I take issue with: "Such issues have been largely overlooked in the Amazon.com discussion, which takes for granted that "glitches" are inevitable and self-correcting, that the market will police itself. (We've seen how well that worked on Wall Street.)"

In a very large, very complex application like Amazon.com, glitches _are_ inevitable. They're not self correcting - there are hundreds of highly competent, highly intelligent software engineers that correct them. It's not Wall Street, where people have an incentive to game the system and naturally collude to do so. It's doing something profoundly more complicated than most people really have a conception of, and not being able to execute flawlessly all the time.

Amazon fully deserves the heat for the 1984 debacle, and their communication left a lot to be desired in the politically sensitive, but probably actually a bug, issue that delisted large swaths of gay literature (while embarrassingly leaving anti-gay rhetoric as the primary search results).

But this author seems to think the reaction is to attribute some sort of malice to the entire culture of software engineering, that we accept some level of imperfection and bug fixing in a major project. This is just not a reasonable expectation, and it does nobody any good. Some companies may not spend enough on QA and testing (and I'm sure more than half of you just thought of the same company), but this stuff is hard, and there will be bugs, and we'll try to fix them, and we really don't need some knucklehead from the L.A. Times trying to infer about a process that's too complicated most times for the managers at Amazon itself to really have a handle on.

Oh noz! Amazon has made a few small missteps over the years. The sky is falling!