You are free to only use the PGP features you need.

But if you want to send secure messages it also matters if your friend has remember to enable --dont-be-stupid and disable --automatically-guess-its-20-year-old-insecure-format

Or your friend should use a software that does all the PGP mumbo-jumbo for you such as Deltachat

I'm not quite sure if you're serious or not. The person you're responding to thinks that a special purpose app would be better than PGP, and your response is that the best solution is a special purpose app which calls into PGP. Why is the latter preferable?

The person I'm replying to says that PGP is hard to use because there's so many options you have to know and use correctly. I completely agree with that, its UX is most probably the main reason it isn't more widespread. But that doesn't invalidate PGP itself: the format is still useful, we have the tools to use it correctly. So I'm saying that using PGP crypto is still worth it, especially if all the bad bits (configuring, tuning) are managed automatically by the application.

The problem with a universal format is it's hard to know that everyone you're talking to does things right. If you have end to end security and the other end leaks, that's not very good. Whereas with a limited format I know a non-malicious counterpart probably has their client configured just like mine.

Here's a good quote from the article at the top of this chain:

> Take AEAD ciphers: the Rust-language Sequoia PGP defaulted to the AES-EAX AEAD mode, which is great, and nobody can read those messages because most PGP installs don’t know what EAX mode is, which is not great. Every well-known bad cryptosystem eventually sprouts an RFC extension that supports curves or AEAD, so that its proponents can claim on message boards that they support modern cryptography. RFC’s don’t matter: only the installed base does. We’ve understood authenticated encryption for 2 decades, and PGP is old enough to buy me drinks; enough excuses.

> You can have backwards compatibility with the 1990s or you can have sound cryptography; you can’t have both.

I am very well aware of the criticisms against PGP-as-a-model and I actually agree with them. The premise that having an open protocol makes changes 10x harder to actually spread is very true. I still believe that it's better to have that than everyone doing the same thing over and over and over again, "but this time it's better".

HTTP, javascript and TLS have also shown that a sufficiently motivated set of actors can move the ecosystem forward. True, organizing the ecosystem is not the same job as actually building stuff, but it's still beneficial to all of us.

> HTTP, javascript and TLS have also shown that a sufficiently motivated set of actors can move the ecosystem forward.

Absolutely. PGP could be decent. Unfortunately efail and sks[1] show it's actors aren't sufficiently motivated.

[1] https://gist.github.com/rjhansen/f716c3ff4a7068b50f2d8896e54...

The latter is preferrable because it uses a widely adopted and time-tested standard, which means another user can use their own special purpose app.

That's the problem. You have n*m possible combinations of clients, leading to many opportunities for bugs, and everyone has to degrade to the lowest common denominator because you can't upgrade in lockstep.

How is that a problem?

If I want to do X, why not use the tool that's built for doing X?