...
Time | Topic | Location | Presenter | Description | |||||
9:00-10:30 | Welcome and discussion | OCR | All | Present yourself and your institution, and share brief thoughts and experience about your Fedora implementation or plans. | |||||
10:30-11:0030 | Short Coffee break (, discussion and transfer to Auditorium) | ||||||||
11:0045-1112:3030pm | Fedora updates and Q&A | MA | Andrew Woods (Duraspace) | ||||||
12:30-1:30 | Lunch | Museum cafe | |||||||
111:30-112:5000 | AIC Updates | MA | Stefano Cossu, Kevin Ford (AIC) | Latest on LAKE, the AIC's Fedora/Hydra Digital Asset Management System | |||||
11:50-12:00pm | Open Science Framework | MA | Rick Johnson (ND) | Overview of Fedora and Open Science Framework integration efforts at Notre Dame, Johns Hopkins, and Center for Open Science | |||||
122:00-122:1030 | Fedora at ICPSR | MA | Harsha Ummerpillai (ICPSR) | Archonnex@ICPSR, Fedora Admin GUI using ReactJS & JEE. | 12:10-12:30 | More? | |||
12:30-1:30 | Lunch | Museum cafe | |||||||
1:30-1:40 | Choose break-out session topics | MA | |||||||
1:40-2:45 | Break-out sessions #1 | MA | |||||||
2:30-3:00 | BreakOpen Science Framework | MA | Rick Johnson (ND) | Overview of Fedora and Open Science Framework integration efforts at Notre Dame, Johns Hopkins, and Center for Open Science | |||||
3:00-4:00 | Break-out sessions #2 | MA | 4:00-4:30 | Discussion and wrap-up | MA | ||||
6:00, 6:30? | Drinks & Dinner | See dining below |
Day 2:
...
In-depth discussion
Time | Topic | Location |
---|---|---|
9:30-10:30 | Karaf/CamelGather use cases for direct query feature | OCR |
10:30- | 10:45Break | |
10:45-12:15 | Content Modeling (part1 & part2) | MA | 12:15-1:30 | Lunch |
1:30-3 | Import/Export | MA |
3:15-4:30 | Karaf/Camel | MA |
Minutes from discussions
12:00pm | Fedora code walk-through | OCR |
12:00-1:00 | Lunch | |
1:00-1:45 | API-X demo | |
1:45-2:30 | Content modeling | |
2:30-3:00 | Fedora documentation |
Minutes from discussions
I -- Specific searches people might like to have exposed in Fedora
Traversing a graph -- getting stuff that's one step away from a resource. This sounds like a sparql query -- two use cases, one based on path, and the other based on date or date range, and response is always a list of uris
Give me all the children of a resource where x property matches in one way or another y value.
Give me all children underneath a resource .. as well as 404 or 405s or 410s, information on the lost resources.
Let me filter the attributes returned. A search returns hundreds but we're only interested in a few. No, clients can't define set of attributes you get back (server attributes and user attributes, clients can specify that they do not want server attributes, but can't manage either list of attributes).
We'd like some very basic counts of things: how many resources are in Fedora in order to compare with what's in the triple store or solr -- we don't have a way to get that, and Fedora is the canonical resource
We'd like quick snapshots of how many binaries vs non-binaries in the repo, the size of binaries .. how many of each RDF type, how many are there in general, etc.. You can do this with LDPath but that will hit every resource, which is a memory hit and not really what is needed.
II -- APIX
(There is a youtube video - marketing-focused; how requests are intercepted, two models for doing so).
On the gihub repo there are dockr containers and steps for getting it up and running and providing feedback.
APIX is an interface to Fedora, with methods for lots of things apps commonly want to do (processing of binaries with ImageMagick or FITS). Apps talk to APIX and it does one of two things: passes data to a service to do something, then puts data in Fedora and then sends back response, or puts something into Fedora and then passes something off to a service and then sends response.
III -- ACLs ... many get created, but does community know much about how and what implications are?
IV -- PCDM. UI might be bigger priority than community collectively realizes.
V -- Jersey is the language Fedora uses to manage REST .. Jersey is an implementation of the JAXRS specification.
VI -- Code walkthrough
Main entry point is Fedora LDP -- there's an http module that is accessed by clients
Other Information
Lodging
Fancy hotels near the museum:
...