Re:Show HN: A Cryptocurrency in 4K Lines of Scala Code
by aaron-lebo on Monday June 29, 2015 @07:55PM
(#9801981)
I love the thought process behind this.
For those who aren't familiar with C++ or cryptocurrencies in general, Bitcoin and similar codebases are daunting to get into. This reminds me of BasicCoin [1], also featured on HN [2].
Re:Show HN: A Cryptocurrency in 4K Lines of Scala Code
by kushti on Tuesday June 30, 2015 @03:34AM
(#9803561)
Curve25519 implementation was taken from the internet. The problem I see with all the Java implementations of it, they're not reviewed by professional cryptographers. This is one the reasons why Scorex isn't production-ready.
RIPEMD implementation was taken from the internet too.
Re:Show HN: A Cryptocurrency in 4K Lines of Scala Code
by j-m-o on Tuesday June 30, 2015 @08:22AM
(#9804653)
That Curve25519 implementation is by Moxie Marlinspike, and that RipeMD160 algorithm is built into BouncyCastle.
There may still be attack vectors on them, but that's about as close to "reviewed by professional cryptographers" as you're likely to get at this point.
Re:Show HN: A Cryptocurrency in 4K Lines of Scala Code
by sscarduzio on Tuesday June 30, 2015 @03:49AM
(#9803616)
This could become much smaller in Scala if you design it using FP style. Now it's more like decaf Java to me! Great project, thanks for making it available, now I have no excuse to understand how a crypto currency actually works.
Re:Show HN: A Cryptocurrency in 4K Lines of Scala Code
by kushti on Tuesday June 30, 2015 @04:09AM
(#9803685)
That's very true. The reason of that is the code was rewritten from Qora cryptocurrency source code(24K LOCs in Java). I will definitely make it more FP-like.
For those who aren't familiar with C++ or cryptocurrencies in general, Bitcoin and similar codebases are daunting to get into. This reminds me of BasicCoin [1], also featured on HN [2].
1. https://github.com/zack-bitcoin/basiccoin
2. https://news.ycombinator.com/item?id=7713795
Any particular reason you chose to roll your crypto rather than build on top of existing implementations?
https://github.com/WhisperSystems/curve25519-java
http://rosettacode.org/wiki/RIPEMD-160#Java
RIPEMD implementation was taken from the internet too.
There may still be attack vectors on them, but that's about as close to "reviewed by professional cryptographers" as you're likely to get at this point.