Orientation of Duplo 11197

[Peter Blomberg]

Let's discuss the orientation of (sloped) Duplo parts.

First of all, sloped Duplo parts have not been oriented as systematically as regular parts. There were some trends (reviewed here https://forums.ldraw.org/thread-28480.html). The existing official parts have several orientations where one major trend seems to be the patterning of a side surface taking precedence over 'slopeness'. This may be partly due to the LDPE editor, where -z is considered the front face of the part and the patterns have been considered the front in common use cases.

Part 11197 is mainly patterned on the side and used in conjunction with part 4198 (https://library.ldraw.org/parts/8177), which is also mainly patterned on the side and oriented in the -z direction.

Therefore, it makes sense to orient part 11197 also in the -z direction.

An alternative would be to rotate all existing sloped parts and their patterns to fit the same convention adopted for sloped regular lego.

[Willy Tschager]

Consider also that the new PREVIEW meta solves a lot of issues but I agree that all sloped/curved parts should share a logical orientation.

Feel free to "moveto" official part with a new orientation as you see fit. Document the reason for the new orientation in the HISTORY.

w.

[Orion Pobursky]

An alternative would be to rotate all existing sloped parts and their patterns to fit the same convention adopted for sloped regular lego.

This is the preferred route. How many parts would we have to change?

[Peter Blomberg]

This is the preferred route. How many parts would we have to change?

13 official parts, plus 3 derivative parts, 9 subparts, and 25 patterns. Additional 9 unofficial patterns.

Part numbers: 3664, 31213, 2301, 2302, 98223, 4198, 72133, 4258, 73162, 98252, 98224, u977, 4197, 11198, 250, and 51704.

These updates/corrections could be made at the same time:
Part 31213 should be subfiled for patterns. 2302 remove invisible lines on inside, already on PT. Parts 98252, 98224, and 4197 curves should be updated to match 11169. Bottom curves of 4198 should be updated to match 11169. Some of the older parts are not using stud groups.

[Peter Blomberg]

Which are considered slopes and which arches?

Back row: 250 and 51704 in light bluish grey.
u977, 4197, and 11198 in light bluish grey.
72133, 4258, and 73162 in dark bluish grey.
98223, 4198, 98252, and 98224 in dark bluish grey.
Front row: 3664, 31213, 2301, and 2302 in dark bluish grey.

   

Is the division indicated by the coloring acceptable?

[Willy Tschager]

Is the division indicated by the coloring acceptable?

Looks good to me.

w.

[Peter Blomberg]

What is the proper procedure for changing the rotation of a part?
Will there be a and b versions with a moveto on the old?

Is this how it would go?
XXXXX.dat with ~moved to XXXXXa.dat as description
XXXXXa.dat with old orientation
XXXXXb.dat with new orientation

Will all three files need to be submitted to PT?
Can admins do this more easily?

Will subfiles also need to be rotated?

[Rene Rechthaler]

The obsolete and new versions are only needed for already official parts...
If the part is still the first time on the tracker (unofficial), it can get changed easily (Rotation , Position)

[Willy Tschager]

What is the proper procedure for changing the rotation of a part?
Will there be a and b versions with a moveto on the old?

Is this how it would go?
XXXXX.dat with ~moved to XXXXXa.dat as description
XXXXXa.dat with old orientation
XXXXXb.dat with new orientation

The Movedto refs to XXXXXa.dat with the new orientation, restoring the old orientation in the matrix:

1 16 0 0 0 0 0 1 0 1 0 -1 0 0 XXXXXa.dat

XXXXXa.dat was rotated 90 degree on Y and the Movedto restores the orientation.

If you also change the geometry, the old part has to be obsoleted.

Will all three files need to be submitted to PT?
Can admins do this more easily?

You have to submit both XXXXX.dat and XXXXXa.dat. No, 'cos they first have to figure out what's going on and then do the same procedure.

Will subfiles also need to be rotated?

No.

w.

[Magnus Forsberg]

What is the proper procedure for changing the rotation of a part?
Will there be a and b versions with a moveto on the old?

Is this how it would go?
XXXXX.dat with ~moved to XXXXXa.dat as description
XXXXXa.dat with old orientation
XXXXXb.dat with new orientation

Will all three files need to be submitted to PT?
Can admins do this more easily?

