Subject:
|
Re: Parts with lengths
|
Newsgroups:
|
lugnet.cad.dev
|
Date:
|
Thu, 9 Sep 2004 19:47:12 GMT
|
Viewed:
|
1684 times
|
| |
|
|
In lugnet.cad.dev, Jindrich Kubec wrote:
> Steve Bliss wrote:
> > I don't think making a separate part file for every known length of a part is
> > the right approach, for at least two different reasons.
>
> i haven't suggested that.
>
> We don't have separate part file for every known color, but instead,
> there is part and color info in the dat. I think the length should be
> put in somewhat similar way in the dat, so that it could be easily
> extracted for inventory purposes.
Jindroush,
True length is a topic that is near the top of the list for LSynth. Orion and
I talked about it at BrickFest. There is at least one implementation out there
by Frederick Glockner (using emacs) that supports true length. Since LSynth is
open source, all we have to do is find the time (the hard part) to incorporate
his algorithm into LSYnth, and then release LSynth under the GNU license (as is
the emacs solution).
The LDraw file format for record type 1 (first non-white character is 1 which
means part usage) provides a field specifically for a color. There is no
provision for true length in type 1 records, so trying to draw parallels between
color and true length doesn't really work.
Lsynth could make guesses about how to split up the total true length between
the pairs of constraints by calculating the point to point distance between each
pair of constraints (with some addditional length if the relative angles between
the constraints is large). From here I think we're going to use Frederick
Glockner's (sorry if I misspelled this) algorithm for generating true length
hoses between any two constraints. We'll want to make provisions for you to
manually specify the true length between each of the constraint pairs, in case
you don't like LSynth's guesses at true length distribution.
Someone also suggested that we apply an interative method for letting the hose
relax into place after we've made a good initial guess. True length
distribution across multiple pairs of constraints is a non-trivial problem (hey
maybe there's a masters thesis in there somewhere :^)
Whatever the algorithm, true length(s) will have to be specified in META
commands.
I'm sure Michael can easily add support for this in his textual parts
inventory in MLCad. In LPub, I already have an image that represents a piece of
hose in the part list image. I could easily add the true length. Creating LEGO
instruction style images showing *actual* length would require knowing the DPI
of the output device where the instructions are being displayed/printed. I
don't think I can solve that problem right now :^)
>
> Jindroush
Kevin
|
|
Message has 2 Replies:
 | | Re: Parts with lengths
|
| (...) Just in case, the source code for lsynth is here. (URL) the CVS Repository link for instructions on how to fetch it all, or follow the Browse CVS link if you only want to examine it. Have fun, Don (20 years ago, 9-Sep-04, to lugnet.cad.dev)
| | | Re: Parts with lengths
|
| In lugnet.cad.dev, Kevin L. Clague wrote: [snip] nod, I agree with everything you wrote. (...) Long term (*very* long term), there's nothing to say the language can't be extended to allow a different linetype for parts-with-lengths. Steve (20 years ago, 10-Sep-04, to lugnet.cad.dev)
|
Message is in Reply To:
| | Re: Parts with lengths
|
| (...) i haven't suggested that. We don't have separate part file for every known color, but instead, there is part and color info in the dat. I think the length should be put in somewhat similar way in the dat, so that it could be easily extracted (...) (20 years ago, 8-Sep-04, to lugnet.cad.dev)
|
11 Messages in This Thread:
 
- Entire Thread on One Page:
- Nested:
All | Brief | Compact | Dots
Linear:
All | Brief | Compact
This Message and its Replies on One Page:
- Nested:
All | Brief | Compact | Dots
Linear:
All | Brief | Compact
|
|
|
|
