I’ve always been interested in enormous, impossible progamming problems.
That’s probably why I’ve steered my career in the direction of machine learning algorithms and statistical AI.
And, for those of you who followed my 30-days-30-ideas thread last summer, that’s why I was so interested in the AI-Coder business idea.
But anyhow, about those impossible problems…
Yesterday, I returned from obscurity to tell you that I’ve been busily grinding away at three major projects. Today, I’m going to tell you a little bit about the first project, with some background information for those who may not have been following along.
Over the summer (and immediately following my failed interview at Google) I announced a project to brainstorm 30 business ideas in 30 days. Although the project slipped a little bit (I ended up developing 28 ideas in 43 days), I came up with a broad range of possible software projects.
Over the next month or so, I arrived at a decision: the AI-Coder project rose to the top of my list, and I started working on an implementation. But within a few weeks, I ran into a number of difficulties (both technical and legal) and I decided to reconsider.
So, I returned to my list of semifinalist projects and began to weigh the options.
And now I’ve chosen a new project.
Well, it’s been a busy two months.
It’s nice to be back here again, sitting in my blog-writing chair and saying hello to all the nice people out in blog-land again.
(No, I will not use the word blogosphere.)
Wouldn’t it be cool if your computer could keep detailed information about all the music you listen to? Then, you could compare your listening profile to the profiles of other people on the internet, finding other people whose musical tastes are similar to your own, and scanning through their playlists for new music that you might like.
My mom and dad arrived in Salt Lake City tonight in anticipation of this weekend’s nuptials (my little brother is tieing the knot), so I spent the evening hanging out with them rather then writing the fifth entry in my series of 30 business ideas.
(No worries: I’ve promised to meet my deadline.)
I thought you’d find this little nuggest entertaining, though:
Yesterday Peter Thomas published a ridiculous J2EE call-stack diagram from an enterprise application he’s been working on. If your head doesn’t instantly explode in a cloud of incredulity after looking at that image (and reading the comments from the readers of Peter’s blog), then perhaps I can interest you in purchasing a General-Purpose Tool-Building Factory Factory Factory.
If you download the PDF version of the stack-trace picture, and zoom waaaaaay in close, you can see that the “Business Logic” portion of this monstrosity is actually a storeItem() method call.
That’s right: more than eighty frames in the call-stack, just to insert a record into a database.
It was a little more than three months ago that I was contacted by a “talent scout” from a little-known internet company called “Google”.
Every software engineer has been told a million times by salty old college professors, or by senior programmers, to always write good hashing functions. We all know that good hashes are critical to writing efficient searches and that if a hashing function produces too many collisions, the search efficiency can degrade from constant-time performance to linear-time performance. Ouch.
Recently, Charles Petzold blogged about a new Microsoft specification called CSAML, which uses XML to directly specify C# parse trees.
Brilliant.
This timely topic is actually very close to my heart. In parallel to the work Petzold has been doing, I’ve been simultaneously working on an XML specification for describing the contents of an XML file. It’s called XXMLDL (eXtensible XML Description Language).
I think you’re going to like it.
My Why I Hate Frameworks article was recently translated into Chinese by Evan Wang. You can read his translation here:
http://www.blogjava.net/evanwhj/archive/2006/03/11/34837.html
Thanks, Evan!!
I’m currently in the planning stages of building a hosted Java web application (yes, it has to be Java, for a variety of reasons that I don’t feel like going into right now). In the process, I’m evaluating a bunch of J2EE portlet-enabled JSR-compliant MVC role-based CMS web service application container frameworks.
