Why rename Layers to Tags?

This has been fermenting as a question for a while now, as the “tag” title has become the way things are called previously, but I’m confused by the purpose of renaming this, especially when there is no particular change or improvement to calling groups of things “layers” like pretty much every other software I use (CAD, BIM, photo, video, illustration, etcetera).

The term “tag” seems to make some sense on the surface. Whatever you draw should be on “Layer 0” (or “untagged” as they now say), while objects (components/groups) get “tagged” to being a member of some group of things (like an old AutoCAD layer). The trouble is, you only get one thing to tag to, unlike a true tagging. For example, this forum message that can have two tags and be filtered into multiple search results. It’s like a web term or hashtag, as if a programmer made the decision, not a person who uses creative software daily or speaks our language.

Now, if the programmers that give us these tools can figure out how to give objects multiple tags and make the interface better (other forum contributors reveal clunky/buggy “tray” functionality), I could get behind the rebranding of layers. But unless I am missing something, I say go back to the common industry name of this function and call it like it is – LAYERS.

Maybe the tagging functions of the object classifier would be another way to do the thing they wanted to do with renaming layers to tags? Maybe there is something else on the developers’ road map? Is there going to be a conflict in language between tagging on the classifier and with layers?

1 Like

I agree that being able to apply multiple tags to an object would be nice.


“Tag” makes you expect that an object ought to be able to carry multiple tags.
“Layer” has long existed in CAD and modelling applications as an abstract concept even if it originated as a way to replicate traditional overlay-based drafting in a computer. It is the industry standard name for the feature, and deviating from it only causes confusion, as the time after the change in SketchUp has shown.


But I don’t want to add another object in the hierarchy simply to be able to tag that object. For example, I have an object that is both a part of the antenna, and is an electrical cable. Might be nice to “label” it with both tags for more flexible visibility control.


That is not at all the same thing. “Tag” is largely synonymous to an “attribute”, and it should be a way to organize and query the data in your model. An object should be able to belong at the same time to “blue”, “$200” and “belonging to the Emperor”.


To my mind, the most important reason for the name change is that far too many new users of SketchUp make a fundamental mistake with “Layers”:

In SketchUp, Tags (old name “Layers”) do NOT prevent geometry (edges, faces, section planes) from interacting with geometry assigned to other Tags (Layers).

Why is this a problem? Many new users are coming from 2D programs where “Layers” DO isolate geometry “on” any layer from interacting with geometry “on” other Layers.

As @DaveR says, there have been Many, many, MANY cases where a user has brought a modeling problem to this forum - and the problem was that the user used “Layers” as they expected them to work - from their 2D program.

Many on this Forum have, for years, advocated changing the NAME of Layer in SketchUp to something else - almost anything else. If I may hazard a guess, the name change has had 98% positive acceptance - especially by the people on this forum who most often volunteer their time to help others.

1 Like

In practice only Photoshop works like this. In CAD/BIM applications, whether objects interact is independent from layers. In AutoCad they don’t interact even when assigned the same layer. In Archicad, walls connect even if they are on different layers. I understand that the majority of serious Sketchup users come from CAD/BIM applications, not Photoshop. And users tend to switch back and forth between these applications and SketchUp.


Agreed there is a confusion of bringing Photoshop and its additional sequential function of layers, where one thing is in front of another in a 2D sense. But SKP is a lot like DWG - it can appear as 2D, but is actually a collection of 3D geometry, and those living in Flatland take a minute to figure out what this 3rd dimension is and how a layer operates in a 3D universe. And it is something that once learned it is understood.

Maybe most of my confusion is just the name being in conflict with how tags are used everywhere else. It is an attribute that should allow multiple choices, but the core programming still seems to be about layers. I hope things will be added soon to have the ability to tag as members of multiple layers. Oops – sorry – TAGS. With that, it’d be great to be able to filter complex models with neat Boolean tools.

Also agreed the term “layer” appearing in both SketchUp and LayOut was confusing things, and those layers in Layout work like they do in Photoshop or InDesign. For my workflow, I still can’t for the life of me figure out how to reliably use LayOut because it can’t easily handle detailed architectural projects at the scales of city blocks and dozens of floors. So we just export images and arrange them in InDesign.

For now, I will have to continue using these “tags” as “layers” and making objects that might be in groups that can be on and off depending on view states I need.

