Apps and Tools Call 20130910

Meeting Notes:

Working Group Goals Discussion

1. UCSF Profiles / Plumage

1. Excited to hear about real world use cases

2. UCSF Profiles data is used by 20+ campus groups — want to share stories of successful (and failed) use cases

2. Share examples of how people are using

1. ingest

2. data coming out of VIVO in the real world

3. Look at ways people build tools on top of VIVO data

1. plum analytics

1. also looking at bringing other data back into VIVO to augment what VIVO has natively -- e.g., metrics

2. via OpenSocial or tools like the Duke widgets

2. looking at data consistency across VIVO instances

1. best practice recommendations

3. CTSAsearch can also be used for assessing data consistency

4. See how people work with different sizes of data

1. see it go in

2. see it come out.

3. find any common solutions

5. Addressing New sources of data

1. UF is talking about an official list of universities

1. where do we get it

2. how do we share it.

2. finding common denominators

3. Finding an authoritative list of universities

1. ORCID is working on ISNIs for organizations -- Melissa wants to address this in the Ontology Working Group

2. UCosmic consortium (http://www.ucosmic.org) has a good list of international universities

Catalog of Tools and Maintainers on the wiki at https://wiki.duraspace.org/x/xusQAg 

• would like to have people add to and update the list

• others to add?

• Cornell’s UriTool

• John Fereira’s Semantic Services

• ORNG/OpenSocial

• following the pattern of other open source projects with a table of tools

SPARQL example pages -- positions as a template for creating examples

Example: https://docs.google.com/document/d/1FlGWCG0nqQ4TAVZmnKXiVMZnGxpygdcoxRcZsxw4UG0

• VIVO 1.4 person and positions

• To understand the relationships between the ontology and the data

• several parts:

• a diagram with a small section of the ontology represented visually

• an actual SPARQL query that can run against the open UF endpoint

• what the results look like in text

• how to get the results in JSON, with an example

• sample JavaScript code to run that query, get back the JSON data, and enumerate it

• is it worth time creating these? ~10 examples people would like to see - will put this out to the list

• positions and titles

• educational history

• contact info

• etc.

• goal is to bootstrap people into creating things

• Anirvan

• a good “Hello, World” example to see a complete small application

• Be able to be making an app in as little as 5 minutes

• we’re still a little high level for people to see the promise

• Example of an older “Hello World” example from UCSF http://base.ctsi.ucsf.edu/profiles/json_api/basic_example.html 

• pulls in JQuery, grabs data from custom simplified JSON API, to paint RNS data on the page. Has comments inline.

• very focused on developer usability -- have a local JSON flavor that are using downstream from VIVO so don’t have to figure out the ontology

• easy things should be easy

• also the Cornell semantic web services (John Fereira) --http://sourceforge.net/projects/semanticservice/?source=directory 

• How did you create the diagram?  was done by Nicholas, Stella, or Amy Buehler

• manual

• Ted -- the sample SPARQL queries on the wiki were very helpful when getting started (https://wiki.duraspace.org/x/lwUGAg)

• Stephen -- having more than just a list of SPARQL queries, especially with sample code, should be a good new thing to do -- people are calling for these kinds of apps at Colorado, as for example a directory of people

• Happy to spit the data out in JSON -- seems more popular than XML now

• An example created from the VIVO widgets using the JSONP feeds coming our scholars data: http://people.duke.edu/~outtenr/vivo_widgets/angular_vivo_widgets/ 

Vagrant -- a virtualization tool that allows you to use a Box file (e.g., an Ubuntu box) that they maintain -- just reference it via its URL to get it

 http://docs.vagrantup.com

• Still finalizing a Vagrant configuration script -- https://github.com/senrabc/vivo-vagrant 

• add additional steps like provisioning bootstrap.sh

• gets Java, Apache, Tomcat, MySQL, and even sets up the database and gets things running

• will be able to have a sample dataset as well, so can start working with the source code examples, run SPARQL queries, etc.

• easier and more lightweight than trying to distribute the 4Gb Virtual Box file

• can synchronize with the Github to get configuration changes, so that if one person fixes an environment variable the rest of the team can get those changes

• collaborators welcome

• has been a real help with LAMP stack projects -- easier to share developer environments across a team

• helps you to codify your configuration changes

• Ted -- would be happy to test it - has used Vagrant but not a bootstrap script

• Chris -- picked the wrong box style -- had openJDK where the tools.jar file is in a place

• should have picked a box without openJDK to begin with

• Stephen -- can get openJDK off and replace with Sun Java

• Chris -- but there’s no longer a Sun Java 6 package for Ubuntu -- have to do a wget to pull down the latest JDK from Sun and install it by hand

• also can pull down the latest VIVO code from Git

• Stephen -- if put things in the /vagrant directory on the virtual machine, can access it via Eclipse from the desktop

• if make changes, can (usually) do a vagrant reload, attempting to reload the configuration without wiping and re-downloading the whole VM

• Next call at 8 am EST -- agenda suggestions welcome

• looking for presenters as soon as the next call