The resource market
There is one significant problem with the way that you describe the mechanisms in both the steem blockchain and EOS:
They don't describe a market.
Economics 101. To qualify as a market, you need certain things to exist.
A resource to be traded.
That we have.
A demand for that resource.
That we have.And finally, some would say absolutely crucially, a price signaling mechanism so that those with the demand can evaluate those offering the resource, compare them, and purchase it via exchange.
That we pointedly do not have. And there is really no interest in implementing price signaling in this market, which is a sad thing. If for no other reason that it would facilitate the means by which you could properly load balance accesses across multiple servers.
In order to have a price signaling mechanism, we would have to have some sort of means by which when you go to make an interaction with the blockchain, you are alerted to how much it will cost in RC. Then you can decide whether to make that transaction or wait (or take that transaction to another platform).
Ideally, each witness would price bytes, storage required, and bandwidth required separately and individually for each witness server. A client would query several (and have the possibility of adding or removing from that list on a user level), check for the best prices, and give the user the option – along with the option to tell the client "make the best deal you can at any given point." But the user has to be presented with a price.
As it stands, the steem blockchain doesn't have a market for resources. It has a dynamic limiter, where transactions have a fixed cost and the pools are communal. The pool costs vary, but there is only one pool per resource and no competing providers.
It's deeply disingenuous to refer to anything of that order as "a market."
Even though I understand your arguments, I am still keeping it like it is.
The simple definition would be: a market is a place where supply and demand meet. It is a market in the most basal form.
The tension between demand and supply will set a price. It works for resources too.
And that's it for me.
Just meeting does not a market make. Everyone wants things, some people make things, but it doesn't matter if I'm supplying something you don't want, and it doesn't matter if you want something I'm not making, even if we come together in the same place.
If there is one supplier, there is no market. There is just a limit on exchange. If the consumer has no choice, that is they have no real indication of what that price is, there is no market. They are not involved. They have no means of engagement. Either they get the thing that they want or they don't and there is no decision-making along the way.
There is an absence of the one thing that is absolutely required for there to be a market: feedback.
This is important for the way that the steem blockchain "prices" RC expenditure because there is literally nothing you can do as a consumer to engage with a provider. There is no competition. There is no price signaling to allow a customer to decide whether they even want to engage in the first place.
Without that signaling, there is no market. There is just a resource pool limit and that's it.
Calling it a "market" gives it a dignity it doesn't deserve.
It's a shame that it's not a real market, because then witnesses might be able to compete amongst themselves to see who could deliver a witness node for the least price which still covers their expenditures. That would be interesting and make the whole system more efficient. It would even incentivize witnesses to potentially offer different classes of service, a high-speed, high responsive node with an expensive API versus a node with slightly less responsiveness but lower cost, perhaps making it more feasible for new and small users to engage with the social media aspects of the system because they don't have a need for sub-second API response.
And that's terrible, because incentivizing innovation is one of the things that blockchain technology is supposed to be about, if you believe the hype.
Your suggestion for a real blockchain resources market is interesting, I think.
Do not know if it can be applied to the current DPOS blockchains, like Steem, Whaleshares and EOS. It might be a next step.
With DPOS the witnesses need to keep up to the standards of the blockchain demands. Telos even lets blockproducers proof if they are up to it.
With the DPOS system a user can decide to put more stake in, leave it like it is, or pull some out. Yet a future innovation might show a solution that makes it possible to integrate your proposal into a blockchain system.
Blockchain tech is still in its early years. Never thought something like Steem would even be possible. Yet here we are.
So, something like you describe might not be that far away. There are things that still surprise me like https://friendup.cloud, also a blockchain innovation. We've not seen it all yet, I guess.
There's no reason that a proper market can't be set up for DPOS systems.
Of course, there has to actually be exchange, which is one of the reasons the RC system also fails to be a market. When I commit an operation to the steem blockchain, I sacrifice the RC. I burn it. No one actually acquires that RC in exchange for what I am allocated temporarily from the pool. It is a pure limiting mechanic.
If, however, instead of burning RC I gave it to whatever provider was giving me bandwidth/disk storage/blockchain overhead, the witness node, and there was a use for RC for witness nodes which would return it into the overall economic ecology, we would have an actual market.
The problem is from the perspective of a witness node, a user is a pure cost. There is nothing in a direct sense that I offer as value to a witness node that they would be interested in trading RC for. Or anything, really. That speaks to a larger problem which is that witnesses have an active and powerful disincentive to want to increase the number of users on the steem blockchain because every new user is implicitly a consumer of bandwidth/disk space/blockchain operations and thus a cost.
You might counter that the more users, the more likely that the crypto-commodity of STEEM will increase in value, and that's true. Simultaneously, the more users who are creating content, not only the more overhead a witness node must maintain but the likelihood of more spread voting which will dilute any given witness's position in the pool, the more likely there will be more competition for being a witness there is, and the nonlinear nature of the means of replication for witness nodes on the steem blockchain very well may more than eat any possible advantage.
This is just a limitation in using the wrong technology for the wrong solution. Blockchains are a terrible solution for providing a database backend for social networks. An absolutely horrible idea. The storage replication issue alone is a nightmare.
Anything is possible in the future, but some things are definitely limited by the axioms of information theory. And by basic real economics.
Could we see an actual market for witness node resources? It's possible. Are we going to? Not on the steem blockchain.
Thanks for the extended reply, will read it tomorrow.
The only thing left for me to reply is that maybe we might agree that we disagree.