Quote from: d5000 on January 21, 2015, 04:39:23 PM

Peercoin seems to be switching to the BTC 0.8.6 codebase with the 0.5 release (see https://github.com/ppcoin/ppcoin/tree/develop and the pull requests). PPC 0.5 has important features regarding PoS (cold-locked minting, duplicate stake detection), but they will be irrelevant if SLM decides to take down PoS.

Perhaps it will be best if we decide first if we want to keep PoS (I don't want to take much part of this decision as I'm not a programmer) or not, because in the case we drop it perhaps it is better to rebase SLM on the latest Bitcoin.

I agree; whether on not PoS is kept will I think mainly determine which one we end up following.

Quote

@all: I have compiled a123's latest version on Linux and the brain wallet code added as last commit probably contains a bug, I cannot get it to compile (can anyone confirm?). The version inmediately prior to it from October 28 (0fc7bc779d82999c99cff44c970871da29c2df38) works fine.

I can't help you with that currently, as my build is based on an earlier (21 Oct) fork -- but I'll try merging in the changes and see what happens. I've currently got a successful build working for Windows; I still need to see if it will build for Linux too (and fix it when it doesn't....)

Btw, can you give details on the errors when you tried to compile? There seemed to be a couple of changes along with the new passphrase function, so it would help to know which bit is likely at fault.

Quote from: d5000 on January 24, 2015, 04:13:11 PM

From an "economical" perspective, I agree with you, PoS can make sense as a component of the economic model.

My favourite would be a mainly PoS/PoB model with PoW slowly declining in importance due to the difficulty model where PoW rewards decline when the hashrate/difficulty grows (I think that's the actual model, or at least it was the original one). What I don't know is if it is harder from the programming/stability perspective if the three models coexist. Now without PoS at least my client works very fine.

I also would like to keep the 3-way generation if possible, but I need to make sure I understand what the issue with PoS was first. If I've got it right, the problem was that stake transactions would end up creating a whole bunch of small credits, and then each one of those would be a candidate for future staking which led to too much processing (searching a huge stack of transactions constantly). If that's the case, then I'll need to understand better how staking actually works to know whether it's fixable or not.

I'm not at all familiar with Peercoin, but I can't help thinking -- did they have this problem? Did they solve it? Huh