The Microsoft Problem

[Blinken]

One problem that seems to be under-discussed is what I call "The Microsoft Problem". The essence of this problem is that the client controls the network. Right now everyone uses the Andresen client, but that might not be true in the future. For example, if Microsoft decides to write a Bitcoin client and ships it with every copy of Windows then their version of the client would become dominant. Sort of like Mosaic versus Netscape versus Internet Explorer. Once Microsoft's client controls the network they can implement whatever policies they want. These policies could deviate from the principles in The Paper. For example, they could decide to increase the number of bitcoins by increasing the Target.

Hmmm....

[1715581949]

1715581949

[1715581949]

1715581949

[Akka]

Not a real Problem. If someone would do this, this would immediately result in a Bitcoin fork. Therefore two Bitcoin Blockchains would exist in parallel. All Clients that use the Original Protocol would not accept transaction from Microsoft Clients and vice versa.

If one user would not agree with the Microsoft Change, he would simply have to switch his client.

[nebulus]

One problem that seems to be under-discussed is what I call "The Microsoft Problem". The essence of this problem is that the client controls the network. Right now everyone uses the Andresen client, but that might not be true in the future. For example, if Microsoft decides to write a Bitcoin client and ships it with every copy of Windows then their version of the client would become dominant. Sort of like Mosaic versus Netscape versus Internet Explorer. Once Microsoft's client controls the network they can implement whatever policies they want. These policies could deviate from the principles in The Paper. For example, they could decide to increase the number of bitcoins by increasing the Target.

Hmmm....

You are assuming that everyone who is currently holding/mining bitcoin will have no choice but to switch to Microsoft client. That's like saying that without Outlook, emailing is impossimple.

[WiW]

Not a real Problem. If someone would do this, this would immediately result in a Bitcoin fork. Therefore two Bitcoin Blockchains would exist in parallel. All Clients that use the Original Protocol would not accept transaction from Microsoft Clients and vice versa.

If one user would not agree with the Microsoft Change, he would simply have to switch his client.

This of course, assumes that everyone despises microsoft, and that everyone will not simply say, "okay, everyone I know uses MS-BTC anyway, so I'll just keep using that".

Kind of like when web developers went for "works best with IE" - traders may well go for "trades wtih MS-BTC". This is assuming someone like MS gets a strong enough hold on the client marketshare.

I think the best solution is that everyone understands that the maker of your client developer may have some degree of control over your trades, which means that you should only use a wallet application developed by someone you fully trust. This is where open source comes in to play.

Either you trust in competition (that there will be no monopoly), or open source has to win the race against all proprietary sources in making the best client software - so that many people will compile and distribute different clients based on the same code. It's a challenge for open source like never before.

[Gabi]

Just like Internet Explorer is the dominant browser today?

[WiW]

Just like Internet Explorer is the dominant browser today?

It doesn't need to hold the number one spot forever. Wouldn't an auto-updating client (or any piece of malware for that matter) running on more than 51% of the hashing power theoretically be able to perform a successful attack on the network?

[Akka]

Just like Internet Explorer is the dominant browser today?

It doesn't need to hold the number one spot forever. Wouldn't an auto-updating client (or any piece of malware for that matter) running on more than 51% of the hashing power theoretically be able to perform a successful attack on the network?

Now you are confusing a 51% attack, with a change in the protocol.

If an auto updating client would introduce a change in the protocol, it would imminently operate on a different network. On one site the Clients with the Update and on the other all different clients and the ones that wasn't updated. Even if over 50% update, there is no attacking the other network. This would be like Bitcoin attacking Litecoin.

If the update would just be introduced to not change the protocol, but to give some kind of remote control over the client (and miners), that could give them the capabilities for a 51% attack. But that would rise the question, why would a company that distributes the Client of a software with an over 50% Marked share destroy this software?

[WiW]

Now you are confusing a 51% attack, with a change in the protocol.

True, however the client in question could perform both.

why would a company that distributes the Client of a software with an over 50% Marked share destroy this software?

Why destroy the software? How about bending the protocol (or the bitcoins themselves, in the case of a 51% attack) to their benefit?

With control of the majority of the users, changing the protocol could mean introducing any form of tax they wish on the network, silently enough to cause enough confusion so that until the network reacts it's not worth switching back to the good protocol. I mean, so what if the new protocol now requires each transaction to give a 0.000000000000001% fee to MS? Are we all going to simply drop and forget all the transactions that were made until we realized this protocol change in order to go back to the original blockchain?

Imagine the chaotic undertaking for everyone to suddenly decide to cancel all transactions from the last 20 minutes and switch to other good clients building on the original blockchain because we just realized everyone's client is running on a new protocol...

[Gabi]

True, however the client in question could perform both.

Do you mean that this client would use the computers to mine for microsoft so that they could do a 51% attack? This is stupid, 1)because with asic mining with a computer will be useless 2)because no one will use it if it bring your gpu fan at 100% 3)because there are much much much easier ways to do a 51% attack than to make a client wich act more like a virus and so totally ruining their reputation.

[Akka]

Why destroy the software? How about bending the protocol (or the bitcoins themselves, in the case of a 51% attack) to their benefit?

A 51% attack gives you the ability to reverse / double spend any transaction made. This is a power a company could only use once.

On the other hand, people are still using PayPal.

With control of the majority of the users, changing the protocol could mean introducing any form of tax they wish on the network, silently enough to cause enough confusion so that until the network reacts it's not worth switching back to the good protocol. I mean, so what if the new protocol now requires each transaction to give a 0.000000000000001% fee to MS? Are we all going to simply drop and forget all the transactions that were made until we realized this protocol change in order to go back to the original blockchain?

Imagine the chaotic undertaking for everyone to suddenly decide to cancel all transactions from the last 20 minutes and switch to other good clients building on the original blockchain because we just realized everyone's client is running on a new protocol...

Any change in the protocol would (like I already said 2 Times) be detected immediately, there is no sneaking in. Even if they include only a 0.000000000000001% fee to MS, all Clients with the old protocol would instantly not be able to transact with the updated clients.

This would result in Chaos, of course. But just taking over the network wouldn't work this way.

[Gabi]

In his theory, no one would use the old client but they would use the new client since it's already installed

[WiW]

In his theory, no one would use the old client but they would use the new client since it's already installed

Bingo.

Any change in the protocol would (like I already said 2 Times) be detected immediately, there is no sneaking in. Even if they include only a 0.000000000000001% fee to MS, all Clients with the old protocol would instantly not be able to transact with the updated clients.

"Realizing" a change in protocol doesn't mean all transactions immediately stop. It means splitting the network. If you "realize" a that most of the world changed protocol (due to updating of the client software) you still can't change the fact that now the majority (it may be 50% and it may be 80%) of transactions will be made with the new protocol, and you're out of the loop - you're the minority building on the original blockchain.
If MS updated all their clients to work on a new protocol, and you're not in it, then you're on the original bitcoin network. But everyone you want to trade with is on the MS bitcoin network (your grocery store, your plumber, your babysitter, your car dealer - they are all using the MS client). So you either convince everyone to stop using their client/protocol (and hence cancel all transactions made until you managed to convince them) and move back to the original protocol using other clients.

This of course requires that MS program their client to manage to update in near perfect synchronicity or else, like you said, many transactions will be rejected.

This is stupid, 1)because with asic mining with a computer will be useless

Yeah... Didn't think of that. I'm assuming miners won't all be dumb enough to all use the same proprietary 3rd party auto updating software...