This is a good, basic, common sense article - thank you. Time to time it is forgotten that no system is without errors, or, maybe the environment, software or hardware, changes and creates new errors. No way out.

Very good points as no system is too small or too large not to benefit from good configuration and source management - easy and not expensive when started from beginning - time consuming and expensive to do later, test just not the functionality but how to break it and save tests for regression tests, have a good communication with users and give them an easy and clear channel to report errors, don't jump to "fix" a bug before you understand why it really happened, document everything even if it looks embarrassing, share the error and change information, keep testing, keep verifiable and synchronized logs of whole systems - not just one application, and so on, the article really mentions most of them.

One problem today are the "nice" end user decorations on pages, windows, etc - keep the vital functional parts separated from those decorations and advertisement. Too many times there is nothing wrong with application itself but the add-ons cause problems and even security problems.

And yes, effective troubleshooting is a skill which comes out of experience but also needs well documented systems. This is often forgotten. There was a time when the first line support was a respected job, today, too often, it is just to record and report the incidents and forget it, not much help to the enduser. And really causes problems when the real troubleshooting is done - the information is often so convoluted going through several levels of support that it may be almost useless - you have to start from beginning, it is a waste of time and in one off errors happening occasionally.. I don't even want to think those, too many sleepless nights trying to find what really happened!

Click to read the article this is in response to.