Before I start to use it, I wonder if there are anyone who doesn't recommend to use Keybase not just because it was acquired by Zoom.
The Keybase backend was never open-sourced and there doesn't seem like a viable path forward for Keybase without the founders.
At the same time, doesn't seem to be any other single solution for the range of things Keybase did.
For secure E2EE chat, I moved to Matrix/Element and it was extremely painful. I am rooting for them, but getting regular people (like my dad and sister) onboarded is super hard. A typical HN reader can probably make it work, and deal with all the bugs and issues, but a "normal" person not so much. The bot story isn't anywhere close to as good as it was with Keybase. Most of my friends can't be bothered to get it set up and working, either.
I really miss Keybase Git, which was a brilliant idea that I haven't seen anybody else do. I set my dad's company up storing their business documents in git backed by Keybase git instead of GitHub. I had to move that git hosting to a server I now have to manage, ugh.
I didn't really use the "secure file sharing between arbitrary people" feature, but I don't know of anything as easy if you did want to do that. I use Syncthing to share sets of files across my own machines, and it works well for that.
The crypto wallet part is well-covered by existing apps. No big loss there.
I really liked Keybase, especially toward the end when the performance problems on mobile with sharing images and basic stuff like that were mostly fixed, and the apps started to show a higher degree of polish.
It is pretty hard to recommend Keybase at this point, though — never fully open sourced, and complete radio silence since the "we've been acquired" blog post. I think it is highly likely that bits will gradually stop working and won't get fixed, and/or Zoom will pull the plug.
(I was afraid that happened the other day, when Keybase Git stopped working and git push operations were failing, but it came back up. Not sure what happened, and AFAIK there was no announcement or anything.)
I feel like Keybase missed a lot of money prior the acquihire. I would've spent $xx/month or year for some of the capabilities or increased storage in KBFS. Additionally, it didn't integrate well with other environments (especially mobile) and applications. Dropbox and others expose their contents to other applications on mobile OSes. If Keybase had done that for KBFS, it would've been fantastic for me. I had been keeping a lot of org files in KBFS (still have them there, but backed up to other locations). Being able to pair something like beorg + KBFS would've been very convenient. Or KBFS + textastic.
But I had to open the files inside a plaintext viewer inside the Keybase app. And that's just one possible UX improvement. If they'd been willing to take customer money for some of the things, they could've afforded to work on those kinds of enhancements.
I only really used the cloud drive type capability, which I've replaced with Sync Thing running on a handful of devices.
The thing I will miss is the identity assertion chain. I'm not sure of the best alternatives for asserting that the same person controls accounts on different platforms.
I didn't really use the chat or git functions.
The moment of Zoom acquisition, I barely used Keybase and started to research about alternatives, but I haven't found one that can match or have better quality than Keybase if ever I've found something good I might say goodbye to Keybase.
You can still use the platform if you find it really useful specially the Chat feature of it, but always be cautious. I personally have doubts and trust issues towards Zoom.
Keybase is great, but a full alternative is needed.
Related Topics/Links that might interest you:
[Zoom Acquires Keybase](https://news.ycombinator.com/item?id=23102430)
[Ask HN: Keybase Alternatives?](https://news.ycombinator.com/item?id=23103386)
[Open source the server components of Keybase #24105](https://github.com/keybase/client/issues/24105)
It was my understanding that Keybase worked like this:
- Alice has a bunch of online accounts (Github, Twitter, LinkedIn, etc)
- Alice wants someone who knows one of her accounts to be able to find all of her accounts
- Alice generates a keypair (public P, private K)
- Alice posts a certain kind of message signed by K somewhere on her public Github, public Twitter, and public LinkedIn
- Alice uploads the account ID's to a server owned by Keybase
- Keybase verifies the signed messages and keeps track of who has what accounts on what platforms
- Bob asks the Keybase server who Alice (Twitter ID) is on other platforms
- Keybase server returns a list of accounts
- Bob can verify the returned list of accounts by running the signature verification himself (i.e. the Keybase server is only used for finding the account names)
Basically in my mind, "Keybase == Somebody duct-taped together a PKI using public posts on social platforms."
But reading others' comments I see that:
- Keybase is a chat
- Keybase is a better Git than Git
- Keybase shares files securely
- Keybase is a crypto wallet
- Keybase is the KBFS filesystem
- Keybase is easy for non-technical users to use
Did I miss something?
I miss encrypted git repos a lot, nothing else I’ve found is as reliable or easy.
I think one way could be the right legal language in an open source license (e.g. acquisition triggers the duty to open source).
Honestly being owned by Zoom makes them more trustworthy at this point, they don't need to scrape the bottom of the barrel to pay for it
In the early days it seemed like a good idea, to improve upon the ideas of the web of trust, but they seem to be a company who lost their way.