[x3d-public] SFRotation default value spec. inconsistency
Joe D Williams
joedwil at earthlink.net
Wed Feb 19 13:06:29 PST 2025
OK, Good, I was worried about x y z all being zero yet have some value for angle not zero.
Thanks,
Joe
-----Original Message-----
From: John Carlson <yottzumm at gmail.com>
Sent: Feb 19, 2025 8:10 AM
To: Extensible 3D (X3D) Graphics public discussion <x3d-public at web3d.org>
Cc: Joe D Williams <joedwil at earthlink.net>, Andreas Plesch <andreasplesch at gmail.com>
Subject: Re: [x3d-public] SFRotation default value spec. inconsistency
Yes, that was the point I was trying to make with 0 0 0 not being an axis or normalized. Thanks for backup.
On Wed, Feb 19, 2025 at 8:51 AM Andreas Plesch via x3d-public <x3d-public at web3d.org (mailto:x3d-public at web3d.org)> wrote:
On Tue, Feb 18, 2025 at 10:16 AM Joe D Williams <joedwil at earthlink.net (mailto:joedwil at earthlink.net)> wrote:
> ... may prefer the equivalent zero-rotation default value 0 1 0 0
..
Thanks Don. So the Transform rotation and Viewpoint orientation
and any other use of rotation and orientation are treated as
default zero when angle is zero, or, if all x, y, and z are zero.
Just a caution that
* https://www.web3d.org/specifications/X3Dv4/ISO-IEC19775-1v4-IS/Part01/fieldTypes.html#SFRotationAndMFRotation
says "The first three values specify a normalized rotation axis vector". So x,y,z all zero is not a valid (illegal) SFRotation value.
-Andreas
-----Original Message-----
From: Brutzman, Donald (Don) (CIV) <brutzman at nps.edu (mailto:brutzman at nps.edu)>
Sent: Feb 17, 2025 9:08 PM
To: Andreas Plesch <andreasplesch at gmail.com (mailto:andreasplesch at gmail.com)>, Joseph D Williams <joedwil at earthlink.net (mailto:joedwil at earthlink.net)>
Cc: Extensible 3D (X3D) Graphics public discussion <x3d-public at web3d.org (mailto:x3d-public at web3d.org)>
Subject: Re: [x3d-public] SFRotation default value spec. inconsistency
Thanks for discussion and for noticing the difference. Related references:
* 19777-1 X3D ECMAScript language binding, clause 7 Function definitions, 7.6.6 SFRotation
* https://www.web3d.org/documents/specifications/19777-1/V3.3/Part1/functions.html#SFRotation
* Table 7.18 — SFRotation instance creation functions
* SFRotation
(numeric x,
numeric y,
numeric z,
numeric angle)
x, y, and z are the axis of the rotation. angle is the angle of the rotation (in radians). Missing values default to 0.0, except y, which defaults to 1.0.
* X3D Architecture, clause 5 Field type reference, 5.3.13 SFRotation and MFRotation
* https://www.web3d.org/specifications/X3Dv4/ISO-IEC19775-1v4-IS/Part01/fieldTypes.html#SFRotationAndMFRotation
* "The default value of an uninitialized SFRotation field is (0 0 1 0). The default value of an MFRotation field is the empty list."
Yes they must be the same, so that X3D models are equivalent no matter what file encoding or programming-language binding are used.
The X3D Architecture takes precedence. So we will change 19777-1 to match and the highlighted y above will become z.
I too think that a default of (0 1 0 0) makes more sense, animation rotations are most often around the vertical (y) axis. However, the community has discussed this many times over the years. We always landed on (0 0 1 0) for backwards-compatibility reasons, not wanting to break legacy X3D and VRML models.
Incidentally here is the X3D tooltips entry. Improvements and refinements always welcome.
