The client ID registers this site with Google, so that Sign in with Google can ask for read access to your buckets. It is set up once, by whoever runs the site, in a Google Cloud project of your organisation. It is not a secret. It takes about five minutes.
https://www.googleapis.com/auth/devstorage.read_only (read only) and save.
.apps.googleusercontent.com) and paste it
into the box behind this window. The client secret is not needed.
To build it into the published site instead, so nobody has to paste it, set it as the
GOOGLE_CLIENT_ID variable of the GitHub repository (Settings ▸ Secrets and variables ▸ Actions ▸ Variables) and run the pages workflow again.
"Access blocked: origin mismatch" means the address in step 4 is not exactly this site's. "Permission denied" on a bucket means your Google account cannot read it.
Open files from this computer. They are read in your browser and never uploaded.
A link to a file: Parquet, CSV, GeoPackage, GeoJSON, FlatGeobuf, Excel, or a Zarr store. Large files are read in parts, as queries need them.
Reading from a database (PostgreSQL/PostGIS, and others) is not available yet.
A web page cannot talk to a database server directly, so this connector will go through the local GeoMarmot server, which can. In the meantime, export a table to Parquet or GeoPackage and open the file.