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

Using timezone: Eastern Standard Time
* dbernstein leaves01:08
* dbernstein joins
* dbernstein leaves01:11
* dwilcox joins07:01
* coblej joins07:44
* bseeger joins09:03
* whikloj joins09:08
* yamil joins09:18
* ajs6f joins09:23
barmintor: Did you see that VITAL question?
* peichman joins09:28
* th5 joins09:59
* coblej leaves10:22
* coblej joins10:39
* bseeger leaves10:44
* bseeger joins10:54
* ajs6f leaves11:07
* bseeger leaves11:21
* kefo joins11:42
* coblej leaves11:45
* bseeger joins12:04
* dbernstein joins12:21
* dhlamb joins12:26
<whikloj>peichman: refresh my memory the deletegatedPrincipleProvider only allows you to delegate to a user with "less" rights than the actual user making the request?12:32
* bseeger leaves
<whikloj>So fedoraAdmin can pass On-Behalf-Of adminuser, but not the otherway around
<awoods>whikloj: correct... and actually, I believe only the fedoraAdmin can make requests with On-Behalf-Of12:33
<whikloj>awoods: right you are, thanks12:35
https://github.com/fcrepo4/fcrepo4/blob/master/fcrepo-auth-common/src/main/java/org/fcrepo/auth/common/ServletContainerAuthenticationProvider.java#L132-L135
awoods: If you have a moment. How many and how complex of configuration options do we want to include in this configuration.xml?12:38
For instance this custom header acoburn uses12:39
https://gitlab.amherst.edu/acdc/amherst-fedora-webapp/blob/master/src/main/webapp/WEB-INF/classes/spring/configuration.xml#L32-35
<awoods>whikloj: absolutely: https://github.com/fcrepo4-exts/fcrepo-webapp-plus/blob/master/src/rbacl/webapp/WEB-INF/classes/spring/auth-repo.xml#L22-L2512:41
<whikloj>awoods: okay...this is going to have a lot of comments in it. We'll see what you think later.12:42
<awoods>whikloj: the assumption is that a single config file with lots of comments is more digestible than multiple config files scattered about.12:43
<whikloj>awoods: and I generally agree, but in XML I am trying to comment each section which means due to not nesting comments. I have to explain what sections to uncomment together.12:45
* coblej joins12:46
<whikloj>awoods: like I said, I'll get something and you all can help make it clearer
<awoods>whikloj: I am not sure there is a good solution to the issue of effective and efficient XML commenting... but you are certainly at liberty to do what you feel is best.
* coblej leaves12:50
<peichman>whikloj: awoods is correct, only admins can delegate, and only to non-admin users
* coblej joins
* coblej leaves12:54
* bseeger joins13:23
* peichman leaves13:49
* peichman joins13:51
* peichman1 joins13:56
* peichman leaves
* dwilcox leaves13:59
* peichman joins14:04
* peichman1 leaves
* dwilcox joins14:34
* ajs6f joins14:45
* tolloid joins
* dwilcox leaves14:56
* dwilcox_ joins
* dwilcox joins15:17
* dwilcox_ leaves
<barmintor>ajs6f: fwiw the constant 2 in that function predicting listener messages was a slf4j config problem.15:21
<ajs6f>barmintor: It's okay, it was just O(some jazz) stuff anything and the important thing is that we understood why-- events are not deltas15:22
<barmintor>ajs6f: I'm running CI one more time, but I think the update to the listener that caches the last op on a statement and short-circuits repeats is about to be PR15:23
<ajs6f>To Fedora, I assume-- or did you want to pack this up and offer it to Jena?
<barmintor>to fedora
<ajs6f>Cool, coll.
<barmintor>it might be a contested merge, so I'll ping awoods et all about it15:24
<ajs6f>PLease do file me that ticket on Jena to improve the docs. I'm doing a release now but I could get to it this weekend if I am reminded. Then it would go with this release.
now = starting Monday
<awoods>barmintor?
<ajs6f>"now modulo weekends"
<barmintor>ajs6f: since it guards this particular situation but at the cost of added overhead to normal updates, too
awoods: the fix for FCREPO-239115:25
* peichman leaves
* peichman joins15:26
<ajs6f>barmintor: That's between you and awoods. Just be careful: merging can be tricky https://safetycompass.files.wordpress.com/2012/06/redoak.jpg15:28
<barmintor>ajs6f: so that's what SPARQL-Update looks like15:29
<ajs6f>barmintor++
All the simplicity and elegance of SPARQL Query + dangerous persistent side effects15:30
In defence of those guys, it's not a trivial problem.
