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

Using timezone: Eastern Standard Time
* jonathangee leaves05:46
* jonathangee joins05:50
* coblej joins08:45
* bseeger joins09:02
* benpennell joins09:04
* whikloj joins09:07
* coblej leaves09:17
* dhlamb joins09:18
* coblej joins
* mikeAtUVa joins09:25
* yamil joins09:34
* peichman joins09:35
* awoods joins09:36
bseeger/mikeAtUVa/whikloj: Good morning, do any of you have the bandwidth to look into this potentially low-hanging ticket:09:38
<whikloj>awoods: so is the goal here to allow people to use localhost:8080/rest/collection/image/fcr%3Aversions/v0/fcr%3Afixity to access the fixity of a version?09:41
<awoods>whikloj: I think there are two things:09:42
whikloj: 1. we should likely not be returning character-encoded URLs09:43
whikloj: 2. we need to determine if the /fcr:fixity service can work on versions
<whikloj>awoods: no guarantees, but I'll have a look09:46
<mikeAtUVa>awoods: I don't have time this week or next.
<awoods>thanks, whikloj, please note anything you discover along the way in the ticket.09:47
<whikloj>I bet bseeger could do it, she's very good at fixing stuff
<awoods>thanks for the response, mikeAtUVa.
whikloj, agreed
<bseeger>awoods, whikloj: just seeing this.10:00
<awoods>bseeger: good morning10:01
<bseeger>awoods, whikloj: good morning. Thanks, whikloj. I could probably look at it next week, if you don't have time whikloj.10:02
<awoods>whikloj/bseeger: I would suggest removing the character-encodings and only keeping the /fcr:fixity suffix on versions if their fixity can indeed be checked.10:03
<bseeger>and whikloj - I've noticed that you're quite good at fixing things as well. :)10:04
<whikloj>awoods/bseeger: I'll see what I can do, it has been awhile since I dug into the codebase and I'm stealing time while monitoring/babysitting some processes.
bseeger: that is a horrible rumour someone started10:05
<bseeger>whikloj: okay, feel free to toss it my way if you need to.10:07
* dwilcox joins10:08
* github-ff joins10:26
[fcrepo-specification] zimeon pushed 1 new commit to master: https://git.io/vQJsv
fcrepo-specification/master ffbdc4c Simeon Warner: Merge pull request #141 from fcrepo/awoods-patch-1...
* github-ff leaves
* github-ff joins
[fcrepo-specification] zimeon deleted awoods-patch-1 at ef1775b: https://git.io/vQJsU
* github-ff leaves
* coblej leaves10:58
* coblej joins11:00
* coblej leaves11:18
* coblej joins
* bseeger leaves11:28
<whikloj_AFK> awoods: ping12:00
* awoods on a call
<whikloj>no worries, not urgent12:04
* mcritchlow joins12:08
<awoods>whikloj: what do you have going on?12:09
<whikloj>awoods: I don't fully understand Jersey, but what is happening is that in the HtmlResourceConverter it take the URL (eg. http://localhost:8080/rest/outside/fcr:versions/v0/fcr:fixity)12:10
It splits the URL on fcr:versions, then gets the JcrNode for the front part (/rest/outside).
Then I tries to get the second half address of a JcrNode inside the Node from step 1.
eg. Jcr.Node.getNode(/rest/outside).getNode(/v0/fcr:fixity) which fails.
awoods: what I'm wondering is because fcr:versions is before fcr:fixity in the URI that it gets called first and actually we should be calling fcr:fixity?12:11
awoods: But I'm not clear how I end up in the HtmlResourceConverter yet, so I could be mixed up
* coblej leaves12:12
<whikloj>This is where it explodes though
<awoods>whikloj: is there a way of getting the version (like it is trying to do) followed by performing the fixity?12:13
* orangewolf joins12:15
<whikloj>awoods: I'm not sure, I can't tell if it sees this as two separate actions. But in the call stack right now, I see the FedoraVersions.getVersion but nothing about fixity.
awoods: and fcr:fixity seems only to be referenced in the HtmlTemplate (which is why I wonder about Jersey)12:16
* coblej joins12:18
<awoods>whikloj: do you know where the character-encoding is coming from?
