Course · ← 0005 Publishing the Map · 0007 Model the System →

Lesson 0006 · ~13 minutes

Choosing the Stack

Esri or open source is not a technology argument in donor work. It is a sustainability argument, and both sides of it can lose.

The two stacks

The functional gap is real but narrower than either camp claims. ArcGIS Enterprise gives governed publishing, multiuser editing, and — the genuine differentiator — server-hosted geoprocessing, with distributed vector and raster analysis tiers, all invoked over REST. The open stack pairs QGIS with PostGIS for multi-user storage and analysis; QGIS Server publishes map and feature services but ships no web processing service, so processing is pushed into PostGIS SQL, GDAL batch jobs, or community WPS bridges.¹

Notice how little of that gap a land registry touches. Parcel-and-rights work is transactional CRUD over polygons with heavy attribute logic — not distributed raster analytics. The premium tiers price capabilities most land administration systems never invoke.

The argument that actually decides it

Donor projects end. The system does not — or is not supposed to. So the real question is not "which is better" but "what survives the day the funding stops?" Both answers have a failure mode, and being able to name both is what makes you credible in the room.

The literature on land governance says both halves out loud: open platforms offer flexibility, transparency and the ability to customise to a country's own legal and administrative framework, and reduce dependence on further funding and licence problems — while the honest counterweight is that maintaining them demands specialised technical expertise that is usually scarce in developing contexts, which is itself a sustainability risk.² Proprietary buys comprehensive out-of-the-box function, vendor support and regular security updates — for a recurring fee.²

The third option people forget

You are not obliged to start from a blank schema. FAO's SOLA Suite is an open-source land administration system, LADM-based, already covering registry, systematic registration, state land, and community capture — on Payara, PostgreSQL/PostGIS, GeoServer and JasperReports.³

Even when you do not adopt it, read its schema before designing yours. And say so in the proposal: "we evaluated SOLA and here is what we took from it" is a far stronger opening than a bespoke model with no lineage, because it tells the evaluator you know the field's prior art rather than being about to reinvent it badly.

How to write the answer

1. State the sustainability constraint first Who operates this in year four, and on whose budget. Everything else follows from that answer.
2. Map requirements to capability tiers If nothing needs distributed raster analytics, say so — that is where the licence premium goes.
3. Name the prior art SOLA, LADM, the national system already in place. Show you looked.
4. Price the whole life, not the build Licences, hosting, training, and the integrator retainer, across five years — not just delivery.
5. Pick, and state what would change your mind An existing in-country Esri estate and support contract is a legitimate reason to go proprietary. Saying so builds trust rather than costing you the point.

Drill 1 — Which way does this push?

Drill 2 — Recall

Your primary source

"Open Source vs Proprietary Solutions in Land Governance" (ON Policy Africa, Uganda case) — short, balanced, and written from inside a real land-governance debate rather than from a vendor's marketing floor:
onpolicy.org — Uganda's digital landscape

Pair it with the FAO SOLA overview (stack and components) so the open option is concrete rather than theoretical.