[benesch]

> This change would break a lot of legacy code for no good reason

Preventing a bug that occurs in 5% of observed codebases (and anecdotally, happens to me during development all the time) seems like about as good as reasons get.

Swapping a perfectly fine print statement for a function, on the other hand… that’s the breaking change in Py3k that’s never seemed worth it to me.

[aflag]

I've never heard from Guido on this, but I've always felt that he created the print keyword in the very early days, just because it was easy and he always thought the language would be a niche small language. But, as the popularity of the language increased, the print keyword just stand out as a sore thumb and he just had to fix that.