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

Using timezone: Eastern Standard Time
* dhlamb joins07:18
* dhlamb leaves07:45
* dwilcox joins07:53
* awead joins08:28
* acoburn joins
* awead leaves08:45
* mikeAtUVa joins08:48
* yinlin joins09:00
* ksclarke joins09:19
* jcoyne joins10:05
* acoburn leaves10:26
* barmintor joins10:32
* dhlamb joins10:46
* github-ff joins11:14
[fcrepo4] yinlinchen closed pull request #732: Fcrepo 1290 (master...fcrepo-1290) http://git.io/AOa2
* github-ff leaves
* whikloj joins11:34
<dwilcox>mikeAtUVa: Ah, I think issues need to have the Fix Version set to 4.1.1 in order to appear on the scrum board11:44
<mikeAtUVa>dwilcox: ah, thanks!
<dwilcox>mikeAtUVa: no problem - Andrew set up the filter so I wasn't sure what criteria were being used11:45
<ruebot>Is this the best documentation on the different export contexts? https://wiki.duraspace.org/display/FEDORA37/fedora-export11:54
<barmintor>ruebot: I have honestly been learning about the difference as I work through the bugs, I think that’s the best we’ve got11:58
ruebot: afaik the difference is whether DS content gets inlined as base64 encoded data or indicated with a URI
or, in the case of AtomZIP, packaged into the zip file as a separate entry11:59
<ruebot>barmintor: kinda the same here. mikeAtUVa hashed it out for me last week when i stumbled across the bug.
barmintor: you can blame me for that :-)
<barmintor>:)
bugs are bugs
<ruebot>we're just trying to link to the best resource here https://github.com/Islandora/islandora_bagit/pull/28#issuecomment-76748991
<barmintor>that was an easy one to fix, though I was pretty annoyed that there were no tests12:00
<ruebot>glad to hear it was an easy fix12:01
<barmintor>ruebot: https://wiki.duraspace.org/display/FEDORA37/REST+API#RESTAPI-export12:02
* acoburn joins
<barmintor>ruebot: https://wiki.duraspace.org/display/FEDORA37/API-M#API-M-exportexport
three locations, all the same description. No indication of difference.12:03
ruebot: the longer descriptions are on the constants in the code12:04
ruebot: excepting ‘default’, which varies by serialization impl
<ruebot>@_@12:05
<barmintor>sorry dude this is literally v3.0 stuff. I’m a v3.5 latecomer
<ruebot>no worries. i'm really starting to understand the whole time and resources thing over the last year.12:06
<barmintor>sorry to hear that
<ruebot>we can only do some much with what we have, and if folks want more, they gotta provide me.
heh
s/me/more
<barmintor>ruebot: I’m happy to put some description in one of those pages, which seems the most easily identifiable to you?12:08
* barmintor inclines to REST out of personal prejudice
<ruebot>barmintor: personal prejudice works for me.12:10
<barmintor>QUOTED
<ruebot>mind if i tag you in that in that issue?
