This depends on who you ask!
If you were to ask me, then no – it doesn’t and never has lagged, crashed or randomly changed the scale.
Actually it has lagged a few times in the last decade that I can remember but that was when I loaded the SketchUp model with some ridiculously high poly count components – which shouldn’t necessarily have been a problem if I had attended to an effective tag structure where I could have turned off those high poly components when not in view in the Layout viewport.
Thanks kevin
Ah. The hairshirt approach.
These applications are here to serve us, not the other way around.
I agree wholly in terms of drawing discipline. Dropping in components from 3D warehouse with 3000 polys, nested groups and random tags is not going to end well. And lazy tagging / class / layer assignment in any drawing package is … well a sackable offence if it were up to me.
However, I think we can also reasonably say that LO has a sensitivity to this which other drawing packages seem to avoid. Why should LO be concerned with objects that are not within the view field? Why should you should attend to tag structure when you have instructed LO to not look at those objects?
This extreme discipline is presented as a fait accompli that everyone should be aware of and work accordingly. Yet this graphics engine is now more robust and perhaps that is part of it. Have you tried committing the cardinal sin of leaving a high poly component in a non-viewed location and seeing if that has an effect?
If it’s already perfect, why would they improve it?
Your welcome…
The discipline was a fait accompli for users decades ago. we learned to work in SU in a more streamlined fashion because machines were not as powerful and bandwidth not as fast. and also, open GL vs new engines.
these days the limitations are virtually gone. you can have a smooth sketchup file with 10M entities and 700mb materials.
in layout however, this will not be fine. layout has to process all that into 2d visuals.
Hi all, we keep this changelog and have been a SketchUp distributor since v2:
With best regards,
Orlando
And they do serve us, within their limitations.
Either you accept those limitations or don’t.
Poly count has never been an issue for me as I like simple proxies for my models and very detailed models for rendering.
I think there are way worse limitations though and I still accept them.
Just thought I would add some thoughts. I have stayed with the last Permanant license 2021 Pro with layout. I have used Sketchup over the years for mainly concepts and 3d details with ArchiCAD for documentation. Coming to the end of my career with a bit more time up my sleeve I have been playing around with full documentation from Sketchup to Layout. I think documentation in layout has and will always required work arounds. Most of the work needs to be done in SketchUp using a good section plugin and workflow. I have managed to get Layout automated really well for Sheet Sets using fields, layers and linked excel tables. At the moment I am looking at leveraging dynamic components in SketchUp for annotation by component in layout and editing in excel. This dynamic component information has been in layout for a while so interested to see if it can save annotation time. I also have a good really simple excel keynote system with a linked table and a scrapbook with keynote tags. I use Curic Section and I just looked at his 3d text plugin today and it is pretty cool and is scalable so you could setup a pretty cool template for quick documents. I have attached a onedrive link to my first nearly complete drawing set in layout if anyone is interested. Sketchup to Layout Example.pdf
Thank you Andy that is interesting. We have a very similar approach, and I have stuck with 2021 permanent license and it’s fine for most, limited, purposes - hence the question, should I consider upgrading OR are things not appreciably better, and the investment would not be justified.
Thanks for sharing your drawing set. May I ask - drawing 505 Joinery details - were they drawn in Layout?
Well, yes, the discipline of not overloading software is blindingly obvious to anyone with a career.
The point of my comment, the actual meaning and question therein was not that. It was about a structural issue - IF LO is calculating poly surfaces or whatever for objects that are not in the field of vision, i.e objects that it does not need to be concerned with… and lagging because of that… that would be a problem that I would say is not acceptable…
Now, whoever said that was what they do (and I have no other evidence, it night be completely untrue) was in view presenting a needless timesuck as acceptable to use this software.
Hence my comment about the hairshirt approach - it’s a total nonsense. Software should be engineered to do what it is supposed to do, and if it doesn’t, the market should deal with it. Investment goes elsewhere.
This is a completely distinct principle to just not overloading drawings. Layout, when given a field of view, is not supposed to look at something else. If it is, that is different issue, that should be rectified.