Graph languages are just hard. There's a reason that we know ten times as much about tree automata as about general graph automata.15:31
* github-ff joins15:33
[fcrepo4] barmintor created fcrepo-2391 (+1 new commit): https://git.io/vDfHM
fcrepo4/fcrepo-2391 d3b8ce0 Benjamin Armintor: FCREPO-2391: buffer unique statement operations to prevent OOM on large sparql update solution products
* github-ff leaves
* github-ff joins15:37
[fcrepo4] barmintor force-pushed fcrepo-2391 from d3b8ce0 to 879ac7e: https://git.io/vDfQB
fcrepo4/fcrepo-2391 879ac7e Benjamin Armintor: FCREPO-2391: buffer unique statement operation when processing PATCH...
* github-ff leaves
* github-ff joins
[fcrepo4] barmintor opened pull request #1169: FCREPO-2391: buffer unique statement operation when processing PATCH (master...fcrepo-2391) https://git.io/vDfQu
* github-ff leaves
* dwilcox leaves15:38
<barmintor>look awoods I made an honest to god code contribution! I bet you thought you'd never see that again.
* dwilcox joins
<awoods>barmintor++15:39
<barmintor>awoods: to be fair you may end up wanting to mark that WONTFIX. I'm not sure what the impact is in the general case, but we have to do something like that if we want to stave off this particular JVM-nuking situation.15:42
* tolloid leaves15:45
* tolloid joins15:48
<f4jenkins>Project fcrepo-module-auth-rbacl build #1294: FAILURE in 45 sec: http://jenkins.fcrepo.org/job/fcrepo-module-auth-rbacl/1294/
* travis-ci joins
fcrepo4/fcrepo4#4930 (fcrepo-2391 - d3b8ce0 : Benjamin Armintor): The build passed.
Change view : https://github.com/fcrepo4/fcrepo4/commit/d3b8ce08f319
Build details : https://travis-ci.org/fcrepo4/fcrepo4/builds/195996017
* travis-ci leaves
<awoods>barmintor: We could certainly run some simple tests for comparing the response time of sqarql-update queries, but my suspicion is that bulk sparql-update performance is not an issue.15:51
* travis-ci joins15:52
fcrepo4/fcrepo4#4931 (fcrepo-2391 - 879ac7e : Benjamin Armintor): The build passed.
Change view : https://github.com/fcrepo4/fcrepo4/compare/d3b8ce08f319...879ac7eb9284
Build details : https://travis-ci.org/fcrepo4/fcrepo4/builds/195997042
* travis-ci leaves
<f4jenkins>Yippee, build fixed!15:58
Project fcrepo-module-auth-rbacl build #1295: FIXED in 2 min 40 sec: http://jenkins.fcrepo.org/job/fcrepo-module-auth-rbacl/1295/
* dwilcox leaves15:59
* dwilcox joins16:00
* dwilcox leaves16:12
* tolloid leaves16:19
* peichman leaves16:41
* peichman joins16:43
* bseeger leaves16:53
* peichman leaves
* peichman joins16:54
* apb18 joins16:57
awoods: Sorry, got tied up today. I put up the delta doc page on the wiki: https://wiki.duraspace.org/display/FF/Fedora+API+Delta+Document16:58
I'd like to clean it up and think about exactly what to include in the call for participation, but won't have time to do that before picking up the kids today
awoods: So maybe I'll pass a draft by you this weekend, then send it out Monday?16:59
Have a good weekend17:01
* apb18 leaves
* ajs6f leaves17:33
* th5 leaves17:34
* whikloj leaves18:00
* peichman leaves18:02
* yamil leaves18:13
* ajs6f joins18:29
* kefo leaves18:50
* ajs6f leaves19:04
* ajs6f joins20:00
* ajs6f leaves20:05
* ajs6f joins21:01
* ajs6f leaves22:03
* dhlamb leaves22:50
* dhlamb joins22:55
* ajs6f joins22:59
* f4jenkins joins23:40
* dhlamb leaves23:56
* awoods leaves00:08
* ajs6f joins00:29
* ajs6f leaves01:04
* downwriter joins01:10
* downwriter leaves01:15
* downwriter joins01:57
* ajs6f joins02:00
* ajs6f leaves02:35
<downwriter>anybody awake?
wondering about import/export utility02:36
it imports into /rest/.well-known/
instead of stated /rest/<desired location>?
* downwriter leaves02:50
* ajs6f joins03:30