Macworld Forums: Twitter explains outage, apologizes for downtime - Macworld Forums

Post your comments for Twitter explains outage, apologizes for downtime here

"Validation via deployment" is okay when you're a small company with no impact, but when you become a serious player, and certainly if you aspire to become publicly traded, validation needs to be done properly. Via controlled, scaled, process testing and repeated auditable exercise, not by just pushing the code out and seeing what happens to the users. Commonplace in pharmaceutics and defense, where lives and freedom hang in the balance. But guess what: the investors financing you will be even more attached to their money. Better start doing things RIGHT.

Only one company in the world can get away with crappy code, and that was only because business got trapped by a dependency on Windows.

TeaEarleGreyHot, on 22 June 2012 - 06:05 AM, said:

I don't care how careful you are or how thorough you think you can be, systems like twitter are incredibly complicated with many interactions and possible failure vectors. Sometimes despite all your best testing, analysis and design @^!& just happens.

Of all services - it's Twitter! Not even as important as email is for most people, but you wouldn't know it for all the weeping, wailing and gnashing of teeth. If it was the nationwide air traffic control system or something I could see some of the angst that was expressed...