The Sayings of Joshua Levy, Software Engineer and Serial Start Up Survivor

• If it is not tested, it does not work.
• If you want a program bad enough, you will get a bad program.
• Programs that are not yet written, do not have bugs.
• The fastest program, is one you do not need.
• The quickest algorithm is to eliminate the need for doing the work.
• People use minutes, engineers use seconds, nerds use milliseconds.
• A small bug will become a big problem, if it is found late enough.
• Many big bugs are thought to be small bugs, when first found.
• The single biggest factor impacting when a project will be finished, is when that project is started.
• Nothing is as perminant as a temperatry solution.
• It is a lot easier to predict the past, than predict the future.
• Is it possible? Of course! This is software: anything is possible. However, some things are a lot more work, and a lot more error prone, than other things.

But there was only one bug! (The "Won't Boot" Story.)

This was in the 1980s a friend of mine worked at HP and was the first guy outside of the UNIX development group to alpha test HP's version of UNIX. He got the software on Monday, filed one bug, and didn't use it again. On Friday, he got an excided call from a guy at the UNIX developement group. "In a week's testing, you only found one bug, on all of UNIX! That's great! We're really happy with all that." My friend said "Did you read the bug?". The other guy said no, so my friend summarized the bug: "Won't boot!".

So remember the two rules:

• Not all bugs are created equal.
• Many bugs hide behind other bugs.