#duraspace IRC Log


IRC Log for 2016-10-26

Timestamps are in GMT/BST.

[6:37] -rajaniemi.freenode.net- *** Looking up your hostname...
[6:37] -rajaniemi.freenode.net- *** Checking Ident
[6:37] -rajaniemi.freenode.net- *** Found your hostname
[6:38] -rajaniemi.freenode.net- *** No Ident response
[6:38] * DuraLogBot (~PircBot@webster.duraspace.org) has joined #duraspace
[6:38] * Topic is 'Welcome to DuraSpace IRC. This channel is used for formal meetings and is logged - http://irclogs.duraspace.org/'
[6:38] * Set by tdonohue on Thu Sep 15 17:49:38 UTC 2016
[12:15] * mhwood (mwood@mhw.ulib.iupui.edu) has joined #duraspace
[12:33] * tdonohue (~tdonohue@dspace/tdonohue) has joined #duraspace
[14:51] * dyelar (~dyelar@ has joined #duraspace
[17:10] * pbecker (~pbecker@ubwstmapc098.ub.tu-berlin.de) Quit (Quit: Leaving)
[17:57] * hpottinger (~hpottinge@ has joined #duraspace
[19:49] * peterdietz (uid52203@gateway/web/irccloud.com/x-mbywcpktqsxbnhzi) has joined #duraspace
[20:00] <tdonohue> All, it's time for our weekly DSpace Developers Meeting. Here's today's (late) agenda: https://wiki.duraspace.org/display/DSPACE/DevMtg+2016-10-26
[20:00] <kompewter> [ DevMtg 2016-10-26 - DSpace - DuraSpace Wiki ] - https://wiki.duraspace.org/display/DSPACE/DevMtg+2016-10-26
[20:01] <mhwood> pbecker did a lot of 1d.
[20:01] <tdonohue> mhwood: I still see a lot of PRs flagged for "6.0". I'm pretty sure pbecker did all the *JIRA* tickets though (moved those to 6.1)
[20:01] <mhwood> No, wait, that was Jira, perhaps.
[20:02] <tdonohue> In any case, getting started here.. first off.. YAY! 6.0 IS RELEASED!
[20:02] <tdonohue> Thanks everyone for your hard work on this release, and for pbecker for helping with the final push / cutting the release
[20:02] <mhwood> A hollow voice says, "whew!"
[20:04] <tdonohue> A "fun fact" about this release, according to the list of 6.0 contributors in the Release Notes, this release had our *most contributors ever*! We ended up with 73 code contributors in 6.0!
[20:04] <mhwood> Interesting.
[20:04] <tdonohue> For comparison, for other major releases (3.0, 4.0, 5.0), I think we maxed out in the mid-50s range
[20:05] <tdonohue> So, we have a lot of folks contributing, and that's great (the big problem is managing all those contributions & the flood of PRs, as we know)
[20:06] <mhwood> :-)
[20:06] <tdonohue> As noted in the agenda, we do have a few last "cleanup" tasks for post-6.0. These all need volunteers:
[20:06] <tdonohue> 1) Someone to update "master" to 7.0-SNAPSHOT (pbecker ran into issues/errors when he tried it, not sure what happened)
[20:07] <tdonohue> 2) Someone to update demo.dspace.org to 6.0 Final (it's still running RC4 or 5)
[20:07] <tdonohue> 3) Someone to quickly/easily reschedule all open GitHub PRs flagged for 6.0 to be 6.1
[20:07] <tdonohue> I believe that's it, unless I'm forgetting anything :)
[20:09] <tdonohue> 2 & 3 should be pretty painless. #1 hypothetically also should be painless, but pbecker said he ran into build issues after he tried doing it
[20:09] <tdonohue> anyone interested in claiming one of those final tasks?
[20:09] <mhwood> I'll take (1).
[20:10] <tdonohue> thanks mhwood!
[20:12] <tdonohue> a lot of silence in here..anyone else around? Or is it just mhwood & I here right now
[20:12] <mhwood> hpottinger and peterdietz were here a moment ago.
