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

Using timezone: Eastern Standard Time
* eddies leaves01:57
* eddies joins
* eddies leaves
* eddies joins
* kaarefc joins03:07
* kaarefc leaves03:20
* kaarefc joins03:31
* kaarefc leaves04:16
* kaarefc joins04:42
* kaarefc leaves06:56
* kaarefc joins
* kaarefc leaves07:02
* kaarefc joins07:41
* unai_ joins08:09
Hi to everyone08:10
I am trying to understand how the versioning works and a I have a really simple question. Think about I have this structure:
I have the top node q1, with 1 child (q11) and this one has 2 childs (q111,q112)
objects/q1/q11/ q111
and q112
At this moment, if I check the first node, I can see that q1 has 1 child.08:11
checking this link: localhost:8080/fcrepo/rest/objects/q1
I want to tag this entity with v1 tag. And after that be able to get the same info but with the version tag, like the api says:
And be able to see that v1 has q11 child08:12
What I have to do, or I am not understanding good?
* fasseg joins08:21
eddies: presentation yesterday was good, and I hope some people will try out the Alpha before OR and give me feedback...08:26
and no wars on that site yet: https://wiki.duraspace.org/display/FF/Downloads? since the site is public and says you can download the wars for 4-0 alpha08:36
* ksclarke joins09:27
* nbanks joins10:04
* barmintor joins10:18
* kaarefc leaves10:32
<barmintor>grr whitespace changes10:38
if I squash someone else's commits, does git maintain the tagged authorship?10:41
<cbeer>barmintor: the first commit "wins"10:51
<barmintor>cbeer++ // thanks
they're all by the same author
* unai_ leaves10:52
* kaarefc joins11:00
<barmintor>merge_commits-- rebase++11:02
<pivotal-bot__>Chris Beer delivered "Document our use of sparql update to update object properties" https://www.pivotaltracker.com/story/show/5156610311:08
Chris Beer added "Add an InvalidChecksumExceptionHandler to provide a reasonable error message when the checksum fails to match" https://www.pivotaltracker.com/story/show/5280463711:24
Chris Beer added "Checksum parameters on FedoraContent should be our urn:sha1:xxxx format?" https://www.pivotaltracker.com/story/show/5280472311:25
<cbeer>eddies: i'd like to finish those tickets before we cut the Alpha 1 release
"that Benjamin and futures are probably friends or at least virtual friends"11:26
* kaarefc leaves11:27
* kaarefc joins
<barmintor>"It would also be impossbile to look at Benjamin's activity stream and not to compare it those of… cbeer"11:30
"With this in mind, it's worth noting that futures speaks more languages." LIKE SCALA AMIRITE?!
<eddies>cbeer: i'd like to see https://github.com/futures/fcrepo4/pull/90 merged in before the alpha gets cut11:33
barmintor, do you want to look at that (because you presumably know something about it)?
<barmintor>I think that's the PR I'm supposed to be reviewing
I'm reviewing another one right now, but it should be easy to get to that one today.11:35
<eddies>ok, i'm going to go tweak the press release stuff for carol. and cbeer thanks for the doc you put up.
<barmintor>actually, should we squash those commits?
<cbeer>that way of framing it makes me less nervous about what we're actually delivering11:36
barmintor: yes.
<barmintor>ugh, that's a bunch of files11:37
<eddies>fasseg: i've put up a notice on the downloads page that nothing is currently ready for download yet
<barmintor>speaing of pre-alpha concerns: awoods was asking a while back whether we should move code into intermediate packages per-module (ie fcrepo-kernel/src/org.fcrepo.XXX -> org.fcrepo.kernel.XXX, etc.)11:38
* ajs6f joins
<cbeer>barmintor: that'd mean re-organizing all the code?11:40
(and having to screw with everything's dependency scanning)
<barmintor>cbeer: pretty mcuh. that's why I didn't just do it.
I wouldn't want to do it while we were, you know, working. But if things get to a quiescent state before alpha, it's worth considering.11:41
Maybe we make awoods do it at OR.11:42
<ajs6f>When is code freeze for alpha? Or are we going to freeze...
