Thanks Mark.
I have written a routine that does exactly that (creating individual strings for each triangle side), but there are a few MX related problems:
1. To not run out of string labels (I am talking about triangulations which consist of numerous millions of triangle sides), I had to revert to using a limited number of strings with discontinuities / gaps in them.
(Even with discontinuities, there seems to be an undocumented limit to the number of points allowed in such a string).
2. Furthermore the 'strings' are duplicated where a triangle side is common to two triangles. To filter those out is not impossible, but cumbersome and slow (I have to check their 'directions' to ensure they are actually duplicates).
3. We use a coordinate sytem in South Africa which is similar to Chech2 (which allows negative values on the Y (East-West) axis).
GENIO throws a hissy-fit when I try to input negative values in the GENIO file, so I have to add a large constant if I have negative Ys.
4. When this model is transformed, MX throws a hissy-fit again, because the Transformation routine does not like negative Ys either!
(This is required to ensure that the MX model is actually correct according to WGS.)
So what you are telling me is that there is no ways of creating a triangulation model from raw data?
I have written a routine that does exactly that (creating individual strings for each triangle side), but there are a few MX related problems:
1. To not run out of string labels (I am talking about triangulations which consist of numerous millions of triangle sides), I had to revert to using a limited number of strings with discontinuities / gaps in them.
(Even with discontinuities, there seems to be an undocumented limit to the number of points allowed in such a string).
2. Furthermore the 'strings' are duplicated where a triangle side is common to two triangles. To filter those out is not impossible, but cumbersome and slow (I have to check their 'directions' to ensure they are actually duplicates).
3. We use a coordinate sytem in South Africa which is similar to Chech2 (which allows negative values on the Y (East-West) axis).
GENIO throws a hissy-fit when I try to input negative values in the GENIO file, so I have to add a large constant if I have negative Ys.
4. When this model is transformed, MX throws a hissy-fit again, because the Transformation routine does not like negative Ys either!
(This is required to ensure that the MX model is actually correct according to WGS.)
So what you are telling me is that there is no ways of creating a triangulation model from raw data?