Ordering of formally (derived?) attributes could use improvement

[Moult]

I recently found that IfcGeometricRepresentationSubContext pages were erroring out. After fixing that (I hope my fix is OK I'm not too familiar with the UML yet) I found the ordering quite unintuitive. I suspect this is coming from the ordering in UML, I assume there is some sort of "explicit" order for express output. However for a user the derived attributes come in a very strange order of 5, 3, 6, 4. Is this something to fix in UML?

[aothms]

Yeah well there are two kinds of derived attributes. I indeed forgot to check up on the "derived in supertype" ones which is really the devil.

These should actually be rendered instead of the attributes in the supertype.

So instead of

We should try to render them in their original place. Prepend the derivation formula. Maybe mark up to this level as non-inherited. And add tons of <blink> tags.

The difference between the cute derived attributes and the terrible ones is this SELF\ in the name.

I guess we can also just append them to the attribute table, but really merging them in their correct order would make it most clear to the readers what's going on and how annoying this is.

Would you be willing to work on this?