Hi,
I noticed I only have on undo in the undo stack, so I looked at the memory and it says 256 so I set it to 1.00 I assume that’s the one gig of memory setting. Anyway I need more than one undo I have 32 gigs of memory what should I set this setting to.
RM
Hi Roland - I think you can reasonably go to 10% of RAM. Are you dealing with a massive mesh? The actual number of undos available will vary of course depending on what is being done to what…
-Pascal
Hi Pascal,
Yep it’s a large file with a large number of meshes, but I meshed to save memory in the first place, most of it’s flat polygon box shapes/standard office cubicles but lots of them. I gathered it was the file, so I can set Rhino to 3.2 gig?
Thanks for your quick help,
RM
Hi Roland - I think when this came up here we decided 10% of Ram is probably safe, You may be able to use more since you have quite a lot to spare.
If you can Join the large numbers of simple meshes, if they have the same characteristics and ca be one mesh, you can save a ton of RAM btw,
-Pascal
Hi Pascal,
Yes I’ve been seeing this come up in a few discussions here. The meshes are just office cubicles and chairs can I just choose them and say join meshes even if they are not touching?
I’m really looking into any saving that will make working possible. I have to render the exterior of buildings that are clad on any visible side to the camera with office cubicles each with a chair and if possible a monitor and a model of a person. I don’t have to fill each cubicle but there are quite a few I am using GH to array the cubicles etc. on the building surfaces.
Thanks for any tips or hints.
RM
Yep. I would join anything that shares a material and mapping into one mesh. Join works on disjoint meshes (SplitDisjiontMesh to reverse that). Joining reduces the considerable per-object overhead that has to be stored and dealt with like layers and display color and on and on - it can add up.
-Pascal
Awesome thanks Pascal
Pascal using the disjoint mesh trick shaved 100 meg off of the file.
That’s a pretty good workaround
Thanks,
RM