Testing 0.14 migration of PartDesign features to Body

[jrheinlaender]

Hi,

in 0.14 all PartDesign features need to be migrated to nest under the Body feature. So I implemented an automatic migration... but now I need parts to test that. And since the parts I devise myself will probably always migrate without a hitch I'd welcome any of your parts that you can spare for testing.

A related question is what the migration should do with parts that have mixed Part and PartDesign features?

Thanks,
Jan

[jriegel]

A related question is what the migration should do with parts that have mixed Part and PartDesign features?
Jan

Uh, thats a tough one!
I think the non-PartDesign features have to stay outside of the Body, as a kind of external reference. Maybe sort them in a group with the same name as the Body.

In Catia this is calles hybrid design.

[NormandC]

I have a bunch of moderately complex parts there. https://github.com/normandc/Lulzbot-TK- ... ster/FCStd

[jrheinlaender]

I have a bunch of moderately complex parts there

Thanks for the parts. I noticed that a lot of the sub-features of MultiTranform are missing references e.g. mirror planes. Specific example: Bearing_Holder.fcstd is missing the mirror plane reference for feature "Mirrored".

[NormandC]

I have no idea why. Those parts were modeled 7 months ago on 0.13.1764, possibly since then the Mirrored feature was changed?

[jrheinlaender]

OK I think I found out what happened. The first version of Mirrored etc. allowed to select just the "vertical direction" as mirror plane in a Property called StdMirrorPlane. Later logari81 pointed out that "vertical" makes no sense because someone might change the Placement manually. So on Jan. 9th he pushed these changes among others:

remove StdDirection,StdAxis and StdMirrorPlane properties
support sketch H_Axis, V_Axis, N_axis as valid references in patterns

So probably your parts made use of the three properties mentioned above.

Meaning that you probably need to check through them all and select "vertical sketch axis" or "horizontal sketch axis" as the reference. Otherwise they fail on recompute.