Will subfiles also need to be rotated?

I think you are mixing 2 different scenarios here.

If all you want to do is change the orientaion of an official part you should make the new position as the b-version, and the old number becomes a Moved to XXXXb. There no need to move the old position to an a-version.
We have some rims and tyres in the library using this method.

An a-version + b-version is only needed if the design is split into different versions. 
Like when there is a different axle hole, in the same rim design.

• If bad geo, or bad orientation, make a ~Moved to XXXb
  
• If design is split, numbers kept, make a ~Moved to XXXa and add the new version as XXXb
  

[Peter Blomberg]

Thanks.

Summarizing:

For each official part that needs to be rotated, create and submit

XXXXX.dat
Description = "~Moved to XXXXXa.dat"
Author = Me
Content = "1 16 0 0 0 {rotation matrix} XXXXXa.dat" (the rotation matrix restores the old orientation)

XXXXXa.dat
Description = Same as before
Author = Same as before
History line = "Rotated part" + any minor corrections
Content = all previous content + any minor corrections

Should the reason for the rotation be documented in the history line or is it enough to state the reason in the PT comment upon submission?

[Magnus Forsberg]

Thanks.

Summarizing:

For each official part that needs to be rotated, create and submit

XXXXX.dat
Description = "~Moved to XXXXXa.dat"
Author = Me
Content = "1 16 0 0 0 {rotation matrix} XXXXXa.dat" (the rotation matrix restores the old orientation)

XXXXXa.dat
Description = Same as before
Author = Same as before
History line = "Rotated part" + any minor corrections
Content = all previous content + any minor corrections

Should the reason for the rotation be documented in the history line or is it enough to state the reason in the PT comment upon submission?

No, that's not what I said.
If the geometry is wrong you should create the new (second) version with a b-suffix.
Note that ".dat" should not be included in the Moved to-file description.


XXXXX.dat
Description = "~Moved to XXXXXb"
Author = Me
Content = "1 16 0 0 0 {rotation matrix} XXXXXb.dat" (the rotation matrix restores the old orientation)

XXXXXb.dat
Description = Same as before
Author = Same as before
History line = "corrected orientation by creating b replacement file" + any minor corrections
Content = all previous content + any minor corrections

[Peter Blomberg]

Thank you Magnus for the clarification!

Sometimes it's a little bit confusing with multiple different answers. I didn't know a and b had different meaning. I just assumed that you used b because I had used a for the old and b for the new orientation in my original post. Now I know that the letter has a meaning

[Willy Tschager]

Thank you Magnus for the clarification!

Sometimes it's a little bit confusing with multiple different answers. I didn't know a and b had different meaning. I just assumed that you used b because I had used a for the old and b for the new orientation in my original post. Now I know that the letter has a meaning

Believe it or not until now I didn't know "a" has a meaning other than create a variant of an already used number. So what meaning has "a"?

w.

[Magnus Forsberg]

   

[Willy Tschager]

Could you please point me to the ref? The only spec I could found in regard was:

https://www.ldraw.org/part-number-spec.html#multi_num

w.

[Orion Pobursky]

@Magnus

I'm not saying that be need to change this practice (since we're already doing it), but this is not what the spec implies:

Occasionally a part is included with geometry that is not quite correct or the part number itself is incorrect. In this instance, in order to preserve backward compatibility, when a corrected part is made, it's part number will follow these rules and the non-alpha suffix part will be made obsolete.

So let's make the spec match what we want to avoid confusion.

[Peter Blomberg]

Does correcting the orientation of a part qualify for fast-tracking a fix? If so, do minor edits count, i.e. affect the fast-tracking?

Should I check the 'official file fix' on submission?

[Willy Tschager]

Does correcting the orientation of a part qualify for fast-tracking a fix? If so, do minor edits count, i.e. affect the fast-tracking?

Should I check the 'official file fix' on submission?

Only changes to the header such as desription or keywords qualify for fasttrack. Changes to the matrix have to go through review.

w.

[Orion Pobursky]

Should I check the 'official file fix' on submission?

The official file fix toggle is for all fixes, not just for minor fixes. It's really there to prevent accidental submission of an official part.