The 50% total hashing power - pooling flaw?

[qdot]

Hi!

Am I the only one concerned by the fact, that deepbit.net's contributors provide nearly 35% of network's hashing power?

By the very definition of pooling, the control server has all it needs to direct all it's peers to hash nearly anything, with no decentralized checks and balances to ensure the block chain it is directing it's computing capability in in fact an undoctored bitcoin broadcast.

[qdot]

Perhaps it's time to add another desired feature to the pooling software?

I believe it should be possible to design the software in such a way, that is to separate the 'shares of work' assignment/reporting (to the pooling server), and the 'block-chain' fetching (which ought to be performed through decentralized means).

Can it? I don't see a reason why not..

I'm not concerned about deepbit.net having a malicious intent, but we should safeguard against, say, domain takeover, server bugs etc?

[kjj]

http://forum.bitcoin.org/index.php?topic=9137.0