Remove reference to "Anonymous Attribute" in socket inspection #91851

New Issue

Dalai Felinto · 2021-09-30T18:01:40+02:00

Dalai Felinto commented

2021-09-30 18:01:40 +02:00

We should not expose the users to the name "Anonymous Attribute".

Instead we can change the text to follow a simpler set of rules that will help users to know where the attributes (and non-attributes) came from:

Sorting criteria:

Geometry
Nodes
Attributes inside node
Then per group we sort by length (smaller to larger)
Example:

Float field based on:

"Group" attribute from geometry
"position" attribute from geometry
Index node
Material Selection node
"top faces" from Cube node
"attribute" from Attribute Capture node
"normal" from Distribute Points on Faces node

We should not expose the users to the name "Anonymous Attribute". {[F10693974](https://archive.blender.org/developer/F10693974/attr.png), size=full} Instead we can change the text to follow a simpler set of rules that will help users to know where the attributes (and non-attributes) came from: --- Sorting criteria: * Geometry * Nodes * Attributes inside node * Then per group we sort by length (smaller to larger) Example: --- Float field based on: * "Group" attribute from geometry * "position" attribute from geometry * Index node * Material Selection node * "top faces" from Cube node * "attribute" from Attribute Capture node * "normal" from Distribute Points on Faces node

Dalai Felinto commented

2021-09-30 18:01:40 +02:00

Changed status from 'Needs Triage' to: 'Confirmed'

Dalai Felinto commented

2021-09-30 18:01:40 +02:00

Added subscriber: @dfelinto

Dalai Felinto commented

2021-09-30 18:08:33 +02:00

Added subscriber: @pablovazquez

Hans Goudey commented

2021-09-30 18:11:44 +02:00

Added subscriber: @HooglyBoogly

Hans Goudey commented

2021-09-30 18:11:44 +02:00

I like using the node name instead of "Anonymous", but I think it's important to keep "Attribute" here, since fields can depend on things besides attributes, like the index.

Dalai Felinto commented

2021-09-30 18:19:47 +02:00

In this case:

Float Field based on:
* Distribute Points on Faces → normal
* Attribute Capture → Attribute
* Attribute: Group
* Attribute: position

+1 for Attributes then. I wonder why "normal" is lowercase in this example by the way.
On a related note Pablo suggested we do drop uppercase for Field, which I incorporated here as well.

Not sure whether mixing → and : works. @pablovazquez what do you think?

In this case: ``` Float Field based on: * Distribute Points on Faces → normal * Attribute Capture → Attribute * Attribute: Group * Attribute: position ``` +1 for Attributes then. I wonder why "normal" is lowercase in this example by the way. On a related note Pablo suggested we do drop uppercase for Field, which I incorporated here as well. Not sure whether mixing `→` and `:` works. @pablovazquez what do you think?

Pablo Vazquez commented

2021-09-30 18:44:49 +02:00

In #91851#1229003, @HooglyBoogly wrote:
I like using the node name instead of "Anonymous", but I think it's important to keep "Attribute" here, since fields can depend on things besides attributes, like the index.

Even though index may not be technically an attribute, the user can use it as such so there is no difference in using the word attribute at the beginning of the tooltip.

In #91851#1229009, @dfelinto wrote:
Not sure whether mixing → and : works. @pablovazquez what do you think?

Attributes and node names are not that long that we need to use shortcuts. I think we can verbose to explain a bit more where it all comes from.

Since it's now possible to use italics in Blender (although I don't think it's used anywhere yet), this could be a good opportunity to use them. To tell the attributes apart from the rest of the sentence and node names.