<whikloj>awoods: I'm not seeing any character encoding... or at least I haven't yet. But I can't use the HTML UI
<awoods>whikloj: have you been forbidden from using the HTML UI?12:19
* coblej leaves
<whikloj>awoods: https://jira.duraspace.org/browse/FCREPO-2525
very funny Mr. Woods
awoods: It's just not working correctly for me, so I'm not sure that any issues I see would be actual issues or otherwise related12:20
<awoods>whikloj: maybe working the HTML-UI bug is step #1
<whikloj>awoods: That whole function is problematic, we are basing our decision on NonRdf or Rdf on the return of Link Headers? Should I just add some Link Header to the fcr:versions response?12:22
<awoods>whikloj: like here: https://jira.duraspace.org/browse/FCREPO-2511 ?12:23
<whikloj>AHHHHHHHH its a verrrrry deeeeep rabbit holeeeeeeeee12:24
awoods: well played mr. woods, ruebot would be proud12:25
<awoods>whikloj: I assume you have reproduced Jonathan Rochkind's comment: https://jira.duraspace.org/browse/FCREPO-1247?focusedCommentId=56142&page=com.atlassian.jira.plugin.system.issuetabpanels%3Acomment-tabpanel#comment-5614212:26
<whikloj>awoods: no I was working towards that, got sidetracked by not even being able to click the fcr:versions link and went to Curl where there is no encoding issue I can see12:28
<awoods>whikloj: another way of approaching this is to character-encode the "hasFixityService" URL on versions of binary resources.
whikloj: see: https://groups.google.com/forum/#!msg/fedora-tech/qw3GlptWBQw/CaGXHMuiRXIJ12:29
whikloj: the fixity service actually works on binary versions if the fcr:versions segment is character-encoded
<whikloj>awoods: ok so this comes back to my original question. With the URL having both fcr:versions and fcr:fixity it appears that fcr:versions takes precedence. So you want to encode fcr:versions on the subsequent hasFixityService Link header12:30
or not subsequent... I guess preceding12:31
<awoods>whikloj: yes... although hasFixityService is a triple, not a header12:34
<whikloj>awoods: right, I've done that. Re-compiling
<awoods>whikloj: you're fast
<whikloj>awoods: you can't have versions inside versions right? I shouldn't expect http://foo/rest/bar/fcr:versions/v1/box/hat/fcr:versions/v12/pictures12:35
<awoods>whikloj: nope12:40
<whikloj>awoods: ok, I'm going back down the hole
<awoods>whikloj: good luck... and be safe12:41
* github-ff joins
[fcrepo4] dbernstein created 4.7.4-RC from 4.7-maintenance (+0 new commits): https://git.io/vQJXp
* github-ff leaves
* bryjbrown joins13:01
* bseeger joins
<dbernstein>awoods: so to create the release candidate, I will run through the commits on master but not in the 4.7.4-RC. If a given commit is not associated with a “breaking-change” labelled jira ticket, it will go in to the 4.7.4-RC.13:03
<awoods>dbernstein: sounds right... assuming your 4.7.4-RC is a branch from 4.7-maintenance.13:08
* travis-ci joins13:16
fcrepo4/fcrepo4#5035 (4.7.4-RC - 23d983c : Andrew Woods): The build passed.
Change view : https://github.com/fcrepo4/fcrepo4/compare/4.7.4-RC
Build details : https://travis-ci.org/fcrepo4/fcrepo4/builds/245018285
* travis-ci leaves
* coblej joins13:19
* benpennell leaves13:23
* coblej leaves13:26
* coblej joins
* bseeger leaves13:29
* benpennell joins13:30
* bseeger joins13:36
* github-ff joins13:42
[fcrepo4] whikloj opened pull request #1202: Add Link headers to /fcr:versions listing and add IT (master...FCREPO-2511) https://git.io/vQJ5p
* github-ff leaves
* mcritchlow_ joins14:29
* mcritchlow leaves14:31
* mcritchlow_ leaves
* mcritchlow_ joins
* mcritchlow joins14:38
* mcritchlow_ leaves14:39
* mcritchlow leaves14:45
* mcritchlow joins14:52
* benpennell leaves14:53
* orangewolf leaves14:59
* orangewolf joins
* orangewolf leaves15:04
* benpennell joins15:08
* github-ff joins15:20
[fcrepo4] dbernstein pushed 1 new commit to master: https://git.io/vQUIh
fcrepo4/master 7887cde Jared Whiklo: Add Link headers to /fcr:versions listing and add IT (#1202)...
* github-ff leaves
* github-ff joins15:21
[fcrepo4] dbernstein pushed 30 new commits to 4.7.4-RC: https://git.io/vQULU
fcrepo4/4.7.4-RC 160b61b dbernstein: Ensures that the object URIs within a SPARQL update are validated bef… (#1154)...
fcrepo4/4.7.4-RC 4623c17 dbernstein: Ensure Content-Type header for HEAD requests on containers mirrors wh… (#1155)...
fcrepo4/4.7.4-RC 84ab8a1 Andrew Woods: Reduce logging level of WebApplicationException...
