This is a written version of the opinion I expressed at the recent gathering of Ethereum 1x in Berlin.
I do not have EIPs yet, and I might not have them in time for the May “deadline”. And that is because I do not want just to create “placeholder” EIP which is under-specified and not particularly thought through. Just to get onto the “train”. I would like us to have a better process with higher quality EIPs, and going against this idea straight away with EIPs full of TBDs would be a step backwards.
My primary focus at the moment is research into semi-stateless client and specification of the necessary data structures (see the latest).
However, it would be a shame if we did not start moving into the direction of the State reduction, simply due to procedural limitations. Currently I think it is still a good idea to prepare the changes A, B, C, D and H from State Fees version 3 proposal, though the next steps would change in the proposal version 4 (which is still in the works).
Since around January-February 2019, my plans was to delegate prototyping and test generation of these changes (and finally preparation of EIPs) within my working group (that is why we are extending it from 2 to 3 people). But, as you all know, growing a cohesive team takes time and effort, and some time ago I realised that trying to get it all done by 15 of May would be impossible. Therefore, we will simply proceed on our best effort basis, and not be anxious about deadlines (in fact I decided not to read too much into these deadlines back in beginning of March, and the work got so much better).
So I will keep the proposed dates in mind, but at the same I will kind of ignore them. Because when the EIPs are ready and good enough quality, we will put them in. Of course, it would be a shame if we then had to wait 9 months until the next planned upgrade, but I am confident that people can be flexible and understand the priorities.