That's a pretty good article, mostly I agree.

Except, not making estimates is not a mistake! Estimates are stupid because they’re never accurate and decisions should be made at the point when you have the most information about a problem (last minute), basically estimates are a waste of time that give a false perception of control.

Anyway the argument in ‘not making estimates’ is not talking about scheduling estimates it actually talks about scalability, using twitter scalability problems as an example therefore it’s kinda conflicting with premature optimisation. But if we are talking about Scalability, then scalability of twitter should have never been an issue because developers should always code in a style that can handle most scalability problems. Whenever I write a application in asp.net it's automatically developed for scalability, it's inherit in my development style I assume other good developers would be the same.

Thanks.

Michael: I suppose part of the reason for writing this "top ten" list is to make people think about what their mistakes might have been. Your point is valid, of course, but there's only so much I could have written in two thousand words for publication.

Cheers, Julian

When you wrote the simile "slower than a Greenland glacier in January", did you actually research glacial velocities or did you just figure they move pretty slow about then? :) I think this requires a fact-finding expedition.

Bryan: I must admit to a certain level of assumption in that Greenland is in the Northern Hemisphere and January is in the midst of winter, ergo glaciers would be slower than in say July. But maybe that's false! Ergo, a fact-finding trip to Nuuk is in order...

Cheers, Julian