Slope 45 2x2 with Spyrius Droid Image (3039px32)

[Philip Peickert]

Ok, so this is my first attempt at parts authoring. I used LDraw Pattern Creator and only had access to a really crappy scan of the part (I am in the process of getting the part in question and will do a better more detailed scan then). I know I have some scale/sizing issues and I haven't figured out how to precisely fit the pattern onto the slope brick. Would you guys mind pointing out a the things I've done wrong or how I can redo the part to make it work better. I don't think it's anywhere close to being ready to put up for verification, but then again I'm not really sure. I've tried to follow the online guides as best I could, but I still managed to confuse myself. Any advice is appreciated. Thanks.


Phil Peickert

[Philippe Hurbain]

Hi Phil,

I find the pattern itself pretty good compared to the images of BL/Peeron. Now the technical issues...
There are 3 surfaces overlaps. This can be seen using LDview in random color mode, or better using Edger2. See Edger2 documentation, bottom of page. This actually detects unmatched edges that generally occur when you have overlaps, and also T-junctions. There are several T-junctions in your design, some of them could be easily avoided (but the one below the long row of dots is more difficult!)
As for fitting the slope brick (with its 45° slope), your pattern must be exactly 40 LDU wide (the width of the tile) and 20*sqrt(2)/2=14.142 LDU high. Then you can place it on the slope subpart top and rotate it 45° (make sure to put MLCad rotation point in the right spot, the hinge point!).

[Philip Peickert]

I think I got most of the T-Junctions and I edited the pattern dimensions so it fits the part. When I look at it in LDview in random color mode it looks like it works. However when I run Edger2 it says I have 21 unmatched edges. I can't figure out where they are, and for some reason Edger2 gives me an error and won't open up a new dat file with the edges marked in white. I have looked through the documentation on both Edger and Edger2 to see where I might have screwed up the installation, and I have tried reinstalling 3 times to no avail. Let me know what you think of the revisions and if you can run Edger and send me the dat file with the marked edges so I can see where they are that would be awesome. Thanks.

Phil

[Philippe Hurbain]

I think I got most of the T-Junctions

Indeed, only 2 remaining (and easy to correct). These are marked as vertical and horizontal red lines in edger2 result, image below and attached file.
Edger2 also show that you have not corrected the 3 overlaps, they are marked as v-shaped lines.

and I edited the pattern dimensions so it fits the part.

Well... almost. The images below show that it doesn't perfectly match on top and bottom. This mismatch also explains why Edger2 displays a red edge around the pattern.

However when I run Edger2 it says I have 21 unmatched edges.

Im afraid it's right... don't forget that a mismatch often result in two almost overlapping lines.

I can't figure out where they are, and for some reason Edger2 gives me an error and won't open up a new dat file with the edges marked in white.

Sorry, don't understand what you mean. Could you attach offending file?

I have looked through the documentation on both Edger and Edger2 to see where I might have screwed up the installation, and I have tried reinstalling 3 times to no avail.

Edger2 beeing a pure command line program, there is no installation needed (or you mean installation of LETGUI?). Edger might be a little more picky since it requires a working Java interpreter working on your machine.

Let me know what you think of the revisions and if you can run Edger and send me the dat file with the marked edges so I can see where they are that would be awesome. Thanks.

Here they are... Good luck

[Philip Peickert]

Thanks for all of your help Philo. I addressed the remaining two T-Junctions and resolved the three overlapping quads. I also aligned the vertexes so that all the parts line up. I couldn't find any more problems in LDView and I'm still having trouble with Edger. If no one sees anything wrong with it, it might be ready to put up on the parts tracker.

Phil Peickert

[Philippe Hurbain]

There is one remaining T-junction (again, though their elimination is best, they doesn't prevent certification of part!) at the right of down arm of green cross (vertical red line in the above scree shots). Otherwise it seems perfect to me.
I'd really like to know what's wrong with Edger2 on you computer!!!
- What happens exactly? any error message?
- how do you use it, from command line or through LETGUI?

Otherwise before submitting to parts tracker, I'd suggest you download and install Datheader. This tool will help you polish the header and make a few automatic corrections such as removing the "0 rotation" statements left by MLCad, reduce precision to 3 decimal places, etc...

[Reuben Pearse]

Hi Philip,

It's great to see you using the LDraw Pattern Creator (LPC) program. Based on your experience so far it would be great if you were able to put some short instructions together on the process you used to construct the part using LPC. This could then be used to improve the current online documentation for the program. Feel free to send over instructions/screenshots and I'll try and formalise this into some form of user documentation.