* github-ff leaves
* travis-ci joins15:35
fcrepo4/fcrepo4#5040 (master - 7887cde : Jared Whiklo): The build passed.
Change view : https://github.com/fcrepo4/fcrepo4/compare/04a56d87a501...7887cdeb1ed0
Build details : https://travis-ci.org/fcrepo4/fcrepo4/builds/245075063
* travis-ci leaves
* travis-ci joins15:36
fcrepo4/fcrepo4#5041 (4.7.4-RC - 3788389 : Jared Whiklo): The build passed.
Change view : https://github.com/fcrepo4/fcrepo4/compare/23d983cb0453...378838903db8
Build details : https://travis-ci.org/fcrepo4/fcrepo4/builds/245075463
* travis-ci leaves
* dbernstein leaves15:49
* mcritchlow leaves16:03
* mcritchlow joins16:05
* mcritchlow_ joins16:09
* mcritchlow leaves
<whikloj>awoods: ping16:23
<whikloj>awoods: Ok I solved FCREPO-1247 by altering the Jersey @Path parameter, but it has revealed that Fixity results also don't return Link headers.16:24
awoods: Should that be in the same ticket or a new one?
<awoods>whikloj: new ticket, please
whikloj: sounds like a prequel ticket
<whikloj>awoods: really it means that the HTML UI still fails, but curl works and a direct access works.16:25
direct access with a browser I mean
<awoods>whikloj: until your prequel goes in?
<whikloj>awoods: right
<awoods>whikloj: thanks for unearthing that additional bug16:26
<whikloj>awoods: which is going to be a copy of the PR I just submitted early today but for FedoraFixity
* dwilcox leaves16:30
* orangewolf joins16:37
* github-ff joins16:43
[fcrepo4] whikloj opened pull request #1203: Skip fcr:versions handler for fcr:fixity requests (master...FCREPO-1247) https://git.io/vQUC7
* github-ff leaves
<whikloj>awoods: should I just add the work for the Headers on to this same PR? ^^16:44
<awoods>whikloj: no, let me merge your work... then you can work off of master.
<whikloj>awoods: sounds good
<awoods>whikloj: oh... I was expecting that you were going to do the headers PR first.16:45
<whikloj>awoods: oh sorry I... um I can do that and then rebase this one16:46
<awoods>whikloj: no worries... it is fine as is
<whikloj>awoods: I can put in the headers PR now though?
your call16:47
<awoods>whikloj: I was thinking the headers PR would be simpler to review and merge
whikloj: ..since it is basically the same as the previous PR
whikloj: I would like to test PR#1203 a little more
whikloj: it looks like rebasing PR#1203 will be trivial... so, yes, please create a PR for the headers first.16:48
<whikloj>awoods: ok, I'll submit the headers change as a PR and then I can rebase 1203 on top
* mikeAtUVa leaves16:51
* yamil leaves17:00
* benpennell leaves17:01
* github-ff joins17:02
[fcrepo4] whikloj opened pull request #1204: Add Link headers to fcr:fixity and integration test (master...FCREPO-2526) https://git.io/vQU85
* github-ff leaves
<whikloj>awoods: I have to go, but will rebase 1203 later17:03
* whikloj leaves
* bseeger leaves17:04
* coblej leaves17:16
* bryjbrown leaves17:18
* github-ff joins17:32
[fcrepo4] awoods pushed 1 new commit to master: https://git.io/vQUE4
fcrepo4/master 4b7d9f1 Jared Whiklo: Add Link headers to fcr:fixity and integration test (#1204)...
* github-ff leaves
* dhlamb leaves17:37
* travis-ci joins17:46
fcrepo4/fcrepo4#5044 (master - 4b7d9f1 : Jared Whiklo): The build passed.
Change view : https://github.com/fcrepo4/fcrepo4/compare/7887cdeb1ed0...4b7d9f17a104
Build details : https://travis-ci.org/fcrepo4/fcrepo4/builds/245124831
* travis-ci leaves
* peichman leaves17:51
* mcritchlow_ leaves18:13
* mcritchlow joins18:15
* mcritchlow leaves18:26
* cbeer leaves18:28
* cbeer joins
* orangewolf leaves18:33
* orangewolf joins18:34
* dwilcox joins20:03
* dwilcox leaves20:39
* orangewolf leaves21:47
* orangewolf joins21:48
* orangewolf leaves21:54
* orangewolf joins21:55
* orangewolf leaves22:00
* orangewolf joins22:11
* benpennell joins22:12
* benpennell leaves22:18
* orangewolf leaves22:36
* mcritchlow joins22:51
* awoods leaves23:03
* mcritchlow leaves23:17