Proposal of new primitives


Proposal of new primitives
#1
I propose the introduction of these six primitives (3 version 16 sides and 3 version 48 sides) that could solve some situations.
The example I have attached is the one with which I came up with the idea, in fact the most correct use would be for patterns with incomplete rings of different colors.
The primitives I propose come from the classic ndis to which, however, I moved the vertex of the circumscribed square to bring it to the vertical of the last vertex of the circle.

They should be quite interchangeable with the current tang primitives, but reducing the number of triangles required (partly because they are already included in the primitives and partly because their shape is more regular).


Attached Files
.dat   npeghol7-official.dat (Size: 792 bytes / Downloads: 7)
.dat   npeghol7-tang.dat (Size: 1.08 KB / Downloads: 3)
.dat   npeghol7-tdis.dat (Size: 810 bytes / Downloads: 3)
.dat   48 1-8tdis.dat (Size: 475 bytes / Downloads: 1)
.dat   1-16tdis.dat (Size: 252 bytes / Downloads: 3)
.dat   3-16tdis.dat (Size: 340 bytes / Downloads: 4)
.dat   1-8tdis.dat (Size: 294 bytes / Downloads: 3)
.dat   48 1-16tdis.dat (Size: 353 bytes / Downloads: 3)
.dat   48 3-16tdis.dat (Size: 601 bytes / Downloads: 3)
Reply
RE: Proposal of new primitives
#2
(2021-07-12, 22:19)Massimo Maso Wrote: I propose the introduction of these six primitives (3 version 16 sides and 3 version 48 sides) that could solve some situations.
The example I have attached is the one with which I came up with the idea, in fact the most correct use would be for patterns with incomplete rings of different colors.
The primitives I propose come from the classic ndis to which, however, I moved the vertex of the circumscribed square to bring it to the vertical of the last vertex of the circle.

They should be quite interchangeable with the current tang primitives, but reducing the number of triangles required (partly because they are already included in the primitives and partly because their shape is more regular).

It would help if you add also a thumbnail to have a general idea before downloading the prims.

w.
LEGO ergo sum
Reply
RE: Proposal of new primitives
#3
(2021-07-12, 22:19)Massimo Maso Wrote: I propose the introduction of these six primitives (3 version 16 sides and 3 version 48 sides) that could solve some situations.
The example I have attached is the one with which I came up with the idea, in fact the most correct use would be for patterns with incomplete rings of different colors.
The primitives I propose come from the classic ndis to which, however, I moved the vertex of the circumscribed square to bring it to the vertical of the last vertex of the circle.

They should be quite interchangeable with the current tang primitives, but reducing the number of triangles required (partly because they are already included in the primitives and partly because their shape is more regular).
Not against it - actually I proposed that when I was a baby parts author a long time ago (Lugnet era...) but tang primitives was said to do the job then. Note that there are many more 48-sided such primitives that make sense (eg. 1-24 or 7-48)!
Reply
RE: Proposal of new primitives
#4
Photo 
Attached is the image with the comparison of the primitives.


Attached Files Thumbnail(s)
   
Reply
RE: Proposal of new primitives
#5
(2021-07-13, 22:00)Massimo Maso Wrote: Attached is the image with the comparison of the primitives.

Would it be agreed that these would all be less than 1/4 circles? In other words, these three, plus the extra 48 ones would be the only possibilities? (I ask because it's unclear what something like 11/16 would look like.)

On the 48 front, the other possible ones would then be:
  • 1-48
  • 1-24 (2/48)
  • 1-12 (4/48)
  • 5-48
  • 7-48
  • 1-6 (8/48)
  • 5-24 (10/48)
  • 11-48
Reply
RE: Proposal of new primitives
#8
(2021-07-14, 0:02)Travis Cobbs Wrote: Would it be agreed that these would all be less than 1/4 circles? In other words, these three, plus the extra 48 ones would be the only possibilities? (I ask because it's unclear what something like 11/16 would look like.)

