currently, axis snapping (rotation, translation, scale, globally-locally) is only available for the extrude tool and for the rotation, scale, translation functions.
as they are snapping functions (and also very practical and useful), they should have a more global dimension like all the other snapping methods and be available for all modeling tools
It is not true. It is also available for push/pull or rip.
And if you change active transform orientation in header of 3D View, you can use it for those tools.
There are lots of modeling tools where it does not make sense : Loop Cut, Edge slide, Grid Fill, etc…
So, please, precise the tools that would need this instead of making a global statement that would always be wrong.
my mistake, I checked better, the snapping on the axes are also bootable on other tools, but some are missing and would be comfortable … for example on the bevel tool and insert faces…
moreover, on some tools only the snapping on global axes is accessible and not on local axes (based on a face or a normal) and rotation and scaling on single axe are not accessible …
in synstesis I was referring to the fact that there is no global support for all the tools that could benefit from snapping
moreover the “new” method of blender 2.8 makes the use of these tools inconvenient because they are based on the click on the tool, left click on the face or other and at the release of the click the tool performs the action and is not easily activated use of the smapping on the axes …
the “old” method that is still accessible via shortcut uses the click = activate tool and another click confirm the action and between these two actions you can activate the use of snapping on the axes …
I think at the moment there is a hybrid of systems that creates a bit of confusion …
Bevel tool is little bit like Loopcut and Edge slide. You expect what is added to follow direction of pre-existing edges. Except for profile, there is no direction involved in this operation.
The tool could be improved.
In many modeling software, there is an ability to define a profile based on a curve.
In blender, you can do it through bevelled curves.
But a selection of edges can correspond to crossed edge loops with Y shapes.
If you make result constrained in axis, it makes result really difficult to predict and to clamp to avoid overlapped faces.
Inset faces is also a tool mainly used onto a complex selection of faces. And its main goal is to define an homogeneous band of faces around scale down selection. Depth option is just an option to improve it but it is not the main goal of the tool.
There is nothing blocking, here. An inset can be followed by a scale or extrude operation. And it would result in what you are currently expecting.
So, there will be a real benefit for user if re-scaling an inset in a constrained axis or extruding selection in a non-normal direction were frequent cases while doing a series of inset.
To sum up current situation is not really annoying the user but adding such level of complexity to the tool would give hard time to developers for same reason that those exposed for bevel tool.
The selection, target of operation, can be a lot complex and tool have to work with it.
Select a face loop or an edge loop of a sphere and try to make an inset or a bevel on it.
What would be the expected result of an axis constraint on that ?
It is the kind of question that a developer would have to face by adding such options.
That is a basic of blender, you have to press 2 times the axis key to use custom transform orientation.
Default transform orientation set to Global in header of 3DView.
G,X -> global X axis G,X,X -> local X axis of selection in object normal X axis of selection in edit mode.
But custom transform orientation set to View in header of 3DView.
G,X -> global X axis G,X,X -> View X axis
It would be ridiculous to cycle through all transform orientation axis by pressing X, 6 times.
So, you have 2 choices, the global axis or the custom axis. But custom axis means all the others.
It works like that for translate, rotation, scale,extrude, etc…
If it does not, you made a wrong manipulation or there is a bug.
There is a method older than the other. But it does not mean that the older will disappear.
The project for 2.8 is to keep both.
One easy to use with tablets and discoverable by new comers.
The other more efficient relative to use of shortcuts.
It is true keeping both methods implies that user have to assimilate distinction between one active tool selected in toolbar with its settings exposed in topbar and activated by left click and all available tools that always can be called and activated by using shortcuts.
In the design of new active tool, user will set settings first and have a gizmo visible in 3DView to help him to perform the action. So, it is not a big problem if it hard to reach MMB axis restriction.
User have a gizmo, manipulator already showing the axis.
He just have to modify Global to View in header of 3DView to see axis of manipulator changed.
So, if you are using a tablet, there is no reason to neglect the gizmo and choose the less practical solution.
It is true that currently, there are some tool requiring a gizmo that don’t have one.
But it is a lack supposed to be filled at the first release.
Can I ask why it doesn’t make sense to have snapping on loop cut, edge slide etc?
I think it would be extremely useful.
The title of the thread can be misleading. Here, we are talking about constraining a movement to an axis.
Main snapping tool is already working with these tools.
By definition, a loopcut & edge slide is a movement that follows direction of edges “perpendicular” to added or moved loop.
These edges may not be really perpendicular but they are crossing the loop and movement is restricted to them.
Imagine that you add a loopcut to a cylinder aligned to Z axis. You add a loopcut to it, following edge slide will be a movement on Z axis.
There is no free movement alternative on a XY plane.
You can modify smoothness parameter to scale the section on a XY plane. It does not mean that this movement is free.
If circle section is moving in XY plane, it means that other edges of mesh are simply ignored.
If they are ignored, movement is no more an edge slide : it is simply a translation.
What would be the goal to add an option to allow the tool to jump from edge slide to translation ?
It is equivalent to call grab tool after first loopcut click.
So, restriction is always there. No reason to add access to other ones that would have no sense there.
For other tools like bevel or inset, main operation on newly created edges is also an edge slide restricted by the rest of geometry.
Because most of times these tools are used to modify boundaries or loop around the whole mesh. Giving a priority to one axis of this selection often corresponding to a section plane, it means neglected the other axis of section. And often, it conducts to a geometry intersecting itself.
I came across a situation where I wanted to slide an edge and snap it to a vertex on an adjoining object. Would be really useful to have that functionality.
I will repeat myself. You want to snap on a vertex. It is different than to constraint movement on an axis.
The fact that vertex snapping is not supported by edge slide is another topic.
Apologies for confusing the thread.