This only works between Coinbase accounts, which kind of defeats purpose of Bitcoin - because it requires both parties to be tied to a single company. Thus I don't see anything revolutionary about this - many Bitcoin services already implement that feature for transfers between their accounts.
What I want to see (and I will try to develop in my upcoming service) is a protocol that lets you transfer Bitcoins between different services instantly. Right now it takes 6 confirmations in most services to receive a payment, but with a mutual trust between them the confirmation time could be essentially cut to zero. Mt.Gox tried to do something like that with their "Green addresses", but it didn't catch on.
I think that if Bitcoin is going to succeed it will be because of bank-like institutions using Bitcoin as the underlying protocol that the average consumer doesn't interact with -- but could if they want to, so this is a fantastic first step. At present money in the bank is not "money" "in" the "bank", it's an entry in a database that is reconciled every so often, Bitcoin should be used in a similar way for it to be compatible with current consumer behaviours -- expectations of protections against fraud (chargebacks), human error etc.
It's the only way Bitcoin can scale to major-currency levels, since every on-chain transaction is seen by every full peer.
If you have a bunch of "banks" who provide checking accounts, they can use on-chain transactions just for periodic settlements. "In the past day you sent me checks worth 10,000 bitcoins, and I sent you checks worth 11,000 million bitcoins, so now I'm sending you 1000 bitcoins."
In ancient times before banks, people paid in gold or silver, but buying something like one apple that cost less the smallest coin presented a problem. The solution was, people left their gold or silver with a gold smith, and were given small denomination notes to spend at the market. Merchants could use the notes to redeem the gold at the end of the day.
It is fun to see 1000 year old ideas getting re-created for the modern world.
The problem with sending micro payments in Bitcoin is not that you can't divide it into smaller coins, but the fact that unless you pay a transaction fee, your transaction may not be included in a block (e.g. confirmed) for a rather long time. Obviously, if the sum of the transaction approaches the amount of fee, it becomes expensive.
Coinbase solves this by having its own DB of transactions and not performing real Bitcoin transactions between its users.
I'm not quite sure what do you mean by saying "losing everything". It's not like anyone ties you to Coinbase. Have your own Desktop wallet for large sums and transactions, use Coinbase to handle small ones.
Isn't it? I have made a habit of finding such items and/or predicting them (with varying success, of course). It's amazing how much stays the same when everything changes.
Coinbase has been doing cool stuff recently (fancy new hire, css tweaks, this story) but their base product needs work, and their support is weak.
I have yet to receive responses to two support emails I sent several months ago, one of which regarded a transaction that disappeared from my account, leaving me with less BTC than I had before.
Also, the last time I checked, their API callbacks were simply not functioning. No answer from support.
Furthermore, their transaction confirmation emails often send days late (and sometimes repeatedly, I think). It's very startling to see that you "just sent" x btc, forcing you to go through your transaction history on the site.
There are other small bugs/issues I have with the product as well, none of which seem like they'd be too hard to fix.
I hope they hire support and fix those bugs before they add the next cool features / hires.
This is a great first step. The next one would be taking transactions between large wallet providers and batching them up to reduce those fees as well.
If you and I do business 100 times a day, we could either send 100 transactions through the system and incur 100 transaction fees, or do a little bit of accounting and only incur one. Using an API / trusted party like coinbase to do the accounting doesn't undermine bitcoin in any way, as it would still be needed for the end of day reconciliation (moreover, the accounting wouldn't make any sense without it.)
Scott McCloud's written a lot of real intelligent stuff about microtransaction/digital currency programs, and implicit in that, why they haven't worked out in the past. I'm hesitant to say bitcoin is The Answer, but it (combined with some other qualities of the internet and modern UX/UI) presents a plausible implementation.
To paint with a broad brush: cost is both financial and temporal. A successful digital payment/microtransaction system has to reflect this fact.
The time cost of paying some amount of money should not outstrip its monetary value. Having six options (or worse yet, having to type in an amount) for payment size and three confirmation screens is infuriating when you want to transmit the digital equivalent of a quarter or so.
My two satoshis: I'm not sure how it will ultimately be implemented, but I think bitcoin (and its subdivided units) is a good way to build low-friction payment schemes.
Oh hells yes. I've been waiting for something like this to make http://www.bitbanter.com/ more interesting. Not sure how I'll handle non-Coinbase microtransactions...but still, this is really, really cool. Thanks Coinbasers!
Cool blog post but I can't get excited about "micro-transactions" when we've heard this same story so many times since the 90s. Zzzzzz. Few weeks of hype (1 cent to read my blog = $120,000!?!) and it never materializes.
This idea of money as simply electronically stored value was actually pushed hard by Dee Hock back in the 70s. He even talked about the ability of someone with a secure network or transactional platform to create new currencies. Hard to do it though :)
(Dee Hock was the "founder"/driving force behind Visa btw).
Well, I would say one large reason that micro-transactions haven't taken off is they are so bloody difficult to implement. Basically you are always dependent on someone's payment network, which usually is not geared towards supporting micro-payments. Take Paypal's fee structure for micro-payments, $0.10 + 10% -- basically kills the idea in the womb.
Also, the idea of paying one penny to read an article might not make sense if it is user initiated, but it could very easily happen if you had an "drip" effect from a connected faucet (i.e. account of virtual currency). I'd be more than happy to pay $.05 for each NYTimes article I actually read, but I have no desire to pay a monthly fee for the Times. If I could have one account that worked across a bunch of different websites and automatically implemented the "drip," I'd consider that a major improvement.
Personally my bet is on other virtual currencies as opposed to Bitcoin (i.e. my own), since the Blockchain stuff ends up being your enemy rather than your friend.
Just like the Zero fees, this has zero thing to do with bitcoin.
Here I go, party and counter-party are in my table and I flip a bit and it ZERO fees! wohooo!
It's implemented by just flipping bits in their database - no Bitcoin transactions or cryptographic functions are used here. I'm don't really understand what's awesome about that - many Bitcoin services already support off-chain transfers between its users.
> Bitcoin is an incredibly efficient protocol for moving money with low fees. However, it does have a “miner fee" ($0.05 - $0.01 at current exchange rates) if you want your transaction to be sent and confirmed quickly.
this look like a trivial step.. but the vision explained in the article are right, and this is quite revolutionary considering we could get into this reality (with widespread adoption of bitcoin)
think about the fluidity of value.. if people could use little cents, imagine what this can cause to donations.. for instance.. open source world..
things will get much more fluid, easy and fast.. money will flow not just for the big guy, that can pay for the big schemes.. but to everybody, anywhere..
this is indeed, a trivial step, with unprecedent consequences.. revolutionary "triviality" :)
The bitcoin protocol doesn't (on purpose) have a way to reverse transactions. Unless coinbase isn't actually moving money, it won't be able to reverse anything. Since they have their own database, though, they could provide chargebacks for any transactions they haven't committed to the blockchain, though I don't know if I'd want to use that as a merchant.
Right; bitcoin does not, but Coinbase is just flipping bits in their database, since these transactions are off-chain. So, just a question of whether or not they support chargebacks, and to what extent. The issue will come up sooner or later, if Coinbase sees more mainstream usage.
In which case people who has compromised accounts wouldn't send it within coinbase's system, they'd sent it on the real network where it can't be reversed.
What I want to see (and I will try to develop in my upcoming service) is a protocol that lets you transfer Bitcoins between different services instantly. Right now it takes 6 confirmations in most services to receive a payment, but with a mutual trust between them the confirmation time could be essentially cut to zero. Mt.Gox tried to do something like that with their "Green addresses", but it didn't catch on.