[20:12] <hpottinger> I'm lurking, playing chicken with the other lurkers, it seems
[20:13] <tdonohue> I'm also interested in hearing any (early) feedback others may have around 6.0 (if anyone has started their upgrades, etc)
[20:13] <peterdietz> thanks for ping
[20:14] <tdonohue> So, from the list above of final post-6.0 tasks, I guess if no one gets to #2 or 3, maybe I can find a 1/2 hour next week. But updating demo.dspace.org and/or rescheduling GitHub PRs should be pretty easy for anyone interested in claiming that task
[20:14] <hpottinger> ok, fine, sign me up for 2
[20:15] <tdonohue> Thanks hpottinger! demo.dspace.org is already running RC4 or RC5, so it should just be a matter of pulling down latest code and doing a quick rebuild
[20:15] <mhwood> There was no 5, and only the artifactId should have changed.
[20:15] <mhwood> So yes, should be easy.
[20:16] <hpottinger> "easy" the time-consuming part for demo is changing the text
[20:16] <hpottinger> but, I'm not complaining, I'll do it :-)
[20:18] <tdonohue> thanks again.
[20:18] <tdonohue> The moving of the GitHub PRs is tiny...if pbecker doesn't get to it, I'll do it in spare time somewhere
[20:18] <tdonohue> So, anyone with early feedback on 6.0 to share here?
[20:20] <mhwood> Not here.
[20:20] <tdonohue> I'll take that as a "no" for now. Well, feel free to bring it to future meetings (or on mailing lists)
[20:21] <tdonohue> So, beyond 6.0, in our agenda, I just noted that we should keep a 6.1 in mind (we still have a lot of open PRs which may be ripe for a 6.1, once they get testing/verification)
[20:22] <tdonohue> Though, unfortunately, many of those PRs may need to be "ported" to the new dspace-6x branch
[20:23] <tdonohue> The other major thing to note (which all may be aware of) is the DSpace 7 / Angular2 UI group is "re-organizing". First small meeting was yesterday. For non-Committers, we hope to open up to anyone interested in contributing in the very near future
[20:24] <tdonohue> (the only thing holding us back is first getting ourselves organized enough to accept contributions & add in others, etc)
[20:25] <tdonohue> If you want to be looped in immediately, get in touch with Art Lowel (artlowel on GitHub, art.lowel at gmail.com) of Atmire, as he's leading the initiative in the coming months. I'll still be in all meetings, but as many are aware, my time is a bit limited until the end of 2016
[20:26] <peterdietz> Once I get passed some projects, I'd might be able to work on some REST work related to new UI
[20:27] <tdonohue> That said, this brings to mind another question to ask everyone. What do we feel is the best usage of *this time* (weekly IRC meetings) in coming months, since we will have separate Angular2 UI meetings?
[20:27] <tdonohue> peterdietz: sounds great, we could use your help. Andrea Bollini has offered to help lead REST API work, but he's already asked for collaborators
[20:28] <tdonohue> Regarding *this* IRC meeting, obviously there are weekly/daily tasks to keep up with... 6.x releases, JIRA tickets, PRs. So, there's plenty we can discuss week-to-week.
[20:29] <mhwood> Next week someone is going to join in to talk about XOAI.
[20:29] <tdonohue> but, I am curious if others want to restructure our time here in any way, now that 6.0 is out the door, and that 7.0 work will start shortly
[20:29] <terry-b3> Just joining in. I think looking at 6.1 stuff makes sense for this group.
[20:30] <tdonohue> mhwood: thanks for the note on XOAI. I saw a brief mention of it, and yes that'd be a good topic to discuss too
[20:31] <terry-b3> Would it be possible to combine the dev meeting and the backlog meeting into one meeting since there will be alot of 7.0 stuff in paralell?
[20:32] <tdonohue> terry-b3: yes, for the time being, it may make sense. Especially since, some weeks, we may not have a lot to talk about in these meetings and therefore it'd be good to jump into backlog tickets/PRs
[20:33] <hpottinger> keep the venue of #duraspace for the combo backlog & dev meeting?
