Discussion on free multiplier with a possible baseline revision as a cocktail solution for problems mentioned in discussion #774

[brian75235]

Please allow this discussion to serve as a child-thread of discussion #774 , as the previous discussion is now quite heavy (and we were not 'allowed' to talk about the possibility of baseline changes). And I do hope this discussion will focus on numbers and logics, less emotional criticism, so that we can move faster to reach a potential agreement.

In this discussion, I would suggest that we discuss specifically on the proposal of applying free multiplier with a possible baseline revision, as a cocktail treatment to the current network.

The content of this discussion so far shall be the ideas and opinions from a lot of contributing community members in discussion #774 and I am doing my best to summarize them in a neutral way, please excuse me if I miss or misunderstand anything and you are all welcome to correct.

Where we are:

1. Data onboarding is important to Filecoin
2. Current FIL+ is problematic and need a lot of changes (or be replaced by one on L2)
3. RBP is important as it affect rewards, and thus SPs' willing to pledge more
4. Inflation is bad and that's something we want to be VERY careful if we were to revise baseline power
5. We want to see a change to happen soon enough.

Rationale:

The idea of free multiplier is well summarized here by @anorth , #774 (comment). This is a pretty straightforward change once applied, it will have the following effects:

1. Current FIL+ will be largely invalid. Since SPs are willing to take part in FIL+ because of the 10X power multiplier, if there is a free multiplier, most SPs will seal CC sectors with 10X power, as it takes extra cost to process deals (bandwidth and extra storage). Though it could still be there, either be optimized or be replaced by later L2 layer, which will not be so urgent.
2. Personally I would expect to see a continuous decrease on RBP (since almost every rational SP will do whatever they can to replace they previous 1x CC sectors with new 10x CC sectors, either through sector extension, if supported, or through re-sealing). The benefit of use 10x CC sectors is straightforward and briefly explained here (FIP-0080: Motivation for a Scheduled Reduction in the Quality Multiplier of Fil+ Deals and Example Timelines #774 (reply in thread)). The drop of RBP may not stop until most of the current CC sectors are replaced.

When most of the 1x CC sectors are replaced by 10x CC sectors, just as @steven004 pointed out, it is similar as all to 1x. Then we might want to look at the RBP vs baseline power by then. I do not currently have exact statistical numbers, but I guess, if I may, it will be pretty left behind and it is possible that SPs may not get enough reward to support continuous pledging. I really hope the CEL team @zixuanzh could do some calculation and prediction for this.

That's why I would suggest we might want to apply a baseline revision along side, which of course, would totally depend on the result of a valid calculations from the CEL team or others. If the RBP decrease is proven to be not problematic, then we can leave the baseline as it is. Otherwise, we might want to carefully adjust the baseline to keep the balance between inflation and block rewards.

Things to do and open questions:

1. Most importantly, the community might need a full and detailed statistical predictive report on RBP/baseline/block rewards change if we were to apply free multiplier. Hopeful the CEL team could offer this?
2. Shall the free multiplier be available for current sector extension?