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

Using timezone: Eastern Standard Time
* jmignault joins07:57
* whikloj joins07:58
* whikloj leaves08:04
* dhlamb joins08:39
* acoburn joins09:05
* jgpawletko joins09:09
* acoburn leaves09:11
* acoburn joins09:12
* ajs6f joins09:18
* ksclarke joins09:21
* |R joins09:37
* dhlamb leaves09:54
* jmignault leaves10:25
* acoburn leaves10:29
* barmintor joins10:50
ksclarke terrellt ruebot are any of y’all in the fedora committers room at OR right now?10:51
<ksclarke>not at OR
* acoburn joins
<ksclarke>my feelings too :-)
<ruebot>barmintor: hola10:54
<barmintor>ruebot: is that a yes?
<ruebot>barmintor: si
barmintor: qui
<barmintor>my grand-boss is asking me to be available for something10:55
<ruebot>barmintor: aacoburn has the hangout open if you want to join us.
barmintor: he is here :-)
<barmintor>okey doke
<acoburn>barmintor: not sure how well you'll be able to hear, though….
<barmintor>acoburn: I’ll do the best I can with what I’ve got
<acoburn>barmintor: right now a. birkland is talking about this: https://docs.google.com/document/d/1UHQMXgnGoIXfe33a6CM6k3IkL3SsYqzkkWBRMaJxMew/edit10:56
<ajs6f>acoburn: I hear nothing.10:57
<acoburn>*** gets up to sit next to a. birkland ***10:58
<barmintor>is there a link to the hangout?
<ruebot>ajs6f: did you hear acoburn typing on the keyboard?
<ajs6f>ruebot: Nope. I hear nothing.
<ruebot>barmintor: https://plus.google.com/hangouts/_/event/c1glu6soq43r1rr6ou17qtobug810:59
ajs6f: acoburn is sitting closer now
<acoburn>yes, but now scossu is talking….
<ajs6f>ruebot: Yes, he is keeping me up to date on the position of his butt via Google hangout.
<barmintor>i hear nothing, but i’ll hangout11:00
<ajs6f>I'm tempted to suddenly unmute and start screaming unintelligbly out of acoburn's laptop.11:01
* ruebot joins the hangout too
<ajs6f>THERE's the sound!
<acoburn>ajs6f: thanks ;-)
<ajs6f>I think ruebot has a better mic or something.
It's definitely his machine that is bordcasting.
<ruebot>whoever said osx is better than linux is WRONG! :-P
<ajs6f>You misheard. It's _sox_ that are better than Linux. Even if you wear sandals, sox are great!11:03
* ruebot gasps
* sdidomenico joins11:05
<ajs6f>This better not involve the core, because we have barely got enough mojo to make that happen now.11:06
<|R>@ OR but in the tutorial room :)
<acoburn>ajs6f: abirkland is talking about keeping this distinct from fcrepo core11:07
<sdidomenico>Is this IRC channel where most of the OR15 backchannel conversation is happening?
<ajs6f>acoburn: Yes, and he and I have already talked about it. I was just emphasizing the point.
sdidomenico: This is the official Fedora room.11:08
<acoburn>ajs6f: great. and I would re-emphasize that
<sdidomenico>Ok great thanks
<ajs6f>acoburn: Don't emphasize it too much, or it will explode in a shower of flame.
<acoburn>ajs6f: or a flower of shame11:09
<barmintor>if we leave aside the discussion of CMA translation, is this an elaboration of microservices on the JVM against LDP?11:12
<ajs6f>barmintor: It's not clear to me that it's going against LDP. It's going against the Fedora Java types, in at least some cases.11:13
* barmintor gets steely-eyed
* ajs6f gets bleary-eyed
<barmintor>ajs6f: I figured you went into defensive shock when you heard ‘OSGi’11:14
<ajs6f>barmintor: AB and I see eye-to-eye about OSGi. I just don't think we will ever summon the ki to do it in the core. The core will always be monolithic.
barmintor: That doesn't mean that things around the core have to be.
<barmintor>right, right
<ajs6f>If that happens, my flying pig will be the main dish at our celebratory BBQ.11:17
* ruebot searches for an apple11:18
<ajs6f>Yes. OSGi is a wonderful plugin-framework, and would serve the role of a "docking station" at the outside of the core very well.11:19
Dynamic modularity and microservices.11:20
Service framework. Say it!
