i am using worldbuilding pasta/ Artifexians methodology to create a fictational planet and i ran into an error. i moved a plate and then connected one plate to another by having it follow a that plate. after i moved the plate i saved the rotation file then opened it in notepad++. after changing the rotation file i saved it and reloaded it in to gplates and then went back to working on the rest of the project. a bit later i hit ctrl-m and saved my changes and notice that the rotation file need to be saved so i saved it too. immediately a bunch of the features that used half stage-rotation jumped around the globe. i realized after ward that i had not move anything since i reloaded the rotation file. i also got a pop up explaining i had a read error for a bunch of the rotation file lines that error is the same for all. “stage rotation takes long path” this has broken my animation and i don’t have an older rotation file to show me what it was like before.
i can’t even find a list of errors explaining exactly what that error means and why it would only apply to some lines not others. can anyone help me fix this?
i did notice that many of the lines have high negative numbers well over -180 and none of the still working lines are like that but i don’t know if that is relevant or just a Coincidence.
That message would have happened when you reloaded the rotation file. And because the reload adjusted the rotations (as explained below) it shows up as an unsaved change. So those reload changes plus whatever subsequent changes you made would require saving (or discarding) before you exit GPlates.
On loading (or reloading) a rotation file, GPlates checks the stage rotation of each finite rotation (each line in rotation file) relative to the previous finite rotation, and if it takes the long path around the globe then it adjusts the stage rotation (and hence the finite rotation) to take the short path. This was introduced in the early days of GPlates to avoid issues that cropped up in users rotation files from time to time.
This short-path-adjustment doesn’t affect (and is no longer needed) for regular reconstruction (ie, non half stage rotations) because when the reconstruction time is between two finite rotations it will interpolate them in a way that always takes the short path (even if the stage rotation between them were to take the long path).
However, if you explicitly calculate a stage rotation between two finite rotations then it’ll be either the long path or short path (ie, whatever the stage rotation between them is). So the (re-)loading code explicitly makes those stage rotations take the short path. And this will affect features that use half stage rotations, which could explain why your features jumped around.
I’m surprised that they jumped around when you saved. They should have jumped around when you reloaded (since that’s when the short-path adjustments were made).
The valid range of pole longitudes is [-360, 360], so that’s fine. And it probably is a result of the short path adjustment.
I don’t know why your rotations are taking the long path - it sounds like it might related be the changes you made to the rotation file in notepad++.
When you make the changes in notepad++ you could try changing the rotation angle (of your modified finite rotation) to be angle - 360 (if it’s positive) or angle + 360 (if it’s negative). This will switch from long path to short path (or vice versa) - see here. Then see if you still get the “stage rotation takes long path” warning when you reload it. That way you can adjust it to remove the warning.