How a small validation could run.
This page exists to answer one question honestly: if a Malaysian public-sector or civic-tech team ever wanted to test this measurement work in practice, what would it actually involve? Very little, probably. It is written to be read in three minutes and ignored without offence.
What it is
- A fresh measurement of public road data for one bounded area, shared with whoever needs to see it.
- A short list of named streets worth a look, drawn from public sources, for a human to sanity-check.
- A written wrap-up: what was measured, what was found, what would (and would not) be worth continuing.
The shape of it, if it ever happens
- Stage 1 · Agree the ground
Which metrics matter to the people doing the work, which streets are worth a look, and what may or may not be shared. Usually one conversation.
- Stage 2 · Look
We refresh the public-data measurements and produce a readable list. Your team looks at it when they have a moment. No deadlines, no milestones, no status calls.
- Stage 3 · Wrap up
A short written verdict: what held up, what did not, whether it saved anyone effort. Share it as widely or as little as you like.
Scale and validation stay separate. Everything here stays inside one neighbourhood-sized sample. The 16-state measurements on the observatory are context, not commitment. Nothing expands without an explicit decision from the people involved, and either side can walk away at any point, for any reason, without awkwardness.
What would be published
- The aggregate measurements themselves: coverage, naming, freshness, connectivity.
- The interactive map and the correction-list CSVs anyone can download.
- The scripts and method, so any of it can be re-checked independently.
What stays private
Anything that touches custodian reference data stays inside the agency until written permission says otherwise. No reference geometry, no address records, no comparisons presented as findings without that permission. Silence and walking away are both fully acceptable outcomes.
No ask on this page
If this is never useful, fine: the page cost nothing to keep and nothing to ignore. If at some point a team does want to see whether this holds up on the ground, khanster.net is the door. A first look at the measurement costs nothing and commits no one to anything.