The real troublemakers are the Rails developers who seem to seriously believe that leaving such subtle security traps in their framework (and then blaming the developers who follow the example code) is a defensible position.
An analogy would be a door that only locks with a special key in a certain sequence. IF you choose not to do so, it's merely a door. Obviously, you could argue that that's a bad default but I think that goes to the crux of the problem.
What do you mean "properly deploy their app"?
The sample code at rubyonrails.org looks like it has the same problem to me, i.e., it would be vulnerable if it were put into production in the right (entirely reasonable) circumstances.
No matter how bad the tool actually is, only a terrible craftsman blames his tools. That's what he meant.
Both Rails and GH are at fault. Rails for not discouraging poor practices and GH for not being more familiar with their own stack.
They/He make it sound like they were proactive about all of this, always one step ahead of the attacker/@homokov
> As soon as we detected the attack we expunged the unauthorized key and suspended the user.
They didn't detect anything. They were informed. It also wasn't an attack. If it was an attack, you wouldn't have seen it coming, or in the very least more damage would have been done.
As far as attacks go, this was probably the "whitest" of white that has ever gone down. Not only did the "attacker" not do any actual damage, but he was continually ignored.
This is a shameful handling of the actual events.