It's not that complex once you have formulas for computing square roots. I've recently implemented it in TypeScript using bigints for browsers & nodejs. Quite readable & performant. See index.ts file here: https://github.com/paulmillr/noble-ed25519
Wish ristretto folks added the library to their website though.
It's fair to say that this is still a theoretical attack. As authors of this paper mentioned, they don't see a way of turning the "flaw" into a real exploit.
"We stress that this is a theoretical attack on the definition of security and we do not see any way of turning the attack into a full plaintext-recovery attack."
... which appears wrong, and even published after the other paper?
Theoretical attacks have a way of turning into weaponized exploits.
For example, check out https://www.openssl.org/~bodo/tls-cbc.txt. This is a document published by Bodo Moeller in the early 2000s that details multiple theoretical weaknesses in the CBC mode used in TLS. Read it top to bottom and see how many practical attacks on TLS you can count.
While it may imply that there's a non-trivial amount of public support to overturn the existing legislation, it doesn't inherently make the regulation "wrong".
>TextSecure folks: instead of ranting that “our stuff exists already, but we got no money and we got no cross-platform support Y U NO USE our protocol?” and using political tricks, go make better protocol and market yourself better.
As DanBC posted[1] in the other thread:
>>You seem to be mistaken about why they do this. It's nothing to do with pushing their app or their approach. They'd welcome good well-formed apps to compete with them. But when they see an app that claims to be secure they have an ethical duty to let people know if it is obviously not secure.
>>Most people are not bashing just for the sake of bashing. Some people need good cryptography software to avoid imprisonment, or torture, or state-killing. This isn't about stopping someone's teen-angsty poetry from being discovered by a sibling, it's about protecting political dissidents from an oppressive regime. In that context pointing out that a software is broken is not mindless bashing, it is a crucial part of the cryptography process.
>Go make your own stuff and don’t listen to HN or any other skeptical community.
Unproven cryptographic systems masquerading as secure need to be criticized. It is very, very dangerous when non-crypto people pretend to be crypto people and call their systems secure.
So, you don't think that just enough people down-voted this bullshit thread? Maybe you are the one wrong on that topic. Maybe you should do your homework.
Just to expand on the switching of axis. It's harder to compare data between graphs, but upon second inspection, it appears the colors from the first graph are not used in the secondary graphs.
Second, trying to get to the projects in the first graph, it's easier to exit the graph from the bottom. Otherwise, short bars (eg. node) are hard to "select" for the "Current" line, and in that case, I have to make a longer trip around the graph to get to the link. Maybe link the labels? Move the "Current" line underneath?
I enjoyed it nonetheless. If you're considering expanding or having a follow-up piece, I'd enjoy combining the three graphs with the format / filtering at http://www.techempower.com/benchmarks/.
Back then I failed. It seemed like Chrome didn’t exposed full console and window api to extensions or so. I doubt things have changed since then. But i’ll try anyway.
I had also advocated for improving built-in Safari web inspector, to make it expose reasonable APIs for extensions like this (radar issue #11653556). Unfortunately, without any result.
As i’ve noticed in repository readme, Chrome coffeescript addons cannot
manipulate DOM or play nicely (or even reasonably) with window properties. They’re just shortcuts for coffeescript.org, which just compiles coffee down to JS. This seems useless to me, because I want coffeescript as first-class browser citizen.
Wish ristretto folks added the library to their website though.