A 3/4-stud or not?

[Magnus Forsberg]

Hi,

On many window and door frames there is a rather odd stud.

   
We usually make them without using a stud primitive, or need to make a very uniqe, specially designed stud.

I had to make eight files for 4209, in normal and low resolution. They all have to be different since there is a logo on top of each stud.

   
Due to the fact that we still need to fool MLCad, and have a lo-res version, (including a Moved to-primitive), to every new stud, the number of uniqe files have to be twelve for every new part with a small hole in the stud. A uniqe set of of primitives that hardly will be reused.

I would like to introduce a 3/4-stud instead.

   
A single set of primitives that could be reused in many of the parts above.
They will help connection seeking software and will place the logo in a correct position.
The small hole is to be added as a single, reusable, non-rotationally sensitive subfile (or primitive).

I've tried this set-up in LDView and MLCad. It works fine.
How will it work in LDCad, SR3D Builder or Bricksmith? Or in any other software with stud substitution ability?

How should I do with studs that have a hole that isn't fully placed in one of the quarters, but in between?

Do all the frames have a logo on top of the stud with a small hole?
I don't have all the frames, but I know that 4209 do have small holes and a logo, and that the two parts with square holes (60594 and 60598) doesn't.

If we ignore the logo, it is possible to use a single file and rotate the gap to were it needs to be.

So, three questions.
Should we continue in making new studs for each part?
Or should we replace them with this 3/4-stud?
Or could we ignore the logo, and simplify it to a single stud?

[Santeri Piippo]

We should keep in mind that we do not currently have an official logo stud implementation. Anything currently in use is unofficial at best. So we don't need to worry about stud logos when deciding on these studs. I agree with the 3-4stud proposal, it could save us a lot of work, assuming we make the remaining 1/4 a non-stud primitive.. or something like that?

[Chris Dee]

How many variations of hole position are there? Just two? I do think we should look to the future to be able to support automated LEGO logo insertion, so stud orientation is important.

I'm not sure what we are trying to minimise here - the number of primitives or the amount of developer time.

One set of primitives (stud19a-d) has already been built and developer time for a second set could easily be surpassed by discussion and testing time (and might already have been).

Did you solve the problems with MLCad modes "line" and "none". Last I read on the Parts Tracker you said

MLCad has four levels: normal, coarse, lines, none.
It looks good at normal and coarse, but strange at line and none.

[Steffen]

This is a very difficult issue.
I do not want to lose the capability to put logos on studs correctly oriented.
I would like to support connection seeking software.
I would like to see the holes in the studs.
And I have no real answer solving this ((

Personally, if I would today encounter this problem, I would fall back to using a shared subpart file PARTS\S
until somebody comes up with a long-term solution...

[Santeri Piippo]

Since all studs should always be oriented the same direction in one part, perhaps we could use a meta-command to apply stud orientation per-part? i.e. provide a rotation matrix for stud-logo.dat. This solution would only work with non-SNOT parts but should be good enough I think?

[Magnus Forsberg]

Did you solve the problems with MLCad modes "line" and "none".

No I did not, but if we are going by how MLCad works, the stud19-familly of studs doesn't work "correctly" either.
It simply removes the lo-res primitives totally, and shows big holes at the corners were the studs should be.
To get a good result in "line" and "none" we can't have the bottom surface included in the stud primitive.

Didn't expect that development. How do I continue?
I have all needed files prepared for both solutions. One new set of stud for 6556, and one set of 3/4-studs.

I think I'll scrap both, and start making studs without integrated bottom surface.
4209s02.dat needs a set of reworked studs too.

[Mark Kennedy]

Makes sense to me to have a 3/4-stud, since the argument about the logos not lining up could also be applied to the stud6.dat, stud13.dat, stud15.dat, stud17, and studel.dat stud primitives