<cbeer>ajs6f: eddies wants to get your PR in, and i just created 2 tickets as i was doing some docs
<barmintor>ajs6f: eddies and cbeer were implying it was after the oauth gets merged11:43
<cbeer>i don't know if there's any other outstanding work
* kaarefc leaves
* nbanks leaves11:52
* ajs6f leaves12:14
<pivotal-bot__>Chris Beer estimated "Checksum parameters on FedoraContent should be our urn:sha1:xxxx format?" as 1 point https://www.pivotaltracker.com/story/show/52804723
Chris Beer started "Checksum parameters on FedoraContent should be our urn:sha1:xxxx format?" https://www.pivotaltracker.com/story/show/52804723
Chris Beer estimated "Add an InvalidChecksumExceptionHandler to provide a reasonable error message when the checksum fails to match" as ..." https://www.pivotaltracker.com/story/show/52804637
Chris Beer started "Add an InvalidChecksumExceptionHandler to provide a reasonable error message when the checksum fails to match" https://www.pivotaltracker.com/story/show/52804637
* ajs6f joins12:29
<cbeer>eddies: are you going to do the nexus setup?12:30
<eddies>unless you're volunteering =)
"oh, you wanted to know why the build failed? let me just tell you about all the line length warnings."12:32
<ajs6f>What really bugs me about them is the warnings for line length on import statements. Breaking import statements across lines is just ugly.12:33
<eddies>how are line length warning breaking the build?12:34
i thought they were just warnings
<cbeer>eddies: they are, but there was a javadoc unused param error in that spew of output12:35
<eddies>oh i see =)
yes, that is annoying
* nbanks joins12:37
<barmintor>this PR… I guess the tests are good, and removing the unnecessary synchronization is good, but the introduction of two unnecessary classes to wrap synchronized access to a Map<String, Lock> is not so great.12:50
* nbanks leaves12:52
<cbeer>barmintor: that's the fcrepo3 thing?12:56
<barmintor>sorry, yes
<ajs6f>barmintor: are you thinking that there's a better way, like something offered by a library already in play, or that it's just cruft period.12:57
<barmintor>ajs6f: I'm thinking the two new classes are cruft, but their intent can be preserved.
<cbeer>ajs6f: is 409 Conflict a reasonable error for a user-provided checksum mismatch?
<ajs6f>cbeer: like the user is trying to mutate a datastream and offers a bad checksum with?12:58
<ajs6f>yeah, 409 sounds right to me, along with an explanatory message12:59
i wish 417 Expectation Failed could be construed more broadly
<barmintor>I get 417s from my parents all the time
<ajs6f>barmintor: so we're wanting to keep track of a global map of Locks?13:00
barmintor: ouch.
<barmintor>ajs6f: we already do
<ajs6f>barmintor: oh, cool. So we can just use our gear.
<barmintor>I think so13:01
bb after lunch
<pivotal-bot__>Chris Beer finished "Checksum parameters on FedoraContent should be our urn:sha1:xxxx format?" https://www.pivotaltracker.com/story/show/5280472313:05
Chris Beer finished "Add an InvalidChecksumExceptionHandler to provide a reasonable error message when the checksum fails to match" https://www.pivotaltracker.com/story/show/52804637
* github-ff joins
[fcrepo4] cbeer pushed 2 new commits to master: http://git.io/DDRV4g
fcrepo4/master f0d3a4e Chris Beer: use the urn form of the checksum instead of passing around the type + digest
fcrepo4/master c170261 Chris Beer: wire in a reasonable HTTP response when user-provided checksums don't match
* github-ff leaves
<pivotal-bot__>Chris Beer added comment: "https://github.com/futures/fcrepo4/commit/c170261f6f714affcf821240fccc6415c01250db" https://www.pivotaltracker.com/story/show/52804637
Chris Beer added comment: "https://github.com/futures/fcrepo4/commit/f0d3a4ee07e6669cb853056017a6c0b50bba016c" https://www.pivotaltracker.com/story/show/52804723
<ksclarke>should someone getting started with ff use fcrepo's master branch?13:20
* travis-ci joins
[travis-ci] futures/fcrepo4#836 (master - c170261 : Chris Beer): The build passed.
