just needs 20 inputs 2 outputs and already that makes $30~ in fee.
Correct. However, saying that "fees" are $30 is nonsensical and a sign of deep misunderstanding of Bitcoin. Fees are to be expressed in satoshi/byte (or per kByte) and should be represented with:
1) Median transaction fee.
2) Average transaction fee.
if you use the word nonsensical one more time, then someone better buy you a dictionary for next christmas
if you read the context. you will see the OP said
"but paying
up to $30 for a $100 transaction would make some BTC users jump to alternatives."
the context was not that fee's are on average $30 now.. nor median now. it was that IF fees continue to rise people will jump ship
i added context (using the byte per tx formulae) to show it doesnt take much to formulate a tx that cost ~$30 in fee.
lauda i understand you cant read more than a paragraph. but things only become nonsense to you. if you fail to read then content of posts to make sense of the overal picture
This again.
It's about the bandwidth, not the storage.
Actually it is a lot of things. A few pointers (they interact with each other at different levels; some are less important/constrictive than others):
1) Quadratic validation time.
2) Bandwidth.
3) Storage.
4) Orphan rates.
5) Propagation delay.
if you cared to check out things..
we are no longer in 2009
we are not stuck with the min specs of Raspberry pi1
we are not stuck with the average internet of 512kbit/s - 3g mobile
we are not stuck with the v0.1 inefficient code.
we are in 2017
we are at the min specs of Raspberry pi3
we are at the average internet of 5mb/s - 4g, .. approaching the fibre/5g era
we are at the v0.12+ efficient code. - EG libsecp256k1 5x efficient
plus other efficiencies all add up that 8mb is deemed min spec safe.. yep even Core admit and accept as such that 8mb is deemed as 'safe' but they still double cautiously prefer 4mb.
as for quadratics.
keep a tx below 4k txsigops and quadratics is not an issue lauda should ever have to cry about ever again.. nor should anyone else.