[LDCad] Part Snapping missing/errors

[Orion Pobursky]

This is intended to be a catchall thread to let Roland know about missing/problematic part snapping info.

Note: make sure that the part you are reporting is official

[Orion Pobursky]

75181 (Turntable Flat Base 4x4) is missing snap info

[Orion Pobursky]

55817, Wheel 21x56 with 8 Spokes Armoured. No snap info for pinhole and the 8 holes on the sides.

[Orion Pobursky]

All the flex-system hoses have no snap info.

[N. W. Perry]

The various components of the old 6x16 train assembly are missing snap info. I've added my own, as follows:

In the 6x16 base plate itself (736.dat), I added female snap cylinders for the two magnet holders, and for the two axles:

Code:

0 !LDCAD SNAP_CYL [gender=F] [caps=one] [secs=R 6 12] [pos=0 39 -140]
0 !LDCAD SNAP_CYL [gender=F] [caps=one] [secs=R 6 12] [pos=0 39 140]
0 !LDCAD SNAP_CYL [group=axleR4] [gender=F] [caps=none] [secs=R 4 80] [pos=-40 42 -100] [ori=0 -1 0 1 0 0 0 0 1]
0 !LDCAD SNAP_CYL [group=axleR4] [gender=F] [caps=none] [secs=R 4 80] [pos=-40 42 100] [ori=0 -1 0 1 0 0 0 0 1]
0 !LDCAD SNAP_CYL [gender=F] [caps=one] [secs=S 6 4] [pos=-50 8 -150] [grid=2 16 100 20]

In the magnet holder (735.dat), I added a male cylinder for mating to the base, along with a pair of point snaps on the magnet itself (for coupling two magnets together):

Code:

0 !LDCAD SNAP_CYL [gender=M] [caps=one] [secs=R 6 12]
0 !LDCAD SNAP_GEN [group=mag] [gender=M] [bounding=pnt] [match=group] [pos=56 -5 0]
0 !LDCAD SNAP_GEN [group=mag] [gender=F] [bounding=pnt] [match=group] [pos=56 -5 0] [ori=-1 0 0 0 1 0 0 0 -1]

And in the wheel/axle shortcut (243c02.dat) I added the male cylinder snap for mating to the base:

Code:

0 !LDCAD SNAP_CYL [group=axleR4] [gender=M] [caps=none] [secs=R 4 80] [pos=-40 0 0] [ori=0 -1 0 1 0 0 0 0 1]

Feedback/suggestions welcome!

[Orion Pobursky]

60478.dat, Plate 1x2 with Handle on End. Mirror doesn't work correctly. Probably missing info

[Roland Melkert]

This is intended to be a catchall thread to let Roland know about missing/problematic part snapping info.

Good idea.

I'll try to include things posted here in the next release, but keep in mind I'm already working on the 1903 parts.

[Roland Melkert]

60478.dat, Plate 1x2 with Handle on End. Mirror doesn't work correctly. Probably missing info

That one isn't symmetrical in the X direction. Mirroring only works without shadow info for parts that are x symmetric.

If you don't want to wait on the update it's easily fixed by adding a mirror meta to the part (shadow) using a 'base flip' of whatever axis is symmetrical (Z in this case).

[Orion Pobursky]

15207 has no bottom snap info

[Philippe Hurbain]

99012 (Technic Rotor 1 Blade 8L with Peghole and Axlehole) is missing mirroring info (and inherited connectivity can be refined).

Code:

0 !LDCAD MIRROR_INFO [baseFlip=Z]
0 !LDCAD SNAP_CLEAR [ID=axleHole]
0 !LDCAD SNAP_CYL [gender=F] [caps=one] [secs=A 6 20] [pos=0 30 0] [ori=-1 0 0 0 1 0 0 0 -1]

[Stefan Frenz]

I have problems with part-snapping of roller coaster tracks on top of other parts, i.e. 25061 (90 degree), 26561 (ramp), 34738 (S-shape), 26599 (S-shape lower)... They snap perfectly "under" another part, but not on top of it - so I let them snap below and move 2 plates (16 LDU) upwards. I don't know if this is my fault or if the parts are missing snapping info.

[Philippe Hurbain]

I have problems with part-snapping of roller coaster tracks on top of other parts, i.e. 25061 (90 degree), 26561 (ramp), 34738 (S-shape), 26599 (S-shape lower)... They snap perfectly "under" another part, but not on top of it - so I let them snap below and move 2 plates (16 LDU) upwards. I don't know if this is my fault or if the parts are missing snapping info.

