"It took 40 minutes to catch up to the blockchain before he headed over to the cafe. And now we have to wait a few more minutes for the laptop to synch up again on the wireless network at the cafe."
The fact that bitcoin.org told people to install the slow-as-molasses Bitcoin-QT software was a big PR mistake. Fortunately they're now recommending fixed software that can sync in a reasonable time.
The Electrum Bitcoin Client doesn't download the blockchain. It uses remote servers to verify transactions. Useful if you need to be up and running quickly.
I think the parent and grandparent are missing the big picture. Having a verifiable digital currency with no central authority was a massive accomplishment, and PR and how fast things sync were (rightly) secondary concerns at the time.
I was using "verifiable" to stand in for "you can perform a full audit of the blockchain, therefore no trust is required." It was a sloppy use of the word. I was not expecting you to come along, betterunix.
We've had lengthy discussions of this before. Where we last left off, you asked me to explain why bitcoin doesn't need "security assumptions" of the kind listed in a sample crypto paper that you provided. That's a fair question I guess, although I suspect it's ultimately based on some mistaken premise [1], but I haven't had time to read the paper.
[1] Bitcoin is a new mechanism that approximates Byzantine consensus, though AFAIK nobody has written a formal paper explaining and proving that. That is why it works. You can formalize the properties it gives you in a similar way to the original Byzantine Generals work, except that each general's votes are proportional to his rate of sha-256 hashing, rather than being one vote each. Another thing that suggests you have a mistaken premise in your dismissals of Bitcoin is that Bitcoin has been working for years now, has received massive investment, and as far as we known has never been compromised in any way.
as far as we known has never been compromised in any way.
Bitcoin suffered a successful double-spend attack [+] during a fork of the blockchain during this March, which necessitated a drop-everything-and-downgrade-your-software response from a cartel of miners to prevent us from having Bitcoin and Bitcoin Prime.
+ It was by a security researcher, not by an adversary, so the community has mostly forgotten it and those that remember would like to pretend it never happened.
You are right to call me on this and given a probably-not-too-reasonable definition of "compromise" for Bitcoin, you are technically correct.
But it's like getting a tour of the Tesla factory and saying, "It can't even run under water? Well gosh, this sucks!"
Blockchain forks are _going_ to happen, because software isn't perfect. That has long been known. It's always been a question of when, not if.
But they do not matter, because it's easy to heal them in precisely the way that this one was healed. Which is part of the beauty of bitcoin.
They also don't matter much because they don't have much of an impact on users unless you're doing something you shouldn't do.
> so the community has mostly forgotten it and those that remember would like to pretend it never happened.
That's pretty cynical. I'd say the community has mostly forgotten it because it wasn't a surprise to anyone in the community that it could happen, it was just a matter of when.
That the whole incident was pretty forgettable is a testament to how easy it is to heal a blockchain fork and how little impact one actually has.
They also don't matter much because they don't have much of an impact on users unless you're doing something you shouldn't do.
This is money we're talking about. If someone finds a way to make millions of euro by "doing something you shouldn't do", by god, someone'll do it. So then what happens?
The most realistic scenario I can think of where that would be possible is if a kidnapper demands that you pay them some huge quantity. You both meet up, you send them the transaction, they drop off the kid and leave, and then it turns out to have been a blockchain fork and you keep all the money.
That is NOT how real business is done.
When you are engaging in trade with someone, someone has to trust the other party, or you have to use escrow. Expecially if a ton of money is involved.
This is why I think wallets like Electrum (or maybe even online wallets) are more practical for a majority of people. The blockchain necessary for Bitcoin-Qt, MultiBit, et al., is currently [over 10 GB][1] and is growing rapidly. It is clearly not ideal for everyone to sync the blockchain, especially for new users that have to download it from scratch.
You don't need to rely on a server. With Simplified Payment Verification (section 8 of the bitcoin paper[1][2]) you can verify payments by querying untrusted peers, without downloading the full blockchain.
Bitcoin Wallet[3] for Android and Multibit[4] do this, which is probably why they are now the recommended clients and Electrum is not.
Edit: apparently Electrum also double-checks what the server is saying with SPV. I'm a bit fuzzy on why it still needs a server.
The right way to do in-person exchanges is to use an escrow service like LocalBitcoins. Zero confirmation transactions are pretty easy to double-spend (and could be even easier in the future [1])
The seller funds the transaction in the escrow service before meeting the buyer. The buyer sees it's funded and can write down a confirmation code that the seller doesn't know. Then when they meet the seller counts the cash and sends a SMS to LocalBitcoins to release the funds. LocalBitcoins responds with the confirmation code which the buyer can check then hand over the cash. No one needs any computers, only the seller needs a phone.
I tried it once and I felt like an arms dealer or something.
[1] edit: IIRC there's some proposal to be able to override unconfirmed transactions to recover coins in limbo due to no transaction fees being included, or something.
There's something about "the future" that brings us closer to film-like situations in real life. Not only the obvious sci-fi ones. Maybe it is the speed at which things are done, don't know...
What, the one on the first page? I don't think it's very problematic. It's belied by the more or less fair actual reporting.
