[PROJ] Pseudo-mercator

[Even Rouault]

Andrew,

for anyone to be able to help you, you should provide more reproducible 
material, and namely the WKT that works and which don't. and possibly 
GDAL or PROJ command line invokations to reproduce issues

Even

Le 01/12/2021 à 21:24, Andrew Patterson a écrit :
> Hello!
>
> We recently made the jump from GDAL 2.4 to 3.3.0 and while we had some 
> major headaches initially, in the end they boiled down to a single 
> (unnecessary) function call that resulted in an issue filed against 
> GDAL. The fix, by the way, was supplied within three or four hours, 
> which is an amazing turnaround time (and greatly appreciated by me)!
>
> Since then, we've released our app (Avenza Maps) and immediately run 
> headlong into recurring issues with pseudo-mercator that I'm not sure 
> how to solve. In our previous version (using GDAL 2.4 & PROJ 6.1) we 
> had various pseudo-mercator maps that worked fine. When we upgrade to 
> our version running GDAL 3.3.0 & PROJ 8.1.1, many of these maps find 
> their referencing shifted. In one case, it was shifted toward the 
> equator ~26km. In another, the referencing is off by ~75km.
>
> The former has 'deprecated' in its name -- though I know that's mostly 
> a cosmetic property -- but it made me wonder if there's some issue 
> with these deprecated versions of pseudo-mercator. We've seen other 
> issues with these 'off brand' pseudo-mercator projections as well, but 
> it's not clear if it's that they're not as well supported any more or 
> if there's something in those WKTs that is problematic.
>
> The latter one seems to be influenced by what appears to be a dropped 
> parameter: the WKT specifies a latitude_of_origin but when I asked the 
> resultant ORGSpatialReference to read the WKT back to me it's not 
> included. If I eliminate that parameter from the WKT in our previous 
> app (using 2.4/6.1) I get a similar-ish result to the one in the 
> current version (using 3.3/8.1.1). That seems like it might be a bug 
> about not considering latitude_of_origin for pseudo-mercator when it 
> should, but even if that's true, the latitude is off by ~0.15 degrees 
> so there's an additional component.
>
> To further confuse matters, we also have other pseudo-mercator maps 
> that work as well as before. I'm well versed enough in GIS from years 
> of exposure that I feel like I'd be able to spot obvious problems, but 
> I'm not seeing anything that jumps out at me. I'm hoping someone on 
> this list has experience and knowledge of how PROJ handles 
> pseudo-mercator and can lend me some guidance on what to do. I've 
> figuratively got management breathing down my neck on this and I'm 
> running out of leads to pursue.
>
> Any help would be greatly appreciated!
>
> ..............................
> Andrew Patterson
> Lead Software Architect
> Avenza Systems Inc.
>
> email: andrew at avenza.com <mailto:andrew at avenza.com>
> phone: 416.487.5116
>
> _______________________________________________
> PROJ mailing list
> PROJ at lists.osgeo.org
> https://lists.osgeo.org/mailman/listinfo/proj

-- 
http://www.spatialys.com
My software is free, but my time generally not.

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.osgeo.org/pipermail/proj/attachments/20211201/ec564a1e/attachment.html>