EIP4844: remove optimistic sync mentions

[terencechain]

This PR removes optimistic sync mentions around is_data_available. As was discussed, we don't want to enable clients to sync optimistically if the blob is not available

[djrtwo]

I like making clear the abstraction of DA in the spec, especially because the sidecar is not a 1st class consensus object.

That said, not too strongly held position

[dankrad]

I kind of like the idea of keeping it as it would make it very easy to transition from 4844 to full sharding just be changing the implementation of that function.

Also, are there no other ways of getting blocks? How about when syncing?

[dapplion]

While not impactful in the implementation, the existence of is_data_available helped me when reading the specs for the first time. I think we should leave it, since it consistently ties to the future roadmap, and hints on its implications.

[terencechain]

@dapplion perhaps we can just remove the processed further optimistically... part, that's ultimately what we decide against

[tbenr]

I'm ok with leaving it. My current implementation is taking in account that data might be retrieved async in the future.

[tbenr]

remove the processed further optimistically... part

👍

[terencechain]

@tbenr @dapplion please take another look 🙏🏼

[dapplion]

LGTM, but you should rename PR to something like "remove optimistic sync mentions", until we have better consensus on it

[djrtwo]

I'm good with this now. @terencechain can you patch the description of the issue given the scope of the issue now?

[terencechain]

I'm good with this now. @terencechain can you patch the description of the issue given the scope of the issue now?

Sorry I missed that. I just updated the description