Welcome! Sorry for the delayed response.
As you know, a layer marked with [bone]
creates a bone at the center of the layer, but I wouldn't suggest sizing a layer solely to get an exact bone position. It would be difficult to be accurate that way. The bone tag is still convenient to get the basic bone hierarchy from the Photoshop layers on the initial import into Spine, then the bone positions can be adjusted in Spine (eg with bone/attachment compensation).
I agree it would be convenient to use a layer solely to mark a bone origin, without saving it as a PNG file. While there's still plenty of configuration that can only be done in Spine (length, rotation, color, transform inheritance, etc), knowing where the bone origins are can be helpful in Photoshop for placing images. We would still use the center of the layer, but an X or other symmetrical image could be used.
It would be nice to not need a new tag, but I don't see a way to do that. We'd not want to use [bone][ignore]
, since ignore really should ignore the layer entirely, plus it's not discoverable. I think origin
is more descriptive than mark
. A layer tagged with [origin]
must be under a [bone]
group, it's center will be used for the bone position, and the layer will not be exported as a PNG. We'll make this addition soon and post in this thread when it's done.
You should not need to start a skeleton from scratch to update it from Photoshop, within reason. The import data dialog can import into an existing skeleton:
Loading Image
Most commonly this is used to add new attachments to a skeleton. Image only updates can be done without import data. Replacing existing attachments can lose work done with the old attachments, like weights or keys.