[20:34] <tdonohue> yes, I'd recommend we stick with #duraspace. It's important for the meeting to be logged (for those who missed it). Not as important for backlog discussions, but it doesn't hurt ;)
[20:35] <mhwood> There was talk of regular, frequent communication about the status of the New UI. That would perhaps be a repeating item for these meetings.
[20:36] <tdonohue> yes, perhaps... though, I actually think that "regular, frequent communication about the status of the New UI" should be via mailing lists & completely public. The goal is not just to loop in Committers, but also all our users on what is going on
[20:36] <mhwood> OK
[20:37] <tdonohue> So, yes, I think it's OK to have brief updates on New UI in these IRC meetings... I think the goal is to make those updates completely public. Questions/brief discussion on those updates are welcome here though
[20:38] <tdonohue> Typing too fast...meant to say this as second sentence "*But*, I think the eventual goal is to make those updates completely public"
[20:39] <tdonohue> So, any other hot topics that folks wish to bring up today? I honestly don't have anything else on the agenda for today, though we could do some JIRA reviews or similar
[20:39] <terry-b3> I hope those meetings do not move to IRC. It was refreshing to meet with audio and video. (This is my continual complaint)
[20:40] <mhwood> Aw. Typing is so much easier.
[20:40] <terry-b3> And slower
[20:41] <hpottinger> +1 typing
[20:41] <mhwood> At least this time the various Google plugins didn't make Firefox unstable.
[20:42] <hpottinger> just don't run top and hangouts at the same time :-)
[20:42] <tdonohue> terry-b3: during the prototyping of the Angular2UI, we kept all meetings as video/audio. I'd anticipate we'd do the same with the DSpace 7 work, as it's much easier to discuss via (minimally) audio, and there are occasions where you need/want to screenshare recent work
[20:43] <tdonohue> The screensharing became very important to use in the prototyping to help us resolve issues we were seeing or show off (quickly) progress on a particular feature
[20:43] <terry-b3> Good to hear.
[20:46] <tdonohue> so, not hearing any other topics. We are running a bit short on time for a full fledged JIRA review...but, I recall seeing something about a 5.6 issue from peterdietz? Should we talk about that here today, briefly?
[20:47] <peterdietz> sure. we upgraded some sites to 5.6, and the external handle resolver we use stopped working. one example of 5.6, and handleresolver not working is https://darchive.mblwhoilibrary.org/handleresolver/listprefixes
[20:47] <peterdietz> (they use old fashioned handleresolver so its fine)
[20:48] <peterdietz> you should get data returned, like these guys: http://escholarshare.drake.edu/handleresolver/listprefixes
[20:48] <peterdietz> I've made the tiny patch
[20:48] <peterdietz> The issue is DS-3366
[20:48] <kompewter> [ https://jira.duraspace.org/browse/DS-3366 ] - [DS-3366] /handleresolver provides no data - DuraSpace JIRA
[20:49] <tdonohue> peterdietz: do you have the "tiny patch" somewhere?
[20:49] <peterdietz> and demo.dspace.org running 6x is fine. so bug didn't exist in 5.5, but exists in 5.6
[20:49] <peterdietz> I've fixed longsight's branch http://escholarshare.drake.edu/handleresolver/listprefixes
[20:49] <peterdietz> oops
[20:50] <peterdietz> https://github.com/LongsightGroup/DSpace/commit/34432c64b3291ac69d5d3129ffa985b06f7769db
[20:50] <kompewter> [ DS-3366 Remove out.close from handleresolverreader · LongsightGroup/DSpace@34432c6 · GitHub ] - https://github.com/LongsightGroup/DSpace/commit/34432c64b3291ac69d5d3129ffa985b06f7769db
[20:50] <hpottinger> I'm happy to test that patch, do you have a PR for us?
[20:50] <tdonohue> yes, sounds like the next step is a PR. wow, that is a tiny patch
[20:50] <mhwood> It makes sense, though.