This is the symptom of no snapping at all. It works from top because the part inherits connectivity of studs, but there is no common structure on plate bottom to allow subparting and inheritance.

[Philippe Hurbain]

Part 6628 is missing bar connectivity inside pin section:
0 !LDCAD SNAP_CYL [gender=F] [caps=one] [secs=R 4 18] [pos=20 0 0] [ori=0 1 0 1 0 0 0 0 -1]
See set 42106 for usage example.

[Roland Melkert]

I've finished the snapping info for the 1903 library.

I also processed all thing people mailed me about and things mentioned above up to message #13.

Thanks for all whom contributed.

You can (always) download the latest seed file at:
http://www.melkert.net/action/download/shadow.sf

Just replace the old one in LDCad's seeds location  (subfolder of the exe file).

Some notes

@Orion
I can't find 75181, did you mean 61485 ? although that one has snapping info.

"All the flex-system hoses have no snap info."
They do have info, but only for the end caps, or did you mean the bended parts. LDCad doesn't support that at the moment, maybe in 1.7

@N.W.Perry
I applied your #5 definitions with some minor changes which should make it somewhat easier to use. Feel free to review.

[Orion Pobursky]

@Orion
I can't find 75181, did you mean 61485 ? although that one has snapping info.

Yes. (75181 is the set I was working on). Unless the info is in this update, then my current shadow library does not have it

"All the flex-system hoses have no snap info."
They do have info, but only for the end caps, or did you mean the bended parts. LDCad doesn't support that at the moment, maybe in 1.7

Not the bent ones, the straight ones. And it would be helpful to have this info for the middle section.

[Stefan Frenz]

The door element 60616a does not snap into 60596 (but for example 60623 does) and the window element 57895 does not snap into 57894.

[N. W. Perry]

@N.W.Perry
I applied your #5 definitions with some minor changes which should make it somewhat easier to use. Feel free to review.

Everything looks good to me!

[Roland Melkert]

The door element 60616a does not snap into 60596 (but for example 60623 does) and the window element 57895 does not snap into 57894.

Did you update to the above shadow?

Because only 57894 doesn't have snap information in that version.

I'll add 57894 to the next update.

[Stefan Frenz]

Did you update to the above shadow?

On the tested installation: not yet, sorry.

[Stefan Frenz]

The hose reel 2585 does not snap into the hose reel holder 2584 (with shadow.sf). Not a problem, but maybe nice to have.

[Orion Pobursky]

There’s no snap info for the center pin in the 2x2 Turntable (3680 and 3679)

[N. W. Perry]

15207.dat is missing bottom stud connectivity. I copied from 30413.

[Johann Eisner]

I am not sure, if this is so wanted.
The middle hole of the Servomotor (99498) have no info.

[Philippe Hurbain]

15301 shooter bottom is missing connectivity.

Code:

0 !LDCAD SNAP_CYL [gender=F] [caps=one] [secs=R 6 4] [center=true] [pos=0 -2 -30] [grid=1 4 0 20]

Snapping for arrow is kind of missing in 15400 shooter assembly. I propose

Code:

0 !LDCAD SNAP_CYL [gender=F] [caps=one] [secs=R 5.5 9   R 4 35.9   R 4 45.1] [slide=true] [pos=0 8.446736 -48] [ori=-1 0 0 0 -0.017452 -0.999848 0 -0.999848 0.017452]
0 !LDCAD SNAP_CYL [gender=F] [caps=one] [secs=R 5.5 9   R 4 35.9   R 4 45.1] [slide=true] [pos=0 8.446736 48] [ori=1 0 0 0 -0.017452 -0.999848 0 0.999848 -0.017452]

...questionnable, but works well!

[Johann Eisner]

Pneumatic valve with axle hole 19474

No information on the hose connections

[Roland Melkert]

This is intended to be a catchall thread to let Roland know about missing/problematic part snapping info.

Not sure if this would need a new thread, but I'm introducing a new (shadow) meta in 1.6d

!LDCAD HINTS

It can (also) be used to override the rest rotation of parts in the part bin.

I used it to rotate panels so their pattern is faced forwards.

If anyone thinks other parts need rotating let me know.

[N. W. Perry]

Microfig 85863.dat (or rather its sub-part -s01) is missing bottom snap info.

