Hi Andrea, thanks for you answers, inline some comments.
> end up with a "EPSG:404000" (a joke on HTTP 404) and mosaic in older > versions of GeoServer could not handle that. Check out 2.14-RC, that should > mosaic files without georeferencing... althought I'm guessing it's not > what you want. > > Also, I'd recomment not to use ArcGrid sources for anything but small > experiments, > reading from text files is slow. > With no requirements against it, I'd gdal_translate everything and add a > CRS into in the process. Geoserver here is just a small part in a tool-chain in which another old (fortran) lib generates/models predictions about the route of a cloud of smoke. The output historicaly is .asc as it was totally not gis-aware. My plan was to use a simple python script which would use the asc files and via Geoserver-REST then would register a mosaic and a time-dimension-aware WMS-T layer. To be picked up by one of this beautifull simple leaflet client http://apps.socib.es/Leaflet.TimeDimension/examples/ Adding gdal to this mix would make it a little more complicated, so I was hoping to not need to translate any of the output. My 'problem' is that if I think that something should work, I tend to mixup the (dutch?) rule: "why make it difficult, if you can do it easy" :( > Apparently the geotools lib is not aware (or does not get) the crs from > the indexer.properties file? > > I believe it does not, and was never meant to. Is the documentation > suggesting that it should? > Maybe there is a misunderstanding on "MosaicCRS", which is the "mosaic CRS", > not the granules one, and it's used when you have multiple crs in input > (as we it has to pretend the mosaic to be uniform, it needs to know > which one to use). > If this is the one, could you please suggest how to reword its > description to avoid confusion? No, documentation/info on google is clear: don't use asc files :-) But indeed reading the release notes of 2.14 I was encouraged to try it (again...)..And as said: for our toolchain it would be easiest if I could point Geoserver straight to the fortran output dir. > The link you are providing does not work, did you use this? > http://spatialreference.org/ref/epsg/28992/ogcwkt/ > > Also, there is millions of ways of "does not work", did not work how? >From http://spatialreference.org/ref/epsg/28992/ I took the .PRJ File option. But I now see that the OGC WKT file is probably better... My 2 hurdles: - [fixed] first I failed to register a non epsg:4326 asc file (as in: the generated index shape did not take up all granules, but just one. But that one is fixed by adding the .prj files ... - [open] the second 'does not work', which I think should work (?) is just creating a WMS from a mosaic from the Geoserver data example .asc file, see https://osgeo-org.atlassian.net/browse/GEOS-8912 You can registre the store and layer, but the WMS fails. My gut feeling first was that Geotools is messing up coordinate transformations, but later I found that even the 4326 example did not create a working wms (as in Layer Preview shows an error). I'll try to do some further debugging. I think for me it is wisest to use GeoTiffs indeed, but I still think GEOS-8912 is a valid issue ;-) Regards & thanks again for all the Geoserver work! Richard Duivenvoorde ------------------------------------------------------------------------------ Check out the vibrant tech community on one of the world's most engaging tech sites, Slashdot.org! http://sdm.link/slashdot _______________________________________________ Geoserver-devel mailing list Geoserver-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/geoserver-devel
