Quote from: I R S Y K E S on July 11, 2017, 07:31:06 AM

as we are know, now everyone is panicking about segwit 1 August, bitcoin and altcoin price go down drastic Angry

So my question why there should be segwit ? Huh

Whether bitcoin can not run without segwit ? Huh

Because bullshit.
Segwit contains some relatively good ideas, but they are not necessary. Bitcoin contains conceptual and technical choices which are less than optimal, and segwit can improve somewhat on them.

So, somehow, segwit is some technical improvement of some aspects of bitcoin technicalities for more sophisticated uses of bitcoin that aren't really used much today. The MAIN reason for segwit is that core wants to push changes, and wants to push the lightning off-chain payment system. Core has been selling a lot of bogus reasons of why segwit is the solution to scaling.

The two main bogus reasons are:
- we cannot afford hard forks, that would be too dangerous, and simply changing *one single parameter* in the existing software, namely the block size limit, is technically a hard fork. Hard forks are not dangerous if they are consensual, ethereum has done many of them, monero has done many of them. A hard fork where a single parameter is simply changed wouldn't have been a great deal, but core made a great deal out of it, to be able to push segwit as a sole solution to the silly block restriction. They do this by allowing (in segwit) "out-of-accounted-blocksize data" (the so-called witness data), and hence allow more ACTUAL data per block than is accounted for (there's only a part of the actual data that goes "into the block", but that block is depending on out-of-block data that gives the full story of the transactions).

- we need small blocks so that there are many voluntary full nodes that "keep bitcoin decentralized". That's probably the biggest bogus reason. Full nodes don't decentralize anything as they have no voting power in any consensus decision (on purpose replaced by proof of work by Satoshi, to *avoid* possible sybil attacks on the number of full nodes running and voting). Full nodes only INFORM their owners. But moreover, a true full node would need just as much network and storage and computing effort to verify the "out-of-block data" too, as if these data were included into the block. Having a block of 1 MB (so, respecting the official limit and no hard fork), with next to that, 3 MB of "out of block" transaction data related to the "fingerprint" in the block, is just as cumbersome to process than old-fashioned 4 MB blocks: the amount of data transiting and processing is the same. The excuse "but you don't NEED to verify this out-of-block data if you take the fingerprinting for granted", is exactly the same reason that one can use to use a LIGHT WALLET, where you don't NEED the full blocks, but just the chain of headers, and the branches of the Merkle trees that concern your transactions.

Finally, once segwit has augmented the effective block size to 4 MB even though it makes believe old nodes that it doesn't need more than 1 MB and hence is not technically a hard fork (old nodes don't understand zilch of segwit, but accept the ununderstandable blocks because they look like funny but legal transactions to them), there is no further block size increase possible EVER without, finally, the dreaded hard fork of changing a single parameter.

So 20 years from now, bitcoin will STILL run on a room of 4 MB per 10 minutes even if at that point, 3 GB blocks would not be any problem. These 4 MB every 10 minutes, which is all you can ever extract out of soft forks in bitcoin (which is what segwit is doing), will by FAR not be sufficient to be able to guarantee any settlements of a large Lightning Network that would run on top of segwit, if ever.

In other words, the soft-forking dogma will restrict bitcoin for ever, and doesn't solve any long-term scaling issues. It will even render off-chain lightning network applications DANGEROUS if ever they pick on because settlement cannot be guaranteed.

Without the soft forking dogma, just increasing the block size would have solved the issues since long ; there's nothing to be won to not do this, whether one wants segwit for technical reasons or not, whether one would like to experiment with off-chain payments or not. In any case, this has no influence on the decentralization of the consensus decision within bitcoin, which is as of now, pure proof of work, and is not much decentralized already.