For the time being, mostly blocksize. As blocksize is all that is needed at this stage.
The BCH community, however, did demonstrate conclusively that:
- generic home computer HW on consumer broadband running bitcoind can handle ~100 tx/s
- above ~100 tx/s there is still plenty of CPU BW
- a fix to bitcoind's naive threading model increases performance
- bitcoind with fixed threading on same HW & net can handle ~500 tx/s
The BCH community is doing the work of re-enabling a host of opcodes that were thrown overboard years ago before any real analysis was performed upon them. Oh, I guess that's not scaling.
Of course, it was that same community (though pre-fork) that first implemented Xthin, which reduces network bandwidth consumption by nearly 2x.
There is discussion of adopting Lightning. Not much traction for that. There's orders of magnitude that can be gained by a simple blocksize change first.
I certainly am not willing to use up that much of my home bandwidth. And even if I was, 500 tx/s isn't nearly enough to be a good long term goal. I mean if that was all that we were capable of achieving than it would be enough for bitcoin to be a useful niche technology, that would even be true if we were stuck forever at 7tx/s, but both are less than ideal.
But ok, so increasing block size is a useful scaling vector. I actually think most people on the main chain side of the divide agree with that, I know I do. Other main chain supporters in this tread, how do you feel about responsible block size increases that correspond to improvements in computer hardware and networking infrastructure? It seems to me that the divide is not between willingness to increase blocksize or not, its a disagreement on whether we should be taking an extremely restrictive uni directional approach to this problem vs a more comprehensive multi faceted approach. I just don't see how someone can argue for the former.
Well, the way I see it, Lightning was two years too late. And it will be three or four years too late before it makes an appreciable difference. In professional engineering, one always does the best bang for the buck before engaging more resource intensive, longer, more complex solutions. Simple blocksize increase first.
Add that the way Core/BS implemented the ugly kluge of shoehorning segwit in via a so-called soft fork has fundamentally changed Bitcoin's security model, and there is not much to love there.
Sure, Lightning allows high frequency txs between a limited set of participants, but it really doesn't do much for increasing the number of participants themselves. You can still only open or close about 5 channels per sec. How long will it take if everyone in the world wanted to open _one_ channel? Three decades.
But we've been over all these points - and more. I will continue to believe that Bitcoin Cash has the right way forward for Bitcoin. Others will continue to think that I am daft. I don't much care. I am happy to let sleeping dogs lie. But when others cast aspersions, I will reply to the slander.
Shill your shitcoin in Rogers forum please... Thanks...
There is only one bitcoin and if there is a "bitcoin cash" then its monero and not your shitty forkcoin.
Bcashers are the MOST annoying people in cryptospace, by far (Only challanged by the Verge-Shills)