* X3D Tooltips: SFRotation
* https://www.web3d.org/x3d/tooltips/X3dTooltips.html#SFRotation
* SFRotation (https://www.web3d.org/specifications/X3Dv4/ISO-IEC19775-1v4-IS/Part01/fieldTypes.html#SFRotationAndMFRotation)
Single-Field rotation value using 3-tuple axis, radian angle form
Default value 0 0 1 0. Model authors and authoring tools may prefer the equivalent zero-rotation default value 0 1 0 0 since rotation about the vertical Y-axis is most common. Example values: 0 1 0 1.57
Warning: initial 3-tuple axis vector cannot hold a zero-magnitude vector.
Warning: comma characters within attribute values are not allowed, and do not pass strict validation.
Hint: see Wikipedia: Radian (https://en.wikipedia.org/wiki/Radian), Rotation matrix (https://en.wikipedia.org/wiki/Rotation_matrix) and Rotation formalisms in three dimensions (https://en.wikipedia.org/wiki/Rotation_formalisms_in_three_dimensions).
XML Schema (https://www.web3d.org/specifications/X3dSchemaDocumentation4.0/x3d-4.0_SFRotation.html), X3D regexes (https://www.web3d.org/specifications/X3dRegularExpressions.html#SFRotation), Java SAI (https://www.web3d.org/specifications/JavaLanguageBinding/Part2/nodeTypeInterfaces.html#SFRotation), X3DJ (https://www.web3d.org/specifications/java/javadoc/org/web3d/x3d/jsail/fields/SFRotation.html)
To keep track of this erratum, I have entered a new issue in the Mantis issue tracker.
* Mantis 1488: (ECMAScript) 7.6.6 SFRotation default value needs to be (0 0 1 0)
* https://mantis.web3d.org/view.php?id=1488
I hope this make sense and sounds good.
Have fun with X3D - or else!! 🙂
all the best, Don
--
Don Brutzman Naval Postgraduate School, Code USW/Br brutzman at nps.edu (mailto:brutzman at nps.edu)
Watkins 270, MOVES Institute, Monterey CA 93943-5000 USA +1.831.656.2149
X3D graphics, virtual worlds, navy robotics https://faculty.nps.edu/brutzman
From: x3d-public <x3d-public-bounces at web3d.org (mailto:x3d-public-bounces at web3d.org)> on behalf of Joe D Williams via x3d-public <x3d-public at web3d.org (mailto:x3d-public at web3d.org)>
Sent: Sunday, February 16, 2025 9:09 AM
To: Andreas Plesch <andreasplesch at gmail.com (mailto:andreasplesch at gmail.com)>
Cc: joedwil at earthlink.net (mailto:joedwil at earthlink.net) <joedwil at earthlink.net (mailto:joedwil at earthlink.net)>; X3D <x3d-public at web3d.org (mailto:x3d-public at web3d.org)>
Subject: Re: [x3d-public] SFRotation default value spec. inconsistency
> I would vote for 0 1 0 0 since rotations around Y seem more common
First,oldest use is in Part 1, way back to origin of VRML standard.
Is Yaw really more popular or common than Pitch or Roll?
For me, Please let's stick with oldest history and update more recent.
Thanks
Joe
-----Original Message-----
From: Andreas Plesch <andreasplesch at gmail.com (mailto:andreasplesch at gmail.com)>
Sent: Feb 15, 2025 9:12 PM
To: <joedwil at earthlink.net (mailto:joedwil at earthlink.net)>
Cc: Holger Seelig <holger.seelig at yahoo.de (mailto:holger.seelig at yahoo.de)>, X3D <x3d-public at web3d.org (mailto:x3d-public at web3d.org)>
Subject: Re: [x3d-public] SFRotation default value spec. inconsistency
I agree, probably the best default value for a unit quaternion is 0 0 0 1.
However, X3D does not concern itself with quaternions, only axis-angle rotations.
So the original question remains: Part 1 has axis-angle 0 0 1 0 as default while the JS SAI spec. has 0 1 0 0. Should both list the same default value ?
If yes, which one ? I would vote for 0 1 0 0 since rotations around Y seem more common.
