Re:New South Wales Attacks Researchers Who Found Internet Vo...
by rtpg on Monday April 06, 2015 @11:53PM
(#9331926)
I realize the inherit issues with rewriting crpyto, but does the javascript-ness of it matter? Serious question, I feel like I've heard saying that it can be a factor.
[ Parent
| on HN ]
Re:New South Wales Attacks Researchers Who Found Internet Vo...
by gtank on Tuesday April 07, 2015 @12:07AM
(#9331954)
People who object to javascript crypto usually mean that in the context of "browser javascript", which is fraught with peril [1]. The javascript language itself isn't necessarily the problem (although parts of it are dodgy by the standards of what you'd like to implement crypto with).
Re:New South Wales Attacks Researchers Who Found Internet Vo...
by RKoutnik on Tuesday April 07, 2015 @01:22AM
(#9332153)
gtank explains the browser crypto problem well - there are a few issues writing JS crypto on the server though.
I've been going through the CryptoPals challenges with Node.js and have hit a few snags, virtually all of them involving types. I've switched to TypeScript and things have gone much smoother.
The crypto module does add some padding unexpectedly, though I'm not knowledgeable enough to say if that's according to spec or not.