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

Using timezone: Eastern Standard Time
* dhlamb_ leaves00:35
* jjtuttle_ leaves07:14
* jjtuttle joins07:19
* coblej joins08:17
* kestlund joins08:53
* coblej leaves08:54
* coblej joins08:56
* kestlund1 joins08:58
* kestlund leaves09:00
* dwilcox joins09:05
* dhlamb_ joins09:10
* ajs6f joins09:13
* coblej leaves09:14
* coblej joins09:15
* whikloj joins09:16
<ajs6f>barmintor: ping09:17
* coblej leaves09:26
* coblej joins09:27
* bseeger joins09:28
* peichman joins09:29
* coblej leaves09:35
* coblej joins09:40
* coblej leaves09:50
* coblej joins09:51
* kestlund1 leaves09:52
<barmintor>ajs6f: slow pong09:54
* coblej leaves09:55
<ajs6f>barmintor: Just checking my assumptions for our github conv. Are you coming from the POV that in order to include something as normative, it MUST be in the tck?
(cuz I'm not)09:56
<barmintor>more or less: if it's not testable, I don't know what the point of saying it's normative is
* apb18 joins
* coblej joins09:58
<ajs6f>barmintor: Um, because it allows people to set expectations for integration and design? The whole point of this exercise is to ensure that code is not what defines Fedora.10:05
* kestlund joins
<ajs6f>Do you intend to test all LDP and HTTP behaviors too?
<barmintor>LDP has a test suite; AFAIK the normative behaviors are exercised10:06
I don't think a client test compromises the exercise, quite the opposite
<ajs6f>barmintor: By no means, and I'm all in favor of tests when they can reasonably be written. And when they can't, that's unfortunate, but it shouldn't warp the process.10:07
Restricting everything we do to a client test is what compromises the exercise.
<barmintor>I don't think saying that something is informative is warping the process? To me normative means it is a structured, demonstrable behavior10:08
<ajs6f>Well, that's maybe the root of our disagreement. That's not what "normative" mean to me, it means "behavior on which I can rely in practice".10:10
"Structured" has nothing to do with it.10:11
<barmintor>I don't know how to describe that reliability if it can't be tested
<ajs6f>You don't have to.
That's not a job for a spec like this
<barmintor>Then you can't rely on it?
<ajs6f>Of course you can.
<barmintor>No, it's a job for the TCK
<ajs6f>Can you describe how to walk?
No, the TCK's job is provide assurance, not acceptance.
<barmintor>I don't have to: The spec just says get from A to B
<ajs6f>EXACTLY.10:12
<barmintor>But I can say: NORMATIVE: You get to B
<ajs6f>Sure, and we can do that too.
And we should.
<barmintor>And I can look at B, & you're there
<ajs6f>No, that's where you've moved out of the spec.
<barmintor>and if I can't describe how I know, it's not normative
<ajs6f>Nope.
* barmintor shrugs
ok
<ajs6f>Look, this is pretty abstruse.
The concrete issue is much plainer: the fixity linked resource thing.10:13
<barmintor>TBQH this is pretty not worth fighting about to me
feel free to PR normative/informative however you like
<ajs6f>The fixity linked resource thing, is def worth fighting about to me.
Okay, we can leave it at that.
<barmintor>ok
As a participant in that convo, I think the fixity link thing is important, I'm sold10:14
As an implementor, I don't know what it means and will not be doing it
and I don't know what it will look like in the TCK, so :shrugs:
there10:15
's an opportunity for more spec there
<ajs6f>Well, what that will mean for the community is a separate question (meaning is there an imprimateur available for impls, and if so how is it awarded, etc.).
That's all for a good while from now.
We hav a lot of much simpler work to do.
(E.g. I think our versioning ideas are really good, and even better HvS thinks they're okay, but we need to prove them out.10:16
<barmintor>Eventually, when we're talking about finalizing, we will want to point to some implementation amongst ourselves as evidence that the requirement is reasonable
Versioning as the draft says I can point to an impl of10:17
<ajs6f>Where?
