Re: [Geoserver-devel] Possible bug in virtual services implementation

2016-06-17 Thread Nuno Oliveira
Ian, You said: "But if I'm using a workspace virtual end point it's for a good reason and I'd expect the user not to be able to bring in other workspace layers." The user will only see the local workspace layers, there is no way to bring a layer from another workspace. When requesting a layer if

Re: [Geoserver-devel] Possible bug in virtual services implementation

2016-06-17 Thread Ian Turton
It feels like a bug to me - presumably if I had set user restrictions on the other workspace they still can't see it. But if I'm using a workspace virtual end point it's for a good reason and I'd expect the user not to be able to bring in other workspace layers. So, not critical but looks buggy I

Re: [Geoserver-devel] Possible bug in virtual services implementation

2016-06-17 Thread Andrea Aime
Hi Justin, thanks for the feedback. Wondering, what others people think about this one? I'm not really hard pressed to fix it anyways, but was wondering if others consider this a bug to start with ;-) Cheers Andrea On Wed, Jun 15, 2016 at 2:23 PM, Justin Deoliveira wrote: > I believe this woul

Re: [Geoserver-devel] Possible bug in virtual services implementation

2016-06-15 Thread Justin Deoliveira
I believe this would have been me and no, I don’t think there was an explicit need for this case, at least none that I can remember. It was just a decision to be lenient in this case. The rationale being that since you specified the workspace/namespace container in the url any namespace prefix on t

[Geoserver-devel] Possible bug in virtual services implementation

2016-06-15 Thread Andrea Aime
Hi, today I was chatting with Nuno and he referred to some code that blindly removes layer prefixes in virtual services, making a request like this one, actually work: http://demo.geo-solutions.it/geoserver/tiger/wms?service=WMS&version=1.1.0&request=GetMap&layers=topp:giant_polygon&styles=&bbox=-