[Awaiting Mainnet25] Clean up database prefixes

[jordanschalm]

---

🚢 This PR is not downward-compatible. It should be merged after the last HCU of Mainnet24 and before the Mainnet25 network upgrade.

---

• change the prefix of ReceiptMetas (previously it had been the same as Results) ⚠️ This is a backward-incompatible change and must be included in a spork / network upgrade.
• move deprecation notes inline with code definitions

[codecov-commenter]

Codecov Report

All modified and coverable lines are covered by tests ✅

Comparison is base (2088ac5) 56.36% compared to head (54ea60d) 56.34%.
Report is 7 commits behind head on master.

Additional details and impacted files

@@            Coverage Diff             @@
##           master    #5134      +/-   ##
==========================================
- Coverage   56.36%   56.34%   -0.03%     
==========================================
  Files         977      977              
  Lines       91915    91915              
==========================================
- Hits        51805    51786      -19     
- Misses      36268    36286      +18     
- Partials     3842     3843       +1     

Flag	Coverage Δ
unittests	56.34% <ø> (-0.03%)	⬇️

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[AlexHentschel]

I like that you clearly assign the Deprecated values.

In the long-term, what is our strategy of reusing those values? I think we desire to re-use them, otherwise we would unnecessarily bloat the badgers key space. In other words, I am assuming that we agree that it is fine to assign previously deprecated prefix values (such as 31) for new purposes.

I would suggest we follow this policy here:

• prefix 31 is not used
• codeExecutionReceiptMeta requires a new prefix value.
• so codeExecutionReceiptMeta could use the prefix 31

Suggested change

	_ = 31 // DEPRECATED: 31 was used for identities before epochs
	codeGuarantee = 32
	codeSeal = 33
	codeTransaction = 34
	codeCollection = 35
	codeExecutionResult = 36
	codeExecutionReceiptMeta = 36
	codeResultApproval = 37
	codeChunk = 38
	codeExecutionReceiptMeta = 39 // NOTE: prior to Mainnet25, this erroneously had the same value as codeExecutionResult (36)
	codeHeader = 30
	codeExecutionReceiptMeta = 31 // NOTE: prior to Mainnet25, this erroneously had the same value as codeExecutionResult (36)
	codeGuarantee = 32
	codeSeal = 33
	codeTransaction = 34
	codeCollection = 35
	codeExecutionResult = 36
	codeResultApproval = 37
	codeChunk = 38
	codeProtocolState = 39

[jordanschalm]

Personally I don't think we should re-use key prefixes until we need to, which could very easily be years away.

There is a benefit to re-using key prefixes only once we are running out of keyspace.

By contrast, although it's possible re-using key prefixes is harmless, I think there is a real possibility it has an adverse impact on some users -- for example the Flow Diver team has recently been indexing data from old sporks. I can imagine how a practice of re-using rather than retiring key prefixes could make that sort of thing more challenging.

[AlexHentschel]

👍