Hacker News Reader

Comment by forgotmypw17 | Hacker News Reader

Excuse me, but does your alternative provide toolchains and user interfaces for every major platform in existence today, including Mac, *nix, iOS, Android, Windows, a library for every major language in existence, and 25 years of attempts to break it?

If not, I don't see how you can claim it is superior in every way, because here are at least two ways in which PGP/GPG are by far superior.

View on HN

I dunno about windows but minisign is on Mac and Linux, yes.

As for 25 years to break it, well, go look at CVEs for GnuPG. There have been many.

The algorithms still stand 25 years later.

Is it also available for iOS and Android? FreeBSD?

Are there libraries for Perl and PHP?

> The algorithms still stand 25 years later.

They don't, though.

Both RSA and ECC still stand strong, AFAIK.

What are you saying here?

I was referring to CAST5

Based on my reading, there are no serious known issues with CAST5, with a minor issue being its small block size. [1]

Is there anything else?

[1] https://crypto.stackexchange.com/questions/24045/is-cast5-st...

This might represent a common misconception about PGP. The algorithm preference information is embedded in the PGP identity. It is signed by the certification/signing key. So a downgrade attack would involve breaking the cryptogrphy used for the certification/signing key. CAST5 is not used for certification or signing and has been nowhere close to the start of the preferences for a really long time. Having CAST5 as some sort of last ditch backward compatibility thing is no more of a weakness than the fact that the computer I am using to write this has an MD5 command.

Having said that, what is wrong with CAST5 for PGP application? What does it have to do with a system that is all about signing things?

Reply on news.ycombinator.com