Hi Baasab,
Can you view a simple/small pointcloud? Did you try to reset the dispalymodes to default?
I must say, the fact that the systeminfo disappears is another issue that seems worth investigating. It should keep a dialog visible if it does not, something else is also wrong for sure.
-Willem
Oh I just now see you have drivers from 2013! You best update them to the latest available and try again.
That is best to update regardless.
-Willem
Are you using the Rhino 6 WIP or is that s happening in Rhino 5. The options dialog that you posted a screenshot of does not look like the latest Rhino 6 WIP version.
Hm⌠Something odd happens in 2D views, see how it becomes transparent when they are far away from the camera:
I guess they should have a minimum size, or a per view setting, depending on whatâs easier to implement.
Itâs really hard to tell if this is good or not based on your screenshots so Iâll have to take your word for it. I can add a minimum pixel size constraint to the command, but that sounds like it is going to make the UI pretty complicated. Do you think a fixed single pixel minimum would work for you?
I fully understand that it isnât easy to see from the images. The point with real world point size is to get a dense pointmesh when zooming in, and therefore it also makes sense that the pointcloud is dense from afar. But I am sure there are users who would like for the point cloud to fade away in the distance, even have view clipping distance for pointcloud only, but i am not among those. (at least not as of today 
)
That sounds perfect!
And now that changed⌠I like that the pointcloud fades. Iâll tinker with it and come back with a matured feedback
Nice to learn about the _SystemInfo command @Willem
Hi Basaab - Are you just typing the command at the command line (presumablyâŚ) ? When it disappears, is Rhino returned to its normal working state? Note, Enter will close the dialog - as stray space-bar or right mouse click perhaps?
-Pascal
Do you think round points instead of square would look better?
I would. Even diamond shaped* would be better than square in my opinion.
I have to say I really like this. The ability to have a dense point âsurfaceâ is really great when I get up close to the pointcloud! Thank you!
I am not sure, the most important thing with real life points is the overlap. The squares work really good, round points might be more visually cluttering. I made an attempt to compare the two styles by using fixed point size and round and square points:
The speed with square points are MUCH faster though, like 60 fps vs 1 fps. So if round points are slower then the answer is easy, its: âNo thank yoouuâ! 
Round points does show some odd point ghost artefacts though⌠Donât know if that slows it down or not.
One wish though: Selecting or snapping to a point is difficult when the points are large, as they donât âlight upâ if they are selected or snapped onto, nor do they hide the points behind from being snapped to, so that makes it more difficult to work with. Turning the point yellow would be great.
Iâll add this as an option to the test command. I doubt youâll notice a speed difference and the edge bleed effect would go away since I wouldnât alpha blend the edges.
Your speed results donât really count since there is a lot of overhead in processing a bunch of point objects versus a single cloud
The next WIP includes a âRoundâ option in TestPointCloudStyle. Let me know what you think.
Sure thing, looking forward to it!
@Holo and @DavidRutten; have you had the chance to try this out? If so, what do you think?
Hi Steve, sorry for the slow response time, I like the round dots for some scenarios and see that the for horisontal and vertical planes they
I see that colorless pointclouds suffer from the size as all details disappears:
UploadingâŚ
I also wish there were some kind of fog option for color intensity reduction based on distance from camera. That would help for monocromatic point clouds, and would give some of the same effect as the current culling of far away points:
And the round points seems quite much smaller than the square ones, so maybe their sizes could mach in area instead of width? (Just multiply size with 1.13 when switching to sircles)
Pretty smooth. Would be even better if the round display had a spherical z-buffer effect rather than being view aligned discs. I appreciate that would require quite a few more triangles per dot and would probably slow things down.
Ultimately what would be ideal for my purposes is a drawing call that accepts an array of points, an array of colours, and an array of sizes, along with a single style setting. Or alternatively, the ability to assign sizes to individual pointcloud items, but that sounds like it would impact a lot of other stuff that might not like being impacted.
This probably wouldnât slow things down since it wouldnât actually add and triangles to the mix; I just need to get the math correct in the shader to do this correctly.
Probably the best way to do this for RhinoCommon would be to create a new class that held all of this information and was passed to the draw functions. This would let me track if any âchangesâ were made to the class members or if I could reuse a cache that I built for the class in the last frame.
Iâm assuming this is for pixel sized point clouds instead of world sized points. Is that correct?
It looks like your uploads didnât work in your post. I would be interested to see what you were trying to show.
