There isn’t much more advice to give, though do check the alternate directory thing - the installation guide lists the possible locations. Also do get a proper copy of python installed - again, the installation guide details how to do this. That is all there is to it I’m afraid - it will work, you just have to set it up right.
I made it work thanks… i just needed to tell blender the python of panda and not the one i instaled ^^.
Lethe I just installed R85 I noticed that the doc files are still at R7x. Any word on when they will be updated or are the changes so insignifigant that its not worth changing anything?
Doc files are always miles out of date I’m afraid - just look at the screenshots! Saying that the docs are updated with each release, and the current docs do go all the way up to R85 in the revision history, plus they have some updates to the main body so they at least mention all the features - are you sure you updated the docs? Could have the previous set read only, or not have pulled them out of the zip archive or something.
Yes I downloaded the full zip from the link and extractd the files into my blender script directory including the docs that were in the zip. I then unzipped just the docs to another directory with the same file structure so that I could read them seperately without going to the blender install directory. They still list version 0.74 as the most recent version.
Well, I just performed the same procedure, and top of the doc file says R85 whilst if you look at the last section the revision history shows R85 as the latest release. At this point I have nothing!
Ok it must have been something I was doing wrongly. Because I just did the same thing again and got current docs 
Hey, thanks for the help with curves again, lethe.
I had found something months ago that in my opinion is a Chicken bug. For some reason I mention it here only now:
discourse.panda3d.org/viewtopic … 6665#46665
Long story short: if you export a model to egg from Chicken and the model uses a png image (and probably other formats that support transparency), the cartoon inking won’t work with that model
filters = CommonFilters(base.win, base.cam)
filters.setCartoonInk(separation=0.5)Now there is a way to ‘fix’ that:
I don’t use png only for transparent images. I can edit each model I export and add that line after each texture, but…
Chicken only indicates to use an alpha channel if the texture is marked as being for the alpha channel in Blender - otherwise it leaves it up to Panda (i.e. writes nothing rather than forcing the issue by writing a format of rgb.). I suspect what is happening is panda is examining the given .png file, finding an alpha channel and hence automatically switching it on. Do you .png files have an alpha channel, even one filled entirely with 1/255? I know if you save a png from Blender it always saves it as a rgba image, in which case you will have to load the image in the gimp, change the format and resave it.
Oh, I get it.
GIMP: Layer -> Transparency -> Remove Alpha Channel should do the trick.
thanks
Looks like there is a scaling issue with nurbs curves / mopaths
When I scaled my object the mopath remained small.
from direct.directutil import Mopath
from direct.interval.MopathInterval import *
model = loader.loadModel('model')
model.reparentTo(render)
model.setScale(100)
ball = loader.loadModel('ball')
ball.reparentTo(render)
# extract the mopath from the model
Path = Mopath.Mopath()
PathNP = model.find('**/path')
Path.loadNodePath(PathNP)
Interval = MopathInterval(Path, ball ,duration=5, blendType = 'easeInOut', name = "Pathterval").loop()The mopath seems very tiny, so the movement is hardly noticeable.
But in blender it is the same size as the ‘model’
Doesn’t surprise me - that code isn’t really in the main pipeline and tends to lack certain features. I’ll have a look when I can find the time.
Don’t rush. I can always scale them in Blender 
But one more thing. I dont know if this is from my lack of knowledge of nurbs curves, but the object I move with the interval doesnt move along the curve you see in object mode, but the ‘curve’ you use in edit mode.
Is that normal?
Not sure what you mean - curves in edit mode and object mode look the same to me. If you mean the control object, i.e. the vertices and connecting lines then it shouldn’t be doing that, but I like to think I would of noticed previously if that was happening!
Yeah, ‘control object’.
I’m pretty sure: mediafire.com/?mmdeyn2h1ym
Though a really ‘smoothed’ version of it…
Well, it looks fine to me - the ball is moving smoothly, exactly as you would expect given the path. I just can’t anything wrong I’m afraid!
You can’t see it? I know the movement is smooth, I did say
But the shapes do not match, it matches more to a smoothed version of the ‘control object’, have a look:
Well, its possible there is a difference between Blenders nurbs interpolation and Pandas, but that seems unlikely - the algorithm is well known and not open for interpretation. The other possibility is the parameters may need some translation between the programs, but again, they are very standard. Its just hard to see how an error could creep in - its the kind of thing where it either works or fails completely.
Could you try putting in some geometry? Its hard to tell what is going on without any - I’ld stick a floor and ceiling in such that the ball should just touch them both. I would also stick in a back wall, to check the camera is in the right position.
EDIT: here is an updated archive. Now I also load a geometry that represents the control object and curve shapes. If this isn’t enough, I dont know what is: mediafire.com/?my2miuo03yh
Doesnt it follow a smoothed version of the control object rather than the real curve?
Sorry I didn’t get back to you sooner - can’t run Blender or Panda at work. Anyway, it looks like its interpreting the endpoints of the nurbs curve differently in Panda to Blender (Note that as the nurbs curve is a smoothed version of the control object we are simply talking about a different interpretation of the data - both are still using nurbs.) Not sure what I can do about this - I can’t change either interpretation method, and I can’t convert from one format to the other easily - my guess is that Panda is repeating the endpoints, so as to get a nurbs curve where the end control points are also the ends of the curve. My best suggestion is to create nurbs curves in Blender where the end knots and end of the curves coincide - the ‘Endpoint u’ button seems to achieve this, though I haven’t tested if the Blender curve and Panda curve then line up.