able to make the 3pm ET Hydra-planning meeting tomorrow with escowles?15:53
<cbeer>that's the plan15:54
<awoods>cbeer: Here's to hoping the plan works.
* ajs6f leaves16:17
* ajs6f joins16:21
* ajs6f leaves16:46
<cbeer>awoods: any idea what Neil means by "USP's"?16:59
<barmintor_afk>cbeer: I looked it up, and think it's "unique selling point"17:04
<cbeer>then i'm even more confused17:05
<barmintor>cbeer: the current "wildcard properties" of the fedora base type is one of FCR4's selling points to Neil17:08
(that's how I read it)
<cbeer>ok. then i agree with him. open world all the way.
(but it's hardly unique... given that we're inheriting it from upstream)
<awoods>cbeer: Yes, it sounds to me like Neil is saying that listing out all of the properties in the CND would be too burdensome, and unusable.18:52