RC3 GetMap Request Renders Differently Depending on Bounding Box

I think I've fixed this one yesterday. You should try a nightly build, I don't think we're going to make another RC (the next one is likely to be 2.1.0 final):

http://gridlock.opengeo.org/geoserver/2.1.x/

Hmmm.. or maybe not.... those are rasters right?
If the nightly build does not fix the issue can you attach the rasters to this bug report?

Looked a bit more, this is not what I thought it was.
I need the data and the raster style you used to investigate it

Thank you Andrea. I am uploading a RAR file with the GeoTiff Raster and SLD that we used for this layer.

For a bit of background it is worth mentioning that we have evaluated each of the RC's (1,2,3). At RC1 we had trouble with the Vendor Option type="intervals", but at RC2 it worked and rendered correctly. Now at RC3 we are using the same sld as we did for RC2, but are encountering this rendering problem.

Thank you for your help with this.

The Geotiff is too big. I will put it on our FTP and send you a private message.

Simone, cc'ed, provided me with a patch that almost solved the problem. I made one further change and as far as I can see now it's working peachy (Simone, the case that was still not working was adding &bgcolor=0x000000 to the request, the green areas would still turn black in that case)

Simone, can you review this one?

More improvements, might want to check if this http://jira.codehaus.org/browse/GEOS-4360 is respected ( I have not been able to test )

See new patch

Simone, tried the patch, it looks good.
Kytt, is the dataset you sent me under some redistribution limit? I don't have a similar dataset and I'd need one to add a test to GeoServer (so that this issue won't appear again in the future). I would use only a small slice of it, somewhere around northern russia where the problem was more evident

The dataset is publicly available so feel free to use it with attribution if possible.

I am sorry that we haven't been able to test the patch yet. It will be another day or so before we are able to. Thank you both for your help. .

We tested the latest nightly build (4/1), but apparently the patch is not there??

The system administrators tell me that they are not able to apply the patch directly because we would need to recompile the application and we aren't configured to do that right away.

Is the patch in a build already that you know of?

Thanks, sorry this has taken some time.

Hi Andrea and Simone,

We have just updated the the RC4 released today and found that it did not fix the problem. Is it possible that the patch wasn't rolled in?

Kytt, correct, got a funeral in family and as a result did not manage to commit the patch

what is the plan for including this patch? when can we expect to see it in a build?

Latest Simone's patch committed on 2.1.x and trunk. Kytt, can you try it out? Either get a 2.1.x checkout and build yourself, or wait for tomorrow's nightly build.

Thanks Andrea.

I will test with tomorrow's nightly build and let you know.

Thank you for all of your help. The problem is fixed.

Thanks for letting us know

Mass closing all issues that have been in "resolved" state for more than a month without further comments