We have a problem, again

[Magnus Forsberg]

Remember adapting 41750 to the shape of 61678 ?
It turns out that the problem is even bigger. The Porsche slope has to fit too.

   

   

The grey are adapted to fit under the trans slope, and the yellow are adapted to fit the trans yellow, but these 2 groups should have the same slope shape.

[Gerald Lasser]

I think we shall check the raw parts from the BI app...

But wasn't the Porsche slope adapted to 61678?

[Willy Tschager]

I think we shall check the raw parts from the BI app...

But wasn't the Porsche slope adapted to 61678?

Evert-Jan Boer is working on it. Who has some fresh BI data?

w.

[Gerald Lasser]

Evert-Jan Boer is working on it. Who has some fresh BI data?

w.

Get me a list of design-IDs I will get you the raw data

[Javier Orquera]

Evert-Jan Boer is working on it. Who has some fresh BI data?

w.

Attached the files, scaled and centered to match the ldraw parts; part 61678 is 11153 on BIA.

[Willy Tschager]

Attached the files, scaled and centered to match the ldraw parts; part 61678 is 11153 on BIA.

This sums-up the problem pretty well:

https://forums.ldraw.org/thread-23953-po...l#pid36418

Affected are:

42022 - https://library.ldraw.org/parts/8218
93606 - https://library.ldraw.org/parts/13997
77182 - https://library.ldraw.org/parts/26315
61678 - https://library.ldraw.org/parts/11107
44126 - https://library.ldraw.org/parts/8672
41750 - https://library.ldraw.org/parts/8015

This involves a ton of patterned parts and sticker shortcuts. Looking at the sheer number of parts, I wonder if we should overhaul them all for the shape of a curve which has no impact on the building experience itself. Speaking for myself, I can live with that ;-)

I'm fine with you fixing the top curve of 46413 and 42022 (as it doesn't involve too man parts), but let's keep the underside untouched.

w.

[Willy Tschager]

This is still unresolved. And this is still at the PT:

https://library.ldraw.org/parts/24932

Opinions?

w.

[Rene Rechthaler]

I think the 41751 (the triple sloped windshield) is also affected, the current underside curve doesnt match any curved brick top, because its kinda hi-res compared to the rest...
(same for the top, but I think there arent much cases where the top matters...)
41751p01 (the only print) also misses the BFC NOCLIP for the pattern

[Chris Böhnke]

Not sure it helps, but here is a comparison of the current PT version (red) and the LDD mesh (blue):

   

LDD raw mesh attached.

[Willy Tschager]

Our approch should be remeasuring/remodelling:

https://library.ldraw.org/parts/8218

(Including pattern/shortcuts)

With a 48er and start from there. Say the same procedure with the minifig legs.

w.

[Chris Böhnke]

   

The main difference is the front is 5 LDU in the LDD one, 4 LDU in ours.

4 LDU results in a neat integer radius for the curve. The angle of the circle sector is 28.0725°.
Using this:

Code:

1 2 10 170 -20 0 -20 0 -170 0 0 0 0 -170 48\1-16cylo.dat

Results in:

  Screenshot 2026-10-15 230823.png (Size: 6.48 KB / Downloads: 77)

Which already has a bit of a difference to our current one.

On a sidenote, this part has no patterns currently made (though there are quite some), we only have the Aliases for the PC and MABS plastic versions.

[Willy Tschager]

How about collisions:

   

w.

[Chris Böhnke]

The plate above the jumper would not fit a real one (just tested). Need to test the others...

Edit:

Hard to tell about the 1x3 Plate above that. It does connect, but you can feel it's wobbly. I'd say that doesn't fit either. The top most plate obviously fits.

[Willy Tschager]

If you wanna tackle it, cleaning the entire mess I'm here to review.

w.

[Willy Tschager]

If you wanna tackle it, cleaning the entire mess I'm here to review.

w.

[Chris Böhnke]

First one:
https://library.ldraw.org/parts/54522

If we can all agree on this, the rest will follow.

As mentioned in the older thread, using the structure of parts 61678 and 93606 (both identical) seems to be the best option we have. Those 2 have a total of 58 sticker shortcuts made, which would all need to be reworked should those 2 parts change (and I think many were already reworked recently...).

I've compared the solution from 61678 to the real part and the collision on the curve seems identical (when plates are placed in front, like in one of the above images). The curve is also a fairly good match to the LDD mesh (except the 1 LDU difference on the front lip). So this seems like good enough?

[Magnus Forsberg]

...... So this seems like good enough?

Yes, I agree to this solution.

[Chris Böhnke]

Currently working on the last part in the smaller curve family (1x4 arc).

Next one up will be the 2x6 arc, like in the top of https://library.ldraw.org/parts/24934

As far as I could find,  there hasn't been any established version yet (correct me if I'm wrong).

Some of the current ones also seem to allow builds that would result in collisions, when done with real ones. The real collision curve would look like this, I'd say:


  Screenshot 2026-10-24 024920.png (Size: 35.43 KB / Downloads: 15)

This curve also has to fit some other parts, for example 41881.

My plan is to rework those starting with 41766 (the most basic variant).

Concept for the curve:

Code:

1 16 -20 126.2219 40 0 40 0 -150.2219 0 0 0 0 -169.7073 48\1-8cyli.dat

   

Match seems pretty good.
Are 6 faces enough?

Any thoughts?


  41766concept.dat (Size: 1.05 KB / Downloads: 1)