These are chat archives for fgpv-vpgf/rcs

5th
Jul 2018
Dan Bowerman
@dan-bowerman
Jul 05 11:11
I think we've gone the "hand over the hacked .py file to Joost" route before and it went OK. A simple file overwrite. It should tide them over just fine until the 3.0.0 upgrade, which will probably be a bigger deal on their end.
i was able to register that service just fine
Barry Yung
@barryytm
Jul 05 17:20
Yes, it was working fine for me too. I had tried the layer Joost provided and it was also working. I wonder if the issue was coming from the metadata service.
Mike Weech
@mweech
Jul 05 17:24
its possible, but since the metadata is sourced from the catalogue itself, it would be odd. at the very least, we have a WOMM / cant reproduce issue
Mike Weech
@mweech
Jul 05 17:35

thought on rcs: we basically validate either strict, or loose right now. its a binary decision. is it reasonably trivial to look at our validation rules and identify ones that would make sense to turn on/off individually?

i.e. when registering a layer, if the metadata url or catalogue url's dont work as expected, the registration fails completely. it may be valuable to allow an administrator to (temporarily?) disable metadata URL validation in order to push through the registration process. this would get the layer added and working, and the outcome would be a metadata URL that MAY not work as expected in the viewer.