Phil, Till, (moving this thread to the dev list),
looks like we are now in a situation where we can get things moving again. terrestris has an OpenLayers core developer (Marc Jansen), and Phil has the insight (see http://trac.osgeo.org/openlayers/ticket/2787). Previously, I have been supporting Phil with advice and patches, but had no time recently to continue doing so. The next thing to do would be for Till to test if the #2787 patches solve the issues, and then Marc can maybe add tests so we can commit #2787. Andreas. On Thu, May 19, 2011 at 11:43 PM, Phil Scadden <p.scad...@gns.cri.nz> wrote: > Welcome to the world of endless frustration. Who would have thought that > in the 21st century, programmers were still living on a flat earth. Back > engineering to fix this flat earth philosophy is a mega hassle and you > must be careful not to break use of mapping on non-earth applications, > ha ha. However... >> first problem: >> The measuring (and drawing) over the dateline does not work. There was >> an issue about exactly the same problem last august on the dev-List with >> an attached kind of "patch", but that one does not work properly >> (http://lists.osgeo.org/pipermail/openlayers-dev/2010-August/006360.html) >> and causes funny behavings. > I think I have more code patched in similar ways than are in that > patches, however there are other tricks that you need too. > Assuming you have put in ALL the patches associated with that ticket > 2782, http://trac.osgeo.org/openlayers/ticket/2787, then your > measureControl should also look like: > measureDistanceControl = new OpenLayers.Control.Measure( > OpenLayers.Handler.Path, { > persist: true, > geodesic: true, > handlerOptions: { > layerOptions: > {styleMap: drawStyleMap}, > citeCompliant :false > } > } > ); > measureDistanceControl.events.on({ > "measure": processMeasurements, > "measurepartial": > processMeasurements > }); > measureAreaControl = new OpenLayers.Control.Measure( > OpenLayers.Handler.Polygon, { > persist: true, > geodesic:true, > handlerOptions: { > layerOptions: > {styleMap: drawStyleMap}, > citeCompliant :false > } > } > ); > measureAreaControl.events.on({ > "measure": processMeasurements, > "measurepartial": processMeasurements > }); > > ie note geodesic= true, citeCompliant=false. In fact, make sure you have > these set for any of your path handlers (eg whatever you use in > DrawFeature). > > If this still isnt working, then contact me, as I may have more patches. > Note also that these patches havent made it to 2.11, - I simply havent > had the time to learn the test environment and make tests. >> Also the WMS-layers are not drawn over the dateline. We worked with the >> "wrap dateline" parameter, but using this requires tiling and tiling >> causes much more requests and images to be loaded by the client. We have >> about 30-40 WMS in the application. A tiling will slow down the >> application (especially the client) and at least lets crash slow >> browsers like IE8 (indeed - IE9 is a lot faster). > I havent really figured out what wrapdateline is supposed to do but > mostly it turns maps into wierdness with vector layers repeated out of > place. The biggest problem however seems to be having a wms server that > understands the dateline. Geoserver 2.1 and a patched version of > geoserver 2.02 are supposed to. I havent got 2.1 up yet but the patch > only sort of worked with a great deal of odd behaviour around polygons > that cross the dateline (depending on native projection) and tiles that > cut in and out depending on whether dateline is to east or west of your > screen midpoint. I'm not convinced yet that this isnt an openlayers bug. > I even requested quote from geosolutions for fixing bug in geoserver but > they have been very slow coming back. > > We had decided to defer this issue till geoserver 2.1 came out and then > work out precisely where blame lies and try and fix it ourselves. We are > in process of getting geoserver 2.1 up (glacial IT processes not > helping). Other WMS servers dont seem better. We've been screaming at > ESRI about 180 line since oh about ArcInfo version 4 (when we became > customers). Some things work but google projection isnt one of them at > least in 9.2. The last comment we had from them implied that since web > mercator didnt have envelopes, it would never work on 180 line. At the > point we dropped it and moved to geoserver which offered some hope and > we could get the source. > > The very best of luck and I am ready to help in any way I can. Feel free > to email direct. > -- > Phil Scadden, Senior Scientist GNS Science Ltd 764 Cumberland St, > Private Bag 1930, Dunedin, New Zealand Ph +64 3 4799663, fax +64 3 477 5232 > > Notice: This email and any attachments are confidential. If received in error > please destroy and immediately notify us. Do not copy or disclose the > contents. > > _______________________________________________ > Users mailing list > us...@lists.osgeo.org > http://lists.osgeo.org/mailman/listinfo/openlayers-users > -- Andreas Hocevar OpenGeo - http://opengeo.org/ Expert service straight from the developers. _______________________________________________ Dev mailing list d...@lists.osgeo.org http://lists.osgeo.org/mailman/listinfo/openlayers-dev