-Andreas
On Sat, Feb 15, 2025 at 11:04 PM <joedwil at earthlink.net (mailto:joedwil at earthlink.net)> wrote:
Doesn't this mean that the _best_ default value, for no rotation, is 'legally' for unit quaternion = 0 0 0 1
because the square root of the sum of the squares of all four components does = 1 ?
0 0 0 0 would not be a legal unit quaternion.
While the axis-angle default is 0 0 1 0 as appears in Part 1 is mostly for convenience
since many value combinations can produce zero rotation?
Also sure shows why axis-angle is so much easier to work with when typing numbers,
pitch and yaw and roll for all.
Thanks,
Joe
-----Original Message-----
From: Andreas Plesch <andreasplesch at gmail.com (mailto:andreasplesch at gmail.com)>
Sent: Feb 14, 2025 11:36 AM
To: Holger Seelig <holger.seelig at yahoo.de (mailto:holger.seelig at yahoo.de)>
Cc: X3D <x3d-public at web3d.org (mailto:x3d-public at web3d.org)>, Joseph D Williams <joedwil at earthlink.net (mailto:joedwil at earthlink.net)>
Subject: Re: [x3d-public] SFRotation default value spec. inconsistency
Let me just add for clarity that the length (or norm) of a quaternion includes the fourth, the scalar component. The norm is the square root of the sum of the squares of all four components.
Only unit quaternions represent rotations. A (0,0,0,0) quaternion does not represent any rotation (but of course exists outside of rotations).
Andreas
PS: To add back in some fuzziness, a unit quaternion can only represent a rotation up to 360 degrees. I think there is a mechanism to use non-unit quaternions for larger rotations.
On Fri, Feb 14, 2025 at 10:22 AM Holger Seelig <holger.seelig at yahoo.de (mailto:holger.seelig at yahoo.de)> wrote:
All quaternions must be normalized to be a legal rotation, so the length of the quaternion must always be one.
Holger--
Holger Seelig
Leipzig, Germany
holger.seelig at yahoo.de (mailto:holger.seelig at yahoo.de)
https://create3000.github.io/x_ite/
Am 14.02.2025 um 15:38 schrieb Joe D Williams via x3d-public <x3d-public at web3d.org (mailto:x3d-public at web3d.org)>:
> So, sensibly, a zero rotation around any axis corresponds to the "zero unit Quaternion" (0, 0, 0, 1 XYZW),
Right. Not sure but 0 0 0 0 for unit quaternion might not be legal. I think the values must compute to1 to be legal?
Thanks,
Joe
-----Original Message-----
From: Andreas Plesch <andreasplesch at gmail.com (mailto:andreasplesch at gmail.com)>
Sent: Feb 10, 2025 12:37 PM
To: Joe D Williams <joedwil at earthlink.net (mailto:joedwil at earthlink.net)>
Cc: Extensible 3D (X3D) Graphics public discussion <x3d-public at web3d.org (mailto:x3d-public at web3d.org)>
Subject: Re: [x3d-public] SFRotation default value spec. inconsistency
https://registry.khronos.org/glTF/specs/2.0/glTF-2.0.html#_node_rotation
has indeed (0, 0, 0, 1 XYZW) as a default value.
This is difficult to interpret as a rotation around an axis since the corresponding axis would be at first glance (0 0 0).
x3dom converts quaternion (0, 0, 0, 1 XYZW) to (axis),angle (0,0,0),0 as does https://www.andre-gaschler.com/rotationconverter/ (https://www.andre-gaschler.com/rotationconverter/) .
On the other hand both SFRotation (0 0 1 0) and SFRotation (0 1 0 0) are also represented by quaternion (0, 0, 0, 1 XYZW).
So, sensibly, a zero rotation around any axis corresponds to the "zero unit Quaternion" (0, 0, 0, 1 XYZW), making it a useful default value for quaternions. But that does not help with choosing a default axis for the SFRotation value.
