scrypt VS security-research-pocs

Check out the scrypt encryption tool.

Given that the encryption algorithm is open source (https://github.com/Tarsnap/scrypt) can you try to explain what you mean here?

I wouldn't use OpenSSL personally. If you just need simple but secure symmetric encryption, checkout the scrypt(1) encryption utility from Tarsnap. If you need support for public keys, check out age(1).

scrypt if you strictly only need symmetric encryption.

^ "scrypt page on the Tarsnap website". Retrieved 21 January 2014.

First of all I know that "implementing your own cryptography is bad". However, at some point, one does stumble upon a use-case that is not (well) covered by existing tools.

Now, assuming one has already done his due-diligence and has read (and hopefully understood at least the main ideas of) cryptography related articles / posts / etc. (especially in the area pertaining to what one wants to build), and thus we can assume one is not a complete newbie in this mater, however, nor is he an expert. Basically we can assume he is an "amateur".

Where would one go with his design to ask for feedback about it, in the hope to at least eliminate some weaknesses that one (as a non expert) might have overlooked. (I'm not speaking here about "proofs" or "audits".)

----

More specifically ---- but please let's not get into this right now, this being just an example ---- I'm trying to implement something similar to `scrypt` (the encryption utility, that uses the `scrypt` PBKDF, ) or `age` (), as a replacement to my current solution that relies on GnuPG.

It entirely does and that's exactly my point. Most "hashes" are designed to be fast, for data validation/checking whatever. For securing data (passwords, anonymisation, etc) you want a "hash" to be as slow as possible. Scrypt for example is designed to be extremely slow and use much memory (making GPU-based parallelisation useless and driving up the cost of CPU-based work). The default settings for five-second hashes changes their 18 hour estimate to a bit over two years... and that's assuming you don't turn it up further.

> Suppose I'm a typical desktop user, how is important information going to be stolen if I have mitigations turned off and JavaScript enabled?

https://github.com/google/security-research-pocs/tree/master...

I don't imagine I'm going to explain it better than the many others who have already done so.

> What state does my browser have to be in, and what actions do I have to take (or not take) for the attack to succeed?

Your browser would have to be pretty old/ outdated since they've been updated to mitigate these attacks.

> What likelihood is it that someone has deployed an attack that meets those requirements?

That's not a simple question. Threat landscapes change based on a lot of factors. As I said earlier, we won't see these attacks because people have already patched and attackers have other methods.

> So we agree it's OK to leave mitigations off and browse the web?

You can do whatever you want, idk what you're trying to ask here. What is "OK" ? You will be vulnerable but unlikely to be attacked for the reasons mentioned. If you are "OK" with that that's up to you.

Those attacks relied on performance.now() https://github.com/google/security-research-pocs/blob/d10780...

Chrome has limited "performance.now" to have a relatively low resolution: https://chromium-review.googlesource.com/c/chromium/src/+/85...

Also, "2018 install of win10", you might have already been patched during install.

Microsoft rolled out specture/meltdown mitigations at the OS level in January 2018.

That said, I'm using 'arbitrary' to mean any algorithm can be downloaded and ran without the user doing much; JS is Turing Complete. Additionally, older browsers' JS engines (at least the configuration of them) were feature-rich enough to do the relevant exploits.