-A A +A

2009-09-24 Conference Call Meeting Notes

Primary tabs

2 posts / 0 new
Last post
Dan Garcia
Dan Garcia's picture
2009-09-24 Conference Call Meeting Notes

Phone Conf Call : 2009-09-24
Attending: Ed, Dan, Lois, Rick, Frank, Greg, Patrick (joined 50min in)

Greg: started working with a grad student on SWENET. I'm new to the harvesting process, but getting help. Is there a "how to" document? "I've got a collection, what do I do?".

Ed: No specific document; we tend to tailor the solution to the problem.

Greg: Ok, I'll dive in, and try to build the how to document.

Greg: Also, I'm an IT guy and have connections to the SIGITE community, and they have no collection. I put together a poster to expost Ensemble and a BoF session. They've both been accepted and I'm hoping to start a conversation in that community, so I hope to connect that group to IT.

Lois: When is the conf?

Greg: October 22-24

Lois: You could carry information from our october meeting there.

Greg: Yes, except I won't be at the october meeting, some of it by telephone, though.

Dan: Take our slides with you to the BoF

Ed: Steve won't make it to the october meeting, I'll try to cover what he does. The most relevant piece is CatSpace and I'll need help from Dan. 

Ed: Lots of changes to the website.

Rick: We've got 3 or 4 things to deliver at the demo. (1) There will be a version of Walden's paths we'll be using. (2) There'll be a dist of VKB 2 that can be downloaded. (3) Steve Carpenter's work in terms of setting up Ensemble in Second Life.

Rick: How do we highlight something that's not a resource or collection?

Lois: I talked to Peter offline. He has tools where the content is produced by the tool. There aren't really documents in the traditional sense, as a poor man's mechanism he could pre-generate some number of "documents" which we'd have available to the OAI.

Rick: When I look at "computing resources" (e.g., walden's paths, vkb) -- where does the tool go?

Frank: Should we have computing tools as well as resources?

Dan: TECH will be there, and we're "technologies"

Frank: We should highlight the tools we're building. Where do you go to download the "Ensemble plugin"?

Dan: Howabout top-level tabs for "Technologies", "Tools", "Portals"?

Lois: Let's go back to the website and look into TECH. The front page used to have a list of groups (e.g., FOCES had a list of 10). We want a list of groups and locally hosted collections. Is that different from the collections we're crawling? Group for CS1, TECH, FOCES?

Dan: Where things live should be invisible (under the abstraction layer) to users.

Rick: We need to make a distinction between collections
within
ensemble or those
affiliated
with Ensemble. \

Lois: But TECH and SWENET have comments/blogs/etc, and we're hosting that. Will that

Ed: The reality is that's not how it works. Forums aren't crawled now. If someone were to give me a metadata record for each web page in Drupal, I can work with that.

Lois: If I put a comment on a blog post, where it's a blog about TECH. Fedora allows documents to have annotations (like a comment). We'd like to be able to have crawlability into comments externally as well.

Frank: Should we have two different search opportunities? One, research-like, fedora. And one, a full-text search of content in ensemble/drupal site, which could be a lucene search. (not requiring metadata).

Lois: We have search now, but that is unavailable to

Dan: We should look to Google as an example and (1) have a single search box and (2) have it search all collections and return values just like Google does. Why not a firefox plugin that sends the search to ensemble?

Ed: Do we want always opt out on search? Wouldn't your search results get cluttered?

Frank: one thing that google does is have different search mechanisms and the user chooses among them

Ed: Howabout lectures, homework, lectures?

Dan: Yes, yes yes. Images / Video / Lectures / Blog posts / etc.  Type AND purpose.

Lois: We've created new types in Drupal

Frank: How big is a resource? Is it individual comments or the blog as a whole? Each of these requires an individual to extarct that subset of things and push them into the index in teh proper way. Is it practical for us to have independent lists like that?

Rick: It's useful to have this information at that level of granularity. Sometimes you want a big resource and sometimes you want the answer to an individual question.

Lois: We've been doing this with subdocuments too (a subset of slides, say).

Rick: How to we collapse the number, in terms of the taxonomy. 500 results is too many.

Lois: This discussion of where things are going to be, fedora or not. In my discussion with Peter, he discussed single-sign-on so the social navigation can be ensemble-wide.  This is related to inst and rewards.  How do we have an "ID" for resource. It might have been born in TECH or SWENET. Let's say it's a PPT file. I put comments on the sub-documents. How does that go back to Fedora? What's the single ID for a document?

Dan: Just go back to the download page.

Lois: But what if you passed the document around; how would it be connnected to Ensemble. If you made a fix and re-uploaded it, could Ensemble know? How would it keep track of versions and give attribution?

Rick: Fedora is flat. Documents and subdocuments and mashups don't easily fit into Fedora.

Ed: We have an object and metadata object. The web has just objects. We insure the quality with tagging and selection. If we start to do annotations, this is a subclass of submissions. Right now the system is collecting from other places, and doesn't have submissions.

Lois: We've explored tools where you can select things in word and there's a plugin that takes can ingest that subdocument, created locally. We'll be demo-ing this. The subdocument is another type.

Ed: We're discussing a range of different types. If we want to incorporate a new feature, we'll work with

Rick: Let's go back to the website. We talked about new tabs (portals, tools, technologies), moving federated search in, about should have ensemble labs. We should be able to mark a forum as an index and made searchable.

