Miners: Time to deprioritise/filter address reuse!

[franky1]

You're an idiot, don't do this!   - 59 (36.6%)
I don't like this, but I agree we need to move forward with it.   - 14 (8.7%)
We should have waited longer, but I guess it needs to move forward now.   - 19 (11.8%)
Great, it's about time!   - 23 (14.3%)
You're a hero, let's get this deployed everywhere ASAP!   - 35 (21.7%)
If it's from Luke, it can't be any good.   - 11 (6.8%)

I read the poll more along these lines. Opposed: 43.2% In favour with varying degrees of support: 56.5% Rounding 0.3%. There is a fair degree of support for this. We must also keep in mind the following

i coloured it because for instance
saying "i dont like this" 14 (8.7%) means literally i dont like this.. in other words NO
saying "we should have waited longer" 19 (11.8%) means literally i dont want this to be done yet.. in other words NOT YET

but both questions are worded to then subtly suggest they agree to implement it. these type of questions are called trick questions.

EG if i said "do you hate termites" and you chose
1) i dont like them, but i agree we to live along side them as they are living creatures

great using that answer, you have agreed to allow me to deliver 500,000 termite larvae to your house as you seem to be ok with it.

[Carlton Banks]

What about donation addresses?

Address Chains, forthcoming in the new version of Bitcoin Qt.

Is it actually being implemented or one of those things that has been sidelined?

BIP list

No. 32

Status is "Accepted", Peter Wuille is the author, it constantly gets talked about as if it is happening, and the (likely) first hardware wallet to market is going to be implementing it. It might actually be the first wallet to implement it period, seeing as the new version of Qt client is currently incubating, no Release Candidates as of yet.

[BitThink]

Apparently no restriction should be done before the clients supporting convenient solutions to avoid address reusing. Otherwise you are trying to kill BTC rather than helping. Please spend more efforts on clients instead of mining softwares. No project will succeed in going to mainstream try to piss off users just for pleasing some genious developers who think they have better vision of the future.

BTC now is no longer the toy of developers as the early days. You are dealing with billions of people's money now. Remember even your decision is correct in the long term, it's side effect will do great harm if not being careful enough. Now it's too dangerous to do something like: just do it and let's see what will happen.

[bee7]

Gavin A proposes adding API callbacks into the bitcoin URI so that retailers do not have to use unique addresses per customer, but using one address and be able to track which TXID belongs to which customer by use of the API callback.

Could you provide a link, please.

[Carlton Banks]

Apparently no restriction should be done before the clients supporting convenient solutions to avoid address reusing. Otherwise you are trying to kill BTC rather than helping. Please spend more efforts on clients instead of mining softwares. No project will succeed in going to mainstream try to piss off users just for pleasing some genious developers who think they have better vision of the future. BTC now is no longer the toy of developers as the early days. You are dealing with billions of people's money now.

This patch isn't designed to restrict, it is designed to discourage.

[killerstorm]

Thin wallets like Electrum and blockchain.info retrieve UTXOs from server, and cost for server is roughly proportional to number of addresses used, so address re-use improves efficiency. Are you against that?

[samurai1200]

 So let me get this right: A mining pool operator implemented a change that, with (wider) acceptance, would delay payments to his own miners? Should his miners change their payout addresses every single mined block?

BIP_32 is not in effect, so that argument is currently moot.

[justusranvier]

Thin wallets like Electrum and blockchain.info retrieve UTXOs from server, and cost for server is roughly proportional to number of addresses used, so address re-use improves efficiency. Are you against that?

Are you saying that address reuse makes the UTXO set smaller?

[Luke-Jr]

So let me get this right: A mining pool operator implemented a change that, with (wider) acceptance, would delay payments to his own miners?

This isn't correct.
If a mining pool paid people more than once per block, there's something more subtle wrong...

Should his miners change their payout addresses every single mined block?

They should use Bitcoin recurring invoices (BIP32), not addresses.

BIP_32 is not in effect, so that argument is currently moot.

Bitcoin is still beta and incomplete.

[Luke-Jr]

Thin wallets like Electrum and blockchain.info retrieve UTXOs from server, and cost for server is roughly proportional to number of addresses used, so address re-use improves efficiency. Are you against that?

Are you saying that address reuse makes the UTXO set smaller?

I think he's saying that these clients are missing bloom filter support, which has been standard for a while.

