> >The transaction costs are a tiny fraction of what a bank or credit card processor charges to send Monopoly bank money.
> Sure, as long as you're buying a car. Of course, you also have to find someone selling a car for bitcoin... Now tell me what it costs to buy your groceries, go to the movies, or get a cup of coffee?
Bitcoin is still generally cheaper for purchases over $30 to pay in Bitcoin even with rising transaction costs. Transactions are getting more expensive for more limited purchases, that's true, due to the limits. But there is a lot of interesting engineering going on to fix this problem, such as the lightning network, which should be able to reduce the cost of transaction fees and allow micropayments for fractions of a cent of transaction fees. Other tech on the horizon includes sidechains, where you could use Bitcoins on a network with a different security model; the more expensive high security chain is not required for day-to-day small purchases, and could ultimately be rolled up into the main chain through a sidechain-to-main transaction, or even a lightning network bridge.
>And you keep your coins in cold storage.
>> Uh-huh. And your average person keeps their keys where? On a computer, because nobody's going to write their access keys down on paper and lock them in a fire safe. In fact, they'll keep their keys in a wallet on their smartphone, where they'll be stolen by hackers. Or they'll use a web wallet, where - in the event they're not defrauded by the wallet provider - a keylogger steals their access codes. Hence the very popular Bitcoin phrase "Sorry for your loss".
There are dedicated hardware devices now that make cold storage very easy and highly protected such as the Ledger and the Trezor. If Bitcoin or other cryptocurrencies get more popular, there's not reason why this type of technology couldn't be implemented safely directly into phones as an independent chip isolated from the rest of the operating system. That would give the benefits of very high security with simple user interfaces for users. There could even be schemes for recovery with multi-factor authentication, key sharding schemes, etc. These are still fairly technical, but it's easy to imagine a world where this is a lot simpler for end users. The current hardware wallets are already light years an improvement over your "write on paper" suggestion.
>>It's a technologically better form of money, far more efficient for moving large sums across borders.
>If you find someone willing to accept it on the far side.
Adoption is continuing to increase. It's small now, but growing relatively quickly.
>>imagine what it will be like in a few years when millions more people grok the tech.
>If millions more ADOPTED bitcoin, it'd crash instantly, since it's totally unscalable.
There are a ton of improvements that have been made and are continuing to be made to increase the scale as the demand is growing. For example, the latest release of Bitcoin Core that came out just yesterday includes a 40%-50% speed-up for sync, significantly reduces memory requirements, and disk space. Another recent improvement, Segwit, increased transaction capacity by at least double while also providing the ability for more future upgrades. Many improvements are on the way as already mentioned for the volume of transactions as well, such as the lightning network, sidechains, etc. There is also a lot of experimental engineering being done that could help scale such as proposals like Mimblewimble.
There are interesting things going on with using bloom filters, and the recently released FIBRE network that is some extremely cool computer science engineering to pass Bitcoin blocks at nearly the speed of light to reduce orphans. There is so much interesting computer science and cryptographic research going on that it's astounding. A lot of this revolves around improving scale. So on a tech site like Slashdot, you probably should keep an open mind about it, read about some of the really neat developments in the pipeline, and not just dismiss it.