Example of the CurveTool workaround:
fr = nuke.FrameRange(1, 100, 1)
ct = nuke.nodes.CurveTool()
ct['label'].setValue("Temp curvetool to update frames")
for frame in fr:
nuke.execute(ct, frame, frame)
# ...vertexSelection stuff...
nuke.delete(ct) # Clean up hacky CurveTool
Theoretically this should be the same as doing nuke.frame(2)
nuke.frame(3) etc, but.. as Frank described, it's not.
> Afterwards don't forget to log a bug report to get this resolved
> finally. Maybe we could ask for a flag to set that forces Nuke to
> re-evaluate during a python loop, then set another flag to turn off that
> behaviour again if it means staying more efficient.
It seems like nuke.frame(framenum) should trigger the update always..
but I guess there might be cases where this wouldn't be desired..?
Alternatively nuke.frame(framenum, refresh=True) or something, or just a
generic nuke.refresh() type method that could be called might be a good
compromise
On 27/02/14 06:47, Frank Rueter wrote:
> Many of us have been fighting this for years.
> One of the quick& (very) dirty workarounds is to create a dummy node
> that can be executed (i.e. CurveTool), execute it to force Nuke to
> re-evaluate the stream, then delete it. Do this for every
> iteration/frame and you should get the correct values if I remember
> correctly.
> I don't think the dummy node has to be connected to the nodes you are
> working with but haven't done this in a while myself.
>
> Afterwards don't forget to log a bug report to get this resolved
> finally. Maybe we could ask for a flag to set that forces Nuke to
> re-evaluate during a python loop, then set another flag to turn off that
> behaviour again if it means staying more efficient.
>
> Cheers,
> frank
>
> On 27/02/14 06:35, Elias Ericsson Rydberg wrote:
>> Maybe you could force it by adding a writegeo downstream. I would imagine
>> that would make nuke evaluate geometry positions upstream?
>>
>> Maybe adding the positions as metadata in the stream could be a valid
>> alternative.
>>
>> Cheers,
>> Elias
>>
>> 26 feb 2014 kl. 12:56 skrev Carl Schröter <ad...@l-rac.de>:
>>
>>> Hi guys,
>>>
>>> I've been experimenting with vertexSelection lately (in nuke7) and noticed
>>> something strange (at least it seems strange to me).
>>>
>>> When I run the code below on selected vertices of an animated abc file,
>>> Nuke does not update the vertex-position values in the loop. However, if
>>> gives me the correct position when I execute it on single frames. To make
>>> it even stranger: when I connect a scene-node to the ReadGeo, the code
>>> outputs 2 points for the same selection on different positions in space.
>>>
>>> ##########
>>> temp = nuke.nodes.CurveTool()
>>> for frame in range(1100,1150):
>>> nuke.execute(temp, frame, frame)
>>> nuke.frame(frame)
>>>
>>> selectedVertexInfo = nukescripts.snap3d.selectedVertexInfos() #it's the
>>> same for selectedPoints()
>>> print nuke.frame()
>>> for point in selectedVertexInfo:
>>> print point.index, point.objnum, point.position, point.value
>>> print
>>>
>>> nuke.delete(temp)
>>> ##########
>>>
>>> To make my confusion complete: it sometimes (very rarely) just works as
>>> expected...
>>>
>>> Is there something I am missing here? A way to force update the 3d-system?
>>> Also is there a way to get the geo containing node from selected points (so
>>> I could temporarily disconnect the scene node downstream)?
>>>
>>> Any help is greatly appreciated!
>>> Cheers,
>>> Carl
>>> _______________________________________________
>>> Nuke-python mailing list
>>> Nuke-python@support.thefoundry.co.uk, http://forums.thefoundry.co.uk/
>>> http://support.thefoundry.co.uk/cgi-bin/mailman/listinfo/nuke-python
>> _______________________________________________
>> Nuke-python mailing list
>> Nuke-python@support.thefoundry.co.uk, http://forums.thefoundry.co.uk/
>> http://support.thefoundry.co.uk/cgi-bin/mailman/listinfo/nuke-python
>
>
>
> _______________________________________________
> Nuke-python mailing list
> Nuke-python@support.thefoundry.co.uk, http://forums.thefoundry.co.uk/
> http://support.thefoundry.co.uk/cgi-bin/mailman/listinfo/nuke-python
>
--
ben dickson
2D TD | ben.dick...@rsp.com.au
rising sun pictures | www.rsp.com.au
_______________________________________________
Nuke-python mailing list
Nuke-python@support.thefoundry.co.uk, http://forums.thefoundry.co.uk/
http://support.thefoundry.co.uk/cgi-bin/mailman/listinfo/nuke-python