Where does the draft say and where are you pointing?
<barmintor>I demo'd a memento-versioning impl on cavendish at HyConn last October10:18
there
's even a LDP-style test suite
<ajs6f>According to the spec? Really?
Which I'm pretty sure hadn't been fully written then?
(And even today is not finalized?)10:19
Not quite what I was thining.
thinking
<barmintor>https://github.com/cavendish-ldp/cavendish/blob/master/cavendish-jetty/src/test/java/cavendish/jetty/VersioningTestCase.java
<ajs6f>I have no idea what that is supposed to prove to me. We just got done arguing out that I don't agree with you about this very issue.10:20
Let's come back to this when we have an actual spec to impl?
<barmintor>I know10:22
I'm just saying
as an implementor that's what will get done
sure
* bseeger leaves10:26
<ajs6f>barmintor: I just glanced through the Cavendish code-- am I missing something, or how are you dealing with binaries?10:33
* coblej leaves10:47
* coblej joins10:48
* kestlund leaves10:50
* coblej_ joins
* coblej leaves10:51
* coblej_ leaves10:54
* bseeger joins10:56
* coblej_ joins10:59
* coblej_ leaves11:03
* coblej joins11:06
* kestlund joins11:13
* coblej_ joins11:29
* coblej leaves
* coblej_ leaves11:49
* coblej joins
* coblej leaves11:53
* coblej joins11:54
* kestlund leaves11:57
* coblej leaves12:08
* kestlund joins12:09
* apb18 leaves12:10
* coblej joins12:12
* kestlund leaves12:21
* bseeger leaves12:30
* bseeger joins12:40
* apb18 joins12:44
* th5 joins12:54
* kestlund joins13:05
* kestlund leaves
* kestlund joins
* kestlund leaves13:18
<barmintor>ajs6f: sorry, just getting back13:27
ajs6f: I haven't needed binaries on Cavendish yet, but I think I'm about to incorporate some of Trey P's work on it in support of some local proposals13:28
<ajs6f>barmintor: Cool, makes sense.13:33
* apb18 leaves14:15
<ajs6f>awoods: ping14:23
<awoods>ajs6f: I will free up @4:30... what's up?
<ajs6f>awoods: Now that work on the spec has moved to github, we should freeze the google docs and put notes on them redirecting peoples interest14:24
<awoods>ajs6f: good idea. I'll do it if you don't beat me to it.
<ajs6f>I think we probably have some wiki gardening to do, too
awoods: I think only you can deal with the google docs, but I will look at the wiki14:25
<awoods>ajs6f: sounds like a plan
* coblej_ joins14:45
* coblej leaves
* kestlund joins14:55
<ajs6f>awoods: done14:59
* kestlund leaves15:05
* kestlund joins15:09
<awoods>ajs6f: docs updated15:33
<ajs6f>awoods++15:38
* coblej_ leaves16:11
* th5 leaves16:17
* coblej joins
* coblej leaves16:21
* dwilcox leaves16:26
* kestlund leaves16:43
* ajs6f leaves16:47
* kestlund joins16:50
* coblej_ joins16:51
* coblej_ leaves16:55
* bseeger leaves16:58
* kestlund leaves
* coblej joins17:02
* whikloj leaves17:03
* dwilcox joins
* coblej leaves17:22
* peichman leaves17:33
* github-ff joins17:51
[fcrepo-import-export] escowles created external-content (+1 new commit): https://git.io/vMPmf
fcrepo-import-export/external-content 964402f Esmé Cowles: Adding option to retrieve and export external content
* github-ff leaves
* github-ff joins17:52
[fcrepo-import-export] escowles opened pull request #74: Adding option to retrieve and export external content (master...external-content) https://git.io/vMPmY
* github-ff leaves
* dhlamb_ leaves18:17
* dwilcox leaves22:02
* awoods leaves22:43

Generated by Sualtam