The comments on here which can be distilled into “you’re doing it wrong” (to which i would say “i am not”) are trapped by their own logic. By that maxim, no software would have changed ever, since IBM DOS. Just do it better.
You could download the trial and give it a go, or we can keep going round and round here. No one posting here can settle this for you.
On the contrary, several people have answered it for me and I have reached a decision… which may or may not be supported by a further reply from @Andy85
You aren’t obligated to keep checking this thread.
The problem, of course, is that to know whether something is not in the field of vision, an application has to do some calculation that involves all their vertices. When I was using AutoCad I often bogged the application down by creating detail sheets with a large number of paper space viewports - with every single of them adding the weight of the whole file to the application’s memory usage.
Hi Ivan I used sketchup for the joinery schedule and they are all copies of the components in the model so they will update if changed. In my opinion, the more you can do in SketchUp the better. FYI all the details in that set were done in layout and I have a line weight scrapbook for detailing see attached.
The work on using taggable attributes using dynamic components is interesting. I have Master Components in my start template ie my Roof, Wall and Foundations are dynamic master components. Oliver from Wisext Tools has done some additional work on his free QS Tools Plugin that allows you to add any number of attributes to dynamic components that will show up in layout for tagging. You can export to excel and reimport changes back to sketchup. I just open the component and start modelling the roof same with walls and foundations. If I need more note organization I create dynamic components inside the master component. It is a work in progress now, but Olivers work is great. This combined with my Keynote system is a good balance.
I am selling my 2 Archicad Licenses and will do any remaining work in retirement using Sketchup 2021 Pro and Intellicad for detailing which is an excellent AutoCAD alternative. FYI Curic’s Section Tool is a bit of a learning curve but gives really good control over iine weights within SketchUp. I still use Nick Zonders layering of vector and raster view ports for sharp drawings in layout. I always offset the size of layered view ports for selection. I have attached a screenshot of my layout system which is another work in progress
Hi @ivan2 … as a long time SketchUp and LayOut user, I’d say whole heartedly that yes, LO has improved since 2021 and continues to do so. Hope that helps and best of luck. I’m rooting for your success!
I’ll probably be shot down here, but no, Layout '24 is seriously problematic! Fortunately for you '21 is probably the last stable version. I wouldn’t upgrade yet, any improvements to the engine making it faster, will be completely wiped by workflow increases!
Damnit Richard I had convinced myself to go for it! Then you come in with this monumental curve ball… lol… what is it with the scaling? I use a clipping mask and a group to avoid inadvertently knocking the scale out, but really, the tickbox to fix the scale should just work.
That is the sort of improvement I was hoping to hear, eg “scale is fine these days”… it seems not…
I don’t buy that, sorry. I know nothing about how apps are coded but I seriously doubt a well-written process would employ resources to make unnecessary calculations. eg. hidden line on a object no where near a field of vision, as defined by parameters which objects either will or won’t be in. That is the structure of LO in a nutshell - take vertices a-z, render any of them that fall within field x, ignore the rest. Why would it do other things that don’t contribute the outcome. It’s not VRay.
I can do full packages on VWorks for example, with hundreds of viewports and no perfomace loss. What will slow VW is things like point clouds or bitmaps, when you are dealing with them.
The latter is an example of where I would agree with some the sentiment here - expecting LO to deal with 10 jpegs on a page is unrealistic.
Impressive. I must try intellicad. This reminds me a bit of a conversation I instigated on reddit that didn’t go anywhere. My thesis was that SU carries enough info for a QS to base a schedule on, in a BIM sense, i.e. a shared dynamic model of the proposal - because SU gives information and volume. So - are you extracting these parameters from your dynamic components, and then applying calculations to figure e.g. joist linear metres etc?
Returning to the detailing on 505 - these are drawn 2D into the SU model and then viewported into LO?
Layout 24 maybe seriously problematic for Richard and others but for me and others it is not seriously problematic.
I can understand Richard’s reasoning for the new inferencing not working for him (as he has explained this on another post) but if Richard can elaborate more – I don’t recognise scaling to be broken, at least in my workflow.
