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

Using timezone: Eastern Standard Time
* bseeger joins08:55
* acoburn joins09:03
<awoods>ruebot_away: ping09:10
* whikloj joins09:13
* peichman joins09:15
* dwilcox joins09:28
* dwilcox_ joins09:33
* dwilcox leaves09:34
<ruebot>awoods: pong09:41
<awoods>ruebot: hello09:42
<ruebot>awoods: hola
<awoods>ruebot: is there another PR in the works for PCDM?
ruebot: post-revert09:43
<ruebot>awoods: i ping you about it yesterday... i can't make a new one, because current one looks like it didn't actually revert.
<awoods>ruebot: https://github.com/duraspace/pcdm/commits/master
ruebot: I see a double reversion09:44
<ruebot>awoods: https://github.com/duraspace/pcdm/blob/master/models.rdf#L19-L20
awoods: and AdministrativeSet isn't there.
awoods: i fully admit i am really confused :-)09:45
<awoods>ruebot: I see things are wacky. Shall I just fix it?09:46
ruebot: I can make a PR... and you can click the button.
<ruebot>awoods: yes, please. i'm afraid i'll mess it up again.09:47
<awoods>ruebot: I am in the middle of something at the moment. Will you be around?
<ruebot>awoods: i'm around all day.
<acoburn>awoods: ping09:54
<acoburn>awoods: question about two (actually four) intertwined tickets
awoods: fcrepo-1983, fcrepo-1498, fcrepo-1742 and fcrepo-175409:56
awoods: in particular fcrepo-1983
<awoods>acoburn: ok
<acoburn>awoods: the last PR #1028 makes ETags work for binary descriptions, but fails for binaries
awoods: that behavior is fixed with fcrepo-174209:57
awoods: I've been trying to keep these PRs small and focused
awoods: but this is suggesting wrapping them all up together
awoods: what do you suggest?
<awoods>acoburn: can you keep them as separate commits in a single PR?09:58
<acoburn>awoods: yes, certainly
<awoods>acoburn: or does that not make sense?
<acoburn>awoods: I'm not sure it makes complete sense though
awoods: they are all related to ETags and last-mod date
awoods: they're just different manifestations of the central issue09:59
<awoods>acoburn: The idea is to keep the commits manageable for initial review and future forensics.
<acoburn>awoods: I completely agree
<awoods>acoburn: beyond that, feel free to do what makes sense for your own sanity.
<acoburn>awoods: the biggest change is with fcrepo-174210:02
awoods: I had been hoping to split off fcrepo-1983 into its own PR (there's not much change there)10:03
awoods: that way, fcrepo-1742 stays small, but they're so entwined I'm thinking I may need to combine them10:04
awoods: alternately, if we move fcrepo-1742 through first, and then deal with fcrepo-1983 afterwards, that might work
<awoods>acoburn: Thanks for raising the issue. It sounds like you have the right sensibilities and goal...10:05
acoburn: I am ready to work on reviews if you need something to go through quickly.
<acoburn>awoods: no, nothing needs quick attention
awoods: I think the best approach will be to do fcrepo-1742 first10:06
awoods: and for that I'd like to wait until after the 4.5.1 release
<awoods>acoburn: to minimize rebasing?
<acoburn>awoods: yes, exactly
awoods: it touches quite a few files, as I recall10:07
<awoods>acoburn: the plan is to get 4.5.1 out Fri/weekend
<acoburn>awoods: I can definitely wait until next week
awoods: as for #1028, should I mark that as "DO NOT MERGE" or something? Or even just close the PR?
<awoods>acoburn: Unfortunately, I am out next week and the following :(10:08
acoburn: I have not looked at #1028
* dwilcox_ leaves
<acoburn>awoods: oh right
awoods: #1028 is just a reincarnation of #1024 with the addition of integration tests
awoods: and targeting the correct branch (DEV)10:09
<awoods>acoburn: It appears to have a lot of tests... which is great.
acoburn: If you are planning on a post-4.5.1 PR, then closing #1028 is a natural.10:11
<acoburn>awoods: ok, I'll do that
* github-ff joins10:12
[fcrepo4] acoburn closed pull request #1028: Fix ETag handling for Binaries and Binary descriptions (DEV...fcrepo-1983.2) https://git.io/vwzD3
* github-ff leaves
<acoburn>awoods: just wait until I issue the PR removing FedoraResource::getNode10:13
awoods: it's a very small change that touches lots and lots of files
<awoods>acoburn: keep'm coming
<acoburn>awoods: likewise, post 4.5.1
awoods: one more step away from JCR
<awoods>acoburn: as a note, the Mode5 work has been pretty smooth
<acoburn>awoods: in the kernel-api at least
awoods: that's great to hear
<awoods>acoburn: the only true change that I have seen is that Federation does not appear to update when files are added to the federated directory... it used to.10:14
* acoburn leaves10:23
<awoods>ruebot: if the goal of the PCDM PR is to: 1) remove adminsets, 2) keep "priorVersion" unchanged: http://pcdm.org/2015/09/28/models , 3) update "versionInfo": 2016/04/18 ...10:26
ruebot: then I think we are good as of now.10:27
ruebot: https://github.com/duraspace/pcdm/blob/master/models.rdf10:28
<ruebot>awoods: yeah, that's what i thought10:30
* peichman1 joins
<ruebot>awoods: which means we can undo https://github.com/duraspace/pcdm/commit/8c3ed39669ac71adf38a31361a6293493be478c2 if you want, since that looks like an empty commit
<awoods>ruebot: we don't want to force-push on master...10:32
* peichman leaves
<awoods>ruebot: and undoing that commit will mean another empty commit :(
ruebot: I would vote for calling it a success and moving on.10:33
<ruebot>awoods: oh, i meant doing a hard reset on that commit :-) kill it with fire.
awoods: but, it is good to own our mistakes too, so that can be in there. just those two commits in a row might be confusing if somebody is looking at the history.10:34
<awoods>ruebot: A hard reset may mess up people's local work, no?
<ruebot>awoods: maybe? we could create an issue to get concensus. but, that might be asking for trouble :-)10:37
<awoods>ruebot: with a open project... it is hard to know who in the world would be affected.10:38
ruebot: I feel like a little messiness is better than community disruption.10:39
ruebot: master is at a good state.
awoods: it is decided!
<awoods>ruebot: I'm happy if master is happy
* arebenji joins10:59
* th5 joins11:19
* peichman1 leaves11:51
* peichman joins11:54
* bseeger leaves12:03
* peichman leaves12:56
* peichman joins13:01
* bseeger joins13:08
<ruebot>awoods: so, 2pm next week, not 3pm?13:20
<awoods>ruebot: yes, Rob was not available @313:21
<ruebot>awoods: ah, ok. well you got jared, and maaaaaaaybe me. but most likely not.
<awoods>ruebot: I saw whikloj was available... no time was good for everyone :(13:22
<ruebot>awoods: right-o!13:28
* arebenji leaves13:48
* arebenji joins13:50
* acoburn joins13:52
* acoburn1 joins14:51
* acoburn leaves14:53
* bseeger leaves14:56
* bseeger joins14:57
* bseeger leaves
* bseeger joins15:03
* bseeger leaves15:10
* dwilcox joins16:53
* whikloj leaves17:01
* dwilcox leaves17:03
* th5 leaves17:20
* acoburn1 leaves17:28
* peichman leaves
* jcoyne joins17:52
* jcoyne leaves18:05
* arebenji leaves20:34
* jcoyne joins21:27
* jcoyne leaves22:32
* jcoyne joins22:34
* jcoyne leaves23:24