(3:00:43 PM) tdonohue: Hey all -- DSpace Devel Meeting is starting now(ish) right here.  Here's the agenda: https://wiki.duraspace.org/display/DSPACE/DevMtg+2010-10-20
(3:01:38 PM) tdonohue: Essentially, the only topic is 1.7 updates -- but, I fully expect it will take us the entire hour, as I know there's lots of features people likely want to get in before the Friday Feature Freeze 
(3:02:15 PM) mdiggory [~mdiggory@ip72-199-217-116.sd.sd.cox.net] entered the room.
(3:02:21 PM) HardyPottinger [80ce6c16@gateway/web/freenode/ip.128.206.108.22] entered the room.
(3:02:51 PM) PeterDietz: Hi All, I'll jump in
(3:03:06 PM) PeterDietz: Here's a list of whats up with 1.7: https://jira.duraspace.org/browse/DS/fixforversion/10306#atl_token=zowX-mpK9E&selectedTab=com.atlassian.jira.plugin.system.project%3Aversion-issues-panel
(3:03:20 PM) tdonohue: Do we want to tackle this as a "round-robin" of sorts?   I'm going to be diligent in cutting off discussion that does not pertain to 1.7 Features (as that is our most immediate deadline) :)
(3:04:14 PM) PeterDietz: Before we jump in, I want to say that I've been really impressed with all the code flying in in the past week
(3:04:24 PM) tdonohue: PeterDietz -- do you want to lead?  Again -- I'd suggest we concentrate on *features* only -- we can tackle bugs in the next few weeks still
(3:05:01 PM) PeterDietz: go ahead tdonohue
(3:06:32 PM) tdonohue: OK -- well, I see mdiggory has added features to our agenda that @mire wants to provide to 1.7.  mdiggory, do you want to start? https://wiki.duraspace.org/display/DSPACE/DevMtg+2010-10-20
(3:06:56 PM) mdiggory: Ok
(3:07:33 PM) mdiggory: So as you can see from the recent activity, listserv email and wiki pages, we have three contributions that we consider important to be in 1.7.0
(3:07:43 PM) mdiggory: These are in various stages of commit to the trunk
(3:07:51 PM) mdiggory: starting with item 1
(3:07:59 PM) mdiggory: New Alternative dri2xhtml and Basic Theme that can be a clearer starting point for XMLUI developers to start from
(3:08:21 PM) tdonohue: https://wiki.duraspace.org/display/DSPACE/New+Base+Theme+For+DSpace+1.7.0 
(3:08:39 PM) mdiggory: We've created an new layout for the structural and other xslts in a new base theme for easing access to the templates
(3:08:42 PM) mhwood: [cheering]
(3:09:17 PM) tdonohue: very nice! :) 
(3:09:21 PM) mdiggory: Likewise, we've created a new Basic Theme for DSpace that pulls forward the relevant templates that we most often customize when working for a client
(3:09:41 PM) mdiggory: this is just being finished up and is expected to be committed in the next day
(3:09:46 PM) stuartlewis: Could we / should we swap to using this as the default theme?
(3:10:47 PM) mdiggory: Our decision to do it in parallel is based on a concern that there are users out there who have altered dri2xhtml directly with their changes, we did not want to break their themes
(3:10:48 PM) tdonohue: we could vote to make this default theme (instead of Reference) in 1.7.0
(3:11:25 PM) tdonohue: (we don't need to decide that now though -- we could always wait to make it the default theme after the Testathon, or just before Testathon)
(3:11:53 PM) tdonohue: mdiggory good decision -- we don't want to replace the existing dri2xhtml, just want to give an alternative
(3:12:29 PM) mdiggory: I agree, while we think that would be great, having an opportunity to test what is about to be committed may be wise
(3:12:35 PM) ben_atmire [~ben_atmir@ip-83-134-115-216.dsl.scarlet.be] entered the room.
(3:12:39 PM) mhwood: But parallel issue and which is default are separate things.
(3:12:54 PM) tdonohue: the only thing that slightly is off to me, is the name 'dri2xhtml-alt'  :)  but, i'm willing to live with it, if we have no other ideas
(3:12:57 PM) grahamtriggs: choosing to make it default doesn't mean not offering the existing dri2xhtml based theme. Question: is there a more suitable name than dri2xhtml / dri2xhtml-alt?
(3:13:47 PM) mhwood: dri2xhtml-tng
(3:13:59 PM) tdonohue: what is "tng"?
