That's the cruft. If they separated those out into separate apps, I'd be happy. It's an E2E-encrypted Slack-clone, public key management system, encrypted file storage, and a crypto wallet.

I guess one man's cruft is another man's crucial feature.

Precisely the reasoning for writing tools that do one thing very well.

The problem with that is that it completely disregards one market, i.e., the one that wants an integrated system. You could probably install six programs that do these individual things, but then you'd have to install six programs. That's a non-starter for those looking for the integrated solution.

What if you made a unified installer that installs those six tools configured in a certain default way?

Certainly there are plenty programs that are designed modularly, componentized, plugin-based, compositional, whatever... They just never feel as good as purpose-built apps with good workflows. If you know of one that does, please do forward it on because I'd love the inspiration.

I think Keybase does what it intends to do, which is to make PKI accessible to non-experts and to cultivate a community of users achieve the usefulness that network effect affords.

If you're really just looking for encryption and signing, encrypted filesystems, encrypted chat, distributed filesystems, and you're an expert, you have the tools you need.

> They just never feel as good as purpose-built apps with good workflows. If you know of one that does, please do forward it on because I'd love the inspiration.

I think the flaw here is that you're (likely) comparing purpose-built with unrelated unix-like apps.

What could have been done here with Keybase, is a purpose built app composed of unix-like tools, all controlled and implemented towards the purpose-built app. Aka the IM would be unique if you wanted it, or bundled with the FS, but both would be first-class Keybase citizens.

Unfortunately most companies don't put the time or effort into letting people consume parts of their offerings. It's all or nothing. Which is the complaint here, imo.

I wonder if Keybase could re-factor in a way that makes the core functions and various features available as small, focused tools, but still allows them to build the comprehensive binary for general users who want everything.

There is a downside to that, though... an integrated product that can do more than one thing can often be easier to use than two separate ones

Teams + git repositories for secrets is great too.

Makes it less likely that our secrets end up on Github where if Github were hacked, or an web account was hacked that our secrets become public.

Sounds interesting, how is this feature useful?

One way we use it is for shared team secrets (e.g. API tokens) in infrastructure scripts.

https://github.com/plyint/encpass.sh/blob/master/extensions/...

Used a similar strategy to great success at a previous startup for various credentials. Everything else had higher friction and was causing devs to do insecure things to avoid that friction.