3D Viewer work

[N. W. Perry]

in Chrome on Andriod phone
The random colour doesn't work.
The logo button doesn't work. Toggle extreme close-up.
Displaying all parttypes.

Confirming the same issues on Chrome for Mac desktop. A hard reload (shift+reload button) solved all except that of subfiles.

[Orion Pobursky]

Sorry, but I still don't get to view subfiles.

Like this one:
https://www.ldraw.org/cgi-bin/ptdetail.c...664s02.dat

Oh, ok.

That was a bug in the partID detection. It has now been fixed.

[Orion Pobursky]

Could you please fix the 3D viewer on the part tracker?  For a long time now the random colour button don't do anything.

All remaining issues should now be fixed. And the stud button toggles correctly.

[Vincent Messenet]

All remaining issues should now be fixed. And the stud button toggles correctly.

Is it me or the 3D View is broken again? It does not work on any official file I tried. No issue with Unofficials

[Orion Pobursky]

Is it me or the 3D View is broken again? It does not work on any official file I tried. No issue with Unofficials

Mordern browser pretty aggressively cache items. You prolly have to force reload the page.

[Magnus Forsberg]

... now work correctly.

Is it just me, or didn't the chosen view stay in focus before?
As a user I want to have the zoomed in view stay in focus when I switch between random colour and normal view.

[Orion Pobursky]

Is it just me, or didn't the chosen view stay in focus before?
As a user I want to have the zoomed in view stay in focus when I switch between random colour and normal view.

Im aware of the issue but solving it has stymied me for the last week or so.

[Orion Pobursky]

Im aware of the issue but solving it has stymied me for the last week or so.

I had some free time today and this issue has been fixed. As usual, you'll probably have to do a force reload (ctrl-shift-R in Firefox) to clear the cached code.

[Philippe Hurbain]

One little glitch: 3D viewer works fine on the texmapped wing https://www.ldraw.org/cgi-bin/ptdetail.c...088p03.dat
But it shows neither texmap nor fallback pattern on mirrored part https://www.ldraw.org/cgi-bin/ptdetail.c...089p03.dat

[Lasse Deleuran]

One little glitch: 3D viewer works fine on the texmapped wing https://www.ldraw.org/cgi-bin/ptdetail.c...088p03.dat
But it shows neither texmap nor fallback pattern on mirrored part https://www.ldraw.org/cgi-bin/ptdetail.c...089p03.dat

Thanks for the bug report. I have replicated the issue locally and reduced the failing file to the following. I didn't find a solution yesterday, but hopefully it will be fixed today.

0 FILE x.ldr
0 Change a.dat and b.dat in the next line to see that it works and fails
1 16 -140 -128 50 0 0 1 0 1 0 -1 0 0 b.dat
0
0 FILE b.dat
0 MIRRORED
0 Name: b.dat
0 !LDRAW_ORG Unofficial_Part
0 !LICENSE Redistributable under CCAL version 2.0 : see CAreadme.txt
0 BFC CERTIFY CCW
1 16 0 0 0 -1 0 0 0 1 0 0 0 1 a.dat
0
0 FILE a.dat
0 Name: a.dat
0 Author: Sticker Builder
0 !LICENSE Redistributable under CCAL version 2.0 : see CAreadme.txt
0 BFC CERTIFY
0 PE_TEX_PATH -1
0 PE_TEX_INFO iVBORw0KGgoAAAANSUhEUgAAAAIAAAACCAIAAAD91JpzAAAACXBIWXMAAAsTAAALEwEAmpwYAAAAB3RJTUUH5AEYChQ1nEYzVAAAAB1pVFh0Q29tbWVudAAAAAAAQ3JlYXRlZCB3aXRoIEdJTVBkLmUHAAAAFklEQVQI12N4KqUh5fWUQcrr6VMpDQAiEgThLKLB4AAAAABJRU5ErkJggg==
1 16 0 -0.5 0 9.75 0 0 0 0.5 0 0 0 9.75 box5.dat
3 16 9.75 -0.5 -9.75 9.75 -0.5 9.75 -9.75 -0.5 9.75 0 0 1 0 1 1
3 16 9.75 -0.5 -9.75 -9.75 -0.5 9.75 -9.75 -0.5 -9.75 0 0 1 1 0 1

[Lasse Deleuran]

One little glitch: 3D viewer works fine on the texmapped wing https://www.ldraw.org/cgi-bin/ptdetail.c...088p03.dat
But it shows neither texmap nor fallback pattern on mirrored part https://www.ldraw.org/cgi-bin/ptdetail.c...089p03.dat

It works now. A fix has been pushed to the repo. I fixed it by creating new "texmap placements" for parts whose sub-parts have texmaps. These new texmap placements apply the placement (position and matrix) from the line of type 1 that uses the sub-parts to the points of the texmap declaration.

However.

What about "a" and "b" for cylindrical and spherical projections? How should these be modified in case of scaling matrices being used on the sub-parts? Right now it works because the failing file used planar projection with a non-scaling matrix. The behaviour for the other cases is going to be a bit messy.

[Philippe Hurbain]

What about "a" and "b" for cylindrical and spherical projections? How should these be modified in case of scaling matrices being used on the sub-parts? Right now it works because the failing file used planar projection with a non-scaling matrix. The behaviour for the other cases is going to be a bit messy.

Thanks! AFAIK no part use spherical or cylindrical projections. And mirrored versions of those inexisting parts are even more scarce
More seriously, I don't think a and b parameters should be impacted by mirroring, since they are used in a symmetrical way (+/- a/2, +/-b/2)

[Orion Pobursky]

It works now. A fix has been pushed to the repo. I fixed it by creating new "texmap placements" for parts whose sub-parts have texmaps. These new texmap placements apply the placement (position and matrix) from the line of type 1 that uses the sub-parts to the points of the texmap declaration.

However.

What about "a" and "b" for cylindrical and spherical projections? How should these be modified in case of scaling matrices being used on the sub-parts? Right now it works because the failing file used planar projection with a non-scaling matrix. The behaviour for the other cases is going to be a bit messy.

I'll try to have this update pushed to the live code soon. Last time I pulled from master, my code didn't play well and I still haven't gotten back to troubleshooting.

[Magnus Forsberg]

Thanks! AFAIK no part use spherical or cylindrical projections. ....

https://www.ldraw.org/cgi-bin/ptdetail.c...287p01.dat
https://www.ldraw.org/cgi-bin/ptdetail.c...287p02.dat

I made two parts using a spherical projection. None of them have a mirrored counterpart, though.

[Lasse Deleuran]

I'll try to have this update pushed to the live code soon. Last time I pulled from master, my code didn't play well and I still haven't gotten back to troubleshooting.

IIRC That was when I pushed the update to BFCGeometry in order to reduce duplicate work whenever LDRGeometry will be changed. I will now keep an "how to upgrade" changelog file ready in case of API changes like that.

[Orion Pobursky]

IIRC That was when I pushed the update to BFCGeometry in order to reduce duplicate work whenever LDRGeometry will be changed. I will now keep an "how to upgrade" changelog file ready in case of API changes like that.

New pull from master works (i.e. renders) with a few JavaScript errors that I'll troubleshoot before I push to the live page. I also want to simplify things on my end.