Add an interactive SPARQL tutorial

[dmcassel]

We have an interactive XQuery tutorial that is pretty popular: it lets people try out MarkLogic without having to download and install. Create a similar one for SPARQL.

The data and responsibility for executing the queries will be on the try server; the tutorial itself will be on developer.marklogic.com. Queries must be read-only and will be subject to a timeout.

• Identify and load a set of triples. (Ideally, the XQuery and eventual SJS interactive tutorials will be updated to show mixed queries. With that in mind, a set of triples that relates to the current XQuery tutorial documents (MailMan.org messages) would be ideal. Perhaps an ontology about Apache projects?)
• Build a tutorial similar to the XQuery interactive tutorial. There should be a set of queries with explanation about what they do and how they work. I think the tutorial itself should assume no knowledge of RDF, but have a page that covers that. People who are familiar with Semantics in general can skip that page and see how it works in MarkLogic.
• Include a couple mixed queries (XQuery + SPARQL, SJS + SPARQL).

[scottrhenninger]

I'm interested in contributing to this tutorial. I have a semantic model on academy awards that includes about 700 films, over 3000 people and industry organization, along with some reference data on locations. That could be utilized as well.

I could also look into converting the MailMan.org messages - do you have a link for where that data lives?

[dmcassel]

I wouldn't think of converting the MarkMail messages (my bad, typo earlier); rather, I'd think about an ontology to supplement them. That way, we'll have docs and triples together. With the right content, we could run some interesting docs only queries (as in the XQuery tutorial), some SPARQL-only queries, and some mixed.

[scottrhenninger]

Hmm, I seem to be missing something. When I go to interactive XQuery tutorial, I don't see a way to log on to the machine.

[dmcassel]

@scottrhenninger are you talking about logging in to make a new tutorial? That would be on a different server. I'll send you a link and set up an account for you. If you're talking about viewing the tutorial, no log on should be needed -- please let me know if that's what you mean.

[scottrhenninger]

I guess I have two questions:

1. For the XQuery tutorial, the second page (http://developer.marklogic.com/try/ninja/page2) has some XQuery statements to enter into the Query Console. But on what server? I'm missing where the tutorial should be executed.

2. Yes, setting up a different server to run the SPARQL tutorial is a good idea. No hurry on that. First I'd like to look at the email db (from 1. above) and do a bit of modeling to see how we can use triples to enhance search. I'll also likely need mlcp access to move the emails and triples to the new tutorial server.

[dmcassel]

It's not intended for Query Console -- each block of code is in a text box with a "Run" button. You click Run, and the code gets sent to a separate MarkLogic server where it gets executed. The results are then displayed on the tutorial page. Are the Run buttons not showing up for you? Are you able to run them?

[scottrhenninger]

OK, I am able to run them. I guess I just didn't anticipate that's how it would work :)