Like Eliminating Half The Autos

Found this blog entry that has an extremely interesting observation:

In our case, we could start with the observation that ninety percent of the aggravation in programming projects comes at the end. Therefore, the way to get rid of most of the aggravation is to chop off the end - the part where you're getting every last bit and twiddle in place. That would be ridiculous if the most important matters were left to the end, so make your job easier by doing the important things first and never doing the unimportant ones. Or hardly ever.

Shocking approach, even more suprising it's most likely to be true. Doesn't it always seem that it takes 80% of the effort to put in the last 20% of the features? No wonder we see so many open source projects which are extremely userful but don't feel just quite finished.

In case you're wondering what does it have to do with autos, here's the quote from the same entry:

It does sound a bit like the proposal to eliminate all the auto collisions by eliminating half the autos - since two cars are involved in each collision.