Re: [bitcoin-dev] Revisiting NODE_BLOOM: Proposed BIP

[Matt Corallo <lf-lists@mattcorallo.com>]

I'll just quote what I said on github:

Neither this pull nor the BIP has any stated intention of phasing out
bloom filtering support in the protocol. As much as I'd love to, I 100%
agree with @mikehearn here, that would break any ability of SPV clients
to operate on the P2P network (either as a way to double-check
centralized servers, or otherwise), and that is really not a good idea
without a replacement in place. This pull/BIP DOES suggest we phase out
REQUIRED bloom filtering support in the protocol - thereby fixing the
peer selection of SPV clients in the face of btcd with some flags/many
patched versions of Core/etc peers, providing a remedy for a potential
DoS attack, etc.

Matt

On 08/24/15 15:19, Tom Harding via bitcoin-dev wrote:
> On 8/21/2015 3:06 PM, Peter Todd via bitcoin-dev wrote:
>> On Fri, Aug 21, 2015 at 05:55:58PM +0000, Matt Corallo wrote:
>>> Anyone have the best reference for the DoS issues?
>> Well actually, we can reference the DoS attacks that Bitcoin XT nodes
>> are undergoing right now - part of the attack is repeated Bloom filter
>> requests to soak up disk IO bandwidth.
> 
> So, to summarize, someone is attacking Mike Hearn's bitcoin fork. 
> Therefore, now is the perfect time to write a BIP and author changes
> that begin the process of dropping support for the most broadly
> successful class of wallets, which Mike Hearn's SPV client library enables.
> 
> 
> _______________________________________________
> bitcoin-dev mailing list
> bitcoin-dev@lists.linuxfoundation.org
> https://lists.linuxfoundation.org/mailman/listinfo/bitcoin-dev
>