It also provides configuration management, dynamic service managment, log management, a lot of the sort of things that a trad JEE app server does for you. But it brings a much more dynamic flavor to it all, because oSGi comes from the world of embedded systems.11:22
You can run an OSGi framework inside a servlet container or app server. In fact, JBoss' and Oracle's big-iron app servers are built _around_ OSGi.11:24
Yes. A monolithic core (sad, but that's where we are) surrounded by a sea of pluggability.11:25
You don't need Karaf.11:26
Karaf is great, but it is not the whole market.
<barmintor>indexing views would be exposed as resources?11:32
<ajs6f>barmintor: They are now. At fcr:transform.11:33
<barmintor>you know how I feel about fcr:transform
<ajs6f>barmintor: No, I didn't know you had a beef?11:34
<barmintor>I ask b/c I thought index processing was our use case for listeners, but I guess TMTOWTDI
<ajs6f>Listeners trigger the work, but they may or may not _do_ all the processing.11:35
* jrgriffiniii joins
<barmintor>ajs6f: fcr:transform is an undiscoverable magic URI pattern, so I look askance at it
<ajs6f>barmintor: Sure, but not everything has to be discoverable for everyone. Did we ever talk about:11:36
<ajs6f>barmintor: My point there was that if your presentation depends entirely on the indexing transforms, you need to make them durable along with the stuff you are indexing.
Get 'em in the repo.
<barmintor>one day we should talk about things again, on a beach, with long chairs and iced tea
<ajs6f>"There are these hacks." —acoburn
<ruebot>a karaf of ice tea?11:38
<ajs6f>Understatement of the day.
A karaf of beer, if you want me there.
<barmintor>it depends on what time of day we get to the chairs. Maybe coffee on a porch, then beer on the beach.
<ajs6f>This sounds nice. We should rent an RV with a porch. Do they make those? They should.11:40
<barmintor>it’s like a barnstorming train, but more relaxing. Probably less bunting, but I’m persuadable.
* rtc joins11:46
<barmintor>ajs6f: so the last couple of days I’ve been forced to look at a lot of JDK8 stuff more seriously. They took a fairly hard turn towards functional syntax, didn’t they?
<ajs6f>barmintor: They didn't jump, they were pushed.
<barmintor>all of those google function libs are basically in the core there11:47
<ajs6f>barmintor: If you look at the ground behind them, you can see the footprints of Scala, Clojure, JRuby… other thugs and goons.
<ajs6f>barmintor: Brian Goetz.
barmintor: But it's all pretty basic stuff, compared to totally lazy or functionaljava.11:49
<barmintor>ajs6f: sure, just a pretty stark lurch in approach. I was surprised.
<ajs6f>barmintor: Good surprised, right?
<rtc>thanks Ben. Lafayette had questions about fedora & audio/video streaming, he may ask questions here11:50
<barmintor>ajs6f: interested surprised. I think if I was going to just start doing functional programming on the JVM, I’d use Scala or Clojure. But it’s nice to have options.11:51
* barmintor waves at rtc
<ajs6f>barmintor: Yes. We _do_ have to spend time with Java.
<rtc>waves back 8-)
<ajs6f>I like where Commons RDF is going, but we need to be cautious. I've had some discussions with Andy Seaborne about doing a Jena impl of Commons RDF, but that's a good ways away.11:52
ruebot: We lost sound in the hangout?
ruebot: Lost you entirely.11:53
<barmintor>ruebot: you dropped
got booted from the wireless
<ajs6f>Was that a question about Jena?11:55
* anarchivist joins
<ruebot>ajs6f: i believe, whether or not the jena api will meet expectations.
ajs6f: acoburn can correct me if i'm wrong.
<ajs6f>ruebot: Expectations?
ruebot/acoburn: Meaning future expectations?11:56
<acoburn>ajs6f: not so much expectations. more like our use case
<ajs6f>acoburn: Our use case of moving RDF back and forth from oersistence?
<acoburn>ajs6f: yes, and handling sparql-update11:57
<ajs6f>acoburn: It definitely will. And Commons RDF isn't going to handle serialization anytime soon, if ever.
Jena moves slow, because it has a giant user base and Andy and Co. have to be careful.