(3:14:10 PM) richardrodgers: the next gen
(3:14:10 PM) mhwood: The Next Generation
(3:14:14 PM) tdonohue: oh -- duh :)
(3:14:55 PM) mdiggory: I'm open to renaming it to whatever the community determines is appropriate, perhaps just "lib" or some other generic name as it is a xslt "library"
(3:15:02 PM) tdonohue: dri2xhtml-2.0  -- though that's not much better
(3:15:15 PM) mdiggory: 2.0 is a dangerous number
(3:15:31 PM) ben_atmire: maybe something that denotes it's split into separate files per functionality
(3:15:32 PM) tdonohue: haha -- i knew someone would bring that up.  
(3:15:40 PM) mhwood: I like "lib" better than my suggestion.
(3:15:44 PM) ben_atmire: but don't know a good short name to say that either
(3:15:53 PM) richardrodgers: dist?
(3:16:21 PM) mdiggory: "includes"
(3:17:11 PM) HardyPottinger: ezdri2html?
(3:17:15 PM) PeterDietz: split / open / blue / prybar
(3:17:17 PM) mhwood: Name to be decided later?  Anything else about it to discuss?
(3:17:28 PM) mdiggory: anyhow, we can certainly discuss it later
(3:17:30 PM) grahamtriggs: although... given the description of the changes - isn't it true that the only problems with simply switching out to the new version going to be with people that have 'chosen' to implement their customisations badly?
(3:17:32 PM) tdonohue: name to be decided later -- agreed, we need to move on
(3:18:12 PM) tdonohue: next topic -- sounds like everyone is excited for 'dri2xhtml-alt', or whatever we call it
(3:18:33 PM) mdiggory: grahamtriggs: it would be advised that in the documentation, we start to talk about best practices and not altering the library directly
(3:19:00 PM) mdiggory: ok, next topic.
(3:19:03 PM) mdiggory: (2) Fixes to Solr performance, additional Statistics customization documentation and one example fix
(3:19:43 PM) mdiggory: Withiin @mire we took the position that we wanted to improve the community involvement on working with the statistics views.
(3:20:11 PM) mdiggory: With this in mind we started an education campaign with Ben doing a presentation on the matter
(3:20:30 PM) mdiggory: and constructing some how to documentation on how to approach these adjustements
(3:21:02 PM) mdiggory: @mire tackled correcting one one of the outstanding tickets
(3:21:33 PM) tdonohue: sounds good -- just need to track all code fixes in JIRA (obviously) if they aren't there already.   Is there any specific feedback/approval you are looking for on Solr/Stats changes?
(3:21:39 PM) mdiggory: as an example, but to further support this effort we are punting the other two that are under my ownership back to the community as projects
(3:22:06 PM) mdiggory: I'll leave it to ben_atmire to comment on the performance improvements
(3:22:42 PM) ben_atmire: I have committed one part of the performance improvements
(3:23:11 PM) ben_atmire: and replied on an existing JIRA ticket for this, as it was already an open discussion
(3:23:28 PM) ben_atmire: and we have a second performance improvement for solr which is a bit more advanced
(3:23:46 PM) ben_atmire: which I also briefly described in my webinar
(3:23:58 PM) ben_atmire: I'm planning to describe that one in more detail
(3:24:05 PM) ben_atmire: but it won't need any commits
(3:24:25 PM) ben_atmire: because it's mainly configuration and crontab scripts afterwards
(3:25:01 PM) tdonohue: ok, sounds like it'd be a good addition to the 1.7 documentation then 
(3:25:02 PM) ben_atmire: I've been drafting a description about that, but it's not finalized yet
(3:25:15 PM) mdiggory: The work will be good material for the Documentation wiki.
(3:25:30 PM) mhwood: [more approving crowd noises]
(3:25:35 PM) ben_atmire: was there a deadline for the documentation?
