JRuby 1.0 Nears Release

In software, nothing sets my heart aflutter like new developments in Ruby/Rails-land. If all you want to do is make a quick, CRUD app, there really is nothing better, and in fact we'll see one of those new CRUD-apps (a [Digg](http://digg.com)-clone for vetting product features) next week.

I am intrigued by developments from Sun on JRuby (Ruby running as a first-class language in a Java virtual machine), and the prospect of JRuby going 1.0.

The current Ruby/Rails isn't particularly mature nor thread-safe, so using it for Enterprise-level development is basically a stretch. JRuby could make Ruby and Rails for enterprise development worth considering. It's definitely worth keeping an eye on.

Anybody interested in joining the hypefest should check out two tutorials: Rolling with Ruby on Rails, and it's successor Rolling with Ruby on Rails, Part 2.

There's a lot of leading-edge work and thinking coming out of RoR, and if you're a professional computer scientist of just a webby hobbyist it's worth your time to check it out.

Managing Software - Part I

There is a certain brand of humor that only a software developer can understand. Writing computer code well requires not only remarkable logical or linguistic skill, but the ability to translate through layers of abstraction (from the "real world" to the machine) as well.

And so, managing software projects and people is a bit different than managing other things. The biggest single risk in any software project is "lack of communication" (see the cartoons above and below), and this blog is one small step in hopes of opening lines of communication.

It is sometimes hard to communicate even the most basic elements of a development effort. In search of better communication I've been working with Darcy to try to implement a means of tracking Development that is gentle on the developer, but rich in the information it provides. We're working to implement a form of painless schedule.

A bit of scheduling can really open channels of communication, as a decent schedule and trends provides a ready, day-to-day answer to some key questions:

• What's the next milestone?
• What's in it?
• Who is working on it?
• When is it expected to be delivered?
• Where are the biggest risks?

It's hard to believe that such fundamental information can be had with about 5 minutes per day per developer, but I've seen it be so. In my experience the ability to answer such questions routinely completely changes the communication dynamic for teams writing software.

More on this to follow...

All in the Family

I miss Accidental Empires - the characters described in it are now in their dotage, but in its day it was the only book on tech that gave the proper human context to Silicon Valley. It's easy to read the news, and see a sterile world of one big company doing business with another. AE blew the cover on that, and showed a world smart, vain people trying to one-up each other.

If you don't know the people you miss the news. Take this story: Oracle claims Yahoo has defected. On the surface, the story looks pretty simple:

• Oracle makes a big move into open source, feeding on RedHat's business - and then
• Oracle snares a big customer win

That's the simple story. Scratch a bit deeper, and you find this: Open Source Business Models - Brent Williams' terrific and influential paper from Eclipse.con. Believe today's press, and Oracle has a big win - believe Brent Williams, and Oracle's entry into Open Source has been a complete failure. Who to believe?

Follow the people, and I think the truth can be found. If you know the people, you might discover that the driver behind the Yahoo! side of the deal was most likely CTO Farzad Nazem. Dig just a bit deeper than that, and you'll know that Farzad and his wife were both (Oracle CEO) Larry Ellison lieutenants in the old days. Know just a bit more, and you'll know that Farzad and his wife were married in Larry's house.

So, Larry was looking at a failure, needed a deal, and called in a favor. Oracle's open source initiative is probably still hurting, but they've got a win in the press, and maybe can build on that.

SO ... what to take out of this? Read Brent Williams' paper because strategy matters...

The Father of Fortran Passes On

Fortran was the first computer language I ever learned, an experience I found so odious that it was four years before I even attempted another language (C - much better experience), and so it is with uneven sadness (miss him, not Fortran) I report the passing of John Backus.

Fortran was fine for scientific programming, but to my knowledge was practically useless for anything else, and gave computer science much of the nerdy, why-would-you-ever-want-to-slog-though-this reputation it still suffers from today. I genuinely lament the passing of any Turing Award winner, but really, we deserved better.

All of this is not so for Backus' contemporary John McCarthy) . Now THERE was a giant!

Godzilla vs. Mothra Part I

I love standards battles - it's the lifeblood of advancement in software technology. I've already written a bit on RIAs (Rich Internet Applications) and this seemed to be (when I wrote it) The Battle of 2007.

Another fine article in the emerging series pits Ajax and Flex.

Dan Webb's blog is one I read regularly - he's always good, and very good at citing other articles.