BUt we've talked about bringing in Java 8 Streams and other nice things.11:58
<acoburn>ajs6f: I was assuming that you'd need jena/sesame for that, but that it would return commons-rdf objects
<ruebot>Andy Seaborn?
<barmintor>No. No. No. No.
<ajs6f>ruebot: Right. Seaborne.
barmintor: You don't like Java 8 Streams?
<barmintor>No to Conneg for this, no, no, a thousand times no
<ajs6f>barmintor: I was kind of liking this. Why do you hate it?
<barmintor>it’s broken.
<barmintor>for the reason scossu just stated, to start with
<ajs6f>Scossu started his renark with "I don't know why I would want to do this…" do _you_ know why he would?
<barmintor>of course, and so do you
<ajs6f>barmintor: I do not, or at least, I'm not telling myself. I'm keeping it a secret from me.12:00
barmintor: Help me convince myself to admit to myself what I already know.
<barmintor>we talked about for days when Aaron was talking about the RDFSource thing
<ajs6f>barmintor: Yeah, and at the time, I don't remember you hating this idea.
barmintor: Maybe I'm just misremembering.
<barmintor>I think conneg for this is hopelessly broken, and I think the issue esme just brought up is also important12:01
<ajs6f>One of the questions at stake here: is a description of a thing a _representation of that thing_.
<ajs6f>I couldn't hear escowles. Dang.
barmintor: See, I would say yes.12:02
<ruebot>ecowles mentioned determing what a thing is from the headers
<barmintor>it is in this we see a rare occassion of our disagreement. But I think I have the weight of much LDP practice behind me.
<ajs6f>What a thing is? Meaning description or representation?
barmintor: My impression (possibly wrongly) was that the other impls do it the way we don't.12:03
<barmintor>marmotta uses a content url, iirc. Lyo uses a separate resource.
<ajs6f>barmintor: So marmotta does the "dual" of what we do? Can you hear Mark?12:04
<barmintor>but marmotta is also busted for rdf binaries
<ajs6f>Ew. conneg via file extension?
<barmintor>one of the weaknesses of all of them is not privileging the ixn model
<ajs6f>barmintor: Not sure what you mean by "privileging"?12:05
<barmintor>but if we preserve the approach of two resources, you can actually change between fcr:metadata and fcr:content without breaking anything
<ajs6f>acoburn/ruebot: Can you bring up the question of whether a description is a reprsentation? I think that crucial here?
<barmintor>ajs6f: it means if I transmit a ixn model header that says “nonrdfsource”, then don’t worry about inferring anything12:06
<ajs6f>That's the homoousian point.
<barmintor>brb googling homoousian
<ajs6f>barmintor: Ah, the header. Okay, that makes more sense. You're saying, "Do what I say, not what you _think_ I want."?
<barmintor>ajs6f: I better not get embarrassing images
ajs6f: yes, exactly12:07
<ajs6f>barmintor: How embarassing is the NIcene Creed?
<barmintor>you can accomplish single subject without resorting to this chicanery
Yes. We should do single-subject no matter what.12:08
I think the idea that description is a totally independent resource from the thing described is a very library-centric POV.
<barmintor>I don’t think that is true, and I love to blame things on library parochialism12:09
<ajs6f>barmintor: Do you see a lot of duality of this kind in other falvors of Linked Data?
<barmintor>ajs6f: see also the Eclipse/IBM impl of LDP, yes12:10
<ajs6f>barmintor: LYO?
<barmintor>ajs6f: yah
<ajs6f>barmintor: They have a "fcr:metadata" equiv?12:12
<barmintor>ajs6f: yes12:13
<ajs6f>barmintor: Hm. I wonder how many of the folks who do that (an "fcr:metadata" equiv) are also doing single-subject.
<barmintor>ajs6f: I’d have to fire up my old tests, but IIRC Lyo’s metadata LDPR has stmts about both the description and the described LDPNR12:15
<ajs6f>barmintor: Interesting. That might be part of why. (Primitive provenance)
<barmintor>ruebot: Can you say for me that we should look at how Lyo does this?12:16
<ruebot>ruebot:: how to i say Lyo? lee low?
<barmintor>b/c it has similar URI patterns and concerns
<ajs6f>Curl your tongue into a canoe shape.
<barmintor>ruebot: there is a vocative click between the y and the o
<barmintor>I am a strong advocate for appropriate subjects, and separate resources
<ajs6f>Fedora is an object repository. Not a triplestore.
