News for hackers. Stuff that matters.
Re:Subspace: A messaging layer for Bitcoin by chrisdew on Tuesday March 10, 2015 @06:40AM (#9176985)
There may be a problem with the "prefix length" parameter.

What the user will want to specify is not the proportion off all messages they want to download, but the number or bandwidth.

i.e. If subspace has 1,024 users, I might specify a prefix which lets me download ~128 users messages to find my own.

When subspace becomes more popular and has 2^20 users, I don't want to download a thousand times more messages.

The prefix needs to be adjustable, or auto-adjusted with regard to the total number of users.

Re:Subspace: A messaging layer for Bitcoin by kubindurion on Tuesday March 10, 2015 @06:13AM (#9176912)
You should take a look at Whisper (independent protocol being developed by Ethereum developers, can be used with Bitcoin).

https://github.com/ethereum/wiki/wiki/Whisper-Overview https://github.com/ethereum/wiki/wiki/Whisper-Wire-Protocol https://github.com/ethereum/wiki/wiki/Whisper

Re:Subspace: A messaging layer for Bitcoin by davidgerard on Tuesday March 10, 2015 @05:03AM (#9176726)
This is an excellent example of the difference between "interesting" and "useful". This should be every bit as successful in its niche as ssh over DNS (which is a thing you can actually do as well).

Seriously, this is a ridiculous stretch to find some use case - any use case! - for Bitcoin.

At least there won't be a spam problem, as you need to pay to get your transactions processed at all: https://blockchain.info/unconfirmed-transactions I wonder if one message would actually cost more than an SMS yet. SMS tends not to wait serious fractions of an hour, though.

Re:Subspace: A messaging layer for Bitcoin by mappu on Tuesday March 10, 2015 @06:09AM (#9176898)
The system described in the article doesn't use proof-of-work. There's no artificial transmission latency at all. It seems unrelated? to bitcoin aside from being inspired by it and borrowing some terminology.

It's a very cool protocol design, clearly a lot of thought has gone into the various tradeoffs, and i hope it sees some uptake.