Loading data

The toolbar button opens the project picker. Settings live under Web > GIMLabs for QGIS.

Picking a project#

The dialog fetches the 100 most recent projects, not all of them. A firm can have thousands, and pulling the lot every time you opened the dialog would be dozens of blocking requests against a rate limit to build a list nobody scrolls.

Type in the search box to search by project reference or name. The search runs on the server, so it looks past the 100 you can see.

The status line always tells you what you are looking at, so a capped list never masquerades as the whole estate. Refresh projects re-fetches it.

Choosing layers#

Tick what you want before loading:

Layer	Default	Scope needed
Locations	On	Locations
Site boundary	Off	Map
Location groups	Off	Map

If your key is missing a scope the dialog says so on the row itself, rather than letting you tick something that will fail.

Everything loads into a layer-tree group named for the project.

Coordinates#

A location can carry its position three ways, and the plugin prefers them in this order:

1. Surveyed latitude and longitude, where your configuration maps those fields.
2. Surveyed easting and northing, which is what most UK ground investigation actually records.
3. Map placement, the position someone pinned on the map in Manage.

Surveyed coordinates beat a pin, because a pin is an estimate made to get a borehole onto a drawing and is often nowhere near the real hole.

The API does not publish which grid an easting and northing are in, so the load dialog asks. Grid for easting/northing defaults to EPSG:27700, British National Grid.

Every feature records which of the three placed it, and the load summary counts locations that could not be placed at all. A layer that quietly drops a third of its boreholes is worse than one that says so.

Refreshing#

Refresh layers re-fetches the project you loaded last and replaces its layer group.

Loading a project that is already on the map replaces its group rather than stacking a second copy, so loading twice refreshes rather than duplicating.

Working offline#

Responses and marker artwork are cached under your QGIS profile directory. A laptop that has loaded a project once can load it again with no signal, which on a site with no reception is the difference between having the data and not.

Three rules worth knowing:

• Only a network failure falls back to the cache. A permissions error or a missing project is a real answer, and serving stale data over it would hide the problem. You need to know your key lost a scope, not quietly keep seeing last week's boundaries.
• When cached data is used, the status line says so first, leading with how old it is, so nobody presents from stale data without being told.
• Cached responses expire after 7 days. Long enough for a week in the field, short enough that "offline" cannot quietly become "last season's data".

What you get in the attribute table#

Locations come in keyed on locationId, which is the LOCA_ID your team already uses. That is also the layer's display expression, so identify results and map tips show the location identifier rather than something internal.

No database identifier is carried as a field at all. QGIS offers every field in the layer when you pick a label, and there is no setting that removes one from that list, so the only way to stop someone labelling a map with an internal id is not to bring one in.

The layer also carries the columns the symbology needs, and QGIS would offer those as label fields too. They are therefore hidden by default in the attribute table and the form. They are still there: the renderer reads them, and you can unhide them if you want to see what it is doing.