<barmintor>ajs6f is an object repository12:19
<ajs6f>Well, I'm full of something, anyway.
The triples in RDF published by the repo actually represent the proerties of objects, not arbirary claims about the world.
<barmintor>It cannot be an opaque RDF source ad also be the value of a describedBy link header
<ajs6f>barmintor: By LDP spec?12:20
<barmintor>unless we want Rob Sanderson to burn us all to the ground
<ajs6f>barmintor: I'd rather be burnt by Rob than anyone else on the Annotation WG.
<barmintor>single subject is the binary does not mean collapsing the resources12:21
single subject is the binary does not mean collapsing the resources
single subject is the binary does not mean collapsing the resources
single subject is the binary does not mean collapsing the resources
single subject is the binary does not mean collapsing the resources
single subject is the binary does not mean collapsing the resources
<ajs6f>What barmintor said, but only the first three times. After that, I disagree.12:22
<barmintor>I retract the last 3
fcr:metadata is a shitty way to model provenance.12:23
That's a second-order task.
<barmintor>that is probably true. really best for technical metadata about the binary
<ajs6f>Right. KISS. _Properties_.
Was that escowles? Tell him what barmintor said above. The description cannot be a binary.12:24
It has to be transparent in the repo, an RDF thing.12:25
<barmintor>ajs6f: escowles said “contains a binary” which is different
<ajs6f>barmintor: Like a literal?
Or like it's a LDP Container?
<barmintor>ajs6f: I think he is interested in the LDPNR being also a container
<ajs6f>barmintor: Oh, cool.
Oh, wait, not the LDPNR!12:26
<barmintor>ajs6f: like a DirectContainer whose proxy subject is the LDPNR
<ajs6f>You already explained why that doesn't work in that ticket.
barmintor: Now I'm totally confused.
And hungry.
<barmintor>well, now we’re down in the weeds
<ajs6f>I can't eat weeds.
<barmintor>the describedBy resource is a LDPR, which Liskov says *could* be a container12:27
<ajs6f>barmintor: Yeah, that's what I said "Oh, cool." about.
barmintor: There might be other problems, but it's not inherently wrong.
<ruebot>ajs6f,barmintor: do y'all want me to keep the hangout open in the afternoon workshop?
<ajs6f>ruebot: Is that a "hands-on" training?12:28
<barmintor>ajs6f: right, so that opens the possibility of it containing on behalf of the LDPNR it describes
<ajs6f>barmintor: via proxy action?
<ruebot>ajs6f: yeah, awoods, acoburn, and mikeatuva doing to stuff
* sdidomenico leaves
<ajs6f>ruebot: No, I'll do actual work. But do me a favor: midway through mikeAtUVa's talk, leap up and run up to him and give him a flying slow motion high five.12:29
ruebot: It's what I would do.
* ruebot looks for his cape
<ajs6f>Way back then, we were trying hard to get an MVP out. Thats not where we are any more.12:32
* cmmills joins12:33
<barmintor>also worth noting that MODE didn’t have a REST API, so our attempts to expose TX through it were fighting the realities of the runtime
<ajs6f>Fedora is too much like a trash can, into which you put things without them ever coming out. We need to make it more like a trash can with a hole in the bottom.
barmintor: MODE did and does have a REST API, don't you remember?12:34
barmintor: BUt I don't think it offered tx.
<barmintor>ajs6f: well, ok
<ajs6f>barmintor: That API, tho, is all about JCR types and ideas.12:35
<barmintor>web based TX is tough, and there’s not a lot of work out there
barmintor: Agreed.
barmintor: Roy Fielding has made remarks to the effect of "If you find yourself reaching for tx in REST, you should question the larger architecture."12:36
<barmintor>how am i not myself?12:37
<ajs6f>I got to leave the hangout, got a phone call elsewhere.12:38
* rtc leaves12:39
* cmmills leaves
<barmintor>ruebot: tell awoods to merge that PR so I can do the release. Or to do the release.12:40
* acoburn leaves12:41
* jrgriffiniii leaves12:42
<barmintor>ajs6f: don’t know if you also hit this issue that awoods and I ran into, but JAXB 2.2.11 produces javadoc that will actually pass the linter validation12:44
you have to install it like apache xerces in the olden days where you run javadoc, though12:45
* barmintor gets a lunch12:46
<ajs6f>barmiintor: G-d be thanked, I have no experience of what you're talking about. (JAXB probs, that is, not lunch. I have too much experience of lunch, which is why I have a big belly.)12:51
* ajs6f out for lunch12:55
* anarchivist leaves13:11
* anarchivist joins13:12
* whikloj joins13:31
* sdidomenico joins
<ruebot>ping me if you want me to fire up the google hangout again. i'm up front close to presenters.13:36
<ajs6f>ruebot: Is it just straight up training?13:37
* rtc joins
<ruebot>ajs6f: yeah. awoods is doing hangs on for about 1.5hrs, then acoburn is doing camel, then mikeatuva is doing migration-utils.