Greg: I have a different perspective, since I came in as a 2nd tier person. Suppose I was the advisory group. It seems pretty clear that we don't understand the architecture as a portal. These are basic architectural questions. We present the current portal as a prototype? (2) Advisors would want some architectural diagrams, how does drupal fit with fedora, and how do the other pieces fit. (3) Rick rattled off good suggestions. I keep wanting some mechanism to record bugs and feature requests. E.g., I've gone in and I click on computing resources and those are just images. That's a usability issue, but we have to have a log for bugs and feature requests. Same thing that Rick suggested.

Ed: Can someone make up an organic group for "Feature requests"

Greg: I'm talking about bug tracking.

Ed: Does anyone know about bug tracking software?

Lois: I'll look into drupal to see if they have a plugin

Ed: what should be under the tools?

Rick: The Greg and Frank collection

Lois: We currently have collections and featured collections. What's that difference?

Ed: Are the featured collections also portals?

Rick: no, they're other front ends.

Ed: The usual way is to mock up the system, so we can discuss it and then put it in

Lois: Greg suggested architectural pictures. I'm willing to give it a try. Would that help to organize it?

Dan: Didn't you have this?

Ed: Yes, but we didn't get any feedback. Do we really want to say "5 of this, 2 of this, etc". I'm frustrated, because we're not getting to the real substance that should be in our collection.  That's not supposed to be in an SDL.

Greg: I want lectures, collections of videos, lab assignments. Each has qualities, who would use them, what ages these are for.

Dan: I don't know of a vast library that exists out there, honestly that we just can tap into.

Rick: We didn't propose to be editors. We have to think of it "meta" because we don't have the staff required to add the content to the collection.

Frank: The way Lisee have done this is to hold workshops (in their case teh earth sciences), then they used the tools to create artifacts. I think this is part of the plan.

Rick: Is CSTA taking the lead on that?

Ed: syllabus isn't quite ready (5-8000 syllabi). CSTA has 5 entries, each of which is a tree. My focus is to make sure these all work and are good order. We want to see mock-ups of what this should look like.

Dan: Where do we find the resources we don't know about.

Ed: We have at least things we've includes so far, and well as what we're all working on.

Lois: Do all the leaves take us away?

Dan: We have to have the organic discussion happen on a leaf page before we're taken away to an external resource? That's what we have on TECH.

Ed: the leaf pages are the metadata record now, we probably want ot filter that a bit.

Dan: Can we spend some time designing the leaf page, with a clear sense of what

Dan: Here's a feature request: clean up the leaf pages to just have the essence, with a "read more" which shows more of the metadata record.

Ed: A descriptive or specification of collections. It would include what kids of objects, where it is, how it's organized, what kinds of metadata, how it should be presented in terms of an interface (is it a tree, 2-level), do we have metadata objects pointing to other objects. As we go through and build these collections, we can describe them in their collection.

Lois, Dan, Greg: Great

Lois: This could help our blog structure too.

Greg: this ties to the "how to" feature -- if we can categorize the collection, the how to is obtainable.

Ed: For the algoviz, they'll do the data provider (dublin core with richer collection), with an XSLT that'll show how to render it. We need a "registry of collections" with a description and XSLT sense. Eventually, we'll get this into a registry structure. We're not close to that but I do want Lois collections and Walden path objects so we continue to grow the list and work with the information.

Lois: my student is on vacation, school starts next week, she has the structure ready, she should be able to put some sample content out there.

Ed: we really need to tease out the comments and collections and when is it added, who sees comments/annnotations, not just random people making nasty comments. Amazon does this well. We should think about that.

Dan: we have moderators who check pages and recaptchas which control spambots

Lois: In Drupal you can have groups, admins, etc. The public could make comments or not, etc. We'll demo what you can do in Drupal (e.g., FOCES group).

Ed: We need to start thinking about schedule and agenda. What will each of us present? Do we have sequence?

Dan: I don't see the list from Peter

Rick: Perhaps steve carpenter can talk to that in terms of look & feel / second life. Ed has his set of things, patrick has his set.

Patrick: I listed what I could, please make corrections to it.

Dan: Your list feels powerpointy

Patrick: that's how I created it

Ed: Are we going to have posters and demos?

Rick: will the board be there for the whole thing? If it's the whole thing, we should go into depth. We have these resources, we have portland state material, We could do it by having the sites demo it. If peter's able to be there, he could show us his stuff. We're trying to find a structure that will work. Should we allocate the schedule with the four main sites plus extra things they want to show. Within each of the sites, we could go top-down.

Ed: Draft Agenda:

  • what is nsdl
  • what's the ensemble vision
  • what's are the portals
  • development activities
  • research issues
  • plan / schedule

Dan: where is the architectural diagram.
Greg: vision has to be more abstract, what it'll accomplish.
Ed: we'll have the ontology into drupal
Lois: that'll make it easy for me to discuss this on teh agenda. In terms of the agenda, the front matter has been presented, then the architecture, what's the research part -- topic by topic (what we have now, what we'll have).
Ed: Can someone take what we have and refine the topic-by-topic section.
Lois: Boots?
Dan: I can tell her in Vancouver
Rick: We need planning time.
Lois: Presenting on day 1, day 2 is board discusses alone (we could be meeting then too), then the board gets back to us at the end of day2.
Greg: One small suggestion. We should move it to computing portal.
Dan: Done!

bootsprojects
bootsprojects's picture
OK - agenda tomorrow, I think
field_vote: 
0
Your rating: None

I will work on a draft agenda tomorrow and try to have it out to you all for comments by the end of the day.

Log in or register to post comments