Krugle: Business and Egosurfing

William Grosso @ June 15, 2006

Krugle went live yesterday. It's a search engine "for code and technical content."

That is to say, vertical search focused at the software developer market. Very interesting to me because I often want such a thing. When you're hacking together a prototype, you often "google for code." Having a tool that's explicitly designed for this purpose is a good thing.

And, once upon a time, I contemplated building such a thing. It's not that hard to build a prototype (mine used a really simple-minded spider, lucene, and some pattern matching). I didn't build a more robust version, though, because a robust version looked like a lot of work (and had some problems that felt unsolvable. For example: figuring out whether a particular code fragement is a high-quality code fragment. I suspect that Krugle's going to get around that problem by relying on the developer community to give feedback).

Even more importantly, I didn't see the business. Quite honestly, I wouldn't pay for access to Krugle. And though that may be a function of where I live, I suspect the vast majority of potential users are similar.

Which means that this is probably an OEM-and-advertising play (OEM? Absolutely. MSDN needs to do much better at search and I could see them wanting to incorporate a good application to solve the problem). But …  8 million in funding? I don't quite see the necessary returns there.

But, for the here and now, Krugle's live and fun. And there's evidence that advertising is big bucks these days, so maybe it'll stick around for a while.

So my first search on Krugle was, of course, about me. I wrote a few articles for O'Reilly a few years back and I keep hoping they've had some impact on the quality of code being written out there.

So I searched for AbstractRemoteMethodCall (a nice little object I wrote to encapsulate remote procedure calls).

And I found clear cut-and-paste usage (from the generics version of the codebase)

I'm not sure what to think. I would have loved for whoever did this to have contacted me. And, of course, for them to have kept my name on my code, or at least said something somewhere in the comments about me (and pointed people to the articles).

What I'm really uncomfortable with, though, is that jhomenet is GPL'd. My code wasn't. And it feels weird to have my code taken, pasted without attribution into another source tree, and then GPL'd.

On the other hand, at least someone's using it. That's good, right? The point of the articles was to teach people to build better software. And they are.

Filed under: Career, Java | Share on Facebook