Forgive my ignorance, but who are all these people setting up these forks, and who is actually buying the forked currency? It makes no sense to me.
The core problem is scalability. Miners are very incentivized to fix it, because they get fees on transactions (so, the more transactions the better). Core dev, on their part, want to keep the closest possible to original design (any change can have dramatic consequences).
In 2016, scalability issues started to become really serious, and miners asked to increase block size (that is, to allow to add more transactions in a single block, so that transactions wait for less time before being validated). In return, core dev proposed segwit as a solution for scalability problems (I still don't exactly get what segwit does). Miners still wanted to go with block increase, so it was a stall (and scalability problems got worse). A few forks tried to take matters on their hand to fix that problem (although, more like competitive softwares, since they did not use the same blockchain history, they started new blockchains at block 0).
In 2017, everybody finally reached an agreement: blocks size will be increased and segwit will be deployed, so both solutions would be implemented. But core devs asked for it to be done in two steps: first segwit, then segwit2x (segwit with double block size). Miners agreed and started to run their mining software with segwit, effectively making it used by the network (Bitcoin cash was born at that time, from miners who wanted block increased, but not segwit). But when came the time to implement the second part of the deal (segwit2x), core devs decided it was not needed anymore (and bitcoin gold was created as a reaction).
I would take two things away, here:
1. each time a decision need to be made, it generates a fork from people not agreeing to the final decision
2. the trust between miners and core devs is probably broken, there probably will be more and more forks, and more stalling on main chain
This wikipedia article explains more thoroughly the scalability issues: https://en.wikipedia.org/wiki/Bitcoin_scalability_problem
This is not true at all. Miners are much better off under bitcoin's crazy fee market (in the short-term at least). No other cryptocurrency is generating anywhere near as much value from transactions as bitcoin
It is indeed the limiting factor for any public blockchain based system. The rest of your comment describes capacity however, which is something else. We can increase capacity without increasing scalability.
> Miners are very incentivized to fix it,
This is false. Miners are the ones who stand to gain from high fees. And before you say more capacity would lead to more transactions and higher fees, run the numbers. Compare with less developed chains. It is quite clear that there is a cliff beyond which fees per block quickly fall off.
There is more to this power conflict than two sides. Explanations that are trivial and wrong will not help us govern this system towards the future.
> (I still don't exactly get what segwit does)
It removes the hard block cap and replaces it with a flexbile limit. Inputs counts as one fourth the size of outputs, so the more inputs a block have the bigger it is allowed to get. This factor was chosen because it could be fit in the same backwards compatible transaction style that was utilized to deploy P2SH, where non-compilant nodes only see anyone-can-spend transactions.
The reason inputs were made to be cheaper than outputs was to disincentivize UXTO fragmentation. This is an increasingly serious problem which will soon rival the capacity problem. It was recently shown that a large Bitcoin exchange have completely unnecessarily fragmented millions of dollars worth into non-movable dust, since there have been no financial incentive to avoid it in the past.
Finally, since the new style transactions breaks compatibility anyway, the opportunity was taken to change the format into one that is not malleable, and avoids the worst case verification times of the old format. The latter is a true scability improvement. One limiting factor for the number of transactions in a block is the latency from verifying the signatures, and it would not be good for the network if people started cheating even more on this.
Bitcoin Core developers did not ask for that. By the way, Segwit2x was entirely unrelated to segwit. The name was adversarially chosen to cause maximum confusion like this.
> But when came the time to implement the second part of the deal (segwit2x), core devs decided it was not needed anymore
Segwit2x wasn't a "deal" that Bitcoin Core developers entered into. Segwit2x was a proposal that was strongly rejected by the community.
I would love to know that as well, I never bought in any fork myself.
Although, the very first time I bought some cryptocurrency, it was dogecoin, soon after its release. I bought for an insignificant amount and considered it lost right away, just like when I buy a pizza or icecream. It was "just in case" (and it turned out to be my first profit, feeling good, albeit just winning a few bucks). Maybe some people are specializing in that?
Would love to hear from anyone buying in those forks.
The core devs are not committed to minimal changes to the original design, but they are committed to decentralization.
> As others have indicated, increasing the block size was not agreed to by the core devs
In fact, segwit is a capacity increase: https://www.reddit.com/r/Bitcoin/comments/6i5sx1/a_reminder_...
I'm pretty sure I read about it like it was a consensus, but I can't find it again. Or maybe I just didn't understand it at the time.
So if trust is broken, it's only between members of that group, and it does not involve devs. Sorry, I can't edit my initial comment anymore (plus, replies would not make sense anymore if I edited it).
The following is my speculation only and not based in any real world data: Many people have found themselves on the winning side of a high risk bet and their Bitcoins are suddenly very valuable. Some of them look to put a part of their winnings in some equally unlikely bet, and it seems like any unlikely bet will do. When enough people do this it becomes self perpetuating.
Sounds like it can go on forever. But the reality is that forever it is most certainly not.