Phil Scadden <p.scadden@...> writes: > Why need a proxy? > > Well because usually you have your application (especially the host > page) on one server and geoserver somewhere else. Since the browsers started > tightening up security on XHR, the restriction has been that XHR request > need to reference an item in same directory as your OL page or in a > subdirectory of that. Obviously mapserver/geoserver will be somewhere > else. What OL needs to do is switch its http requests through a proxy. > In my system, I have > proxy.jsp in my main directory, with proxyhost pointing to it. So > instead of GET to http://myserver, it generates > proxy.jsp?url=http://mygeoserver > Proxy.jsp checks the URL to see if it is one my allowed hosts, opens the > url, fetches data and pipes back to OL.
This much I understand, except instead of writing the proxy script myself. I am having the webserver (Nginx) doing the proxying for me. I redirect /geoserver to http://otherbox/geoserver. This effectively "fools" OpenLayers into thinking my geoserver requests are in a subdirectory, but I'm not writing the proxy. This is dead standard web server configuration. With proxying implemented on the server, why would I need ProxyHost set in OpenLayers? Again, I don't think the problem is a proxy problem (but I am not 100% sure) because the same server (Ubuntu) serves the same page on the same FF 12.0 on a different machine (Windows) just fine. The identical page on FF 12.0 on a Windows client has a clickable map. In addition, a known working page (renders just fine with a clickable map on Ubuntu 11.04) does not have a clickable map on Ubuntu 12.04. The only time I have trouble is with FF 12.0 on Ubuntu 12.04 Danny _______________________________________________ Users mailing list [email protected] http://lists.osgeo.org/mailman/listinfo/openlayers-users