(3:26:14 PM) mdiggory: Oct 29th
(3:26:15 PM) mdiggory: https://wiki.duraspace.org/display/DSPACE/DSpace+Release+1.7.0+Notes
(3:26:23 PM) tdonohue: "tentative" documentation deadline (for just initial versions of docs) is Oct 29th.  But, in reality, they mostly just *need* to have some version on the Wiki Documentation area in time for Testathon (Nov 8)
(3:26:55 PM) tdonohue: the Oct 29th deadline was much earlier to allow for Jeff Trimble to have time to do editing and ask for clarification, etc
(3:27:03 PM) ben_atmire: I'll see what I can do by next Friday
(3:27:25 PM) ben_atmire: My draft is currently on our server at http://atmire.com/statistics_performance
(3:27:37 PM) ben_atmire: but it's work in progress
(3:27:46 PM) tdonohue: ok -- in essence of time, we should move along (unless there's anything else needing immediate attention on this)
(3:27:59 PM) mdiggory: Ok
(3:28:40 PM) mdiggory: (3) We have been working intensively the last couple weeks to finalize a version of Discovery that fits well into the dspace trunk
(3:28:41 PM) mdiggory: https://wiki.duraspace.org/display/DSPACE/DSpace+Discovery
(3:29:04 PM) mdiggory: This process invloves
(3:29:30 PM) mdiggory: (1) moving a copy of the dspace-discovery provider, block and webapp sources under the trunk
(3:30:07 PM) mdiggory: and adding configuration details to the xmlui.xconf, solr distribution and config directory to make it easy to enable discovery in the XMLUI
(3:30:25 PM) mdiggory: https://wiki.duraspace.org/display/DSPACE/Discovery+Configuration
(3:31:00 PM) mdiggory: as outlined in the wiki page
(3:31:03 PM) tdonohue: Are there also Theme level changes to make Discovery "appear"?  Or will it work with all existing out-of-the-box themes?
(3:31:33 PM) mdiggory: it will work with preexisting themes
(3:31:39 PM) tdonohue: excellent :)
(3:32:06 PM) mdiggory: However, there are some modifications we have to the aspect tier we want to discuss so there are no surprises
(3:32:23 PM) grahamtriggs: I have a bunch of caveats with discovery - which are largely catered around it being optional - but worth discussing
(3:32:29 PM) mdiggory: if you look in the example xmlui.xconf presented above
(3:32:47 PM) mdiggory: <aspect name="Browsing Artifacts" path="resource://aspects/BrowseArtifacts/" />
(3:32:47 PM) mdiggory:         <aspect name="Searching Artifacts" path="resource://aspects/SearchArtifacts/" />
(3:33:17 PM) mdiggory: We have worked on breaking apart the aspect definitions for ArtifactBrowser so that different features can be enabled/disabled via aspects
(3:33:44 PM) mdiggory: this allows one to enable discovery and disable search and still retain browse capablities
(3:34:04 PM) mdiggory: or disable both search and browse and enable discovery to replace them
(3:34:46 PM) mdiggory: we are working further on breaking off RecentSubmissions into its own aspect as well, since recent submission in Discovery is powered by solr instead
(3:35:06 PM) mdiggory: and can prvide facet lists on the Site, community and collection pages if desired
(3:35:07 PM) mhwood: It was probably time to step back and look at how the work is divided among the stock aspects.
(3:35:47 PM) tdonohue: sounds good.   (let's hurry explanations along, so that we have time for questions from graham & others -- also need to give others time to go over last-minute proposed additions to 1.7)
(3:36:04 PM) mdiggory: if you remember back to richardrodgers presentation on "bricks", this is an important change in that direction
(3:37:53 PM) mdiggory: ok, well, that summarizes most of the work, we will continue to enhance the features once we have them in place.  And now I would like to open the floor to other topics and or questions
(3:38:12 PM) tdonohue: 1.7 + Discovery questions?  grahamtriggs? others?
(3:39:19 PM) PeterDietz: its good to see lots of performance fixes coming in, big or small, I see that mhwood hates new Integer(...) as those are all be squished out
(3:39:47 PM) mhwood: The new code-quality reports are a big help.
(3:40:10 PM) grahamtriggs: my points won't come as much of a surprise - I've already covered them before. But nonetheless...
(3:40:30 PM) mdiggory: It will be interesting once we get the Bamboo instance generating those reports in CI builds
(3:42:35 PM) grahamtriggs: I'm concerned about the resource usage of Solr at the lower end of repositories.... also, I'm concerned that we don't have enough guidelines / experience around maintenance of Solr indexes, in particular combining the nature of browse/search indexes and statistics data in a single Solr application
(3:43:36 PM) PeterDietz: does discovery use a separate core?
(3:43:40 PM) tdonohue: grahamtriggs -- all valid concerns, but mdiggory had proposed having Discovery disabled by default in 1.7 -- and that people would need to actively choose to enable it (by a series of documented configs)
(3:44:38 PM) grahamtriggs: As I say, this is mitigated by it being optional - but we should have guidelines about it's use/maintenance, or explicit statements that we aren't providing those guidelines
(3:46:13 PM) mdiggory: grahamtriggs: as this is Open Source, these guidelines need to evolve over time and by experience, we intend to be using this quite a bit in the future and will be vetting those concerns
(3:47:06 PM) tdonohue: makes sense to me -- the docs should try and provide guidelines, and make it clear that its the initial release of Discovery, and as such, it is more for early adopters until we feel comfortable enough that we'd want to make it the default setting (should we ever decide that)
(3:47:11 PM) mhwood: I would suppose that the coming documentation will begin to address this.  No doubt it will be possible to refine and elaborate as we gain experience.