[travis-ci] Change view : https://github.com/futures/fcrepo4/compare/7a5c02d3e716...c170261f6f71
[travis-ci] Build details : http://travis-ci.org/futures/fcrepo4/builds/8704366
* travis-ci leaves
<cbeer>ksclarke: as long as by fcrepo's master branch you mean https://github.com/futures/fcrepo413:28
<ksclarke>ah, thanks, I was looking at the wrong repo... fcrepo/fcrepo13:29
<cbeer>eddies: what are the edit permissions on the FF space like?13:52
<eddies>cbeer: they are weird13:53
confluence-users can add 1) pages 2) comments 3) attachments13:54
Anusha has explicit permissions that match confluence-users
jonathan, tom and i have administrative perms. that's it13:55
someone is clearly screwing around w/ permissions :-(
* nbanks joins13:58
<eddies>i just granted wider permissions to the fcrepo-committer group…but i have no way of seeing who is actually in that group
or at least i'm not aware of how to see the members13:59
<cbeer>eddies: i also wondered if we should let all registered users or something edit it during OR
i was just about to ask ksclarke to update the wiki if he noticed anything wrong, etc
<eddies>i only see page add/remove…i don't know what controls edit14:00
<eddies>well, you can edit pages, right?14:01
<ksclarke>I have an edit button on that first steps page
<eddies>(rather, you were able to edit 5 mins ago?)
<ksclarke>haven't actually tried it though
<cbeer>oh, cool.
eddies: i'm not sure i'm in that fcrepo-committer group.14:02
<eddies>yeah, i have a feeling it's from years ago
probably something cwilper maintained
* nbanks leaves14:03
<pivotal-bot__>Chris Beer delivered "Checksum parameters on FedoraContent should be our urn:sha1:xxxx format?" https://www.pivotaltracker.com/story/show/5280472314:05
Chris Beer delivered "Add an InvalidChecksumExceptionHandler to provide a reasonable error message when the checksum fails to match" https://www.pivotaltracker.com/story/show/52804637
Chris Beer added "Fix HTML validation errors" https://www.pivotaltracker.com/story/show/5282191714:49
<cbeer>eddies: there's another blocker for a release, imo14:50
<eddies>we're not spitting out valid html?14:51
<ajs6f>That's a blocker?
<cbeer>some minor issues, and a content escaping one
<ajs6f>We don't have i-tests for the views, do we?
(I'm still not sure it's a blocker.)14:52
<cbeer>we have some tests that prove they're created
but nothing validating it
<ajs6f>Right, fair enough.
<pivotal-bot__>Chris Beer estimated "Fix HTML validation errors" as 2 points https://www.pivotaltracker.com/story/show/52821917
Chris Beer started "Fix HTML validation errors" https://www.pivotaltracker.com/story/show/52821917
<ajs6f>There's always JTidy.14:53
<eddies>ajs6f/barmintor: how's https://github.com/futures/fcrepo4/pull/90 ?14:55
<ajs6f>barmintor is reviewing the living daylights out of it.
<eddies>i'm writing up oauth in the release notes, so it'll be a bit odd if this isn't merged
ajs6f: github is claiming there are merge conflicts14:56
<ajs6f>I'll check it.
<cbeer>velocity: because why would you want to escape data anyway?14:57
<ajs6f>When what you really want is escape velocity.
eddies: Needed the Maven repo for Oltu15:04
* github-ff joins
[fcrepo4] ajs6f pushed 1 new commit to OAuthIntoWebapp: http://git.io/BwY7rg
fcrepo4/OAuthIntoWebapp 477f5ca ajs6f: Link Apache Maven repo for Oltu
* github-ff leaves
<ajs6f>eddies: which was in the module, but I forgot to take with when I cut the module into fcrepo4.
now I gots to rebase.15:05
* kaarefc joins15:06
<pivotal-bot__>Chris Beer finished "Fix HTML validation errors" https://www.pivotaltracker.com/story/show/5282191715:07
* github-ff joins
[fcrepo4] cbeer pushed 1 new commit to master: http://git.io/4zeFjQ
fcrepo4/master 70c57e4 Chris Beer: make things validate as HTML5
* github-ff leaves
<pivotal-bot__>Chris Beer added comment: "https://github.com/futures/fcrepo4/commit/70c57e4f5da4f7e03887b27dbd05d80b64004b44" https://www.pivotaltracker.com/story/show/52821917
<cbeer>can't say I did a good job of it, but the couple pages i tested validated
<eddies>ajs6f: that's some nice xml formatting you've got there :P15:09
<cbeer>ajs6f: was this ticket just about rdfa? https://www.pivotaltracker.com/story/show/52229267
<pivotal-bot__>feature: Add microdata RDF to HTML API responses (unscheduled) / owner:
<ajs6f>eddies: What's wrong now?!?