<barmintor>I kid, I kid12:11
nope. tag away
<ruebot>barmintor++12:12
<barmintor>aw geez what the hell is this crap https://jira.duraspace.org/browse/FCREPO-84912:13
<ruebot>ouch.12:17
<barmintor>voila.12:27
ruebot: sorry, that was opaque- I updated the REST docs12:31
<ruebot>barmintor++12:57
thank you!12:58
* awead joins13:18
* Nianli joins14:54
* dwilcox leaves15:08
* dwilcox joins15:09
* dwilcox leaves15:41
* dhlamb leaves15:56
<ruebot>awoods: is this the cla you need me to sign? https://wiki.duraspace.org/display/DSP/Contributor+License+Agreements16:08
...or dwilcox16:09
* github-ff joins16:32
[fcrepo-camel] acoburn opened pull request #59: added transaction support for fcrepo-camel (master...fcrepo-1349) http://git.io/xgY5
* github-ff leaves
* dhlamb joins16:39
* github-ff joins16:53
[ontology] acoburn closed pull request #21: draft fcdm vocabulary (master...fcdm-vocabulary) http://git.io/xIjU
* github-ff leaves
* dhlamb leaves
* Nianli leaves16:58
* mikeAtUVa leaves16:59
* jcoyne leaves17:00
* github-ff joins17:04
[fcrepo-camel] acoburn opened pull request #60: this fixes a nonsensical error message (master...fcrepo-1389) http://git.io/xgzt
* github-ff leaves
* github-ff joins17:07
[fcrepo-camel] acoburn closed pull request #52: added support for transactions (master...support-transactions) http://git.io/NkW2
* github-ff leaves
* acoburn leaves17:12
* mikeAtUVa joins17:14
* whikloj leaves17:52
* ksclarke leaves18:37
* dwilcox joins19:03
* dwilcox leaves19:10
* dhlamb joins20:49
* jcoyne joins21:31
* jcoyne leaves21:55
* ksclarke joins22:02
* dhlamb leaves22:53
* jcoyne joins23:03
* jcoyne leaves23:38
* dhlamb joins23:50
* dhlamb leaves23:54
* dhlamb joins00:51
* dhlamb leaves00:55
* ksclarke leaves01:21
* ksclarke joins01:27
* dhlamb joins01:30
* ksclarke leaves01:32
* dhlamb leaves01:34
* ksclarke joins01:49
* ksclarke leaves02:08
* dhlamb joins02:31
* dhlamb leaves02:36
* dhlamb joins03:31
* dhlamb leaves03:36
* dhlamb joins04:32
* dhlamb leaves04:37
* dhlamb joins05:33
* dhlamb leaves05:37
* dhlamb joins06:34
* dhlamb leaves06:38
* dhlamb joins07:34
* dhlamb leaves07:39
* dwilcox joins07:51
* dhlamb joins08:00
* jcoyne joins08:18
* mikeAtUVa leaves08:49
* mikeAtUVa joins09:02
* whikloj joins09:25
* ksclarke joins09:31
<whikloj>dwilcox: I don't seem to have access to any of the tickets, other than commenting. So I am creating a PR for FCREPO-1357 which I will add to the comments for now.09:42
<dwilcox>whikloj: Ah, I just need to adjust your JIRA permissions. Just a minute.09:43
whikloj: Ok, try again after refreshing - you should have the correct permissions09:44
* github-ff joins09:45
[ontology] whikloj opened pull request #22: Addresses FCREPO-1357 (master...FCREPO-1357) http://git.io/xonX
* github-ff leaves
<whikloj>dwilcox: That did it, thanks.
<dwilcox>whikloj: np09:46
<whikloj>dwilcox: one more thing, is the correct JIRA flow to "start work", then "close issue" when your done or do I need to "start review" to shift it to an "In Review" state?09:49
<dwilcox>whikloj: I believe you need to "start review" to move it to "in review" but let me know if that doesn't work09:51
* Nianli joins10:06
* acoburn joins10:16
<Nianli>dwilcox: yesterday, I was working on FCREPO-1371. I deployed the war file on the same environment as report. I did see the errors as report at first time. But I don't have performance issue on federated files. Details I will put in JIRA comments. The interesting thing is the error gone after I turned the log to debug mode for fcrepo.log.http.api. Today: I will continue working on this JIRA issue.10:19
dwilcox: Help/Blocker: Please help on StreamingBaseHtmlProvider.java line 237, have error: Exception rendering #parse(views/common-metadata.vsl) at /views/node.vsl[line 37, column 13 ] . But use debug setting in logback.xml, error disappeared,10:22
<mikeAtUVa>dhlamb: let me know if you have any time today or tomorrow morning to have a quick chat about migration tooling designs and tasks
<dhlamb>mikeAtUVa: i'm in the middle of a project meltdown atm, but i should be able to free up some time this afternoon or tomorrow morning10:24
<barmintor>mikeAtUVa: I would like in on that- it’s what I’m supposed to be working on during our pilot migration this month10:25
<dhlamb>barmintor: the more the merrier
<barmintor>dhlamb: tomorrow morning better for me personally10:26
<mikeAtUVa>barmintor, sounds good. dhlamb, barmintor how is 10am eastern tomororw?
