[Bf-committers] Joining objects and merging UV Layouts has Changed in Behaviour (2.69)

2013-10-15 Thread Gaia
Hi; For 2.69 when joining objects, then the UV maps are joined by name instead of by index. While this sounds like a good change unfortunately this results in an unpleasant follow up issue: Assume you have 2 objects, each with one single UV Layer, but both UV Layers are named differently. Now, af

Re: [Bf-committers] Joining objects and merging UV Layouts has Changed in Behaviour (2.69)

2013-10-15 Thread Bastien Montagne
Hi Gaia, Working on adding the option to the Join operator. But I really don’t think this would make it into 2.69, this is not strictly speaking a bug, and that kind of change would be more than risky to introduce at this point. Bastien On 15/10/2013 10:49, Gaia wrote: > Hi; > > For 2.69 when

Re: [Bf-committers] Joining objects and merging UV Layouts has Changed in Behaviour (2.69)

2013-10-15 Thread Brecht Van Lommel
Why not rename the UV layer before joining, so that it matches? There's also vertex groups, shape keys, vertex colors, should all those have options to join by index too? I think it's better if they all just join by name instead of adding options here. If you choose a UV layer in a material that i

Re: [Bf-committers] Joining objects and merging UV Layouts has Changed in Behaviour (2.69)

2013-10-15 Thread Bastien Montagne
Why not simply add a toggle to join cd layers either by name or by indices (with first option by default)? I think it can be useful. And we also use CustomData_merge in other contexts than join op, where we do not need name comparison. So as the change is pretty simple (will have a patch very s

Re: [Bf-committers] Joining objects and merging UV Layouts has Changed in Behaviour (2.69)

2013-10-15 Thread Bastien Montagne
Here is proposed patch: http://www.pasteall.org/46507/diff On 15/10/2013 12:20, Bastien Montagne wrote: > Why not simply add a toggle to join cd layers either by name or by > indices (with first option by default)? I think it can be useful. And > we also use CustomData_merge in other contexts th

Re: [Bf-committers] Joining objects and merging UV Layouts has Changed in Behaviour (2.69)

2013-10-15 Thread Brecht Van Lommel
The reason why not is that the index of layers is not supposed to have any meaning, so there also should not be any option to join by index. The question should not be "why not" but "why", I still don't see a good argument to join by index? It's not because an option is simple to add or can be use

Re: [Bf-committers] Joining objects and merging UV Layouts has Changed in Behaviour (2.69)

2013-10-15 Thread Gaia
I understand that joining by index is not so good. And yes, i can rename the UV Layers before joining to avoid the problem. But honestly i believe this behavioral change of Blender is not completely implemented. I believe that adding a "Combine UV Layers" function is necessary to allow this change

Re: [Bf-committers] Joining objects and merging UV Layouts has Changed in Behaviour (2.69)

2013-10-15 Thread Dalai Felinto
Hi Gaia, What would the "Combine UV Layers" operator do exactly? How to solve conflicts in a predictable way when the same face has UV information in multiple layers. > this would allow old behavior which works very well in most cases (since i guess most objects only have one layer) For most of

Re: [Bf-committers] Joining objects and merging UV Layouts has Changed in Behaviour (2.69)

2013-10-22 Thread Bastien Montagne
Another complain about new behavior (this time, he’d like to *not* merge his UVMaps…): https://projects.blender.org/tracker/?func=detail&atid=498&aid=37182&group_id=9 I really think we should allow that as well (so merge by name, merge by index, no merge), does not really complicate that much t

Re: [Bf-committers] Joining objects and merging UV Layouts has Changed in Behaviour (2.69)

2013-10-22 Thread Dalai Felinto
The report doesn't make much sense. I replied it there. Not that we shouldn't have per-id merging (that be still be discussed) , but his example doesn't makes a valid point for that in my opinion 2013/10/22 Bastien Montagne > Another complain about new behavior (this time, he’d like to *not* me