OffsetSrf bug?

I tried to run a OffsetSrf (thickness 0.5 to vertices sharp edges) but there is a defect in the surface.
To check please.
OffsetSrf bug.3dm (99.4 KB)


Your modeling tolerance is 0.05 and it is used by _OffsetSrf. If you set it to Tolerance=0.001 you will get a better result (which is still not 100% correct). To get a usable result, split the side surface by isocurve as shown in below file, then perform the offset and when done, run _MergeAllFaces.

OffsetSrf.3dm (280.0 KB)


How ever the shell (thickness 0.4) of one of the faces (for example, the rear) works only with tolerance of 0.3? if you increase the shell it does not work anymore.
I think there are still many things to fix in this command.

If i set my tolerance globally or in the _OffsetSrf command to 0.001 and have the side surface splitted by isocurve as shown above, the offset works. _MergeAllFaces then cleans up the mess (which i do not dispute).

Apart from this i really like to understand why users are messing with tolerances if things do not work as expected. If you are modeling a small part like this and have chosen the factory template for small objects and millimeters as units, why do you change the absolute tolerance from the default = 0.001 to 0.05 and start modeling with this tolerance ?