True, its a workaround I use when things really hit the wall and I get super frustrated.
But its another example of silly LO method that people have to use.
a similar method is to Export 2D DWGs from SketchUp (they’re flattened) and then Import them into LO.
It would be better if SketchUp could flatten the image prior to (or during) export, retaining the “referencing” aspect.
How about it views came through from SketchUp as “flattened” and then within LayOut you see a flat image… nice and quick. But then if you want to style tags or change the model, it would require you to “activate” it (meaning it then downloads the full SKP file). This would be similar to LayOut’s present option to “enable sketchup model editing”, except it would be “enable sketchup tag editing” (or something like that)
Yeah that’s a more accurate description of what I (and others) mean by the difference between a “surface” or “face” modeler and a different type of modeler that uses more parametric code (inc solids or nurbs) to generate geometry.
As far as I understand…SketchUp calculates then commits the surface, represented as polygons/edges, whereas other models retain the mathematical formula that creates the surface.
I’m aware that, in reality, it’s not that simple. A SketchUp arc can be regenerated with more segments, and the adjacent face will update it’s area automatically. So that’s somewhat advanced. With LayOut, Im not really sure how much computation is carried over - a curve can’t be regenerated, so it’s sort of like adopting the static properties as though it is flattened (non editable) but it’s not flattened because it can be changed (camera view or style)…and the full 3dmodel information is stored in LO.
Context-aware generation is perhaps another possible pathway to SKP+LO efficiency. By this I mean geometry that has a “Level of Detail” adjustment based on view type or proximity to camera. Example, a cross section sliced through a wall will only generate the internals of that wall (timber, insulation, etc) IF and when the wall is viewed in cross section. If it’s not being viewed in cross-section, then it remains as a simple box-like object.
In SKP+LO, context-aware calculation could involve an example where a 2d wireframe image sent from SKP to LO would only transmit that data, not the faces or other invisible tags. You would have to “stream” those as and when required.
All this reminds me of a mid-2000s gaming design concept called the “crate problem.” In that, you can have basically as many wooden supply crates in your scene (level) as you like, but only some of them can be opened or manipulated (because the contents are computationally heavy). To get around it, when smashing open a crate, the contents within and the broken pieces of timber will quickly melt into the ground afterwards.
We used SketchUp since 2007 to design large, complicated projects. Layout could NEVER do what we needed to make presentable construction documents (CDs). We’ve always had to export to 2D DWG, import to CAD, and clean things up there.
Over the years, we developed plugins to make the translation to CDs easier - first FlatText, then DPLineStyler, and most recently, PrettyPrint. There is still more work to do to get where we really want to be, but these plugins have made life much easier for us.
I would love to see Layout fulfill its promise as a CD generator, but in the meantime, we have work to do!
Well said Sam. We are a small architecture firm that would love to do more in Layout but the lack of speed is crippling. I know this shortcoming comes up regularly in this and other forums, so Trimble is well aware of the issue. It seems to me that if a software company is going to thrive, they should listen and respond! Every time there is a new release I get my hopes up and every time they are dashed.
Agreed. I have many times asked for an option in the export dialogue to “export as vector all viewports”. It would take a while to export as it would need to re-render all viewports but it would be a wait at a predictable time, and in my terms. Then one could work in raster all time and be spared the constant wait to convert all viewports to vector, export, and convert them back to raster so one can continue working In Layout.
Yes I hear the voice that raster is good enough for large printed plans. But nobody prints my plans anymore, it’s all digital. They are posted on giant monitors and carried around on handheld devices. The expectation is of infinite digital zoom on details without getting pixelated. I need vector, but only on output.
I do love layout, I rely on it, and I don’t expect it to be magical, other graphics programs bog with massive detail too but I wish there were better tools to keep the display quality optimized for speed to keep the program moving on large docs.
Adding a whole whearted, ‘HERE, HERE’! Our office also has an Apple laptop with the M1 chip and we are about to get the iMac with the M1 chip. The new iteration of SU is very frustrating when using LO.
Would like to add, it would also be helpful to be able to guidelines in LO too.
When I move my work over to PowerCADD, that also effectively breaks the link, so this achieves some of that advantage inside LO.
I’ve wondered about duplicating a viewport onto another layer before exploding it and just turning off the original for safe keeping. Repeat the process to update the exploded version. Doesn’t keep file size done, but help with speed?
“Awareness” here is evidently quite different from “Solutions” in the LayOut camp. Judging from recent progress, very little of Trimble’s “vast resources” seem to be trickling down to Adam’s LayOut group. It would appear that fundamentally improving LayOut is not a high priority …
Interesting perspective I assume your talking and reference to Multi-Core and Multi Processors and mega amounts of Memory.