<dhlamb>barmintor, mikeAtUVa: sounds good. i'll block it off now10:27
<barmintor>dhlamb, mikeAtUVa ditto
<mikeAtUVa>barmintor, dhlamb, alright, I'll find you here tomorrow at 10 and we can choose a venue then.10:28
<ruebot>i'm around if y'all need me too.
<barmintor>no, no ruebot
<ruebot>no project directors!
<barmintor>no team players allowed
* ruebot goes back to designing an abstract shared community data model naming model into order to properly name the shared community data model10:29
<barmintor>i see what you did there ruebot10:30
<yinlin>- Yesterday: Work on FCREPO-1290, add unit test casts, check FCREPO-1373 and post a new comment and wait for response.10:40
- Today: FCREPO-1290
dwilcox: I use a new client, not sure how to directly msg u yet10:41
dwilcox: does it work?
<whikloj>hey dwilcox: generated a vagrant machine for FCREPO-1377 with the desired components, not sure if awoods wants F4 pre-configured to index to Fuseki and/or Solr. So I am leaving it open till he confirms.10:51
<ruebot>barmintor: i had a thought the other day about the openseadragon and djatoka issues you'd told me about once.10:52
barmintor: maybe it has something to do with our setting for generating jp2 and our openseadragon settings.10:53
<whikloj>dwilcox: working on FCREPO-1357 now with the guidance of one acoburn
<barmintor>ruebot: go on10:55
<dwilcox>whikloj: Sounds good, thanks.10:56
<ruebot>barmintor: https://github.com/Islandora/islandora_solution_pack_large_image/blob/d85bc1076d9e0bd52c56d8db7df1d7b98fb471f1/includes/derivatives.inc#L273-L325 && https://github.com/Islandora/islandora_openseadragon/blob/7.x/includes/admin.form.inc
barmintor: worth looking at some beautiful php maybe? (i'm gonna disappear for the rest of the day here in about 15 minutes. but if you have any questions, ask away, and i'll reply when i can.)10:57
<dwilcox>Nianli: It's off that the errors disappear when you switch to debug mode. Are you able to investigate further?10:59
* Nianli leaves
* Nianli joins11:05
* dhlamb leaves11:24
* ajs6f joins11:45
<barmintor>ajs6f: you know Java. you have some familiarity with fcr3.12:27
<ajs6f>barmintor: Neither by my own choice or fault.12:28
nor
Or maybe it is or?
Hm.
<barmintor>ajs6f: i just want you to review PRs :P12:32
<ajs6f>barmintor: Which one(s)?
<barmintor>ajs6f: all of them. just all of them. i just want it to be over.12:33
<ajs6f>barmintor: "The only way to win is not to play."12:34
<barmintor>tecoripa is going to try to look at the ones I submitted.
ajs6f: yes, that was a wise Amiga that said that
<ajs6f>barmintor: Do you want me to also look at yours?12:36
<barmintor>ajs6f: if you have time, it would be helpful. it’s bugfix stuff including mikeAtUVa’s problem12:37
<ajs6f>barmintor: sure. Can you give me some links (maybe the issues in Jira)?12:38
<barmintor>ajs6f: all of the open PRs from me on fcr3 are named for their corresponding jira issues12:39
ajs6f: and you’re a prince :)
<ajs6f>barmintor: so Fcrepo 849, Fcrepo 1233, Fcrepo 1384?
<barmintor>ajs6f: yes12:40
<ajs6f>barmintor: I'll examine them now.