(3:47:27 PM) mdiggory: As @mire had products and is vested in solr as solution, we will be vetting such concerns for our clients
(3:47:33 PM) mdiggory: had = has
(3:48:20 PM) jefftrimble [~jefftrimb@maag127.maag.ysu.edu] entered the room.
(3:48:32 PM) tdonohue: Do we all agree with the move to add Discovery to 1.7 out-of-the-box -- I'm assuming from the silence that most do, but perhaps we should do a very quick vote on this?
(3:49:27 PM) mhwood: I for one don't yet know enough about it to object or encourage.
(3:49:30 PM) mdiggory: solutions, such as the performance enhancements, will propagate to the community when they solve problems with community solr work, as we've seen with Ben's recent work
(3:50:11 PM) mhwood: Yes, but we do need to capture what experience there is, so that the community has a starting point.
(3:51:03 PM) mdiggory: We've tried to do some of this by improving the wiki for the discovery module
(3:51:03 PM) bradmc left the room (quit: Read error: Connection reset by peer).
(3:51:04 PM) tdonohue: ok -- more silence.. I'm assuming that there's no one who really objects, so we should release Discovery in 1.7 as a optional, beta-level component.  If there are objections, speak up here or on listserv later on
(3:51:08 PM) stuartlewis: +1
(3:51:28 PM) stuartlewis: I'm sure once it is in, a lot of us will be turning it on.
(3:51:48 PM) stuartlewis: Can I ask a quick question about [dspace]/bin/ ?
(3:52:04 PM) tdonohue: quick -- sure
(3:52:15 PM) ThomasAJohnson left the room (quit: Ping timeout: 252 seconds).
(3:52:19 PM) stuartlewis: Now that most of the scripts have been deleted, we're left with a few such as dispace_migrate
(3:52:27 PM) stuartlewis: or dspace-info.pl
(3:52:41 PM) stuartlewis: They are much more visible now that there are very few scripts left.
(3:52:42 PM) robint: +1 to Discovery, but I'll let others be the early adopters :)
(3:52:45 PM) stuartlewis: Do we want to keep them?
(3:52:49 PM) stuartlewis: Or deprecate them?
(3:53:02 PM) ***tdonohue notes we need to give someone else the floor very soon -- MIT has a lot of stuff on their burner still  -- this is a pre-prompt for richardrodgers to get ready :)
(3:53:03 PM) jefftrimble: Many of us are still using them.
(3:53:07 PM) stuartlewis: Or move them into a 'suplementary' or 'unsupported' directory?
(3:53:25 PM) jefftrimble: if we have replacements, they can do....
(3:53:31 PM) tdonohue: i'd say deprecate if anything -- they shouldn't be removed unless there's a replacement in Java or we know they are unused
(3:53:35 PM) stuartlewis: jefftrimble: Which do you do use?
(3:54:01 PM) jefftrimble: Both....I print the daily stats off and give them to staff
(3:54:21 PM) jefftrimble: and the migrate I use from a processing server to a production server for some of our departmetns.
(3:54:44 PM) stuartlewis: dspace_migrate has been replaced (I think) with the -m (migrate) option in import/export scripts.
(3:54:48 PM) mdiggory: I thought the perl work was moved to java in a previous verion?
(3:54:52 PM) jefftrimble: that's right.
(3:54:58 PM) stuartlewis: mdiggory: Most were - a few remained.
(3:55:14 PM) tdonohue: stuartlewis & jefftrimble -- in essence of time, can we move this to dspace-devel or #dspace?   I'd like to give richardrodgers the floor to talk about 1.7 features still being worked on by MIT
(3:55:22 PM) stuartlewis: tdonohue: Sure.
(3:55:24 PM) jefftrimble: absolutely
(3:55:42 PM) richardrodgers: OK - curation looks on track (although feedback welcomed!)
(3:55:58 PM) ThomasAJohnson [~tom@216.160.210.151] entered the room.
(3:56:15 PM) richardrodgers: Also - new CC licensing (using their current web service API - not iFrames)
(3:57:20 PM) richardrodgers: we are still up in the air about Google SEO - not clear if what we have will make Google Scholar happy enough...