s/hangs/hands on/13:38
<ajs6f>ruebot: It's cool then. I'm eating really good Canadian kippers with rye bread. I don't want to be distracted. Thanks, tho', and thanks for yeoman work this morning keeping barmintor and me in the loop. I hate to miss a chance to publicly disagree with barmintor.13:39
* jrgriffiniii joins
<barmintor>ajs6f: even semi-publicly!
<ajs6f>barmintor: Any which way you can.13:40
<barmintor>Every which way but loose.
<ajs6f>Which one of us is the monkey? I'd rather be the monkey than Clint Eastwood.
<barmintor>wasn’t it an orangutan?13:41
<ruebot>is it not obvious i am the monkey/gorilla?
<ajs6f>Oh, no. Not only was it indeed an orangutan, but its on Youtube. There goes my productivity.
<whikloj>I think the chimpanzee was BJ & The Bear
<ajs6f>ruebot: I know you have long hair, but that's ridiculous.
whikloj: Ooh, doesn't that one have CB lingo in it?! I _love_ CB lingo!
<whikloj>ajs6f: yes, yes it does13:44
<ajs6f>whikloj: Oh, I'm going straight down that rabbit hole. Youtube, take me away!
* dhlamb joins13:52
<ruebot>dhlamb: you in toronto now?13:53
* whikloj leaves13:58
* acoburn joins14:07
* jrgriffiniii leaves14:09
* dhlamb leaves14:10
* jonroby joins14:14
* jonroby leaves14:16
* whikloj joins
* dhlamb joins14:23
* dhlamb leaves14:28
* acoburn leaves14:53
* barmintor leaves14:58
* whikloj leaves14:59
* ksclarke leaves15:04
* ksclarke joins15:05
* sdidomenico leaves15:11
* sdidomenico joins15:36
* rtc leaves15:41
* whikloj joins15:43
* whikloj leaves16:17
* rtc joins16:26
* whikloj joins16:27
* rtc leaves16:31
* whikloj leaves16:39
* rtc joins16:46
* rtc leaves16:49
* rtc joins16:57
* ajs6f leaves17:00
* sdidomenico leaves17:05
* jrgriffiniii joins17:12
* dhlamb joins17:14
* whikloj joins17:21
* jrgriffiniii leaves17:24
* whikloj leaves17:38
* rtc leaves17:51
* dhlamb leaves
* jrgriffiniii joins17:52
* whikloj joins18:21
* whikloj leaves
* ksclarke leaves18:31
* whikloj joins19:21
* whikloj leaves19:22
* whikloj joins20:22
* whikloj leaves
* ksclarke joins20:27
* ksclarke leaves20:55
* ksclarke joins20:56
* ksclarke leaves21:00
* ksclarke joins21:15
* whikloj joins21:23
* whikloj leaves
* dhlamb joins21:39
* dhlamb leaves22:00
* whikloj joins22:16
* whikloj leaves
* whikloj joins22:17
* whikloj leaves22:40
* dhlamb joins22:57
* dhlamb leaves23:05
* anarchivist leaves23:06
* jgpawletko leaves23:08
* rtc joins23:21
* dhlamb joins23:45
* dhlamb leaves00:26
* dhlamb joins00:28
* dhlamb leaves00:33
* dhlamb joins00:59
* rtc leaves
* dhlamb leaves01:05
* dhlamb joins02:01
* dhlamb leaves02:06
* anarchivist joins02:15
* jrgriffiniii leaves02:42
* dhlamb joins03:02
* dhlamb leaves03:07
* dhlamb joins04:03
* dhlamb leaves04:08
* dhlamb joins04:26
* dhlamb leaves04:30

Generated by Sualtam