any word on whether she’s been saving her Jing screenshots to a personal Dropbox public folder?

This video is processing – it'll appear automatically when it's done.

If you believe that anything about organizing this fight happened “just by random chance”, then I have a bridge to sell you (incidentally, the original London Bridge resides today in Lake Havasu City, Arizona, part of the greater Las Vegas metropolitan area)

This video is processing – it'll appear automatically when it's done.

Again, give me some examples, or at least an example!

In my own experience, unit tests have been helpful in a few specific scenarios, for example I recently submitted a pull request to the open-source twitter-text library that attempted to fix a bug when processing hashtags in languages that are written from right to left (Arabic, Hebrew, etc)

Although my first attempt fixed the bug I had been dealing with, it broke an existing test on an Arabic hashtag. It was really helpful to see that test break: I had to change my approach, but it was a fairly trivial 1-liner anyway. It’s a good thing that test suite was in place because without it I would have broken something!

This video is processing – it'll appear automatically when it's done.

IMO this post could benefit from a small handful of concrete examples – both times when tests were good/helpful, and times when TDD led him astray.

I don’t have a horse in the TDD race, so I’m down to be convinced either way, but without anything specific I don’t feel like I can have an informed opinion about whether he’s right

This video is processing – it'll appear automatically when it's done.

It’s one of my favorite gags to apply the Madden “create a player” interface to real life: you have N points to spend across K categories, and you can’t max out every category, so you have to make some tradeoffs.

This video is processing – it'll appear automatically when it's done.