Adding custom projection to have NEU doesn't work in versions after 3.8

[antoniolocandro]

Describe the bug

Custom projection worked in QGIS 3.8, I can't add it in newer versions included dev it doesn't matter what I do
How to Reproduce

Additional context

[nyalldawson]

Try doing what the error suggests, and changing the format combo box from "proj" to "WKT".

[antoniolocandro]

Tried to WKT also get the same issue, tells me something about EPSG 4326 exists, actually I only need this as a workaround because a GML i am loading is written in lat,lon but apparently QGIS doesn't honor axis order

[antoniolocandro]

Error message trying WKT, only thing is that it worked before in 3.8 which I downloaded for the time being

[nyalldawson]

Yep - follow the suggestion from the error message again.

it's there for a reason!!
😉

[antoniolocandro]

My problem is related to this issue #35393

As a workaround in QGIS 3.8 I used to be able to save this custom projection to swap coordinates of layers that were not correct, what would be the option in latest version, I would want to add +axis=neu for this layer while I can see what the issue is for the swap in the first place

I tried the suggestion

GEOGCRS["unknown",
DATUM["unnamed",
ELLIPSOID["WGS 84",6378137,298.257223563,
LENGTHUNIT["metre",1]],
ID["EPSG",6326]],
PRIMEM["Greenwich",0,
ANGLEUNIT["degree",0.0174532925199433],
ID["EPSG",8901]],
CS[ellipsoidal,2],
AXIS["latitude",north,
ORDER[1],
ANGLEUNIT["degree",0.0174532925199433,
ID["EPSG",9122]]],
AXIS["longitude",east,
ORDER[2],
ANGLEUNIT["degree",0.0174532925199433,
ID["EPSG",9122]]]]

test works I want to invert 2,3 to 3,2

I am able to save it

But when I apply it to the layer, nothing happens to it (east and north should switch for visualization and this worked in QGIS 3.8) my data is in EPSG:4326 BTW. When I did this in QGIS 3.8 the features axis would swap

Thanks for any ideas

[gioman]

QGIS 3.8

QGIS 3.8 shipped with GDAL 2, QGIS 3 LTR and above ship with GDAL 3. It is a major change, you should first check if the issue lies in this underlying library.

[antoniolocandro]

Ok so reporting in the GDAL tracker would be the way to go I guess. Anyway this is happening at that level to be visualized in QGIS correct?

[gioman]

Ok so reporting in the GDAL tracker would be the way to go I guess

if is confirmed the issue is there.

[antoniolocandro]

My use case is no longer needed as it was fixed upstream on the rendering part Just for general knowledge fixed upstream OSGeo/gdal#2356