(3:58:03 PM) tdonohue: so, nothing is better than something with Google SEO afterall?
(3:58:12 PM) stuartlewis: WIll it make them unhappier though? If it is an improvement, even if not perfect, can it be included?
(3:59:12 PM) richardrodgers: we need to discuss with them further - I'm not sure. I'd like to include what we have already
(3:59:47 PM) richardrodgers: CGI is 1.8 for sure, although it will be better for it
(3:59:50 PM) mdiggory: richardrodgers: do you have a pointer to that work?
(4:00:12 PM) bradmc [~bradmc@207-172-69-79.c3-0.smr-ubr3.sbo-smr.ma.static.cable.rcn.com] entered the room.
(4:00:13 PM) mdiggory: it may be wise we assure its in the dri2xhtml-alt library as well
(4:00:24 PM) richardrodgers: mdiggory: yes, sands does, but he's out - I'll get it for you later
(4:00:29 PM) mdiggory: thx
(4:00:59 PM) ***tdonohue thinks Trunk is going to have a very busy next few days :)
(4:01:05 PM) richardrodgers: I think that's mostly it - a bunch of bug fixes as well
(4:01:33 PM) tdonohue: sounds great -- glad that most of it will be ready in time
(4:01:34 PM) mdiggory: tdonohue: good reason for me to get that Bamboo build working ;-)
(4:02:00 PM) tdonohue: mdiggory: agreed
(4:02:39 PM) tdonohue: OK -- any other 1.7 features that we need to discuss from anyone?  Any general questions?  Are we all ready for the great 'feature' push for the rest of this week :)
(4:02:58 PM) mdiggory: oh... its fixed now
(4:02:58 PM) mdiggory: https://bamboo.duraspace.org/browse/DST-TRUNK/test/case/4030465
(4:03:58 PM) mdiggory: hmm, and its broke again...
(4:03:59 PM) mhwood: DCDateTest.testGetCurrent() seems unstable.  I think it needs a small fuzz factor in case the clock ticks during the test.
(4:04:21 PM) mhwood: I've seen it fail before with apparently identical times.
(4:04:31 PM) tdonohue: silence -- so I guess that means we are all ready, and everyone should head off to finish up their coding, etc.
(4:04:46 PM) tdonohue: mhwood -- yea, I've noticed it fail occasionally as well -- not consistently though
(4:05:10 PM) robint: mhwood: I'll have a wee look, cheers
(4:05:14 PM) mhwood: Probably depends on whether the VM clock ticks in the middle.
(4:05:20 PM) mdiggory: robint: excellent
(4:05:24 PM) mhwood: Thanks.
(4:05:42 PM) tdonohue: We'll close up this meeting with one annoucement.   Instead of our meeting next week, we have a "Intro To Fedora" session scheduled on Weds @ 20:00UTC: http://www.mail-archive.com/dspace-devel@lists.sourceforge.net/msg04565.html
(4:05:54 PM) mdiggory: :-)
(4:06:03 PM) jefftrimble left the room (quit: Quit: Leaving).
(4:06:08 PM) mdiggory: looking forward to it
(4:06:19 PM) tdonohue: I'd encourage those who are interested to attend --we've specifically scheduled it at a time where we all usually meet for that purpose
(4:06:43 PM) tdonohue: it will also be recorded, if you are unable to attend
(4:07:15 PM) tdonohue: that's it!   Go off and make 1.7 great!  I look forward to seeing the stream of 'dspace-changelog' emails over the next few days.
(4:07:22 PM) richardrodgers: Thanks to @mire for all their 1.7 efforts...
(4:07:37 PM) mhwood: Seconded.
(4:07:41 PM) mdiggory: richardrodgers: thanks, looking forward to the curation work as well
(4:07:41 PM) tdonohue: exactly -- thanks to all of you!  Thanks to @mire & MIT for these last additions
(4:08:31 PM) mdiggory: I will pass it onto the team, thanks
(4:08:55 PM) richardrodgers: Gotta go - bye all
(4:09:01 PM) tdonohue: also thanks to mhwood & grahamtriggs for their diligent cleanup work this week
(4:09:01 PM) richardrodgers left the room (quit: Quit: richardrodgers).
(4:09:23 PM) HardyPottinger left the room.
(4:09:48 PM) robint left the room.
(4:11:03 PM) ben_atmire left the room (quit: Quit: ben_atmire).
(4:11:43 PM) mhwood: You're welcome.  Must go, thanks all.
(4:14:17 PM) mhwood left the room.