Tuesday, April 18, 2006


Herds of Lispers?

Ever VNC into a computer that has a VNC window open to the computer that you're on? That's how I feel when I link to a blog that's linked to mine.

About twice a week I search for 'lisp' on technorati, and it's still a little surprise for me when I see someone else writing about one of my articles.

Bob Congdon's entry is mainly a discussion about my entry about Java vs. Lisp language expansion, but at the end he talks about the likely number of Lisp programmers on one project - he himself worked on a project with five. I can't imagine much more than that, both because you wouldn't need that many and we seem to be so few and far between I'm not sure how you'd get that many together.

What's the biggest number of Lisp programmers you've seen on a project?

My group at work has, let's see, seven people who do much of their work in Common Lisp. Some of the non-Lisp part has been outsourced, with both good and not-so-good results.

It's not that hard to hire lispers; getting lisp talent offshore (in the places where the talent is cheaper) seems harder.
As someone who is starting (again again) to learn Lisp, the same thoughts come to my mind as to Bob's. The current success stories in the Lisp arsenal are generally of the form: "Very small team tightly collaborates to wild success. Big company buys them, abandons Lisp, subsequently fails."

Repeatedly, I hear that Lisp is the raw material from which the actual language of the application is constructed. A regional dialect seems inevitable.

A strength is its own weakness, of course, and to an equivalent degree. The problem is essentially how to propagate and maintain the shared cultural knowledge of the application. It's easy in a small team working closely together, but J. Random Java Programmer can walk in off the street and read any existing Java code for any company. Nearly any two people on the planet can share Simpsons quotes. Culture can be enforced by ubiquity. That's always been the goal of Sun with Java.

I'm also interested in how well a Lisp team can scale. Ultimately that doesn't factor into my decision to write in Lisp. Personally, I'm insterested in tools that grant the most power to a lone developer, and signs currently point to Lisp as a good candidate. If it has properties that make it distasteful to the masses, all the better. It better serves as a secret weapon :)
In grad school, we maxed out at 4 lispers.

My current work team had three lispers at one point. It's too bad we were working in Java.

We've all heard that Lisp expands your programming horizons, etc, blah, I was wondering if all the Lispers produced better Java code than the "pure" Java developers.
Re `getting lisp talent offshore (in the places where the talent is cheaper)': there is a nascent talent for Common Lisp growing in India, but it is not much concentrated in one city or state. Some of the young people who have taught themselves Lisp here seem very talented. There is also a lot of interest in Lisp. Our Yahoo! Group bangalore-lisp has over 50 members. However, the actual number of people who really take the plunge and try to learn Lisp here is much smaller. It would be great if we could get people to visit Bangalore to try to build a Lisp team. I think it would certainly be possible to build a `herd of Lispers' within about 6 - 12 months here. How does a team of very bright 20 Lispers, each extremely happy with a salary of $1000 a month, sound? It might even be possible to get pretty good, Lisp-knowing talent at $200 a month per programmer.

Tom Elam, tomelam NOSPAM at gmail dot com
rickar cook,

i wonder if "pure" java coders produce better lisp code. or any code for that matter.
Post a Comment

<< Home

This page is powered by Blogger. Isn't yours?