<barmintor>ajs6f: i love you, despite or perhaps because of everything
<ajs6f>barmintor: Both and nor neither.12:41
* dwilcox leaves12:47
* dhlamb joins12:54
<barmintor>ajs6f: it’s the ternary13:03
* dwilcox joins
<ajs6f>barmintor: It can't inference over the ternary?
barmintor: That's weird. I could swear I've written that sort of thing before.13:04
<barmintor>I’m looking at the compiler errors right in front of me
<ajs6f>barmintor: I believe you. I'm just really surprised. It actually calls out the ternary?13:05
<barmintor>oh no: It says that it can’t cast List<?3 extends Object> to List<Triple>
<ajs6f>barmintor: Well, that is more like what I would expect. Although again, it's weird, because emptyList() is generic and ought to take on List<Triple> in that context.13:06
<barmintor>ajs6f: fixed in Java 8
<ajs6f>barmintor: Oh well. It's a choice between an ugly annotation and extra verbosity. Who cares?13:07
barmintor: I'm pretty happy with fcrepo-849.
barmintor: 1233 looks a lot bigger.
<barmintor>ajs6f: Collections.<Triple>emptyList()
<ajs6f>barmintor: Oh, right. Call the shot explicitly.13:08
barmintor: I think that's a little more readable than an supress-warnings, but whatevs.
<barmintor>ajs6f: I do, too about to push it up
<ajs6f>barmintor: Cool. TYPE-SAFETY 4EVA.13:09
* acoburn leaves13:13
<ajs6f>barmintor: Looking at 1233, is this tested by running with -Ddemo.format=mets, -Ddemo.format=atom, etc.?13:14
<barmintor>ajs6f: the thread safety of the de/serializers is unit tested, and the IT for different format existed for everything but an AtomZIP roundtrip, that I could find.13:15
<ajs6f>barmintor: Okay, cool.
<barmintor>ajs6f: (the AtomZip test I added, aslo)13:16
ALSO
* acoburn joins13:19
* dwilcox leaves13:21
* dwilcox joins13:22
* acoburn leaves13:25
* acoburn1 joins
<barmintor>it’s grimly comic that one of the most frequent items of feedback we’re getting on the hydra side lately is about not having JDK 7 installed, here in the last month of its public support.13:26
ajs6f: as an aside, I would note the bug that branch was aimed at (rightfully tagged major) persisted since 3.4 b/c of reluctance to change public Java API of a module for which most customization is at a separately config’d store underneath13:39
<ajs6f>barmintor: Who, besides other Fedora code, is possibly using that "public" API?13:40
<barmintor>ajs6f: “implementers”
<ajs6f>barmintor: So it's an SPI?
* edInCo joins13:41
<ajs6f>barmintor: Wait, wtf implementors? Who are these people?
<barmintor>ajs6f: this is why I’m so whiny about scoping version commitments
<ajs6f>Can we name any?
barmintior: Not whiny. Not whiny at all.
barmintor: If anything, you're too kind.13:42
<barmintor>Bah.
but I was flummoxed to hit this bug, find the ticket, and see that it was about 5 years old.13:43
<ajs6f>barmintor: I don't know why. There are bugs in there filed by Sandy Payette.13:44
Possibly on her own code.
<barmintor>we’re doomed.
<ajs6f>We're done. (With F3, that is, unless you keep being generous.)
* dwilcox leaves13:46
* dwilcox joins13:47
<ajs6f>barmintor: Do you know of any site that actuall _used13:51
_ the Atom serialization?
<barmintor>yes: the institution employing the reporter of the bug
<ajs6f>barmintor: I thought the report was on METS?13:52
Wow, indeed. Atom
<barmintor>1233 is AtomZIP
<ajs6f>Yeah.
* acoburn1 leaves13:53
<ajs6f>I was wondering why so much diffing to Atom stuff.