[Orion Pobursky]

28326 (Car Mudguard 4x3x1) has no bottom snap info

[Gerald Lasser]

30301 has wrong bottom info

[Magnus Forsberg]

30301 has wrong bottom info

??
Can't find that number.

[Gerald Lasser]

??
Can't find that number.

Typo: 30201

[Roland Melkert]

Typo: 30201

These kinds of mistakes happen when I go trough too many parts in one go

[N. W. Perry]

Round tile 2x2 (14769.dat) is missing the connection on the center underside stud.

For that matter, I've noticed that snapping info is not generally included for any of the in-between underside studs/tubes on 2x or larger parts (not to mention the space between studs on the top surface). I'm sure this is intentional and I can fully understand the reason.

But I think it's worth including for the 2x2 round tile, since having that connection is fundamental to the (re-)design of the part.

[Roland Melkert]

Round tile 2x2 (14769.dat) is missing the connection on the center underside stud.

For that matter, I've noticed that snapping info is not generally included for any of the in-between underside studs/tubes on 2x or larger parts (not to mention the space between studs on the top surface). I'm sure this is intentional and I can fully understand the reason.

But I think it's worth including for the 2x2 round tile, since having that connection is fundamental to the (re-)design of the part.

Yes those are missing on purpose.

The main reason being the LDCad snapping engine only searches for a single connection pair when dragging stuff around. So the middle 6ldu hole would make it snap 'wrong' 95% of the time.

That said, you might be right about this part being used more often in that way.

This is the exact reason I'm considering adding a 'priority' property to the snapping metas in 1.7, which should reduce the false matches considerably.

[Roland Melkert]

I've processed all up to post #34

It will be in the def shadow of 1.6d

[N. W. Perry]

4x4 Dome (86500.dat) has no info. I put anti-studs at all four corners (under the notches), but I'm not sure that's the right approach. You may need more studs inside the rim, or maybe the whole thing has to act as one giant cylinder snap?

[Roland Melkert]

4x4 Dome (86500.dat) has no info. I put anti-studs at all four corners (under the notches), but I'm not sure that's the right approach. You may need more studs inside the rim, or maybe the whole thing has to act as one giant cylinder snap?

The corners is probably good enough.

I've also added the ring, so if something happens to use the same radius it will snap to that too.

Code:

0 !LDCAD SNAP_CYL [gender=F] [caps=one] [secs=R 6 4] [grid=C 2 C 2 60 60]
0 !LDCAD SNAP_CYL [gender=F] [caps=A] [secs=R 38 4] [pos=0 -4 0] [ori=1 0 0 0 -1 0 0 0 -1]
0 !LDCAD SNAP_CYL [gender=M] [caps=one] [secs=R 39.5 4] [pos=0 -4 0] [ori=1 0 0 0 -1 0 0 0 -1]

[N. W. Perry]

Part 30046 (window lattice) snap just needs to be nudged -1 on Y. (The matching window frame 30044 is correct.)

[N. W. Perry]

Part 4740 has overlapping cylinder snaps, one in the part and one in the sub-part. (Neither is quite correct for this part's weird dimensions; the cylinder sections should be [ R 6 5.3  R 4 6.7 ].)

[N. W. Perry]

So I've got a few more in here:

• Wheel rim 11208 has no snap info.
  
• Inverted dome 15395 has no snap info.
  
• 95109 could use a cylinder snap on the non-pointy section. However, that may be getting renumbered as a subpart.
  
• EDIT: The various 90462 (Egyptian headdress) parts don't have bottom snap info (I guess because it's a tube prim)
  

[Roland Melkert]

I've processed up to #40

Latest shadow is at:
http://www.melkert.net/action/download/shadow.sf

It also includes a whole lot of shadow files donated by Alex Taylor, some of which are for unofficial files.

I haven't had the time to review them all but this is a work in progress version of the shadow anyway

I will not include the unofficial ones in LDCad releases so some snapping functionally might 'disappear' when you upgrade to an official version.

Wheel rim 11208 has no snap info.

What kind of pin is supposed to snap to that?

[N. W. Perry]

What kind of pin is supposed to snap to that?

Good question, actually—the only set I've used it in uses a 3.18 bar. But it seems to pair with the city/town scale wheel pins, such as 4488.dat.

