Medeek Wall

Another very good question:


How can you release drawn geometry (wall plug-in) from being parametric?

Is there a global setting to deactivate or is exploding the only way to achieve this?


Exploding the main wall group is one way to do it, then group it back together if you wish.

If you do this you essentially eliminate the attribute library of the wall group so this action is irreversible once you save the file and exit the model.

Another method is to rename the Instance Name of the group (remove the RECT_WALL_ASSEMBLY keyword string). If you do this the plugin will not recognize the wall panel group as one of its own and will essentially ignore its existance.

The good thing about this method is that if you at some time in the future do want to restore parametric ability to this wall panel you can do so by renaming the Instance Name to include the keyword string above.

Just make sure the trailing digits or text after the keyword is unique to that wall panel, so that the plugin does not somehow confuse it with another wall panel.

1 Like

Just a side note:

I’ve dug myself in pretty deep with the garage door module. Yet another fun rabbit hole (ie. arched doors, portal frames, separate garage door plugin within a plugin).

If I’m a little slow to get back to you on a specific bug or question or potential fix this is why. I’m trying to keep focused on the garage door module so that I get it right. I appreciate your patience.

I understand that many designers don’t really care about the framing details (ie. portal frames vs. standard framed opening) however I have had enough requests for this feature that I cannot ignore it.

Yes, please! To my mind, how the actual wall is framed is the highest priority. I figure I could always customize cosmetic stuff like glass panes, muntins and arched glass later where it’s needed.


The detail below shows most of the basic elements of the portal frame that will be drawn by the plugin:

The PF King Stud and PF Trimmer Stud parameters will allow the user to increase the quantity of the king and trim studs accordingly.

The PF Bottom PLT parameter will allow PF Bottom Plates to vary from zero, one or two.

Optional head and jambs studs can be specified (shown in red).

The detail shows a double portal frame, a single portal frame is very similar with one side conventionally framed and one side portal framed as shown.

The light blue elements are the regular studs (or cripples) on their typical 16" o/c spacing.

PF C2 and PF C1B are probably optional elements but I like to include them, please let me know if there are any objections.

For now only strap holdowns will be available since this seems to be the prescriptive path most people employ. However in my professional opinion other holdown types (HDU) could be permitted provided that the proper engineering has been done.

Eventually I will include a PFH calculator in the plugins engineering module based on my own Excel based Portal Frame Calculator:

Please feel free to chime in with any comments, suggestions, concerns or additional options or parameters you would like to see added to this module.

1 Like

PF Framing algorithm is complete:

Double, Single (Left), Single (Right)

In this case a 10’ wall creates a pony wall above the headers.

Now I just need to add in the logic for the optional jamb studs.

Getting closer to being able to do this sort of thing with the plugin:

1 Like

A garage door with optional head and jamb studs enabled:

Also note that this feature can be utilized in the CMU or no framing mode:

1 Like

I’ve been thinking about the parametric ability of the plugin and the fact that manual edits to the plugin geometry are for the most part discouraged since they will be blown away when the wall panel is updated by the plugin.

With that in mind does it really make sense to use components rather than groups, even for wall studs?

Components are essentially the same thing as groups and actually involve more overhead and hence a heavier model.

You’ll know the programming end better than I do. However, if you create a cube, texture it, group it, copy it 10000 times and do the same thing except you create a component of it instead of a group. Compare the two SKP file sizes. The 10000 component file size is much lower, not so?

1 Like

You would think that would be the case but it does not appear to be, maybe I’m missing something here with components vs. groups. The component file is actually larger.

Interesting. That’s not what you’re led to expect, but as Richard Feynman always said, “The test of all knowledge is experiment. Experiment is the sole judge of scientific ‘truth’.”

I did the same experiment a while ago. With the same results.

Speculation/Guess as to the explanation:

While a group is (so I’ve been told) internally a subclass of a component, so when you copy groups, they are essentially instances of a compotent - they’re still linked to their “progenitor” group - perhaps more like a identical siblings until you edit them. As soon as you edit one of the groups, it becomes completely independent of its progenitor/siblings.

