feat: use chunk groups feature

[rklaehn]

caution: this is not compatible with existing iroh/sendme

[dignifiedquire]

what's missing here, given this is a draft?

[rklaehn]

It was a draft because the crate was not published yet, and the tests were failing. Let's see, maybe they work now...

[dignifiedquire]

/perf

[dignifiedquire]

/netsim

[github-actions]

body: |
.
Perf report

test	case	throughput_gbps
iperf	1_to_1	67.20
iperf	1_to_3	107.60
iperf	1_to_5	163.50
iperf	1_to_10	187.60
iperf	2_to_2	71.60
iperf	2_to_4	125.30
iperf	2_to_6	163.70
iperf	2_to_10	192.00
iperf_udp	1_to_1	5.68
iperf_udp	1_to_3	-0.00
iperf_udp	1_to_5	25.47
iperf_udp	1_to_10	41.69
iperf_udp	2_to_2	10.92
iperf_udp	2_to_4	20.93
iperf_udp	2_to_6	30.24
iperf_udp	2_to_10	41.10
iroh	1_to_1	1.80
iroh	1_to_3	4.99
iroh	1_to_5	4.83
iroh	1_to_10	4.28
iroh	2_to_2	3.45
iroh	2_to_4	6.89
iroh	2_to_6	9.47
iroh	2_to_10	12.06
chuck_tls	1_to_1	5.78
chuck_tls	1_to_3	17.83
chuck_tls	1_to_5	27.48
chuck_tls	1_to_10	39.90
chuck_tls	2_to_2	15.22
chuck_tls	2_to_4	21.87
chuck_tls	2_to_6	36.76
chuck_tls	2_to_10	54.87
chuck_http	1_to_1	27.71
chuck_http	1_to_3	52.52
chuck_http	1_to_5	66.20
chuck_http	1_to_10	72.24
chuck_http	2_to_2	49.77
chuck_http	2_to_4	80.46
chuck_http	2_to_6	104.02
chuck_http	2_to_10	179.27

[dignifiedquire]

@rklaehn do we want to ship this? did you compare the perf numbers?

[rklaehn]

Well, perf is basically identical, a tad faster. The main benefit is that the outboard is smaller, so we can share larger files with outboard in memory.

As far as I am concerned it is ready to go - provided that we're clear that this might change in the future, together with the protocol.

[dignifiedquire]

@flub any concerns merging this before DC release?

[flub]

@flub any concerns merging this before DC release?

i guess there's no problem with merging this?