[N. W. Perry]

Two updates to official parts:

24866.dat
Needs the following additional cyl snap (to connect with various stem parts):

Code:

0 !LDCAD SNAP_CYL [gender=F] [caps=none] [secs=R 2 2] [pos=0 2 0]

63965a.dat
Has no snap info; should be:

Code:

0 !LDCAD SNAP_CYL [gender=M] [caps=none] [secs=R 4 16   R 6 8   R 4 96] [slide=true] [pos=0 16 0]

[N. W. Perry]

What kind of pin is supposed to snap to that?

So yeah, it's meant to just go onto wpin.dat. The wheel pin clicks into the empty space in the middle, so it only extends halfway through the wheel.

Here's what I ended up going with:

Code:

0 !LDCAD SNAP_CYL [gender=F] [caps=none] [secs=R 4 11   R 5 3   R 4 11] [center=true] [slide=true] [ori=1 0 0 0 0 1 0 -1 0]

[Magnus Forsberg]

So yeah, it's meant to just go onto wpin.dat. The wheel pin clicks into the empty space in the middle, so it only extends halfway through the wheel.

Here's what I ended up going with:

Code:

0 !LDCAD SNAP_CYL [gender=F] [caps=none] [secs=R 4 11   R 5 3   R 4 11] [center=true] [slide=true] [ori=1 0 0 0 0 1 0 -1 0]

Yes, but on both sides.

This rim should have used 2 wpinhol2
1 16 0 0 -3.25 -0.707107 0 -0.707107 -0.707107 0 0.707107 0 9.25 0 wpinhol2.dat
1 16 0 0 3.25 -0.707107 0 -0.707107 -0.707107 0 0.707107 0 -9.25 0 wpinhol2.dat

[N. W. Perry]

Yes, but on both sides.

Well, the single cyl snap extends all the way through. A wheel pin will snap on either side, but a 318 bar will still slide all the way through. But perhaps this is not the preferred method?

This rim should have used 2 wpinhol2
1 16 0 0 -3.25 -0.707107 0 -0.707107 -0.707107 0 0.707107 0 9.25 0 wpinhol2.dat
1 16 0 0 3.25 -0.707107 0 -0.707107 -0.707107 0 0.707107 0 -9.25 0 wpinhol2.dat

The geometry of the rim is funky overall, and doesn't match the wpin's dimensions well. Maybe the snap info should match the rim's geometry rather than the pin's, allowing some realistic "play", but in my models I find that more a hindrance than a help. I like things like wheels to just snap to one place, and one place only.

[Stefan Frenz]

I have no idea if this is my fault or maybe I missed something, but perhaps it has to do with part snapping: using 35563 (Cone 4x2x4 Half with Roof Tiles, I think it was in the last LDraw part update) as upper part and trying to snap this with the bottom to the upper side of a below 48310 (Cone 8x4x6 Half), the part is 1/2 stud (10 LDU) beside the "real" studs..

[N. W. Perry]

I have no idea if this is my fault or maybe I missed something, but perhaps it has to do with part snapping: using 35563 (Cone 4x2x4 Half with Roof Tiles, I think it was in the last LDraw part update) as upper part and trying to snap this with the bottom to the upper side of a below 48310 (Cone 8x4x6 Half), the part is 1/2 stud (10 LDU) beside the "real" studs..

Yes, that's because the upper part has no snap info of its own, but it is inheriting the snap for the bottom stud tube. So it's connecting that to the hollow studs of the lower cone.

Adding the following pair of lines to 35563 should work:

Code:

0 !LDCAD SNAP_CYL [gender=F] [caps=one] [secs=R 6 4] [pos=0 96 -30] [grid=C 4 1 20 0]
0 !LDCAD SNAP_CYL [gender=F] [caps=one] [secs=R 6 4] [pos=0 96 -10] [grid=C 2 1 60 0]

[Magnus Forsberg]

I think there is something wrong in the shadow file s\31111s01.dat


  31111.png (Size: 21.98 KB / Downloads: 698)

My log-file say this:
Circular reference detected to "s\31111s01.dat" from "31111.dat", reference will be cleared and rendered as "missing".

[Roland Melkert]

I think there is something wrong in the shadow file s\31111s01.dat

My log-file say this:
Circular reference detected to "s\31111s01.dat" from "31111.dat", reference will be cleared and rendered as "missing".

Indeed,

"s\31111s01.dat" references to 31111 for its snap data, while 31111 references "s\31111s01.dat" mesh wise.

Solution is to move the snap data from "31111.dat" to "s\31111s01.dat", and delete the shadow for "31111.dat" entirely.

Is "31111.dat" using a subpart a recent change?