Blender 2.8 is now fully usable with left-click object selection

news

(Piotr Adamowicz) #569

Well, yeah, this requires a measure of practice and estimation. But to activate it, you need to consciously throw the switch, it’s not like it’s implicit or context sensitive. Other tools such as Bisect and Knife carry similar risks. This will probably mostly be used in ortho views. I imagine displaying the switch on the topbar in an angry red when on would help if someone forgets, but we’re getting into bikeshedding here.

And with this you only have to undo if you make a mistake, whereas display-mode switching was something you had to do no matter what, so as long as you succeed most of the time, that’s less clicking anyway :slight_smile:


(Thinking Polygons) #570

I wouldn’t call it a mode, it’s just a regular feature of the selection tools.

It might look like, but in reality it’s not.


(English is not my native language) #571

As long as you realize quickly that you spoiled something, I guess.

Anyway I’m more worried about they have put Pie Menu for Z (instead of putting pie menu in some combination of keys). I take longer to find the options in Pie Menu than to go directly to the header.


(BTolputt) #572

The theory is that you share the clicks across two fingers - half clicks on the index finger, half on your middle finger. It is an interesting theory, provided you are not a tablet user, but I think the keymap choices made never allowed for it to be fully realised. For example, in no workflow I’ve ever seen is the 3D cursor used as much as selection.

FWIW, I think this is another case proving that familiarity & muscle memory is, to a great many people, more important than (theoretic) speed improvements. There is a reason that, despite there being no mechanical need for it, nearly everyone on this forum uses the QWERTY keyboard and not Dvorak’s simplified one. :wink:


(Antaioz) #573

I don’t know about the ergonomic benefits, but I’m young and don’t generally notice that stuff yet.

But seriously, the cursor isn’t the LMB’s main use in that setup, because action and confirmation is used a ton and that is what evens out the buttons.


(BTolputt) #574

Action and confirmation can be spread out over both buttons regardless of right-click or left-click select. It is also not unique to Blender is the way RMB select is.


(zeauro) #575

Let me reformulate.
You don’t expect that behavior to work only for transform, move gizmo.
You expect that behavior with extrude, spin, shear gizmos, too.

That is exactly what I am trying to explain. In blender, select tool is a tool like the others.
So, the dissatisfaction of 2.8 design is not related to keymap or lack option.
It comes from the fact that select tools are on the same level of other tools.
if we are coherent, we don’t expect a Topbar only showing one more option to display a gizmo for box select but a Topbar showing box select options + gizmo’s active tool options at same time.

I prefer to say that worked as long as tablet use for 3D was neglected.


(Piotr Adamowicz) #576

Yeah, this is a design fail in my view as well. Selection is special since it occurs in between almost every action.

I’d happily relegate tweak to some keyboard combo, or even just lose it completely from the mouse and always use G, if it meant we could have sane always-on box select. Likewise for the context menu. 0rAngE’s setup kinda does that, but it’s so unlike what we’re used to, devs would have a revolt on our hands if they tried to pass that off as the new LCS defaults :smiley:


(Antaioz) #577

It’s not a design fail, because blender wasn’t designed to work this way. Now we’re at this point it’s just something that hasn’t been adapted to integrate with the brand new active tools workflows.

In blender’s original workflow, any tool can be activated at any time, so there is no need for selection to be singled out, to be overtly special, the design wasn’t an issue. Actually from a design perspective it was a good thing, keeping the selection tools working within the same systems that the other tools use, reusing more common code, keeping things modular and reducing special unique cases.

It’s only a problem now because they’re integrating things onto LMB and the active-tools, they’re trying to adapt blender to a workflow it wasn’t designed for. This new workflow is completely different, you can’t just activate any tool at any time, but it’s assumed you can select things at any time. Blender’s tools weren’t designed for that, give it time to get this new one working.


(Piotr Adamowicz) #578

Oh I’m giving it time. But I can only comment on what’s there now.

I love the blender devs, don’t get me wrong, a more competent, patient, and thick-skinned bunch you’ll be hard pressed to find. But unless you tell them very explicitly what’s wrong and why, they tend to keep their first implementation, or finish before something is actually finished :smile:

(Cycles baking, I’m looking at you, with your missing bias, displacement and antialiasing, and laughably clunky UI)


(English is not my native language) #579

I was thinking that the feature would be really useful if that selection mode/state would only select external faces (in ortho view mainly), this would allow to select external part quickly in a single step and then hide it and have internal part available to work on it.

By the way, I’m not sure if it’s possible to do it in one step on 2.79 anyway, I usually do what I show in the video.

Edit:
is possible with Select Mirror in this case, but I think that are more steps anyway.
And yes, stupidly I took one more step in the video. I was imagining having non-symmetrical more complex geometries than a simple sphere.


(Thinking Polygons) #580

Phong break selection

How something like this could be implemented in blender? :thinking:


(Ace Dragon) #581

It sounds like being able to select external faces would be useful, but I wouldn’t call the operator Phong break.

Though it should also rely on a smart algorithm so it can be useful for more cases than spheres (as many real world objects have somewhat complex shapes).


(Thinking Polygons) #582

No, it doesn’t work just like that, this was just a quick demo to replicate yafu’s example.
This selection tool really works on phong angle/break, I can adjust the angle in the settings. So yes, this is useful for million of things.


(Piotr Adamowicz) #583

You don’t actually have any internal faces in your model there, though I suspect you’re just not using that term precisely. (If you were then Select->Non-Manifold would work for you)

You could do that selection in a single step by just doing a circle-select from top-down, though I guess that’s not really what you want either. Or you could use the Select Similar by Perimeter operator.

Anyway, the feature you want, while neat, is unrelated to the one that was discussed. I suggest starting a topic on devtalk and trying to sell it to the devs by making a use case for it.

It looks a bit like the Select Similar by Normal operator except that one compares the face normal to one face, and here you’d want each to be compared to neighboring faces. Maybe that’s a good place to start.


(English is not my native language) #584

Did you see the video? Please tell me how in a single step with circle select on the same type of mesh.

For me it is absoultamete related to what you are requesting, even more useful for what I have already explained.


(Piotr Adamowicz) #585

I’m not questioning that it’s useful. It’s just a different thing, you know?

Can you at least mathematically define an ‘external’ face? What test would a face have to meet to be included or excluded?


(Ace Dragon) #586

You could perhaps imagine a shrinkwrap applied to the mesh and then testing if nearby faces are facing other faces.

If another 3D app. has this operation, then it is indeed possible.


(English is not my native language) #587

You call it as you want, we are offtopic anyway. But in my view what I request would avoid making mistakes when accidental selecting internal non-visible geometries if you select in solid mode with what you proposed. So if that were possible, it could be an additional option to what you are proposing.


(Piotr Adamowicz) #588

You could certainly cast a ray from each face or face-vertex to see if it hits something, but is this really all you want? Other than that sphere with a hole in the middle, what would be the use case for it?

The thing Thinking Polygons demonstrated happened to work for the example, but it’s really very similar to Blender’s combo of Select Sharp, and then Select Loop Inner Region. I wouldn’t mind having this, but I also suspect it’s not really what you want.