Remove option to use RESP2 protocol

[adam-fowler]

This means we don't need to add additional support to the channel handler to support RESP2 subscriptions
Also a number of commands return different types based on whether you are using RESP2 or 3. Sticking to one return type would make life a lot easier.

[github-actions]

✅ Pull request no significant performance differences ✅

Summary

New baseline 'pull_request' is WITHIN the 'main' baseline thresholds.

Full Benchmark Comparison

Comparing results between 'main' and 'pull_request'

Host '2be58c2ba704' with 2 'x86_64' processors with 7 GB memory, running:
#12~24.04.1-Ubuntu SMP Mon Mar 10 19:00:39 UTC 2025

ValkeyBenchmarks

GET benchmark metrics

Malloc (total): results within specified thresholds, fold down for details.

Malloc (total) *	p0	p25	p50	p75	p90	p99	p100	Samples
main	4	4	4	4	12	24	24	11
pull_request	4	4	4	4	12	24	24	11
Δ	0	0	0	0	0	0	0	0
Improvement %	0	0	0	0	0	0	0	0

HashSlot – {user}.whatever metrics

Malloc (total): results within specified thresholds, fold down for details.

Malloc (total) *	p0	p25	p50	p75	p90	p99	p100	Samples
main	0	0	0	0	0	0	0	21
pull_request	0	0	0	0	0	0	0	20
Δ	0	0	0	0	0	0	0	-1
Improvement %	0	0	0	0	0	0	0	-1

ValkeyCommandEncoder – Command with 7 words metrics

Malloc (total): results within specified thresholds, fold down for details.

Malloc (total) *	p0	p25	p50	p75	p90	p99	p100	Samples
main	0	0	0	0	0	0	0	745
pull_request	0	0	0	0	0	0	0	743
Δ	0	0	0	0	0	0	0	-2
Improvement %	0	0	0	0	0	0	0	-2

ValkeyCommandEncoder – Simple GET metrics

Malloc (total): results within specified thresholds, fold down for details.

Malloc (total) *	p0	p25	p50	p75	p90	p99	p100	Samples
main	0	0	0	0	0	0	0	1957
pull_request	0	0	0	0	0	0	0	1951
Δ	0	0	0	0	0	0	0	-6
Improvement %	0	0	0	0	0	0	0	-6

ValkeyCommandEncoder – Simple MGET 15 keys metrics

Malloc (total): results within specified thresholds, fold down for details.

Malloc (total) *	p0	p25	p50	p75	p90	p99	p100	Samples
main	0	0	0	0	0	0	0	359
pull_request	0	0	0	0	0	0	0	363
Δ	0	0	0	0	0	0	0	4
Improvement %	0	0	0	0	0	0	0	4

[fabianfett]

I think this is mostly about not supporting RESP2 today. We should reconsider this once we reached 1.0.