Yeah, these kids will go far:
Not the Java crap that NXT uses (with all the code in one file lol)
thanks for telling me about this. i definitely don't want to put all of my money in 1 basket so if other talented developers surface with the intention of copying the ideas of nxt with all new code ill definitely be investing in that. its the ideas that are important not the specific implementation. more competition in this space is definitely good.
with that being said i already see some evidence in their post indicating that they may not really understand how decentralized networks work. anyway its all going to come down to how well they can answer my questions.
First we need as solid specification.
Then we need run some simulations.
The current situation is that all the fixes are like duct tape holding together an implementation that was written in haste.
Let's for second imagine a better solution that has
(1) Addresses that Bitcoin addresses with checksums. The current implementation is a java long which is only 2^63. Clearly that is too small and is hackable via brute force GPU systems.
(2) Using arbitrary-precision numbers in the code rather than using a long to represent a number.
(3) A more secure wallet, where wallet access is not exposed to the internet.
(4) Code that come with tests to verify functionality and perform regression tests.
(5) An open code review, non of this funny business where code has features removed and have supposedly 'injected bugs'.
(6) Standard JSON interfaces to allow other exchanges to easily integrate the coin.
.... ( there are a boat load of other improvements that tells me that everything needs a re-boot )
http://bitcoin-ng.boards.net/
