[Bitcoin-development] replace-by-fee v0.10.0rc4
alex.mizrahi at gmail.com
Thu Feb 12 14:32:26 UTC 2015
> "The approach" is how Bitcoin has always worked.
Mike, you're making "it worked before, and thus it will work in future"
kind of an argument.
It is an extremely shitty kind of an argument. And it can be used to
justify any kind of bullshit.
E.g. any scamcoin which haven't yet collapsed will work forever.
As I mentioned, it depends on scale. Highly sophisticated attacks are only
feasible when scale is sufficiently big.
I.e. when you have millions of dollars transacted each day it is one thing,
but if you process billions of dollars, it becomes a whole another matter.
The best way to profit from zero-confirmation payment disruption is through
derivatives: short-sell Bitcoin while performing this attack. But this kind
of an attack depends on a number of conditions:
1. highly liquid and reliable derivative market
2. sufficiently stable exchange rate
3. significant attack impact: lots of merchants relying on
zero-confirmation payments, and lots of customers paying this way
4. significant amounts of capital available to the attacker
These conditions are not yet met, and were never met in the Bitcoin's
history so far.
This is why I wrote "5 years from now", I believe that we might reach those
conditions around that time.
Direct impact of an attack might actually be low (but even if it is just
0.1%, 0.1% of 1 billion is 10 million, which isn't bad), but attacker might
profit from the panic it causes.
Note that I'm talking about situation where Bitcoin-aware PoS solutions are
deployed on a big scale, so cost of upgrade might be huge.
So anyway, in my opinion, it is actually great that Bitcoin is still
relatively small: we have an opportunity to analyze and improve things.
But you seem to be hostile to people who do that (and who do not share your
opinion), which is kinda uncool.
Also, you do not bother to back your intuition with rigorous reasoning,
while also attacking people who offer alternatives with non-rigorous
slipper-slope kind of arguments. Which is doubly uncool.
-------------- next part --------------
An HTML attachment was scrubbed...
More information about the bitcoin-dev