rBc6cbcf83d015 caused to early out when not all faces were selected (but
surrounding faces were, so implicitly all vertices were selected). Now
take face selection mode into account.
Details
Details
Diff Detail
Diff Detail
- Repository
- rB Blender
- Branch
- T71554 (branched from master)
- Build Status
Buildable 5669 Build 5669: arc lint + arc unit
Event Timeline
Comment Actions
Wouldn't this need an edge check as well? For this case? (white edge is not selected)
Comment Actions
"All edges" selected, but one (or more) verts aren't:
"All faces" selected, but one (or more) verts aren't:
Sorry, I really hate to be "that guy" commenting over the shoulder, but, looking at the logic, I think these cases are still not handled (i.e. it'll bail too early):
"All faces" selected, but one (or more) edges aren't:
"All edges" selected, but one (or more) verts aren't:
"All faces" selected, but one (or more) verts aren't:
...and etc. with all three selection modes active.
Comment Actions
Sorry, I really hate to be "that guy" commenting over the shoulder, but, looking at the logic, I think these cases are still not handled (i.e. it'll bail too early):
"All faces" selected, but one (or more) edges aren't:
"All edges" selected, but one (or more) verts aren't:
"All faces" selected, but one (or more) verts aren't:...and etc. with all three selection modes active.
To me, having the "that guy" commenting over the shoulder is actually the strength of OSS [ this way we catch issues early on :) ]
Will check again if the logic in EDBM_mesh_hide will actually handle your cases well (or if we'd need changes there, too...)
Comment Actions
If the implementation hasn't changed since 2.79, it should, as these cases work there.
Comment Actions
Since modes can be mixed, the order of these checks should be reversed and converted to else if.