<barmintor>it’s a much more lightweight deserialization, as it happens13:54
you can implement that deserializer statelessly w/o recourse to SAX content handlers and crap
<ajs6f>barmintor: Yeah, as such things go, it's not bad. But I'm very happy to leave that whole end of things to a decent RDF parsing library, which we could now do.13:55
<barmintor>yay RDF
<ajs6f>Really Decent Format.
<barmintor>speaking of, I have to go talk bibframe with the metadata librarian
<ajs6f>Sorry to hear it. Bibframe is a pretty lost cause, for my money.13:56
MARC-in-triples.
<Nianli>ajs6f: help please!14:02
<ajs6f>Nianli: Are you on fire? Are you falling from a great hight? Are you about to be hit by a speeding train? What's the matter?!14:03
<Nianli>yes, I am smoking right now! Please look at fcrepo-1371, I put comments there.14:04
ajs6f: Do you have any idea about this? It is so weird.14:05
<ajs6f>Nanli: Velocity is the engine that renders HTML templates for the views of resources.
Nianli: It's not obviously connected with federation at all.
Nianli: But you may want to see if it appears accessing non-federated material. To be honest, I wouldn't worry too much about it. At most, I would file a ticket and let awoods decide whether it's worth pursuing.14:07
<Nianli>ajs6f: Yes. I agree with you. I am not so sure I should put my efforts there. Since I didn't see it is related with the federated performance issue.14:08
<ajs6f>Nianli: I can almost guarantee you that it is not.
<Nianli>ajs6f: but the weird thing is, I can't duplicate the error now :(. I don't know when it will come, and why it is gone.14:09
<ajs6f>Nianli: Velocity has a habit of throwing a lot of stuff into the logs that appear to be errors but turn out simply to mark the processing of missing information or other non-problematic occurences.
<Nianli>OK. Great! thanks a lot!
<ajs6f>Nianli: You can file a ticket, then immediately mark it "Can't duplicate— Won't fix" and close it. Then you've tied up all the loose ends.14:10
* Nianli_ joins14:11
ajs6f: thank you, thank you...
* Nianli leaves14:14
* Nianli_ leaves14:15
* Nianli joins14:16
* jcoyne leaves14:23
* dwilcox leaves14:32
* jcoyne joins
* dwilcox joins
<Nianli>ajs6f: did you look at the catalina log file attached to the ticket? Is it the same error with mine? It is so similar but "Caused by: java.net.SocketException: Broken pipe".14:33
<ajs6f>Nianli: BUt did you see the delays?14:34
<Nianli>ajs6f: what delays?14:36
<ajs6f>Nianli: The delays that caused @scossu to file the ticket.14:37
<Nianli>ajs6f: No, I didn't.
<ajs6f>Nianli: Hm. I think we need some more data. Are you able to write a script (using curl, perhaps) to repeatedly time a retrieval of HTML pages from your test repository?14:38
Nianli: We need to either reproduce @scossu's delay or convince ourselves that we can't.14:39
<Nianli>ajs6f: Yes. I can run a script to retrieve files. Actually I did, but I didn't see any problem with it.14:41
<barmintor>ajs6f: I don’t understand what you mean by new HashMap<>() in place of declaring the types14:42
<ajs6f>NIanli: You were retrieving HTML pages (because @scossu says it only happened for him when they were HTML pages), and did you time the retrieval of the pages? And you didn't see any delays?
<barmintor>ajs6f: is it just that the types are declared on the variable, so they’re redundant on the constructor call?
<ajs6f>barmintor: right, since Java7. Just the diamond <> is sufficient.
<Nianli>ajs6f: I didn't. I will try now.14:43
<barmintor>ajs6f: I don’t have a strong sense of style on these issues, that’s good to know
<ajs6f>barmintor: You've seen the way I dress. You can't accuse me of a sense of style.
<barmintor>ajs6f: it’s not a lack of style. there’s no shortage of style there.14:44
“a strong point of view”
<ajs6f>Nianli: Ah, that may the crucial point. It's about the HTML pages. @scossu says in his first comment "I cannot replicate the exception using curl. It only appears when using the web UI."14:45
Nianli: That means HTML.
