4 ms·
Yes: that it's not universally accepted (yet?) and highly volatile.
by mathgeek 9y ago
Yes: that it's not universally accepted (yet?) and highly volatile.
- dawhizkid 9y agoDo you mean universally accepted as payment for goods? The fact that it takes an hour+ for one transaction to be confirmed (i.e. behind ~6 blocks) seems like a huge UX problem for buyers/sellers. If I'm going to use bitcoin to buy a TV, I'm not going to wait around an hour for the merchant to say I'm good to go. And transaction fees are so high now that it makes no sense to use bitcoin for small transactions. Other issues include refunds/disputes. If you receive something in the mail from a seller that is not as described, then no equivalent of a dispute/chargeback process to get your bitcoin back.
- Lerc 9y agoThe problem is this perspective mixes levels of certainty that are not comparable. A trader can accept a transaction as soon as it shows up, even with zero confirmations. Doing so creates the risk of not actually receiving the payment, The circumstances under which this can happen are probably much rarer than the possibility of a credit-card transaction being reversed. You want the high certainty on large sums, For small payments you could easily just take the hit on the tiny (if any, in-reality given the minimal reward for effort of fraud) transactions that don't go through. The issue of transaction charges is a growth pain. When Bitcoin came on the scene, It was expected that the future incarnations of the protocol would handle vastly larger numbers of transactions, that view hasn't really changed. The only issue has been on the time-frame of the expansion. If anything that can be considered as being too successful. It's a little system that might be a big system one day, but people anticipating the future state of the big system is causing strain. It will be interesting to see where things stand 30 years from now.
- eric_h 9y ago> Doing so creates the risk of not actually receiving the payment, The circumstances under which this can happen are probably much rarer than the possibility of a credit-card transaction being reversed. When everyone and their mother is using bitcoin as currency, I'd agree with this statement. For now, and for a reasonable short term timeline (~1-5 yrs?), I'd argue that the probability of an immediately accepted bitcoin transaction being a fraudulent one are much higher, simply because there are more opportunists as a percentage of total 'users' of bitcoin than there are in a more established market. Of course most of the new money that's come into the bitcoin market (from the say ~$2k/BTC - $16k/BTC time period) isn't using it as a currency, they're speculating on it growing in value even more. > It will be interesting to see where things stand 30 years from now. It most certainly will. Personally, I don't think bitcoin is going to be the go-to cryptocurrency of 30 years from now, but who knows. I missed my chance at being a bitcoin billionaire already(could have stashed away a few 1000s of bitcoin back when it was well under $1), so I'm not playing the speculation game anymore (I sold what I had at under $10k/BTC) - I hope everyone on here who is holding onto loads of BTC cashes out before it crashes, or that it doesn't crash and actually becomes useful currency! I'm just going to keep putting money into my 401k.