[MC84]

The models display properly, and their clipshape works fine, however I cannot select them in 3d mode.

Last working revision was 7657. I'm using polymost. I noticed that in the changelog for the subsequent revision (7743) changes were made to improve mouse selection in the editor.

Thanks

[MC84]

I'm currently using r9257 and I can select 3d models again! Thanks guys, keep up the good work

[Mark]

In latest revision 9310 model aiming works in Polymost but has issues. I have to move the cursor above the model to select. Some models have a huge capture area which makes it impossible to select anything else near them. For the first time I am in a project that is not Polymer and thats why it took me so long to notice the issue is present again.

[Mark]

I decided to try 9257 in hopes the problem wasn't there. But it is. Model selection is fine in Polymer. Hopefully a pic will help to show the severity of the problem and get some attention. The white rectangle behind the trash can model is it's selection area. It starts about 20 percent of the way up from the bottom of the model and extends way above and out to the sides. Being so large is making it impossible to properly select other models and textures in the area. Especially bad in map interiors.

Attached thumbnail(s)

• 

[Mark]

For a test, I lowered the dummytile dimensions to smaller than I would normally use and the model's selection area shrunk down closer to where it should be. So that tells me that Polymost sets selection area by dummytile value but exaggerated larger, and Polymer sets selection area by the model's actual polygons for more precise aiming. For instance, in the above pic there was an error on my part due to copy/paste of the def lines. The trashcan accidently had a dummytile value of 64 64. But the selection area looks much larger than that.So maybe its a matter of finding out if somehow the code is detecting or using the dummytile values at twice what they actually are.

Or, I could be way off and have no clue what I'm talking about and its something else I'm doing wrong. But for now, to keep from pilling my hair out from frustration I'm going to lower my dummytile dimensions on the worst offenders.

[Mark]

YOW!!!! Replacing the C-9 cannister sprite with a model. The model selection area is humungous. See the arrow up top for the upper limit for selection.

Attached thumbnail(s)

• 

[Doom64hunter]

Okay so I do experience the inaccuracy of the model selection with the cursor when comparing polymost to polymer, but not to the extreme degree that you find.

Can you type "setrendermode 0" in the console and check how large the dummytile actually is? Because I think what's going on here is that regardless of the model's dimensions, the editor still uses the bounding box of the underlying sprite for the selection.

[Mark]

You may be on to something here.

For the case of the C9 which had a real large selection area, it was a replacement of the original tile and therefore had no dummytile dimensions in defs.
I used scale 1 in the defs for the model and that forced me to greatly increase the the size adjustment for the model in mapster. This caused the huge underlying sprite size. I increased the scale value of the model in the defs so I could reduce its size in mapster and that brought it in line with proper selection area. So I have to go back and increase scale in the defs for the offending models and then remember to make them larger in the first place next time.

Thanks for pointing out what should have been so obvious in the first place. I'm embarrassed I forgot to compare to sprite size. I've done it in the past. Working with Polymost for the first time clouded my judgement and I wanted to blame it right away. Shame on me.