abci: codespace

[ebuchman]

Closes #1858

• Updated all relevant documentation in docs
• Updated all code comments where relevant
• Wrote tests
• Updated CHANGELOG_PENDING.md

[codecov-io]

Codecov Report

Merging #2557 into develop will decrease coverage by 0.1%.
The diff coverage is n/a.

@@             Coverage Diff             @@
##           develop    #2557      +/-   ##
===========================================
- Coverage    61.33%   61.22%   -0.11%     
===========================================
  Files          202      202              
  Lines        16677    16708      +31     
===========================================
+ Hits         10229    10230       +1     
- Misses        5585     5611      +26     
- Partials       863      867       +4

Impacted Files	Coverage Δ
evidence/pool.go	52.08% <0%> (-4.06%)	⬇️
consensus/reactor.go	70.93% <0%> (-1.68%)	⬇️
libs/db/remotedb/grpcdb/client.go	0% <0%> (ø)	⬆️
libs/db/remotedb/remotedb.go	41.52% <0%> (+4.68%)	⬆️

[melekes]

🍰 🌮 🍉

[ValarDragon]

I'm not really a fan of codespace as a string on every Check/Deliver Tx response. We probably don't have time prelaunch, but I think it would be worth benchmarking the performance overhead vs the int based codespacing. I know in the ADR it said that it was complex and hacky, but I completely disagree. It seemed like a very natural and simple thing to me. (If the performance overhead is negligible, then my concern is gone)

[ebuchman]

but I completely disagree.

You think the way we're hacking namespaces into the Code in the SDK isn't hacky? It's also going to run out. I wouldn't worry much about a small string here :), but sure we can always benchmark. More important is to enable the SDK to clean up code around this.