When you copy a component, you’re immediately creating a full instance including how it’s transformed (location, rotation, scaling) AND copies of attributes including custom attributes. But when you copy a group, you’re immediately creating only a skeleton instance - possibly transformation only.

Thus as long as you don’t edit a copied group, it’s “overhead” as a separate entity is LOWER than the overhead of each component instance. But once you edit it, that “overhead” advantage is gone as the file must store it - including its included primitives - separately while components (until they are made unique) only store the primitives once and redraw them as needed from the component definition - as modified by the transformations of each instance.

End Speculation.

My usual disclaimer on this sort of stuff: I haven’t done any ruby programming. Thus the only insight I’ve gained into these sorts of issues are those I’ve gleaned from reading this forum. I will happily accept correction from those who know better!

1 Like

This is what I’m getting at, since I don’t plan on editing the studs (on a regular basis) then they are just as well off being groups with a lower overhead (and lower file size).

It would appear that they are already pseudo components unless you edit them and force them to be unique, so they already have the major advantage of component like behavior.

Another reason is Sketchup creates 2 internal graphic images but just for components. This adds overhead both in size and in creation time. Sketchup wants the images to use in the components tab.

1 Like

A typical raised panel and flat panel garage door (10’x7’) with the panels parameters set to “AUTO”:

I may need to fine tune the spacing parameters a bit (I have not exposed the parameters to the user, it would probably just over complicate things) depending on feedback from designers.

Working on the window versions of these two variants now.

Then the all glass version for commercial applications (ie. Fire Station overhead doors etc…)

I will probably not release a carriage door style just yet, not until I have further researched this type of door and determined what configuration and additional parameters might be needed.

Below is a preview of the Garage Door Edit Menu (note that most of the advance options have been disabled so that if fits on the screen):

I’ve enabled the Portal Frame option so that you can see the available parameters.

Exterior and interior of a 10’x7’ garage door (raised panel with glass and 2x2 grille):

The number of horizontal and vertical panels can be specified or be set to AUTO so virtually any number of combinations are possible:

Full glass door exterior and interior. However I’m noticing that these types of doors usually have a kick panel (no glass) installed.

1 Like

A glass door with a threshold and door lever (hardware):

Note that the grey material applied to the door was first created in the material manager in the global settings and then applied in the garage door edit menu. I am very happy with the implementation and added functionality of the new material manager.

The blocked out areas on each side of the garage door indicate the extent of the portal frames when enabled:

It seems you’re envisioning every garage door ending up at the top (or above) the opening and (mostly) horizontal when open.

But what about roll-up doors? In my experience, they’re mostly in commercial/industrial settings, so probably don’t need anywhere near the number of options you’re currently considering.

I’ve got two of them in a very sketchy model in my mind for a warehouse type space dedicated to providing leased indoor build space for tiny homes. At 12 wide by 15 feet clear height opening, I’ll be using roll-ups.

Framing details would, I’m sure, be nearly identical, but the dotted lines indicating the “footprint” when open would enclose far less space.


I got curious and started searching Google for insulated roll up doors. I stumbled on a neat feature offered by one company: A custom drawing generator!

Might help in designing a roll up door feature for the Wall plugin.

Just because I could, I asked for a drawing of a 12’ wide by 15’-6" clear opening height, tracks mounted to inside wall, wood framing, motor operated, motor on right:
Sample Roll Up Door Drawing.pdf (281.3 KB)


I used a roll up on my own house. I put a car port on my house, with the thought that most people fill their garage with stuff and the cars end up in the driveway anyway. Having clients ask me for three car garages because of this made me think I’d go the way of a carport, and maybe I wouldn’t collect stuff in it and and actually park my cars there. (I was only half right in the end; cars are parked in it, but there is some clutter too)

Because there’s all this other stuff like lawn mowers, ladders, and garden tools, etc, I made a 6’ or so deep “garage” off the house side of the carport. Being only 6’ deep, a standard garage door wouldn’t fit, so we used one of these roll ups. It works, but boy is it noisy to operate.

1 Like