MapPrism

v1 → v2 Migration Guide ​

Why the change ​

MapPrism v1 sent job_type webhooks directly to n8n — one workflow per job type. That became rigid fast. v2 adds Ordo as an orchestration layer between MapPrism and n8n.

Ordo owns job state. n8n workers read pending steps from PostgreSQL and run them. The frontend doesn't talk to n8n at all.

From a product perspective, nothing changes:

  1. Upload to MinIO
  2. POST to start processing
  3. Poll for status
  4. Read output from MinIO

The change is in steps 2 and 3.

What changed: API ​

v1 API ​

POST /webhook/n8n
GET  /progress/{job_id}

Submit a job (v1):

json
POST /webhook/n8n

{
  "job_type": "POTREE_CONVERSION",
  "payload": {
    "input": "development/pineydam/mbes.las",
    "output": "development/pineydam/mbes"
  }
}

Poll status (v1):

json
GET /progress/{job_id}

{
  "job_type": "POTREE_CONVERSION",
  "status": "completed",
  "progress": 100,
  "output": "development/pineydam/mbes"
}

v2 API ​

POST /ordo/jobs
GET  /ordo/jobs/:id
GET  /ordo/jobs
GET  /health

Submit a job (v2):

json
POST /ordo/jobs

{
  "recipe": { ... },
  "inputs": {
    "job:input_las": {
      "type": "las",
      "uri": "development/pineydam/mbes.las",
      "hash": "abc123"
    }
  },
  "params": { ... },
  "outputs": { ... }
}

Response: { "id": 4 }

Poll status (v2):

json
GET /ordo/jobs/4

{
  "job": { "id": 4, "status": "success", ... },
  "progress": { "percentage": 1, "completed_steps": 2, "total_steps": 2 },
  "steps": [ ... ],
  "artifacts": [ ... ],
  "outputs": [ ... ]
}

v1 job_type → v2 recipe mapping ​

v1 job_typev2 recipe nameNotes
POTREE_CONVERSIONpotree-conversionSee build-potree
ENTWINE_CONVERSIONept-conversionSee build-ept
COPC_CONVERSIONcopc-conversionSee build-copc
POTREE_CONVERSION_AND_REPROJECTIONreproject-potreeSee reproject-and-build-potree
ENTWINE_CONVERSION_AND_REPROJECTIONreproject-eptSee reproject-and-build-ept
DATASET_INFO(not a standalone recipe)Dataset info now runs in parallel within all recipes automatically

Submitting the equivalent of a v1 POTREE_CONVERSION ​

v1:

json
POST /webhook/n8n
{
  "job_type": "POTREE_CONVERSION",
  "payload": {
    "input": "development/pineydam/mbes.las",
    "output": "development/pineydam/mbes"
  }
}

v2:

bash
curl -X POST https://dev.mapprism.com/ordo/jobs \
  -H "Authorization: Bearer <API_TOKEN>" \
  -H "Content-Type: application/json" \
  -d '{
    "recipe": {
      "name": "potree-conversion",
      "version": "3.0.0",
      "definition": {
        "recipe": [
          {
            "id": "build_potree",
            "type": "BUILD_POTREE",
            "inputs": { "input_las": "job:input_las" },
            "outputs": { "output_potree": "step:build_potree.output_potree" }
          },
          {
            "id": "dataset_info",
            "type": "DATASET_INFO",
            "inputs": { "input_las": "job:input_las" },
            "outputs": { "metadata": "step:dataset_info.metadata" }
          }
        ],
        "on_exit": {
          "id": "call_webhook",
          "type": "CALL_WEBHOOK",
          "inputs": { "waits_for": "step:dataset_info.metadata" },
          "param_keys": ["webhook_url"]
        }
      }
    },
    "inputs": {
      "job:input_las": {
        "type": "las",
        "uri": "development/pineydam/mbes.las",
        "hash": "abc123"
      }
    },
    "params": {
      "call_webhook": {
        "webhook_url": "https://myapp.example.com/hooks/done"
      }
    },
    "outputs": {
      "step:build_potree.output_potree": {
        "path": "development/pineydam/mbes"
      }
    }
  }'

Submitting the equivalent of a v1 POTREE_CONVERSION_AND_REPROJECTION ​

v1:

json
POST /webhook/n8n
{
  "job_type": "POTREE_CONVERSION_AND_REPROJECTION",
  "payload": {
    "input": "development/pineydam/mbes.las",
    "output": "development/pineydam/mbes",
    "source_epsg": "EPSG:2271",
    "target_epsg": "EPSG:3857"
  }
}

v2: Use the reproject-potree recipe and add source_epsg/target_epsg under params.reproject. See reproject-and-build-potree.

What's new in v2 ​

Dataset info is automatic. All preset recipes now run DATASET_INFO in parallel with the main conversion step. You no longer need to submit a separate job to get metadata.

Webhook is built in. All preset recipes include CALL_WEBHOOK in on_exit. Pass webhook_url under params.call_webhook to receive a notification when the job completes.

Full job state on poll. The GET /ordo/jobs/:id response includes step-level status, per-step timing, all artifacts, and output destinations — much richer than the v1 progress endpoint.

Composable pipelines. You can write custom recipes that combine any registered executors in any order. See Writing a Recipe.

Base URL change ​

Environmentv1v2
Devhttps://dev.mapprism.com/webhook/n8nhttps://dev.mapprism.com/ordo/jobs

Auth change ​

v1: auth was embedded in the webhook endpoint configuration.

v2: all requests require Authorization: Bearer <API_TOKEN> header (except GET /health).