Quote from: MoonShadow on December 22, 2013, 10:03:53 PM

Quote from: jinni on December 22, 2013, 09:13:45 PM

Quote from: taltamir on December 22, 2013, 08:59:35 PM

Quote from: Eri on December 22, 2013, 09:59:28 AM

1. Thats not what a 51% attack is and doesnt let you do that. do research.
2.bitcon isnt centralized, pools are false centralization. ie. if someone pulls some shady stuff, the people 'get out of the water' and they lose the hashrate and peoples faith in them, pool owner then loses their income stream. any other form of centralization is just as much BS. do research.
3.no it doesnt... do research...

3. If this was the case we would have prunning already

As far as I have understood, Gavin and the developers have promised that pruning will be ready "when it is necessary". I have no reason to doubt the developer team, but when is it really necessary to prune, at 1TB? Has anyone come up with a timeline? What proof is there that pruning will be possible when it really is necessary?

I recently inquired about this issue, and the general response was that pruning is already emplimented in the reference codebase, but isn't yet enabled for the offical client. This is mostly because of concern about what kind of impact widespread pruning would have on the "many copies keeps data safe" stragedy that the blockchain employs. Some of the developers don't believe that there are enough full clients at the moment, and that enabling pruning would result in a significant drop in the number of compete copies of the blockchain from which new clients could bootstrap. I'm not a programmer, but it was strongly implied that certain power users with the programming skills to enable this code themselves already do prune. So when they say that pruning will become available when it's necessary, they kind of mean that they will turn it on when they feel that the burden of running a full client is reducing the number of full clients, and that we are not to that point yet.

Sounds fair. So if we trust the developers on this, then scalability is pretty much solved?