Reuben
[email protected]

[Philip Peickert]

Let's see...

1. I fixed the last T-junction. When I run Edger 2 it shows no T-Junctions or overlaps, and the only unmatched edges are where the parts line up with each other. Which leads to...

2. I fixed the problem with Edger 2. I hadn't installed LETGUI and like a noob I didn't realize it until your last post. I installed LETGUI and it works fine.

3. I installed and ran Datheader so the header and file are now cleaned up.

Philo, I just wanted to say again thank you for all of your help. Your advice was able to point me in the right direction every time and while it will take me a while to master the authoring software, I feel comfortable using it thanks to you.

[Philippe Hurbain]

The part looks plain perfect now! Submit it to PT
But 3039px32 is a Peeron ref., this part should be named differently, see http://www.ldraw.org/library/tracker/ref/patterns/ and http://www.ldraw.org/cgi-bin/ptpatterns.cgi?p=3039
Maybe 3039pc7?

[Philip Peickert]

Changed the part number and submitted to the Parts Tracker. Link is below. Thanks again for all of your help.

http://www.ldraw.org/cgi-bin/ptdetail.cgi?s=3039pc7.dat

[Philip Peickert]

Hello Reuben.

It took me a while to start getting comfortable with LPC (actually I find I get better at it every time I use it). At first I was translating the French documentation through babblefish.com, but not only did that waste a lot of time, I found I learned more simply through trial and error. I'll put together a little narrative on how I use the program and any tips I've picked up along the way and send them to you.

Phil

[Reuben Pearse]

Hi Phil,

I started working on a translation of the French manual, but didn't get very far :-(

My working copy of the translation is here:
http://www.pearse.co.uk/lego/links/ldraw...tor/en-GB/

This was based on tidying up the rough translation provided by Google. A copy of the Google translation can in found in this Open Office document:
http://www.pearse.co.uk/lego/links/ldraw...nslate.odt

I look forward to your narrative.

BTW, I've tried to contact Nils Schmidt to get permission to publish the source code for LPC, as this would allow others to contribute to the programs development. I'll let you know when I hear back from Nils.

Bye for now

Reuben

=====================================
Download 3D Lego models and other resources from:
http://www.pearse.co.uk/lego
=====================================

[Philippe Hurbain]

Reuben,
I am somewhat missing time to do a full translation, but if you are stuck by a French expression, I certainly can help!!!

[Philip Peickert]

Reuben,

I have a little over half of the French to English translation completed and when I send you my notes and narrative, I'll send it along to. In my recent dealings with SlicerPro I have learned a few things about LPC that I didn't know before regarding the import features and I'll be sure to include my findings. I am working a full 48 hour shift this weekend so in between running calls I am going to try to finalize my notes and the translation and get it to you.

Phil

[Reuben Pearse]

Phil,

Thanks for the update. I hope your shift goes well.

Reuben

[Philip Peickert]

Reuben I have the translation and my ideas ready. My email is [email protected]. Send me an email with something about LPC in the subject and I'll email you what I've got so far.

[Reuben Pearse]

Hi Phil

I emailed you the other day, but haven't heard anything back.

When you're ready you can email me directly at:
[email protected]

Bye for now

Reuben

[Reuben Pearse]

You may be interested in the following post regarding J.C. Tchang's latest update to French manual for the LDraw Pattern Creator

http://forums.ldraw.org/showthread.php?tid=3572

Bye for now

Reuben

=====================================
Download 3D Lego models and other resources from:
http://www.pearse.co.uk/lego
=====================================

[Daniel Goerner]

Then you can place it on the slope subpart top and rotate it 45° (make sure to put MLCad rotation point in the right spot, the hinge point!).

Could you please explain that? I don't know how to place the pattern right. I took the 3039 part, threw out the subpart and rotated the surface 45° and started painting. After exporting I ended up with this:



How do I get it in the exact right position?

[Philippe Hurbain]

The first thing to do is to place top of pattern precisely on top of the brick. Look at Y coordinate of top element of the pattern, let's suppose it is -8.284 (chances it is the exact value, but not sure since I don't have the file...). Select complete pattern, open the move dialog in MLCad, and set Y position value at +8.284. Top of the pattern will be exactly at top of brick. Set rotation center on world origin, and now rotate pattern by 45° to match slope front. Now you just have to move the pattern by 10LDU in Z direction to get it correctly placed.