Furthermore, I think that people who would immediately conceive of Bitcoin enthusiasts as "anti-government hippie-hacker drugdealers," as well as those people who would similarly conceive of any of those things as inherently pejorative, already have their minds made up and don't need help from the supremely subtle implications of a photo.
Oh geez,the first page has a pic of a mac laptop with a QR code sticker. I just had a hipster gag reflex but thankfully my fresh kombucha is calming my upset stomach.
I don't understand how this is supposed to work. When I installed the standard Qt client for the first time, it took a couple of days before it had downloaded the blockchain and was ready for use. What moron thought this was going to scale?
The article even points out that Litecoin clients need several hours to catch up with that currency's version of a blockchain.
Is someone ever going to build a true P2P cryptocurrency, or are all of the current contenders doomed to collapse under their own mathematical weight?
Buying in person links your address to your physical person, getting good images, possibly your address by following you, even perhaps getting a DNA sample. That seems somewhat traceable.
It seems much better to send cash via mail, much more anonymously (watch out for cameras near post boxes and use gloves).
34 comments
[ 5.3 ms ] story [ 89.3 ms ] threadSounds like a great way to spend money...
http://electrum.org/index.html
What exactly do you mean by "verifiable" here? What is being verified?
http://en.wikipedia.org/wiki/Protocol_of_Bitcoin#Payment_ver...
We've had lengthy discussions of this before. Where we last left off, you asked me to explain why bitcoin doesn't need "security assumptions" of the kind listed in a sample crypto paper that you provided. That's a fair question I guess, although I suspect it's ultimately based on some mistaken premise [1], but I haven't had time to read the paper.
[1] Bitcoin is a new mechanism that approximates Byzantine consensus, though AFAIK nobody has written a formal paper explaining and proving that. That is why it works. You can formalize the properties it gives you in a similar way to the original Byzantine Generals work, except that each general's votes are proportional to his rate of sha-256 hashing, rather than being one vote each. Another thing that suggests you have a mistaken premise in your dismissals of Bitcoin is that Bitcoin has been working for years now, has received massive investment, and as far as we known has never been compromised in any way.
Bitcoin suffered a successful double-spend attack [+] during a fork of the blockchain during this March, which necessitated a drop-everything-and-downgrade-your-software response from a cartel of miners to prevent us from having Bitcoin and Bitcoin Prime.
+ It was by a security researcher, not by an adversary, so the community has mostly forgotten it and those that remember would like to pretend it never happened.
But it's like getting a tour of the Tesla factory and saying, "It can't even run under water? Well gosh, this sucks!"
Blockchain forks are _going_ to happen, because software isn't perfect. That has long been known. It's always been a question of when, not if.
But they do not matter, because it's easy to heal them in precisely the way that this one was healed. Which is part of the beauty of bitcoin.
They also don't matter much because they don't have much of an impact on users unless you're doing something you shouldn't do.
> so the community has mostly forgotten it and those that remember would like to pretend it never happened.
That's pretty cynical. I'd say the community has mostly forgotten it because it wasn't a surprise to anyone in the community that it could happen, it was just a matter of when.
That the whole incident was pretty forgettable is a testament to how easy it is to heal a blockchain fork and how little impact one actually has.
This is money we're talking about. If someone finds a way to make millions of euro by "doing something you shouldn't do", by god, someone'll do it. So then what happens?
That is NOT how real business is done.
When you are engaging in trade with someone, someone has to trust the other party, or you have to use escrow. Expecially if a ton of money is involved.
[1]: http://allchains.info/
Bitcoin Wallet[3] for Android and Multibit[4] do this, which is probably why they are now the recommended clients and Electrum is not.
Edit: apparently Electrum also double-checks what the server is saying with SPV. I'm a bit fuzzy on why it still needs a server.
[1] http://bitcoin.org/bitcoin.pdf
[2] https://en.bitcoin.it/wiki/Thin_Client_Security#Header-Only_...
[3] https://play.google.com/store/apps/details?id=de.schildbach....
[4] https://multibit.org
The seller funds the transaction in the escrow service before meeting the buyer. The buyer sees it's funded and can write down a confirmation code that the seller doesn't know. Then when they meet the seller counts the cash and sends a SMS to LocalBitcoins to release the funds. LocalBitcoins responds with the confirmation code which the buyer can check then hand over the cash. No one needs any computers, only the seller needs a phone.
I tried it once and I felt like an arms dealer or something.
[1] edit: IIRC there's some proposal to be able to override unconfirmed transactions to recover coins in limbo due to no transaction fees being included, or something.
There's something about "the future" that brings us closer to film-like situations in real life. Not only the obvious sci-fi ones. Maybe it is the speed at which things are done, don't know...
Furthermore, I think that people who would immediately conceive of Bitcoin enthusiasts as "anti-government hippie-hacker drugdealers," as well as those people who would similarly conceive of any of those things as inherently pejorative, already have their minds made up and don't need help from the supremely subtle implications of a photo.
But haters gonna hate.
The article even points out that Litecoin clients need several hours to catch up with that currency's version of a blockchain.
Is someone ever going to build a true P2P cryptocurrency, or are all of the current contenders doomed to collapse under their own mathematical weight?
It seems much better to send cash via mail, much more anonymously (watch out for cameras near post boxes and use gloves).