[Impaler]

I've argued with Luke before on this obsession of his, your never going to go mainstream with something that dose not give people DURABLE identities.  Disposable identities are never going to fly in mainstream commerce where law enforcement must be conducted.  Also you have the grandmother factor at work here, a users needs a durable identity to give to other people and they can barely handle the technical hurdles of BTC as it is now.  This kind of secrecy and techno-elitist obsession will always fail.

[killerstorm]

Thin wallets like Electrum and blockchain.info retrieve UTXOs from server, and cost for server is roughly proportional to number of addresses used, so address re-use improves efficiency. Are you against that?

Are you saying that address reuse makes the UTXO set smaller?

No. They ask server about each address they use. If client uses N addresses, there are N queries against server's database.

[Luke-Jr]

I've argued with Luke before on this obsession of his, your never going to go mainstream with something that dose not give people DURABLE identities.  Disposable identities are never going to fly in mainstream commerce where law enforcement must be conducted.

Bitcoin does not provide identities of any sort.
If you need identities, use PGP.

Also you have the grandmother factor at work here, ...

Payment protocol should make things easier.

[killerstorm]

I've argued with Luke before on this obsession of his, your never going to go mainstream with something that dose not give people DURABLE identities.

Address isn't supposed to be an identity. It is more like an invoice number.  Invoice numbers are not supposed to be durable, are they?

[justusranvier]

I've argued with Luke before on this obsession of his, your never going to go mainstream with something that dose not give people DURABLE identities.  Disposable identities are never going to fly in mainstream commerce where law enforcement must be conducted.  Also you have the grandmother factor at work here, a users needs a durable identity to give to other people and they can barely handle the technical hurdles of BTC as it is now.  This kind of secrecy and techno-elitist obsession will always fail.

Maybe you can use grandma as a hostage against privacy preservation today, but there are projects in the works to make anonymous commerce just as safe (or safer) as non-anonymous commerce, in a grandmother-friendly way.

After that's done, law enforcement can shove their desire to spy on everyone and everything up their ass.

[samurai1200]

Should his miners change their payout addresses every single mined block?

They should use Bitcoin recurring invoices (BIP32), not addresses.

Can a miner use a BIP32 recurring invoice as an Eligius login right now?

If not, perhaps your patch rollout timing was ill-informed.

[killerstorm]

I think he's saying that these clients are missing bloom filter support, which has been standard for a while.

Are you sure Bloom filter is the right solution? It creates more load on Bitcoin nodes than UTXOs-for-address queries would.

[mdopro1]

I have no flipping idea what anyone is talking about. Am I going to need to create new address everytime I get paid from my pool? Because that would be plain stupid. Please correct me in simple English because this is how I'm seeing it.

I recently lost all my coins because I created a new receiving address and didn't back it up. Wouldn't this create more similar problems?

I may be barking up the wrong tree here... but what gives? Bitcoin is fine the way it is.

[DeathAndTaxes]

the other thing to note is:

Gavin A proposes adding API callbacks into the bitcoin URI so that retailers do not have to use unique addresses per customer, but using one address and be able to track which TXID belongs to which customer by use of the API callback.

so we have Gavin A trying to reduce the need for unique addresses. and we have Luke Jr trying to increase the need for unique addresses.

can we find a room for the main dev's to go in and slug it out for a few matches as to something that can be agreed on that is best for the community..

Link please.  Nothing I have read about the payment protocol implies static addresses.

[Luke-Jr]

... there are projects in the works to make anonymous commerce just as safe (or safer) as non-anonymous commerce, in a grandmother-friendly way.

Bitcoin is never anonymous, no matter how you use it.

After that's done, law enforcement can shove their desire to spy on everyone and everything up their ass.

Law enforcement can do forensics the same way they do with cash (actually, they can do better with Bitcoin).
This only buys you privacy - without it, everyone would be able to see everything you do.

I think he's saying that these clients are missing bloom filter support, which has been standard for a while.

Are you sure Bloom filter is the right solution? It creates more load on Bitcoin nodes than UTXOs-for-address queries would.

"Lookup by address" is only less load than "lookup by bloom filter" if every node carries an address index (which adds to disk and memory requirements).
In any case, Bitcoin was designed with addresses being single-use, and it becomes unusable with too much reuse, so this side-topic is irrelevant.