the first approach should work but you can not draw all the indices at their location at the same time
it would be a bloody mess
of course you won't recognize much if you view the whole model but you can simply zoom in with the camera. it shouldn't be a big problem to find the vertices you want then. i did this a few times already.
The second approach seem to be faster but I wonder whether ět really works
I mean, the coords which you read in your editor and the figures stored in the "Vertices" array are, of course, related numbers but are they also the same numbers?
If I import the same model from different file formats I get different scale factors and different global coords
Did all file format store exactly the same local coords for the same model ?
what do the coords have to do with it? scale doesn't matter.
if the indices stay the same depends on your modeler and pipeline. i guess with some it works, with some not.
by the way, i think most mentioned examples better get done with bones and not by manipulating vertices.