On the 48 front, the other possible ones would then be:
  • 1-48
  • 1-24 (2/48)
  • 1-12 (4/48)
  • 5-48
  • 7-48
  • 1-6 (8/48)
  • 5-24 (10/48)
  • 11-48
Travis, any plan to support these prims in LDView?
Reply
RE: Proposal of new primitives
#9
(2022-01-16, 16:47)Philippe Hurbain Wrote: Travis, any plan to support these prims in LDView?

I would be willing to support them if they were created and used in parts (unofficial is fine). This thread doesn't appear to have anybody answering the original question, which is whether or not they should be added. I won't preemptively spend the time to support them on the off chance that they might be created at some point in the future (with possibly a completely different name).

Also, I don't understand the proposed name. I'm fine with having 2 appended if we also like the inverse, but I don't know what the "t" stands for. These are effectively ndis primitives that have been cropped to the highest z coordinate along the curve.
Reply
RE: Proposal of new primitives
#10
Looks like we have 6 tdis currently on the PT they are definitely in use.
Reply
RE: Proposal of new primitives
#11
(2022-01-17, 6:43)Orion Pobursky Wrote: Looks like we have 6 tdis currently on the PT they are definitely in use.

I guess t is for truncated (truncated ndis). But this could apply to any other truncated thing. Maybe tndis would be better?
Reply
RE: Proposal of new primitives
#12
(2022-01-17, 8:37)Philippe Hurbain Wrote: I guess t is for truncated (truncated ndis). But this could apply to any other truncated thing. Maybe tndis would be better?

For now I am the proponent and the largest user of these primitives but I have seen that others are also beginning to use them.
If no one has clearly opposed their introduction, I believe that they can now be considered accepted.

For me the name does not need to be changed.
The t was born as assumed by Philo from the word "truncated".
I don't think there can be any other way to truncate negative discs, except for the ones that I have listed below as tdis2, but which are easily obtainable from ndis and do not give real advantages.
For positive discs we already have disc and chrd, and I don't think we will ever need to introduce truncated disks: you would save a line of code to add a triangle to a chrd, a bit like for tdis2.
Reply
RE: Proposal of new primitives
#13
(2022-01-17, 6:43)Orion Pobursky Wrote: Looks like we have 6 tdis currently on the PT they are definitely in use.

Someone should add tdis to the primitives ref.
Reply
RE: Proposal of new primitives
#14
(2022-01-31, 0:22)Travis Cobbs Wrote: Someone should add tdis to the primitives ref.

It's in the queue. Updating the Prim Ref is a manual process and it hasn't been done in a bit so it may not happen until next release tentatively scheduled for the beginning of March.
Reply
RE: Proposal of new primitives
#15
(2022-01-17, 6:43)Orion Pobursky Wrote: Looks like we have 6 tdis currently on the PT they are definitely in use.

I just released LDView 4.4.1 Beta 1 with support for tdis. I would greatly appreciate it if people would test my tdis support, and either report that it is working, or report any problems found. Note that I support them in the viewer and in POV-Ray exports.

(I'm not posting a release announcement in the Announcements forum because tdis is the only new feature.)
Reply
RE: Proposal of new primitives
#16
(2022-01-31, 0:59)Travis Cobbs Wrote: I just released LDView 4.4.1 Beta 1 with support for tdis. I would greatly appreciate it if people would test my tdis support, and either report that it is working, or report any problems found. Note that I support them in the viewer and in POV-Ray exports.

(I'm not posting a release announcement in the Announcements forum because tdis is the only new feature.)
I tested on all tdis primitives currently on PT, they all work fine. I also generated two oddball 48\tdis (48\5-48tdis.dat and 48\11-48tdis.dat), they substituted fine too.
Reply
RE: Proposal of new primitives
#17
(2022-01-31, 7:40)Philippe Hurbain Wrote: I tested on all tdis primitives currently on PT, they all work fine. I also generated two oddball 48\tdis (48\5-48tdis.dat and 48\11-48tdis.dat), they substituted fine too.

