Great question about the costs. As you’ve said Orbs is a decentralized network meaning it must rely on an economy that makes sense for it to work. As we all know, designing decentralized economies is not an easy task.
The first thing the architecture needs to address is what is the cost of operating the network for a validator. Everything starts there because if validators can’t recover their operation costs, they won’t participate, or they will have to increase the fees in order to do so.
The architecture is horizontally scalable with the concept of virtual chains - meaning that unlike many blockchains there’s no limited amount of resources so there’s no competition between apps over resources and no bid wars over execution. What this means is that the cost for validators for adding a new virtual chain is fixed and equals the standard cloud cost of running this app separately (this is what makes the cost predictable).
Now that we’ve established that the cost is fixed, let’s examine the price. In an efficient market where demand for the token comes primarily from app developers purchasing computing power, token price is expected to become stable and correlated with the cost of computing power.
If the market is inefficient and token price becomes volatile or uncorrelated with the cost of computing power, validators should uphold the virtual chain pricing set in the initial design of the network in order to stabilize the market price over time.
That being said, since the project is decentralized and the Orbs team doesn’t have actual control over validators or the prices they charge, the network validators and guardians may deviate from the original design. It is possible that guardians take a vote and change the original protocol and adjust its pricing (like they can do with any other feature of the network). This could be justifiable in extreme situations of unfair pricing that could prevent apps from using the network.
However, the Orbs team does not think that this is the desired approach since it would undermine the predictability of the model.
For a deeper dive read through the details in https://www.orbs.com/pricing