You can add more than one classification tag to an object (from different classifications) but you cannot ‘show all’ or ‘show only’, let alone select in the outliner, for instance.
Or apply a linestyle…

It would be far more obvious to have added ‘locking’ to layers, IMO, then renaming into something that’s evenly confusing for ‘newcomers’ that are using other software like:
Google mail, Freshdesk, Trimble Connect, etc.

I never mention layers in the beginning of our course and when it does come in play,
I warn ‘adobe-addicts’ that layers can’t be locked, so you would need to isolate edges and faces with groups.
(Like they just learned)That’s usually enough.

They are called Tags now, they may acquire new abilities in the future (like tag folders), move on.


SketchUp Tags have never behaved the way layers behave in other programs, and the named caused a lot of confusion with users having different expectations.

In many programs, layers make up a document hierarchy. An object can be placed on a layer, and sometimes the layer can be placed inside another layer. Sometimes an object can be parented by another object which in turn is parented by a layer.

In SketchUp Tags don’t make up the model hierarchy, but are properties assigned to objects separate of the hierarchy. A building can contain an apartment which in turn contains a room that contain a table that contains a leg. In SketchUp these objects would be groups and components, and they can all have tags, not just the top one.

1 Like

Quite a large proportion of serious SketchUp users have an AutoCad background. SketchUp Tags behave exactly the same as Layers in AutoCad.
Geometry stickiness is a totally different matter. In all applications I know with sticky geometry, the feature is independent of layers.


I haven’t used AC much, but can you both apply a layer to an object and have it parented by another object (e.g. some sort of group) at the same time? In 3dsMax for instance you have to chose between one or the other.

Yes. Objects inside AutoCad blocks can have any layer. Like in SU, it is often desirable to keep the contents of blocks on layer 0. I never used AutoCad groups much but I seem to remember that they are not the kind of separate entities like in SU, merely a way to make a bunch of objects move together.

1 Like

Other than enabling a number of objects to move at once, AutoCAD groups are, IMO, pretty useless. They don’t even survive a cut and paste into another drawing. The pasted objects don’t stay grouped. When I started with SU, I remember thinking, “Wow, a program with architectural uses where groups have a useful purpose!”

Are AutoCad groups more of “selection groups” (a saved selection) than an object in its own right? Rhino has selection groups, and it is quite different from SketchUp groups.

It seems there are three organizing concepts that can’t be handled simultaneously by a single mechanism, no matter what you call it:

  • structural/physical/“part-of” hierarchy
  • logical relationship, e.g. all the things that are plumbing
  • ad hoc gathering, e.g. a selection set

No, they’re just to keep objects together. You can copy them in the same drawing and have numerous copies but nothing else. If you want to save something to reuse/edit you basically are going to have to go to blocks, like SU’s Components.

In fact, right now I happen to be using a Lisp routine I wrote long ago to enter the site boundary data (in surveyor units) from a paper plat and have AutoCAD draw the boundary lines. The routine creates a group for each boundary line that contains the line, the circles at the ends for the pins and the text label. If I select the group and look in the Property Pallette all it says is that I’ve selected multiple objects- 1 line, 2 circles and an MText object. Though they’re surrounded by a bounding box on the drawing screen, nothing on the pallette gives any indication I’ve selected a group rather than 4 random objects.

1 Like

Indeed! Sketchup’s Group and Component features (including the Outliner and Component Browser) address the structural hierarchy pretty well. I have never really wanted a way to persistently identify a collection of unrelated (to SketchUp) entities - a selection set - but I can imagine it being useful (analogous to Photoshop’s saved selection feature dare I say).

Regarding the logical relationship between entities, more flexibility would be nice, for example the ability to associate multiple tags to a given entity. I haven’t needed to define a hierarchy of tags (which would be independent of the structural relationships between the entities to which the tags are associated), but I can imagine this being helpful also.


A further thought that occurred to me: the existing tags are sort of in the logical relation type, but they are permanently tied to specific set of rendering behaviors - visibility, color, and dash type. So, there is an entanglement between organizing and viewing. That makes it hard to envision what would happen if objects could have multiple tags. Would all tags have possible associated behaviors as well as “what this is” relationships? Would the existing tags stand out as a misfit in a system that was otherwise category-based?

1 Like