This will rarely matter. A use case may be when a SFRotation is initialized to the X3D default in a Proto but then only the angle is modified by a Proto script with the assumption that the axis is (0 0 1).
-Andreas
On Mon, Feb 10, 2025 at 12:57 PM Joe D Williams <joedwil at earthlink.net (mailto:joedwil at earthlink.net)> wrote:
not zerp, zero
-----Original Message-----
From: Extensible 3D (X3D) Graphics public discussion <x3d-public at web3d.org (mailto:x3d-public at web3d.org)>
Sent: Feb 10, 2025 9:50 AM
To: Extensible 3D (X3D) Graphics public discussion <x3d-public at web3d.org (mailto:x3d-public at web3d.org)>, X3D Graphics public mailing list <x3d-public at web3d.org (mailto:x3d-public at web3d.org)>
Cc: Joe D Williams <joedwil at earthlink.net (mailto:joedwil at earthlink.net)>, Andreas Plesch <andreasplesch at gmail.com (mailto:andreasplesch at gmail.com)>
Subject: Re: [x3d-public] SFRotation default value spec. inconsistency
I think the 0 0 1 0 should be used everywhere for default axis-angle.
I thinkI see 'zerp' unit quaternions (as in gltf as 0 0 0 1 XYZW)
Thanks,
Joe
-----Original Message-----
From: Extensible 3D (X3D) Graphics public discussion <x3d-public at web3d.org (mailto:x3d-public at web3d.org)>
Sent: Feb 10, 2025 8:44 AM
To: X3D Graphics public mailing list <x3d-public at web3d.org (mailto:x3d-public at web3d.org)>
Cc: Andreas Plesch <andreasplesch at gmail.com (mailto:andreasplesch at gmail.com)>
Subject: Re: [x3d-public] SFRotation default value spec. inconsistency
The JS SAI spec. at https://www.web3d.org/documents/specifications/19777-1/V3.3/Part1/functions.html#FieldServices (https://www.web3d.org/documents/specifications/19777-1/V3.3/Part1/functions.html#FieldServices)
does not list functions for SFColorRGBA, only SFColor. There may be an earlier report. -Andreas
On Mon, Feb 10, 2025 at 11:31 AM Andreas Plesch <andreasplesch at gmail.com (mailto:andreasplesch at gmail.com)> wrote:
https://www.web3d.org/documents/specifications/19775-1/V4.0/Part01/fieldTypes.html#SFRotationAndMFRotation
specifies that (0 0 1 0) is the value of an uninitialized SFRotation given as (x y z a).
However, the JS SAI spec. in table 7.18 in https://www.web3d.org/documents/specifications/19777-1/V3.3/Part1/functions.html#SFRotationInstanceCreationFunction (https://www.web3d.org/documents/specifications/19777-1/V3.3/Part1/functions.html#SFRotationInstanceCreationFunction)
effectively makes (0 1 0 0) the default value if no parameters are provided.
The inconsistency does not have much of an impact since both are 0 rotations but I believe (0 1 0 0) - a 0 rotation around y - may have been intended as the default.
-Andreas
--
Andreas Plesch
Waltham, MA 02453
--
Andreas Plesch
Waltham, MA 02453
--
Andreas Plesch
Waltham, MA 02453
_______________________________________________
x3d-public mailing list
x3d-public at web3d.org (mailto:x3d-public at web3d.org)
http://web3d.org/mailman/listinfo/x3d-public_web3d.org
--
Andreas Plesch
Waltham, MA 02453
--
Andreas Plesch
Waltham, MA 02453
--
Andreas Plesch
Waltham, MA 02453
_______________________________________________
x3d-public mailing list
x3d-public at web3d.org (mailto:x3d-public at web3d.org)
http://web3d.org/mailman/listinfo/x3d-public_web3d.org
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://web3d.org/pipermail/x3d-public_web3d.org/attachments/20250219/a8e9c8c2/attachment-0001.html>
More information about the x3d-public
mailing list