[20:51] <peterdietz> i never debugged to see WHY is was crashing. But mhwood and I were speculating, that of all the things changed in Andrea's refactoring (the only commit to touch that file between 5.5 and 5.6) the out.close is the most likely to affect things
[20:53] <tdonohue> ok, so sounds like we need a PR for more testing. We might want to see why this works on 6.x too (does it not have that close()? Or is something else happening here we aren't seeing?)
[20:53] <tdonohue> Glad that, at the very least, there's a "quick fix". But, it would be good to double check why 6.0 works but 5.6 doesn't
[20:54] <tdonohue> hpottinger: 6.0 *is* on demo (saw your comment on DS-3366). It's just running 6.0RC4 instead of 6.0 final
[20:54] <kompewter> [ https://jira.duraspace.org/browse/DS-3366 ] - [DS-3366] /handleresolver provides no data - DuraSpace JIRA
[20:54] <hpottinger> https://github.com/DSpace/DSpace/blob/master/dspace-xmlui/src/main/java/org/dspace/app/xmlui/aspect/handle/HandleResolverReader.java#L153-L154
[20:54] <kompewter> [ DSpace/HandleResolverReader.java at master · DSpace/DSpace · GitHub ] - https://github.com/DSpace/DSpace/blob/master/dspace-xmlui/src/main/java/org/dspace/app/xmlui/aspect/handle/HandleResolverReader.java#L153-L154
[20:55] * peterdietz is learning that he shouldn't try to "do it live" by trying to create a PR from github UI.. (I've accidentally made a branch on DSpace/DSpace instead of peterdietz/DSpace
[20:56] <hpottinger> it's in a try/cactch block, so... out.close throws an exception... what exception did you see, peterdietz?
[20:56] <peterdietz> I got NoException
[20:56] <hpottinger> o.O
[20:57] <mhwood> I hate it when code throws NoException.
[20:57] <hpottinger> would you happen to have the stack trace you could paste into that ticket?
[20:57] <hpottinger> or, I could try to reproduce...
[20:58] <peterdietz> it doesn't produce errors or exceptions, it just dies quietly. Swallowing the error, in 5.6 atleast
[20:58] <tdonohue> So, it does sound to me like we should... (1) Document this tiny patch as a "quick fix", (2) do some further investigation as to whether it is the *right* fix...especially looking at why 6.0 seems to work
[20:59] <mhwood> My guess is that AbstractReader doesn't expect 'out' to be closed when generate() returns, but isn't able to tell us about it.
[21:00] <tdonohue> it's a good guess, but then the obvious question is why isn't 6.0 acting the same way?
[21:01] <peterdietz> okay. PR created https://github.com/DSpace/DSpace/pull/1560
[21:01] <kompewter> [ DS-3366 Fix handleresolver by removing out.close by peterdietz · Pull Request #1560 · DSpace/DSpace · GitHub ] - https://github.com/DSpace/DSpace/pull/1560
[21:02] <tdonohue> ok, in any case, we have a plan here, and an initial PR for folks to play with. Seems good enough for today
[21:02] <peterdietz> hmm. yeah, weird that 6.0 has the same code. the out.close, and its working. and 5.6 has out.close, and not working
[21:02] <tdonohue> We're also at the top of the hour, so I'm going to close up the official meeting for today. Discussion is welcome to continue though, either here or in #dspace. :)
[21:03] <mhwood> Gotta go, thanks all.
[21:03] <peterdietz> I wonder how many people would be using handleresolver anyways. Gotta create a dspace-spymode, that phones home to our feature monitoring and data gathering dashboard
[21:03] * mhwood (mwood@mhw.ulib.iupui.edu) Quit (Remote host closed the connection)
[21:27] * dyelar (~dyelar@ Quit (Quit: Leaving.)
[21:50] * tdonohue (~tdonohue@dspace/tdonohue) has left #duraspace
[21:50] * hpottinger (~hpottinge@ Quit (Quit: Leaving, later taterz!)

These logs were automatically created by DuraLogBot on irc.freenode.net using the Java IRC LogBot.