<eddies>eliminates blank lines but appends whitespace…i don't even know how you get the latter =)
ajs6f: i'm just looking at https://github.com/futures/fcrepo4/commit/477f5ca431c87811db9c4c1c2822b089b6db10ef15:10
but it's ok
<ajs6f>eddies: I don't know. I'm using the same formatter you are.
<eddies>unfortunately, i don't think there's a way to share xml formatter settings
i think you have "clear all blank lines" selected15:11
<ajs6f>eddies: Do you know where that is?
<eddies>XML/XML Files/Editor
<ajs6f>nm found it.
* github-ff joins15:13
[fcrepo4] ajs6f force-pushed OAuthIntoWebapp from 477f5ca to d52d0b8: http://git.io/bqbUiA
fcrepo4/OAuthIntoWebapp ee08b2d ajs6f: Linked in fcrepo-auth-oauth
fcrepo4/OAuthIntoWebapp d4ad7fa ajs6f: Bringing fcrepo-auth-oauth under fcrepo4
fcrepo4/OAuthIntoWebapp 384ab1a ajs6f: Link Apache Maven repo for Oltu
* github-ff leaves
* nbanks joins
* nbanks leaves15:18
* travis-ci joins15:23
[travis-ci] futures/fcrepo4#837 (master - 70c57e4 : Chris Beer): The build passed.
[travis-ci] Change view : https://github.com/futures/fcrepo4/compare/c170261f6f71...70c57e4f5da4
[travis-ci] Build details : http://travis-ci.org/futures/fcrepo4/builds/8708412
* travis-ci leaves
<eddies>ajs6f: can you review https://wiki.duraspace.org/display/FF/Authentication+and+Authorization for correctness?15:30
i think some of it is wildly off now
<ajs6f>That's looks fine to me. What are you notcing?15:31
<eddies>ajs6f: didn't you stub out some simple authn endpoint for testing? i don't remember you doing any testing w/ container-level authn15:37
<ajs6f>eddies; Thjose are two different things. Maybe we need to reformat that page to make that clear?
I'lldo that now.15:38
* bljenkins leaves15:55
* bljenkins joins15:58
<cbeer>ajs6f: i might have missed it, but https://www.pivotaltracker.com/story/show/52229267
<pivotal-bot__>feature: Add microdata RDF to HTML API responses (unscheduled) / owner:
or something more
<ajs6f>cbeer: Sorry— yes, RDFa, you totally got the right idea, even though I said the wrong thing. Microdata? WTF?15:59
<pivotal-bot__>Chris Beer deleted "Add microdata RDF to HTML API responses" https://www.pivotaltracker.com/story/show/52229267
<cbeer>ok. we have that already :)
<cbeer>and i've actually checked that it works now
<ajs6f>Do we have validation on the way out, or in tests?
If not tests, make me an issue.
<pivotal-bot__>Chris Beer added "add tests for html responses to actually see if the content is sane" https://www.pivotaltracker.com/story/show/5282758916:01
Chris Beer edited "add tests for html responses to actually see if the content is sane" https://www.pivotaltracker.com/story/show/52827589
<cbeer>i think i asked barmintor if there was a good way to do that.. the only tools i found were unwieldy16:02
<ajs6f>You sure JTidy won't do it?
<cbeer>ah, that's a new one16:03
<ajs6f>In theory, it's a complete port of Tidy.
<cbeer>i saw http://htmlunit.sourceforge.net/
<ajs6f>Hm. Don't we want just rock bottom validation?16:04
(For now.)
Actually, what about doing XHTML strict and just letting the XSD do the work?
Aw, foo. JTify has no Maven presence. Goodbye JTidy.16:05
cbeer: Actually, thinking harder about it, were any of the things you were finding such as not to be caught by XHTML XML validation? Do we really need an HTML-specific tool?16:14
* github-ff joins
[fcrepo4] cbeer pushed 1 new commit to master: http://git.io/JYT_nQ
fcrepo4/master 405f99d Chris Beer: change jgroups bundler type to 'new' (see ISPN-3126)
* github-ff leaves
<cbeer>well, the type wasn't declared. and i've declared it as HTML5 now16:15
so.. no.
<ajs6f>cbeer: Cool. So we should just add the appropriate declarations, and have a test that tries to parse every kind of response as XML.
