feat: switch back to protobufjs

[twoeths]

Motivation

• protons is too slow to use, see Slow compared to protobufjs ipfs/protons#51

Description

• Switch back to protobufjs
• Copy generated files from v1.0.0
• Change rpc.js to rpc.cjs, new rpc.js to reexport RPC from rpc.js

[mpetrunic]

@tuyennhv have you tried with new versions of protons?https://github.com/ipfs/protons/releases/tag/protons-v4.0.1
ipfs/protons@protons-runtime-v2.0.1...protons-runtime-v2.0.2

should have performance optimizations

[wemeetagain]

We still need this file for the subpath export (import ... from '@chainsafe/libp2p-gossipsub/message')

[twoeths]

yeah this is exactly what I got when I build lodestar 👍

[codecov-commenter]

Codecov Report

Merging #310 (bef99b6) into master (e0d3a7d) will decrease coverage by 1.70%.
The diff coverage is 71.02%.

@@            Coverage Diff             @@
##           master     #310      +/-   ##
==========================================
- Coverage   81.53%   79.83%   -1.71%     
==========================================
  Files          43       44       +1     
  Lines        9501    11044    +1543     
  Branches      918     1014      +96     
==========================================
+ Hits         7747     8817    +1070     
- Misses       1754     2227     +473     

Impacted Files	Coverage Δ
src/metrics.ts	23.08% <6.66%> (ø)
src/index.ts	70.82% <68.75%> (+0.03%)	⬆️
src/message/rpc.cjs	71.28% <71.28%> (ø)
src/message-cache.ts	79.09% <100.00%> (ø)
src/message/rpc.js	100.00% <100.00%> (ø)
src/types.ts	95.13% <100.00%> (ø)
src/utils/buildRawMessage.ts	84.39% <100.00%> (ø)
src/utils/create-gossip-rpc.ts	100.00% <100.00%> (ø)
test/message-cache.spec.ts	100.00% <100.00%> (ø)
test/utils/index.ts	100.00% <100.00%> (ø)
... and 2 more

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update e0d3a7d...bef99b6. Read the comment docs.

[twoeths]

@tuyennhv have you tried with new versions of protons?https://github.com/ipfs/protons/releases/tag/protons-v4.0.1
ipfs/protons@protons-runtime-v2.0.1...protons-runtime-v2.0.2

@mpetrunic I've tried it, below is the result:

• current protons

rpc
    ✔ decode Attestation message                                           7089685 ops/s    141.0500 ns/op        -    2045205 runs   30.1 s

• new protons (v4.0) & protons-runtimes (v2.0)

rpc
    ✔ decode Attestation message                                           6807815 ops/s    146.8900 ns/op        -    1977187 runs   30.1 s

• protobufjs

rpc
    ✔ decode Attestation message using protobufjs                      1.315789e+8 ops/s    7.600000 ns/op        -   28030718 runs   30.0 s

=> there is no improvement made with new protons and it's still 20x slower than protobufjs

[dapplion]

LGTM!

[achingbrain]

there is no improvement made with new protons and it's still 20x slower than protobufjs

Can you please share the benchmarks you're using to make these measurements?

[twoeths]

there is no improvement made with new protons and it's still 20x slower than protobufjs

Can you please share the benchmarks you're using to make these measurements?

@achingbrain the benchmark is in #318