Hacker News Reader

HNHacker News

Top New Best Ask Show Jobs

Comment by shuntress | Hacker News Reader

Parent Full thread

shuntress·

Git

View on HN

Am I misinterpreting you or are you saying Bitcoin would make a better, faster, cheaper Git?

If you are, I am already laughing.

They're saying git would not have been better or faster or cheaper if implemented in a classical centralized way.

That's the thing about blockchain/"distributed". They are such vauge terms they can apply or not apply to anything depending on what point you need to make in your argument.

If we compare the traffic of Github vs Bitcoin, Github is likely doing 1,000+ writes per second and Bitcoin is doing what, 5-7 maybe higher with specialized stuff?

Github is nowhere near the world's "central and only" service for Git, so what am I missing to not laugh about?

The downside of a global distributed database (no matter what) is the speed of light, if you need ordering in any transaction you are in trouble, and no classic service requires that for all transactions in its scope, we figured out partitions, row locks, and shards a long time ago.

Isn't git most of the time used centralized? And that offers better user experience than doing it some decentralized way? It seems to me like most prefer centralized use of git. Be it private server or some large server.

Git is nice distributed tech. It's permissioned, though. Good old permissioned distributed tech. Which predates Bitcoin (obviously, as git is older than Bitcoin).

Yeah, and I always say git with commit signing is a cryptographic block chain in the loosest sense. But in this context I was of course referring to the proof of work/stake BS. In git the proof of work is the work you put into writing the source code. There is actual value in it, not just fictional speculative value.

Reply on news.ycombinator.com