12-30-2013, 11:08 PM
Yeah, it's possible that something (including me) updated the drivers since the last time I knew it worked. I just tried to roll back the driver, but that doesn't appear to be an option right now. So unfortunately, I can't experiment with that.
In my last post, I said that after I restarted the computer, it seemed to work the second time. I think I can understand why now.
The time it worked, I used a sphere primitive instead of a cube primitive.
When I went back in again with a cube primitive, I had the same error: All edges look like plain edges, regardless of whether they're selected or hard edges.
So this is where it gets really interesting. I started transforming my sphere into a cube, in the off chance that there was something unique about the sphere vs. cube shape itself. I cut off the top with a loop cut. I could still see the edges. I cut off the bottom with a loop cut. All of a sudden, I couldn't see the edges any more. I undid that last step and the highlighted edges came back. I started deleting the edges one at a time from the bottom, and it worked, until I got towards the end. It seemed like once the model had no triangles left (only polygons with 4+ sides) I couldn't see my edges any more.
If I start from a cube (where I can't see highlighted edges) and connect two corners so that there is at least one triangle, I can see my highlighted edges again.
So it takes a very specific set of circumstances (an object with no triangular faces) to make the edges disappear.
This behavior doesn't feel like a device driver problem to me at this point (but I could be wrong). It seems like something about the way the drawing code is structured maybe. I don't know. I'd go dig into the code myself, but I don't know anything about Erlang.
At any rate, including at least one triangular face seems to be the workaround, and that's something I can live with. (Though I'd much prefer getting it fixed, and if I can help with that somehow, let me know.)
As I start using the new version of Wings, if I discover anything else about this bug, I'll report it to this thread.
In my last post, I said that after I restarted the computer, it seemed to work the second time. I think I can understand why now.
The time it worked, I used a sphere primitive instead of a cube primitive.
When I went back in again with a cube primitive, I had the same error: All edges look like plain edges, regardless of whether they're selected or hard edges.
So this is where it gets really interesting. I started transforming my sphere into a cube, in the off chance that there was something unique about the sphere vs. cube shape itself. I cut off the top with a loop cut. I could still see the edges. I cut off the bottom with a loop cut. All of a sudden, I couldn't see the edges any more. I undid that last step and the highlighted edges came back. I started deleting the edges one at a time from the bottom, and it worked, until I got towards the end. It seemed like once the model had no triangles left (only polygons with 4+ sides) I couldn't see my edges any more.
If I start from a cube (where I can't see highlighted edges) and connect two corners so that there is at least one triangle, I can see my highlighted edges again.
So it takes a very specific set of circumstances (an object with no triangular faces) to make the edges disappear.
This behavior doesn't feel like a device driver problem to me at this point (but I could be wrong). It seems like something about the way the drawing code is structured maybe. I don't know. I'd go dig into the code myself, but I don't know anything about Erlang.
At any rate, including at least one triangular face seems to be the workaround, and that's something I can live with. (Though I'd much prefer getting it fixed, and if I can help with that somehow, let me know.)
As I start using the new version of Wings, if I discover anything else about this bug, I'll report it to this thread.