Hello,

Is there a way to re - assemble the SI behavior, or more exactly, something similar to behavior of Maya tweak node. Let's say, once I have something like in attached pic, that every next edit, invoked by view port action, became in existing Edit node, even this one is somewhere in middle of chain - not in new Edit node appended after displayed node.
For some (unknown) reason, I believed this is purpose of ‘selectable template’ flag. But it doesn't work ( for me, for now) in this way.
I'm asking after entire day of searching docs and googling, without results.

You can continue to manipulate an edit SOP in the middle of the chain as long as it is the current SOP (i.e. the one selected SOP in the worksheet), regardless of where the display flag if the mode of the viewer is Show Current Operator (the default, can be changed in the right mouse button menu of the geometry select button on the left sidebar).

There is no mode that will automatically search your entire chain for an edit to manipulate.

The selectable template flag does allow you to interactively select from that geometry and then manipulate it, but as soon as you manipulate it, the new node (or the edit SOP itself if it is being reused) will become the current SOP.

Thank you for clarifying. Had Show Display Operator as default.

Just out of curiosity,

What does the intended workflow look like if you DO have the show display operator mode set?

G

keyframe
Just out of curiosity,

What does the intended workflow look like if you DO have the show display operator mode set?

G

Had show display operator in H, just to be able to work with attribute vops and such, without constant changes of displayed result. Just been curious how it works with H ‘edit’ node, have to admit that my view port skills in H are somewhere around drawing the curve, not much more.
In Softimage, ‘construction mode’, for example, allows to move (deform, exactly) object deformed by curve, along curve, if ‘deform by curve’ op is above modeling, in ‘animation’ region. You just move points along some axis of object space. Another example is shrinkwrap op (like H Ray SOP) applied above tweaks. And so on. In some cases (envelope (skin) op), SI calculates inverse deformation, so it's possible to edit the corrective shapes on object deformed by envelope, while SI takes care of ‘baking’ the deltas down into reference pose. If I'm correct (could be wrong) Modo has some similar capabilities.
By the way, ‘inverse view port feedback’ is not strange for 3d apps - uv editing could be a common example.

Yeah, this was marketed as “non-linear workflow” (alluding to the destructive workflows in Maya/Max mainly) before ICE became the new baby of the marketing department.

An XSI user after skinning (capturing) a character which means adding an “envelope” operator in the animation stack (which sits above modeling) could go at any time and change point position or even the topology long after the envelope weights have been adjusted.

I haven't tried this in Houdini (will get my hands deep to the neck in this at one point), but I'd be very surprised if it wasn't possible due to the non-destructive Houdini's approach to everything.

McNistor
An XSI user after skinning (capturing) a character which means adding an “envelope” operator in the animation stack (which sits above modeling) could go at any time and change point position or even the topology long after the envelope weights have been adjusted.
.

It was tied only to traditional, linear blend skinning, never worked together with ICE based Dual Quaternion skin or something else ICE based. Somehow is hard to believe anyone will go, these days, with such built-in, so focused refinement.

Beside that, for ‘model in one space, see it in another’, perhaps most comfortable, long run approach is two objects setup, using H Object Merge, Softimage clone, or Max ‘reference clone’. Sow some Max setups of this kind, long time ago.

amm
It was tied only to traditional, linear blend skinning, never worked together with ICE based Dual Quaternion skin or something else ICE based. Somehow is hard to believe anyone will go, these days, with such built-in, so focused refinement.

I do. It is very helpful to make corrections after posing the model if you don't want to change your weights.

OneBigTree

amm
It was tied only to traditional, linear blend skinning, never worked together with ICE based Dual Quaternion skin or something else ICE based. Somehow is hard to believe anyone will go, these days, with such built-in, so focused refinement.

I do. It is very helpful to make corrections after posing the model if you don't want to change your weights.

I meant, how hard is to believe *any developer of 3d app* will go…