* ajs6f awed by the might of barmintor's POV, questions his own strength of purpose14:46
afk bbs14:48
<Nianli>ajs6f: I am confused here. I thought WebUI is through localhost:8080/fedora/rest/federated/, right?14:52
<ajs6f>NIanli: It's through whatever the URI of the resource in question is. If you put your federated stuff at that URI, then yes.14:54
<Nianli>ajs6f: that is what I did. But I didn't see big delay either.14:55
<ajs6f>Nianli: And you're sure that you are retirieving HTML?
<Nianli>ajs6f: what you mean retrieving HTML?14:57
<ajs6f>Nianli: That you are not retrieving the RDF representations at those URIs.
Nianli: What exactly are you scripting? Is it the retrieval of different URIs under the federation point?14:58
* dhlamb leaves
<Nianli>get the content of the files14:59
ajs6f: I wrote scripts to get the content of those federated files.15:05
<ajs6f>Nianli: Okay, now we're getting somewhere. scossu wasn't having any trouble doing htat.15:08
Nianli: He was having trouble retrieving the HTML representations of those resources.15:09
Nianli: See the first comment on the ticket: "I cannot replicate the exception using curl. It only appears when using the web UI."
"the web UI" means it must have been HTML.
<Nianli>ajs6f: How could I do it?Could you give me any example for that?15:10
<ajs6f>Nianli: So you need to write the script so that, instead of retrieving the files, it retrieves the HTML pages in that section of your repository.
Nianli: If you upload your thing to /foo, then /foo/fcr:metadata would be the URI of the description of that thing. Thats what scossu was finding the delays in.15:12
<Nianli>ajs6f: thank you so much! I will try it see if I can replicate the problem.15:13
<ajs6f>Nianli: Good luck.15:14
barmintor: It looks like 1384 is mostly either destatic-izing or stuff that appeared in 1233. I can't find the string bounds thing?15:17
<Nianli>ajs6f: thanks. Although I didn't write scripts to do that. I did look at the fcr:metadata through browser. I didn't find any delay. But I will write scripts to check. Thanks again!
<ajs6f>Nianli: Sure. I like to write a nice simple script that just read URLs from a file and uses "time curl $file". Then you can grep the results.15:18
<Nianli>ajs6f: Thanks! Big help! :)15:21
<ajs6f>Nianli: Sure. If you haven't used time, just be aware that it sends its output to STDERR, so you can pipe that alone through grep.
<Nianli>ajs6f: OK.15:26
* github-ff joins15:35
[fcrepo4] yinlinchen opened pull request #741: Fcrepo 1290 (master...fcrepo-1290) http://git.io/xizN
* github-ff leaves
* jcoyne leaves15:48
<Nianli>ajs6f: I can't replicate the error and the delay for for:metadata either. Any suggestion for next step?15:56
<ajs6f>Nianli: Check with scossu as to whether he only saw the delays in an actual web browser. If he did, it might be because there is some connection with Javascript being used in the web UI. Curl wouldn't execture Javascript, and you might not be seeing it because of that.15:57
* dwilcox leaves
* dwilcox joins15:58
<Nianli>ajs6f: OK. Will do that.
* ajs6f leaves16:22
* dwilcox leaves16:24
* dwilcox joins16:48
* dwilcox leaves16:52
* mikeAtUVa leaves17:02
* whikloj leaves
* dwilcox joins17:04
* Nianli leaves17:08
* barmintor leaves17:15
* barmintor joins
* dwilcox leaves17:16
<barmintor>ajs6f: the String bounds stuff in the 1384 branch is in StringUtility; it’s two lines. The rest is making it testable.17:18
oh wait: the StringUtility stuff is actually a prior commit from another branch, the 1384 stuff is17:19
wait… no, nvm. The first one. I’m exhausted, have mercy.17:20
* scossu joins17:37

Generated by Sualtam