Thanks, Philo. Did you test POV export in addition to the viewer, or just the viewer? (The code for the POV geometry is totally different from the code for the viewer geometry.)
Reply
RE: Proposal of new primitives
#18
(2022-01-31, 22:14)Travis Cobbs Wrote: Thanks, Philo. Did you test POV export in addition to the viewer, or just the viewer? (The code for the POV geometry is totally different from the code for the viewer geometry.)
Just the viewer... But now I tested POV export too and it works  Smile. Below the comparison between 4.4 and 4.4.1 (difference is subtle for 48-prims so I rendered both to be sure!)


Attached Files Thumbnail(s)
       
Reply
RE: Proposal of new primitives
#19
(2022-01-31, 0:59)Travis Cobbs Wrote: I just released LDView 4.4.1 Beta 1 with support for tdis. I would greatly appreciate it if people would test my tdis support, and either report that it is working, or report any problems found. Note that I support them in the viewer and in POV-Ray exports.

(I'm not posting a release announcement in the Announcements forum because tdis is the only new feature.)

Per the discussion here on the forums, the tdis primitives have been renamed to tndis.
Reply
RE: Proposal of new primitives
#20
(2022-02-05, 5:51)Orion Pobursky Wrote: Per the discussion here on the forums, the tdis primitives have been renamed to ntdis.

*tndis !  Big Grin (rename on PT is correct)
Reply
RE: Proposal of new primitives
#21
(2022-02-05, 7:44)Philippe Hurbain Wrote: *tndis !  Big Grin (rename on PT is correct)

Yes, there's a typo in the post from Orion. Everything looks OK to me, at the PT.
Reply
RE: Proposal of new primitives
#22
(2022-02-05, 7:44)Philippe Hurbain Wrote: *tndis !  Big Grin (rename on PT is correct)

Typo fixed.
Reply
RE: Proposal of new primitives
#23
(2022-02-05, 5:51)Orion Pobursky Wrote: Per the discussion here on the forums, the tdis primitives have been renamed to tndis.

LDView 4.4.1 Beta 2 changes from tdis to tndis. Note that this removes recognition of tdis primitives. I intentionally am not supporting that as a fallback.

As before, I would greatly appreciate it if people could validate this Beta release.
Reply
RE: Proposal of new primitives
#24
(2022-02-06, 2:48)Travis Cobbs Wrote: LDView 4.4.1 Beta 2 changes from tdis to tndis. Note that this removes recognition of tdis primitives. I intentionally am not supporting that as a fallback.

As before, I would greatly appreciate it if people could validate this Beta release.

Works fine, both for direct view and POV!
Reply
RE: Proposal of new primitives
#25
(2022-02-06, 2:48)Travis Cobbs Wrote: As before, I would greatly appreciate it if people could validate this Beta release.

As far as I understand to use POV, (and I never use the 64-bit version, since I always open LDView from LDDP) everything works fine.
Installed this beta 2 version anyway and tested it, works fine.
Reply
RE: Proposal of new primitives
#6
Since 1-4tdis would be equal to 1-4ndis, it makes no sense that these primitives exceed 1-4 of a circle. In case you need it, use a normal ndis and tdis for the remaining part.

As for the 48-sided versions, they are useful because they save some lines of code and neighboring but not coincident points for rounding.



In the meantime it occurred to me that the complementary part could already be foreseen (with a lot of imagination I indicated it as tdis2); even if these are already obtainable now with the ndis and adding a triangle would only reduce the number of lines and the possibilities of error.


Attached Files Thumbnail(s)
   

.dat   1-16tdis2.dat (Size: 274 bytes / Downloads: 0)
.dat   1-8tdis2.dat (Size: 311 bytes / Downloads: 0)
.dat   3-16tdis2.dat (Size: 353 bytes / Downloads: 0)
Reply
RE: Proposal of new primitives
#7
I am for these primitives.

I've found this part Duplo arch u977 has an underside center divider (missing in the part model), and these primitives will fit nicely.
https://www.ldraw.org/cgi-bin/ptdetail.c...s/u977.dat
Reply
« Next Oldest | Next Newest »



Forum Jump:


Users browsing this thread: 1 Guest(s)