We're running what I'm going to call an "invitational" hiring process. Basically it means that we'll find you, don't find us. The thing about hiring is that it's very hard to do. I've heard that—at least for software—you might as well pick a resume from the stack at random as much as use traditional HR methods (post a job, receive a bunch of resumes, filter them based on some process like keywords or something, interview a bunch of people, make a few offers, pick someone). Resumes don't really say much except for the job experience section, and even there it's thin. I think you can probably explore plenty on the topic at Joel on Software since he talks about it all the time.

I believe that hiring the right person for software can make the difference of something like 100x to 1000x in terms of creating quality code. In other words, the superstars are just so much better that they can turn a project that could take years into a matter of weeks. And it will work a lot better, you can continue to improve it, and so on. Not only that but if you're working on "hard" problems it will mean the difference between success and total abject failure. Therefore I'm not going to bother hiring someone to write code for Semacode (which is "hard") unless they are a superstar.

Here's an example. Apple has a brilliant file system search feature called "spotlight" that allows you to do a full content search of all your files in seconds. Microsoft doesn't (although they planned to and spent years working on it for the next version of windows, it's been abandoned). The difference? From what I know, it's basically this one file systems engineer that Apple picked up—if my memory is right—it's been a while— from Be .

And so it goes, if you've hung around at the top end of software development you've probably seen similar effects many times, although maybe not at such a grand scale. It could just be the team of "web developers" that spends years and millions of dollars on a system that never works, and then a small team of quality people steps in and makes it all happen overnight. Software development is a world with genuine superheroes.

I've been reading people's opinions about how you create and nurture a silicon valley-like tech hub. Waterloo is struggling to build/maintain it's status as such . It really seemed to be headed in that direction in the 80s (with watcom and others) and then things were weak for a while until RIM started really making it good in the late 90s.

What "they say" is that in order to get something like silicon valley you need three elements:

• a leading university to act as a stabilizer, base, and talent source
• a city/place that people like to live in
• successful companies that spawn a generation of rich executives who reinvest in the new up-and-coming companies

As a talent source and stabilizer, the University of Waterloo is great. I think that most people put it in the top 10 list of computer science schools in the world. Certainly Bill Gates'many visits to the school indicate what he thinks (recently he make UW his only Canadian stop in an international tour).

And that's good, because as someone who's running an invitational hiring program, you need people to give the invitations to. At a university, you can hook up with a network of professors, students, organizations and so on and, if you have a valuable opportunity for them, you can probably find the right people.

Of course you have to have something to offer. I think that Semacode's software is cool to work on because it has many unique and interesting technical challenges and problems. In addition, the end result—the product—is also interesting and it's fun to see how people want to use it. The superstar programmers will only be interested by superstar-level challenges. If you have one, then you can find a match.

On the subject of my previous post about the "dna" in the path at the BBC. Apparently "DNA" is the name of the software that runs h2g2 and also The Collective. I'm sure it's named after this DNA .