<cbeer>sorry, that no should have been yes, maybe.
we need an html-specific tool
because it's html5 now
<ajs6f>oh, didn't catch that.16:16
Do we need HTML5 for this? I'm not against it, just wondering if we can save ourselves some time here.
<cbeer>doesn't everyone need HTML5?
i think it's probably the right thing to do, though16:17
<ajs6f>Which is the right thing?
<ajs6f>Okay. In that case, we have to find a validator.
* ajs6f1 joins16:26
Well, it seems that HTML5 doesn't do validation.
* ajs6f leaves
<ajs6f1>But there's this:
and maybe I can pick up a list of errors from it?16:27
Hang on! There are schamta for HTML5:16:29
<ksclarke>though not popular, isn't there an xml form of html5 too?16:31
I'd think that'd have a way of validation
<cbeer>s/not/popular/probably dead/16:32
<ajs6f1>It likely would have XSD. But I don't know. Seems like in either case, the road to follow is declarative— get a schema.
cbeer: Objections to the WHATWG's schema?
<cbeer>go for it, if it works16:33
<ajs6f1>If it doesn't work, we just need to get involved with yet another community— WHATWG!
<ajs6f1>Fedora 4: Our views are so sweet, your browser can't render the,16:34
* kaarefc leaves
<ksclarke>https://raw.github.com/sideshowbarker/html5-schema/master/xhtml5.rnc with the claim "The xhtml5.rnc schema is automatically generated from schema used by the validator.nu andvalidator.w3.org/nu services."16:37
but html5 seems to be against validation... that's yucky xml stuff :-P
* ksclarke leaves
<ajs6f1>That makes testing so much easier. :016:38
http://syntax.whattf.org/relaxng/ comes straight from WHATWG. I'll trust 'em.16:39
So it seems to use it, we would need to be creating well-formed XML. That doesn't seem too burdensome.16:40
* ksclarke joins16:42
* ksclarke1 joins
* ksclarke leaves16:43
<ajs6f1>cbeer: That action in http-commons where we pass around the type of the resource to create (in AbstractResource), that's interesting… that seems to be the point where we couple the Fedora object model to the Web via identifiers. I don't have any problem with it, but that's the kind of thing we want to notice if we want to eventually explicate our architecture explicitly and expectorate it as documented diagrams.16:45
<barmintor>ajs6f1: I swear I'll get to that PR today. This FCR3 one is a bugbear.17:11
Vim's in LA, so I'll just have a whiskey in one hand while I go over your code.
<ajs6f1>barmintor: no prob. I'm still not convinced we should merge it anyway. I only PR-ed it because eddies/cbeer held my head in the toilet and gave me a swirly.17:12
Review of my code deserves all the whiskey I gave it. Let's just leave it at that.
Why do Java and XML hate each other so much...?17:14
<barmintor>http://shop.oreilly.com/product/9780596101497.do 480 pages17:15
<ajs6f1>I hope that book was produced with a Java XML framework, or it's WORTHLESS>17:16
<barmintor>ajs6f1: cover on the 1st ed is more appropriate: http://shop.oreilly.com/product/9780596000165.do17:17
<ajs6f1>Square pegs, round holes. Story of my career.
afk bbi1017:24
* ajs6f1 leaves
* ajs6f joins17:32
<ajs6f>Urg. Never mind that RNG schema. It looks like WHATWG never finsihed the job. They never defined the data types they're using.
barmintor: maybe we need a MultithreadedTestsIT?
<barmintor>ajs6f: that's the name of a test method in this crazy-ass PR
<ajs6f>Yeah, I get it.
Oh, wait. That's fcrepo3, right.
<ajs6f>^&*# it.
^&*# it in the ear.17:50
<barmintor>we said we would do it, we said we would get it in for 3 point-SEVEEEEEEEEEEN
in my head that a little tune
<ajs6f>You mean 3.LastReallyLastThisTimeWeMeanIt?17:51
<barmintor>III-POINT-VII : The Crushening!17:52
<ajs6f>I'm out of here. Enjoy the pain, barmintor. It builds character. :)17:54
* ajs6f leaves17:55
* ksclarke1 leaves17:56
* barmintor leaves18:36
* cbeer leaves19:23
* cbeer joins
* ksclarke joins20:43
* barmintor joins21:26