| |
In lugnet.cad, Willy Tschager wrote:
> dear LSC members,
>
> LEGO Digital Designer http://www.lego.com/eng/create/digitaldesigner/ has only
> one feature I really miss in the LDraw system: the snap-in behavior. guys,
> I'd like to ask you - no I beg you down on my knees - to define a working
> standard for a connection database. there is currently a proposal at ldraw.org:
>
> http://www.ldraw.org/modules.php?op=modload&name=News&file=article&sid=135&mode=thread&order=0&thold=0
>
> but it is all theorical. to launch this thing we would just nead 2,3 working
> examples (bricks, plates) of these .cdl files
>
> http://www.ldraw.org/OLD/reference/specs/lcd/#Appendix%20B
>
> with the proper coordinations and a prog (LDDP plug-in or stand-alone) where we
> could test new definitions. a paralell PT could collect those .cdl files.
> I can also think of a restriction for submitting new parts. authors would have
> to submit a .cdl file first, before they are alowed to submit a .dat file. this
> would boost the build-up of the library. could you make this happen?
I like the idea of connections. I think it's really neat and would like to see
it realised. I support asking LSC to look into this further.
But I fear that requiring authors to submit a .cdl file might be increasing the
barriers to entry for new parts authors, which are already higher than we might
like, for valid reasons, so I'd be very careful about having that requirement in
place. At least initially. Perhaps someday down the road, maybe.
++Lar
|
|
| |
I'm not sure where this standard definition is at these days, or whether it
could be used to automate what I'm doing by a clever MetaWiki robot coder, but I
(and others) started capturing some connection information in BrickWiki (
http://brickwiki.zapto.org/ )
You can see what we've created so far here:
http://brickwiki.zapto.org/index.php/Category:Connnection_Types and the articles
in that category...
Do people think this is valuable? Would people like to take a crack at adding
more types? Or rendering the needed images for the types there already? or
refining the ConnectionTypeBox?
Wikis are editable by anyone... if enough people capture what they know on
BrickWiki, this one will be really cool.
|
|
| |
On 7/17/05, Larry Pieniazek at@at dot.dot
<larmiltontrainworkscom@qs483.pair.com> wrote:
> I'm not sure where this standard definition is at these days, or whether it
> could be used to automate what I'm doing by a clever MetaWiki robot coder, but
> I
> (and others) started capturing some connection information in BrickWiki (
> http://brickwiki.zapto.org/ )
>
> You can see what we've created so far here:
>
> http://brickwiki.zapto.org/index.php/Category:Connnection_Types and the
> articles
> in that category...
>
> Do people think this is valuable? Would people like to take a crack at adding
> more types? Or rendering the needed images for the types there already? or
> refining the ConnectionTypeBox?
>
> Wikis are editable by anyone... if enough people capture what they know on
> BrickWiki, this one will be really cool.
While a connection database could be helpful, I just cant help
thinking that a relationship/constrain system like solidworks and
Pro/Engineer use could acheive a lot more.
It could certainly be a start, then simplified - so the CAD
applications (MLCad etc), then try (with confirmation) to make
assumptions that you want to place a peice on that one - with the
studs attached as so. It would be nice to tell it to center an axle in
a hole and so on.
This could get interesting though - as there are no real curves and no
parametric definitions, and therefore no center points defined - so
creating a "concentric" or "tangent" constraint could be difficult.
This may be contraversial (very) but an extension to the LDraw
standard could allow more advanced clients to work with parametric
definitions, and then render them into simple meshes for use in
others.
Just my two pence worth, I used to be a software engineer with a large
CAD corp - so I have a lot of ideas on this stuff.
Danny
--
http://orionrobots.co.uk - Build Robots
Online Castle Building RPG -
http://www.darkthrone.com/recruit.dt?uid=V30311I30328J30379X30379E30260X30277
|
|
| |
In lugnet.cad, danny staple <orionrobots@gmail.com> wrote:
>
> On 7/17/05, Larry Pieniazek at@at dot.dot
> <larmiltontrainworkscom@qs483.pair.com> wrote:
> > I'm not sure where this standard definition is at these days, or whether it
> > could be used to automate what I'm doing by a clever MetaWiki robot coder, but
> > I
> > (and others) started capturing some connection information in BrickWiki (
> > http://brickwiki.zapto.org/ )
> >
> > You can see what we've created so far here:
> >
> > http://brickwiki.zapto.org/index.php/Category:Connnection_Types and the
> > articles
> > in that category...
> >
> > Do people think this is valuable? Would people like to take a crack at adding
> > more types? Or rendering the needed images for the types there already? or
> > refining the ConnectionTypeBox?
Note that Rosco already did take a crack at a few, including generating some
images for them, and in doing so, found a problem in one I created, and caused
me to decide to add fields to the connectionTypeBox, which is goodness. Thanks!
> > Wikis are editable by anyone... if enough people capture what they know on
> > BrickWiki, this one will be really cool.
I saw a couple other people adding comments (corrections, expansions, links to
good external articles, etc) to the talk pages associated with the entries,
While that's awesome in its own right, I'd like to encourage people to be
bold... Edit that info right into the main article if you're sure it adds
value. (after peeking at the article to grok how the tagging and layout works)
Wikis can be edited by anyone. (Just please consider setting up an ID first so
we can tell who did it and so that communication back to you can happen)
> While a connection database could be helpful, I just cant help
> thinking that a relationship/constrain system like solidworks and
> Pro/Engineer use could acheive a lot more.
>
> It could certainly be a start, then simplified - so the CAD
> applications (MLCad etc), then try (with confirmation) to make
> assumptions that you want to place a peice on that one - with the
> studs attached as so. It would be nice to tell it to center an axle in
> a hole and so on.
>
> This could get interesting though - as there are no real curves and no
> parametric definitions, and therefore no center points defined - so
> creating a "concentric" or "tangent" constraint could be difficult.
> This may be contraversial (very) but an extension to the LDraw
> standard could allow more advanced clients to work with parametric
> definitions, and then render them into simple meshes for use in
> others.
>
> Just my two pence worth, I used to be a software engineer with a large
> CAD corp - so I have a lot of ideas on this stuff.
I agree with all that and certainly would love to see it come to pass.
However, my thinking on the stuff being put in BrickWiki though, is more
"encyclopedic", that is, it's an explanation of what's possible rather than a
constraint system. New builders often ask "how did you do that, how did you get
those things to connect that way" and perhaps these entries would be a resource
on how to do things. More techniques in the toolbox means better builders,
faster. And that means builders who enjoy themselves more, I think.
If LSC actually does get a lot of data, I'd be after getting someone to try to
write a bot to inhale it and create the connection box info.
(there could be a parallel of sorts in setting up information on how to do SNOT,
what sorts of things turn what angles, and information on how to do offsets,
what sorts of parts arrangements give you a 1/10 brick horizontal offset, or a
half plate vertical offset, etc...)
|
|
|
