Ok, here goes. I’ve been trying to get my head around the best way to build support for some planning processes we need to invent to support an infrastructure build project.
Today I’ve got my hands on a load of GB Ordinance Survey GML data which we want to do our planning around. I’m trying to determine what the “best” way of managing this data is and how to make it available to the team. The team are currently using a KML based planning tool, and so I ultimately need to offer a dynamically windowing KML presentation layer for them to use. How do I go about doing that? Rolling up my sleeves, my go-to solution is to write a dedicated importer for the OS data GML and create Neo4j spatially indexes nodes for each feature, and then feed that into a KML rendering layer that we’ve already developed in Java (Groovy). But before I head down that route I’m wondering whether I’ve missed a trick. Should I be using GeoServer/GeoTools as a platform from which to manage the heavy lifting? We already have an existing graph which describes a whole load complex data connected to a “GIS” spatial layer in Neo4J. We serve that to KML and that works well. I suspect that in time I’ll want us add a WFS API to that data set, so that we can use some standardised mapping tools, or export a view of our network a bunch of third parties who are working with us with their own tools. I imagine I’ll do that by implementing a GeoServer data-store which interacts with the REST API of our “Neo4j”-backed “Database”. But today, I’m wondering about this O/S GML base maps. It doesn’t look like GeoServer can import GML today - there is a plug in but it appear to be unsupported and probably broken. So, to it into GeoServer will require writing a load of code in GeoServer to make it work, or converting my GML to something that it can import (What?) with either a hand crafted tool or something preexisting (what? I can’t believe that others haven’t already done this). If that’s easy then it’s a no brainer - I’ll use a GeoServer to export KML for the base maps that underlie our planning process. If it’s going to be fiddly I’d probably be better off just implementing it all from scratch in our own code with Neo4J spatial. I’ve no idea what direction I should take. I would really appreciate anyone’s thoughts on this - has anyone here been down either of these routes before and can advise on the potential pitfalls? Many thanks, Joe -- You received this message because you are subscribed to the Google Groups "Neo4j" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. For more options, visit https://groups.google.com/d/optout.
