LispCast

I would like to respond to Peter Christensen's recent post about the inability of "language snobs" to finish anything.

While I don't like language snobs (and I have been guilty of a little snobbery myself), I also don't think that his analysis is fair. I don't think the (effective) business model of "Always be shipping" is the end-all-be-all of software engineering.

This is an idea I've been bouncing around in my head for a while now. And it's funny that Peter quotes Joel Spoelsky on this point, because I think Joel is one of the leading proponents of this view.

Here it is in a nutshell:

You won't make money if nobody uses it.
Nobody uses it if you don't finish.
Therefore, at all costs, finish the product.

I don't think there's anything wrong with this argument. And it certainly is important with a software business.

But there's another argument that is not as commonly shouted on the net.

I have a vision for a product that will change the world. // note: lots of people get here
The product won't change the world if I don't finish it. // fewer get here
Therefore, I will finish the product. // rarely does anyone get here

This is the kind of thinking that Alan Kay says made Xerox PARC what it was. And what did they ever invent?

• The mouse
• Overlapping windowing system
• Ethernet
• Personal workstations
• Smalltalk

You can read all about the environment at PARC in this paper. He also argues against safety-oriented business thinking that trades big risky ideas for producing mediocre short-term results. One model produces MS Windows. The other creates a new way to compute.

But finishing is important. The difference between the two ideas is mindset. If you've got a big idea for a project, that project will drive you. You are an artist with a message for the world. And you need a language that can let you express that idea. And you won't be able to rest until it is completed.

The difference between the two arguments is in the conclusion of each. The business-oriented one (the first one) has the phrase "at all costs". That phrase is not present in the second one. If you have a vision, you trim off the fat and focus on what's important. But you don't compromise it. If you have to ship, you have to compromise.

Where would computing be without the serious, dedicated, and finished (!) efforts at Xerox PARC? I use all of the things on that list except Smalltalk on a daily basis. Macintosh and Windows copied them and made billions. It's too bad they didn't get a chance to monetize it themselves. But Apple and Microsoft couldn't have done it without the big thinkers.

More to the point: would Perl exist if Larry Wall had just decided to hack things out in bash scripts---totally focused on the next release? Where would Python be without far-thinking ideas?

Bottom line: finishing is always important. Business (shipping) is important. Big ideas are important, too.