Zach Copley

There were two presentations. The first was given by Jon Seidel, the creator of 4MyPasswords.com!, a service designed to be a secure, websessible central repository for passwords and other secret bits of data.

Jon developed it partly to scratch an itch — as a way to always have access to a password manager from any computer — and partly as a way to learn Rails.

He demoed the application and walked us through the evolution of its development. Then he gave sort of a laundry list of the things he likes and doesn’t like about Rails.

Some of the things he likes about Rails: the overall framework, code generation (scripts), database migrations (which he especially enjoys as he hates being a DBA), having a built-in Web server while developing, the tools, documentations, and some of the libraries — he mentioned ActionMailer as being really cool.

Dislikes: pluralizations, a lot of the conventions, and in general he seems a bit skeptical of some of the hyped features of Rails. For example, he doesn’t find Rails to be as intuitive as it’s supposed to be.

It also sounded like he has had a lot of trouble with his PC-based tools and development environment. The IDE he used, RadRails, frustrated him a lot. It’s broken in a bunch of ways and he’s received little or no support for it. Several attendees of the Meetup had good suggestions for alternative tools to try.

His final verdict, however, is that Rails is awesome and he loves it.

I’m sure for some people 4MyPasswords.com! will be just the ticket, but I’m far too paranoid to trust a hosted service with that kind of information, even though I’m sure Jon is an upstanding guy. During the Q&A for his presentation, I asked whether he had posted any sort of privacy policy. He had not. I’m glad to see he’s added one.

My own paranoia aside, 4MyPasswords.com! is a nice little Rails app, and I’m glad I got a chance to hear the story behind it. One of the cool things about the East Bay Ruby Meetup is there’s a nice mixture of experienced Rails developers, people just getting started (like me) and everything in between.

OWL

Next up was Tom Atwood, the CEO of DATA-GRID, a startup in Oakland developing a DBMS based on the W3C’s Web Ontology Language (OWL) standard. My understanding is that it’s a database optimized for handling ontologies, which could replace current DBMS systems in some (or all, I think they hope) circumstances. He gave a whizz-bang PowerPoint presentation, which I found a little hard to follow, but the technology looks neat. Part of the presentation was a mock-up of a futuristic sports news site covering the Tour de France in real time with lots of annotated video feeds. Tom made the point that video data is getting really complex now (heavily annotated) and meshes well with OWL.

It sounds like the OWL DBMS is a huge system, “analogous to Oracle 10g.” The schemas for it will be in OWL, although it wasn’t clear whether they will be developed using OWL or some higher-level intermediate language or tool. Identifiers for records (keys) will work across the Web, instead of only being local to a single database. They’ve developed something called ActiveOWL to “replace” ActiveRecord. I’m assuming it’s an object-to-ontology-mapping (OOM?) layer that will sit in for ActiveRecord. DATA-GRID is looking for Rails developers to test OWL DBMS and ActiveOWL, and perhaps build applications that can take special advantage of an ontological DBMS. Maybe this is my imagination, but I felt like Tom was hinting that applications with user-defined ontologies — tagged stuff ala flickr and del.icio.us — might benefit from having an OWL DBMS backend. Web 2.0-style mashups are apparently a good candidate for some reason as well. (Perhaps because the feeds they are comprised of already have a considerable amount of semantic structure?) It will be interesting to see more of the nuts and bolts of ActiveOWL, and hear more about what they have in mind.

What the special advantages of an ontological database are, I’m not clear on. I guess the bet is the future will be all about “The Semantic Web.” If application data and Web content will be organized in an heavily ontological way, then it makes sense to have a persistence layer to put it in that understands ontologies, instead of shoehorning ontological data into relational DBMSs. The problem would be similar to what we have today with persistence in object-oriented languages and applications. Ideally, at least it seems to me, we would store objects in object-oriented databases (ODBMSs), but for various reasons they haven’t caught on much (yet), and instead objects generally get flattened into relational databases.

It also sounded like for ontologically-organized data, OWL DBMS can be faster than relational databases. Tom mentioned that the NSA is testing OWL DBMS with huge datasets and has had some favorable results.

License

This work is published under a Creative Commons Attribution-Noncommercial-Share Alike 3.0 License.