How about the following:

Float field based on attributes:

normal from node Distribute Points on Faces
Attribute from node Attribute Capture
Group from geometry
position from geometry

> In #91851#1229003, @HooglyBoogly wrote: > I like using the node name instead of "Anonymous", but I think it's important to keep "Attribute" here, since fields can depend on things besides attributes, like the index. Even though index may not be technically an attribute, the user can use it as such so there is no difference in using the word *attribute* at the beginning of the tooltip. > In #91851#1229009, @dfelinto wrote: > Not sure whether mixing `→` and `:` works. @pablovazquez what do you think? Attributes and node names are not that long that we need to use shortcuts. I think we can verbose to explain a bit more where it all comes from. Since it's now possible to use italics in Blender (although I don't think it's used anywhere yet), this could be a good opportunity to use them. To tell the attributes apart from the rest of the sentence and node names. How about the following: --- Float field based on attributes: * *normal* from node Distribute Points on Faces * *Attribute* from node Attribute Capture * *Group* from geometry * *position* from geometry ----

Dalai Felinto commented

2021-10-01 11:50:23 +02:00

Added subscriber: @JacquesLucke

Dalai Felinto commented

2021-10-01 11:50:23 +02:00

I was with Pablo when he posted, so +1 for me. But what @JacquesLucke and @HooglyBoogly think?

Jacques Lucke commented

2021-10-01 12:17:22 +02:00

I'm generally fine with this. Don't have a strong opinion about the exact details of the wording in the tooltip. If that helps people the most, great. Some notes:

No idea how difficult it will be to use italics here, especially when you say that we don't use them anywhere yet. Maybe it's easy, but more often than not, these small changes are rather complex in the ui code.
It assumes that every input to a field is an attribute. That can be true if we redefine an attribute to be everything that can be an input to a field, but usually the definition of an attribute for me is "Data that is stored per element of a domain on a geometry". For now it may be fine to call all of the field inputs attributes, but I'm not sure if that will work in the future. Maybe this particular issue could be solved by removing "attributes" from the first line.
What would the field input coming out of the Material Selection node be called?

I'm generally fine with this. Don't have a strong opinion about the exact details of the wording in the tooltip. If that helps people the most, great. Some notes: * No idea how difficult it will be to use italics here, especially when you say that we don't use them anywhere yet. Maybe it's easy, but more often than not, these small changes are rather complex in the ui code. * It assumes that every input to a field is an attribute. That can be true if we redefine an attribute to be everything that can be an input to a field, but usually the definition of an attribute for me is "Data that is stored per element of a domain on a geometry". For now it may be fine to call all of the field inputs attributes, but I'm not sure if that will work in the future. Maybe this particular issue could be solved by removing "attributes" from the first line. * What would the field input coming out of the Material Selection node be called?

Dalai Felinto commented

2021-10-01 12:47:38 +02:00

A proposal based on Jacques' remarks:

Float field based on:

normal from Distribute Points on Faces node
Attribute Capture node
Material Selection node
Group from geometry
position from geometry

If the node has more than output (e.g., Distribute Points on Faces) we refer to the attribute name.
Material Selection at the moment is hardcoded to always say "Material Selection", regardless of the node name/label.

A proposal based on Jacques' remarks: --- Float field based on: * *normal* from Distribute Points on Faces node * Attribute Capture node * Material Selection node * *Group* from geometry * *position* from geometry --- 1. If the node has more than output (e.g., Distribute Points on Faces) we refer to the attribute name. 2. Material Selection at the moment is hardcoded to always say "Material Selection", regardless of the node name/label.

Dalai Felinto commented

2021-10-01 14:34:50 +02:00

Final proposal:

Sorting criteria:
- Geometry
- Nodes
- Attributes inside node
Then per group we sort by length (smaller to larger)

Example:

Float field based on:
* "Group" attribute from geometry
* "position" attribute from geometry
* Index node
* Material Selection node
* "top faces" from Cube node
* "attribute" from Attribute Capture node
* "normal" from Distribute Points on Faces node

Final proposal: * Sorting criteria: * Geometry * Nodes * Attributes inside node * Then per group we sort by length (smaller to larger) Example: ``` Float field based on: * "Group" attribute from geometry * "position" attribute from geometry * Index node * Material Selection node * "top faces" from Cube node * "attribute" from Attribute Capture node * "normal" from Distribute Points on Faces node ```

Jacques Lucke commented

2021-10-15 15:14:22 +02:00

Changed status from 'Confirmed' to: 'Resolved'

Jacques Lucke closed this issue

2021-10-15 15:14:22 +02:00

Jacques Lucke self-assigned this 2021-10-15 15:14:22 +02:00

Dalai Felinto commented

2021-10-15 18:33:08 +02:00

@JacquesLucke when was this addressed? What we have in master now seems still different than #91881. I will check this again next week