Honua
// the honua belief

The most powerful thing a GIS server can be is boring.

Spatial work earns its reputation for difficulty honestly — projections, predicates, topology, the cartography that makes a map true. None of that should leak into how the thing runs. A geospatial server should deploy like any other container, price like any other infrastructure, and answer a where-question the day it's asked without ceremony. The craft of the map was never the problem; everything underneath it should be too dull to mention.

// everything follows from that

Five principles — and what they make possible.

The first five guide the product. The final card is the outcome they are meant to create. Follow the links for current evidence and limits.

01 · Cost scales with workload, not adoption

Infrastructure you already know how to price, and a flat license. No named-user seats, request credits, or data-volume billing. Adding a user creates no Honua seat fee; added workload may require a larger capacity band.

The pricing model →

02 · It runs like the rest of your stack

One lightweight container beside the PostGIS you already run, with documented paths for Kubernetes and ECS plus serverless evaluation paths. Upgrades follow your normal redeploy process rather than a separate GIS installer.

Operations →

03 · Standards are the exit, not the checkbox

Supported ArcGIS Pro, QGIS, Excel, Power BI, and browser workflows connect through familiar interfaces. The public matrix shows what works, what needs adaptation, and where standards preserve an exit.

Compatibility → · The evidence →

04 · Spatial work belongs to everyone

Developers can work with curl, analysts can use familiar tools, and spatial professionals can spend more time on analysis. The server should not turn every contributor into an accidental administrator.

Zero to a map → · Migration →

05 · Automation proposes. Policy decides.

Private-beta operator tools prepare plans, dry runs, and recorded actions. Models propose; explicitly configured policy may permit deterministic low-risk actions. Risky and irreversible changes remain approval-gated, and broader autonomous recovery is planned.

Operate with agents → · The MCP quickstart →

+ · What this unlocks: the craft is not the cage

Making geospatial ordinary doesn't make spatial expertise less valuable — it makes it count for more. When the plumbing disappears, the judgment is what's left.

See supported QGIS workflows →

// what ordinary makes possible

It stops being rationed.

When geospatial becomes ordinary, more teams can ship maps, more dashboards can answer where, and more questions can be handled in the tools people already use. Making GIS boring was never about less of it. It is how spatial expertise reaches more work.

// the whole belief in five beats

One container beside PostGIS. Documented protocol surfaces.
Boring to run. Easier to leave. Yours to build on.