I build and maintain a ‘base’ model (for each architectural project) as a component (inserted model) in other SketchUp files for the same project, as is generally the recommended technique. I then set up different files, typically one for sections, another for plans, reflected plans, sections, ‘axonometric’ views ('bird’s eye and worm’s eye), etc., and these are then all inserted into various Layout files where the actual drawings are compiled and annotated.
This system really helps with making the individual SketchUp and Layout files more manageable (especially given the memory constraints of SketchUp on the Mac!).
The biggest problem I am experiencing is that whenever I need to update the ‘base’ model, and then ‘reload’ the component in the other SketchUp files, elements that I have hidden in Scenes in those files (such as ceilings when using an ‘axonometric’ view) are not remembered, and I have to work through each Scene, resetting items; sometimes these will also be Tabs in Groups that have been forgotten (such as stories/floors).
In some of the SketchUp files I may have 25 Scenes, and this is a laborious process across five or six files!
I have tried just copying the top grouped model in the ‘base’ model, deleting the existing model in the other files, and pasting in the updated version, but same problem.
what an odd way to work with your model… I could get it if you separated all the technical drawings (plans, sections, iso…) and made a separate one for animation / rendering with extra surrounding stuff. maybe. because a technical section might not need the whole neighbourhood. maybe.
I have to say, in 17 years of sketchup on mac and PC, I think this is the first time I’ve read that this is the “recommended technique”. recommended by whom?
memory constraints ? what memory constraints ?
well yeah, when you reload a component, it will take the state of the external one.
If in your Component you’ve hidden the walls, then everywhere you “update” it, the walls will be hidden. because that’s the state it is now. reloading a component is pretty much like deleting / reimporting. except it’ll keep the same position, scaling, rotation…
working the various views in scenes within a single file, not across many.
Do you have the exact same Tag structure for all the models?
I don’t think using hide / unhide to control visibility works for this workflow - I believe you could do it all with Tags though, and it should work with minimal effort to correct errant scenes.
oh yeah, people on this forum keep dropping his name, never actually read anything from him though. Well, if it floats your boat, and you get the expected results in the end, that’s the main thing right ?
(I guess I can see the value on big projects, with dozens and dozens of plans, sections… these projects are behind me now, and most people around me use archicad or revit for those)
So yeah, if you want to use this method, I would do as mike says. You need the same tag structure in all of them, and hide stuff using tags, not the outliner.
and in the master file, you need to have EVERY group/component turned visible whenever you save (prior to synchronisation), to avoid problems. That way, the visibility is dictated not by the master file’s outlier but by the tags in each file.
It might be worth making a specific “synchronisation” scene in your master file, with everything (tags, groups, components…) turned on.
that way, you can model as you want, touch all you need to touch, and everytime you need to update the rest, simply pick the synchronisation scene, save, and go on.
Your input is much appreciated and, agreed, I need to look at Tags to hide elements.
It’s just a pity that the only option is to “Reload”, rather than “Update” the referenced model (as is the case in Layout when updating the referenced model… if only there was better harmonisation between tools, menus, and commands in SU and LO…).
On the memory issue: probably due to the way I am working, I will often have three (or more) SU files open, with each model being in the order of 20-30MB, and then running LO with three or more files open, each running at around 30 - 50MB. And I’ll get system memory warnings on my MacBook Air, until I quit the likes of Safari, Mail, and LibreOffice… but, even then, SU will often crash unexpectedly. And Layout slows right down.
It’s not a train smash, just needs better management on my side, but I’d hope a M3 MacBook Pro with more RAM would cope better. Actually, what I really need to look at is a Mac mini - the M3 must be coming soon
I’m just not that into layout to adopt your prophet yet
well off course, if you leave the RAM-eater open…
Browsers will just grab it and won’t let go, I use LibreOffice too, and just opening it idle already takes 250mb.
Here, I just turned on my computer, opened firefox (I’m on it right now) and opera (I use it for my pro email / accounts, I separate both in different browsers), and after 30 min, it’s already like that :
(and logitech, quietly taking 700mb for my wireless gear… I should try bluetooth to see if it’s the same)
edit : yeah, dongle is nice on laptops because it use less battery than bluetooth but man, I just saved 700-800mb RAM by switching. I’ll keep the dongle at hand just in case, but it’s worth it.
From my archicad years, I learned the hard way that opening too many softwares is never a good idea.
héhéhé, same.
I’m on a 2021 mini M1 with 16gb RAM, I’ll probably upgrade it to a M3 mini with 32 GB when it’s updated.
So logitech can discretely hog 1,5GB
That’s what I do. I have a “working model” scene that has all tags on. The style is quick and simple, no shadows or profile edges. All modeling is done in that scene and the scene is always selected prior to saving.
Do you have a fix for the problem of ‘reloading’ the model in other SU files, set up with various scenes, and losing hidden objects (and, sometimes Tag visibility) across the saved scenes?
It’s driving me nuts (to the point I am now manually doing the base model updates in every inserted model!).
It would seem the issue is the way the base model updates each ‘clone’ - it replaces the entire model (on “Reload”), rather than simply update it (as a referenced file normally would).
I am about to issue a big set of drawings, but the client has requested a change in the design - so now I have to work through all 50 (or thereabouts) scenes across the five files with ‘clone’ models, identifying what is now visible, and was meant to be hidden… I guess I’ll just have to suck it up and work with it :-))
I’m just finding my way through this workflow. The one thing I learnt is all outliner views are turned on in the clone, so anything you do for visibility needs to be managed by Tags. As has already been mentioned you should save the Master file using an active Scene that has all Tags turned on. The Tag structure etc need to be duplicated in the clone host files. This best achieved by duplicating the Master file, deleting the 3D model data and then importing the cloned Master as a component.
There are pros & cons to this workflow. Two that I think stand out… Pros: It gives additional flexibility to presentation, especially if you apply colour by tag. Cons: it isn’t so easy to manage if you are developing a flexible project in terms of tags & surfaces.
The key is n the templates. I specialize in high end custom homes, so I constantly deal with changes. I never have to mess with the tags in any file. The only thing I end up doing in the other SU files is updating the reference. If you are doing anything more than creating required additional scenes in the other building files then your template is not setup correctly.
