chifra chunks overview version 2

[tjayrush]

• We don't know how to manage Pinata (remove unused pins, ensure wanted pins are present, list pins, etc.)
  • things to do for remote pinning service: list all, download one chunk, upload one chunk, remove one chunk, remove all unused pins (what is an unused pin? preserve mainnet new, mainnet old, sepolia, and gnosis?)
• We don't know how to --check --deep
  • Does every address in a chunk report "not no" for inclusion in the bloom?
  • Does every chunk contain every block in the filename's range (or is allow_missing true)?
  • Do the file name ranges have missing blocks? (They should not.)
  • The file size is as expected by both the manifest (already done) and the calculated size of the file (see below).
  • The hash of the manifest is the same as the hash stored in the smart contract
  • The hash in each header is identical to all other headers and is as expected from the spec
• During regular (non-deep) checks
  • check file size (in fact, we probably already do).
  • Is there a snap_to_grid after first_snap everywhere it's supposed to be?
  • For non-snapped chunks, nApps ≥ apps_per_chunk.
  • For snapped chunks, nApps < apps_per_chunk.
• Add an option to chifra init called --dry_run. Prints to the screen but does nothing else.

[tjayrush]

See #3319