More often than I would expect, I run into conversations about certain IT and web related projects where someone will say, with a bit of authority:
– This is not a big deal! It can be done over the weekend.
Once I was told that a certain feature that needed to be implemented was “only two hours work”! And this came from a person that had no knowledge of the software stack in use.
Another time upon protesting (and basically saying “Put your money where your mouth is!”) I got the best exit:
– I did not say that I can do it over the weekend. I said that it is possible to be done in a weekend by you!
Like I do not have better things to do in my weekends…
So if you have ever claimed that something is possible over a weekend (and can be delivered as a working beta) the weekend is not far. Shut up and prove your point. You can even use Friday afternoon as extra time.
But on Monday please call the guy that said it will take him a week, a month or more and apologize.
Managing organized complexity 
A related gripe about myself is that estimating how much time a software-related project will take is something I don’t know how to do well :-(
I almost invariably discover that something is more difficult or more tricky than what I first thought.
Here is a trick that might help you.
Heh, yes that’s a trick I have heard from multiple sources. One of the most surprising sources was the VP of Engineering at a previous job, who said exactly the same thing over dinner at a place far far from here :)
obligatory Linus reference: http://lkml.org/lkml/2000/8/25/132
ObRef: Hofstadter’s law