XI'AN · HYPERSCALE FOR RADIX

Hyperscale Weekly

Week #6 · 15–21 Jun 2026 · Milestone 1 · Adaptive Sharding · Week 8 · ~month 2 of 4
Split the work, not the world
MILESTONE 1 PROGRESS47%

Previously: flightofthefox had just landed the settled-waves fence for straddlers, the first end-to-end shard split and merge in the simulator, and per-shard boundary QCs anchoring crossings in the beacon — work moving from parts t

A heads-down, no-chat week focused on closing the gap between the simulator and the real network. flightofthefox made the sim mirror production at the same epoch length and with the same follower pools, multiplied the end-to-end tests on the production side, and quietly introduced the first on-chain knob validators can vote to retune. The foundations are starting to feel production-shaped.

78
-52% vs last wk
commits
+10k 4.1k
-26% vs last wk
lines changed
159
files touched
30
day commit streak
Commits / day
M
T
W
T
F
S
S
peak 16/day · 7/7 days active
Momentum
Lines changed / week
last 8 weeks
98% Rust2% Config
What flightofthefox built

Validators get a live tuning knob

flightofthefox added NetworkParams to the beacon at genesis and a way for stake-pool votes to retune them at runtime. It's the first self-governing dial in Hyperscale — a thermostat the community can turn, not a setting frozen by the developers.

→ see “The Governor” on hyperscale.rs

The simulator catches up to production

flightofthefox rebuilt the sim around the same epoch length and follower pool the real network uses, made message iteration deterministic, and added request retries and simulated latency. The crash lab is now a faithful rehearsal space, not a smaller parallel world.

→ see “The Crash Lab” on hyperscale.rs

Reshape gets production-grade tests

End-to-end tests for the split, the merge, and straddler settlement across the reshape boundary all turned on in production. The pacemaker now starts the moment a new shard is seated, cohort seeds lock at split time, and a merge parent composes cleanly from its child terminals.

→ see “The Will” on hyperscale.rs
The bottom line
Standout change
The simulator and the production test suite are now running the same epochs, the same follower pools, and deterministic iteration — the first realistic dress rehearsal for a public testnet.
What's next
Likely next: stabilise the production reshape test suite under fault scenarios before moving on to the lottery-style validator shuffling also promised in M1.
This week's concept
The Crash LabReplay every bug byte-for-byte
This week's biggest theme was the simulator and production test suite converging onto the same epochs, the same follower pools, and deterministic iteration — turning the crash lab into a realistic dress rehearsal for a public testnet.
hyperscale.rs
Jargon, decoded
NetworkParamsThe on-chain settings that describe how the network runs — like timeouts and committee sizes — stored in the beacon so validators can vote t
Stake poolA group of XRD holders who delegate their stake to a single validator so they can vote and earn rewards together.
Cohort seedA random number that locks in exactly which validators get assigned to a brand-new shard the moment it splits off.
Where the work landed
simulation · The Crash Lab
mirrors production epochs and follower pool
beacon-follower pool and route gate
NetworkParams and stake-pool voting
Reference — concept map, roadmap & links
The full picture
The System
The whole sharded design at a glance.
The Journey
One transaction's trip through the network.
The Clock
Consistent time across independent shards.
The Census
The registry of validators and shards.
The Overlap
Why the chain can never split in two.
The Generals
All-or-nothing commits across shards.
The Archive
How the network remembers its past.
The Library
All state in one clean, reshardable tree.
The Will
Protects transactions caught in a reshape.
The Lottery
Random validator shuffling, ungameable.
The Triage
Detects and removes bad validators.
The Governor
Auto-prices the stake to run a validator.
The Crash Lab
The simulator that replays any failure.
The Proof
Proving safety with maths, not just tests.
The Asterisks
Trade-offs rivals accept — this avoids.
The road to mainnet
M1
Adaptive Sharding
~4 mo
M2
Radix Engine
~5 mo
M3
Gateway & API
~3 mo
M4
Validator GUI
~3 mo
M5
Migration
~3 mo
M6
Live support
12 mo
Now: Milestone 1 · Adaptive ShardingWeek 8 · ~month 2 of 4 · 47%. Since day one: 2,324 commits — one developer, fully in public.
Links
Written by AI from public commits, the community chat and hyperscale.rs — may contain mistakes. · generated 2026-07-09