Quote from: smartwart on February 05, 2016, 07:50:10 AM

Quote from: Come-from-Beyond on February 04, 2016, 04:52:22 PM

Hey, guys and gals. I'd like to get your feedback on the following:

In it's current implementation Iota's ledger is based on inputs and outputs like Bitcoin. There is another way - balances of accounts like in Nxt and Ethereum.

Now I see that if Iota used the latter it would be more efficient because:
1. No need to send the change back to myself which makes the tangle smaller
2. A lot of dust inputs could be spent with a single payment and this would be more secure because every address reuse leaks the private key
3. A new address wouldn't be needed for every incoming payment (this would make acceptance of Iota donations simple, in the current design it's PITA for humans)
4. Off-tangle payments would become simpler
5. RAM requirement for full nodes would be relaxed

The only problem that might arise in balance-based Iota is worse consensus convergence, but after analyzing the issue I don't see what could break.

I'm thinking if it's worth to do a little redesign that could take few days of extra work...

As stake holder I vote for waiting and implementing the best from beginning.
From my experience its mostly not a good idea to start quick and dirty and try to fix it after release.
And in distributed systems where each confirmed transaction matters, its even more critical not to break the past!

Anyone who need a working client now can try to play with alpha release.
We talk about days or maybe a week!?
IOTA, please don't satisfy the folks who want to see iota on the exchanges as soon as possible for quick profits.
Please develop with respect to sustainability.

And what is the best from beginning balances or input/output ?