-snip-
The block limit is the technological limit that constrains the on-chain capacity, the arise of fees is the free market reaction.
People representing 90% of global hashrate disagreed during their panel in HK.
The idea that 1MB, "scaling" up to 1.75MB equiv with the gradual rollout of segwit, well into 2017... is the technological limit? That's
BS and you know it.
The important part of their plan is that hard forks
remain "controversial".
Currently the network node operators are overwhelmingly supporting 1MByte block limits, so that is the reality of the technological limit on the network today.
How that changes in the future is the subject of much debate, as you well know.
I was attempting to get causality straight in BJA's mind.
You have to admit that Core has a certain sort of
inertia. There's a big chunk of operators still plugging along on previous versions of Core. I also haven't seen an alternative that has the level of ongoing support that I'm comfortable with... so far. I really had high hopes for some kind of good will compromise between the opposing forces (ala garzik), bringing us back together with a common purpose... but it appears that is not going to be the case.
Yes, there is a huge amount of inertia and for a very good reason, that is very organic and was to be expected if you have the right experience.
Most people will not be aware that actual operational software in critical infrastructure changes very slowly, much slower than bitcoin Core (which is very high risk by comparison but it is still 'beta'). 90% of fortune 500 companies run RedHat operating systems on their back-office servers and they have very long term support for old versions for those reasons. Apparently the Banks still have some COBOL code running on old mainframes that still do the actual monetary-base and settlement calculations, so-called "green screen" functions because they are too scared to touch it.
Nuclear power plants and oil-rig safety shutdown systems that include software hardly ever make changes to their code after it has been commissioned and the plant "goes live" with active material. It takes committees of programmers and managers poring over every line to get even the simplest changes into an active system.
The disconnect between what is happening here and what the public have been told to "want to happen" is astounding for anybody with any experience in high-risk industrial software systems. The bitcoin protocol is almost complete now and will hardly ever change from now or else it will risk catastrophic failure. It is just very, very unfortunate that Gavin and Hearn "went there" with the whole hard fork MAD power grab, against the vast majority of the development community's technically better judgement and in an entirely reckless manner for critical infrastructure software management.