(no subject)

classic Classic list List threaded Threaded
59 messages Options
123
Reply | Threaded
Open this post in threaded view
|

Re: Missing operators

Brian Korsedal
I used a python program to generate a .dxf and then imported that .dxf
to OpenSCAD.  I'm not the one who wrote the python program, but I can
find a reference to how they did it.  If you want to graft python into
OpenSCAD it might we worth checking out.

On Mon, Sep 13, 2010 at 12:12 PM, Tony Buser <[hidden email]> wrote:
> Has anyone seen this? http://pygts.sourceforge.net/ Would this + an
> STL output function get us most of the way towards a Python OpenSCAD?
> _______________________________________________
> OpenSCAD mailing list
> [hidden email]
> http://rocklinux.net/mailman/listinfo/openscad
>

Reply | Threaded
Open this post in threaded view
|

Re: Missing operators

tbuser
No what I mean is, as I understand it after a brief look at their
webpage (I haven't actually tried it), pygts could be used in a
similar way to how we're using OpenSCAD.  ie. write code -> generate
STL using most of the same CSG boolean operations we use in OpenSCAD.
Except, you use a more full featured language. (python, even though I
don't particularly like python)  Of course there's a lot more to
OpenSCAD than that, but still...

On Mon, Sep 13, 2010 at 4:51 PM, Brian Korsedal
<[hidden email]> wrote:
> I used a python program to generate a .dxf and then imported that .dxf
> to OpenSCAD.  I'm not the one who wrote the python program, but I can
> find a reference to how they did it.  If you want to graft python into
> OpenSCAD it might we worth checking out.
>
> On Mon, Sep 13, 2010 at 12:12 PM, Tony Buser <[hidden email]> wrote:
>> Has anyone seen this? http://pygts.sourceforge.net/ Would this + an
>> STL output function get us most of the way towards a Python OpenSCAD?

Reply | Threaded
Open this post in threaded view
|

Re: Missing operators

Owen Taylor
In reply to this post by Giles Bathgate-2
I'm not going to take a position on what features the OpenSCAD
language should have and should not have - it's clearly a judgement
call.

But I think there *is* legitimate use for a division of labor - not
making OpenSCAD do everything itself. Say you wanted to use OpenSCAD
to create a model that included some engraved text in Hindi.

 - You could add complex text layout as a built-in feature of OpenSCAD
 - You could add object orientation and a complete module system to
the OpenSCAD language so you could access a complex text layout engine
(Pango, ICU, Uniscribe, whatever) from your OpenSCAD scripts

Or:

 - You could call out to Python. Python already has existing bindings
for text language engines - as near to hand as an import statement.

You could make the same argument for rendering SVG vector graphics, or
using complex data structures, or many other things. Even if a ton of
effort is put into making OpenSCAD fully capable as a general purpose
language, it's not always going to be the right tool for the job.

(And really, I think people who who want to mix OpenSCAD with other
languages should look at my recently posted patch. It's a nice way to
work! ;-)

- Owen

On Mon, Sep 13, 2010 at 12:40 PM, Giles Bathgate
<[hidden email]> wrote:

> It seems ridiculous to me that people write python scripts to generate
> OpenSCAD scripts that generate STL.
>
> This is surely a sign that the language is lacking necessary features.
>
> I am all for your "declarative" way of doing things, but how do you
> extend the power of assign to solve problems such as andy is having?
>
> Its really up to the maintainer which direction the language goes, but
> I personally think that an assignment operator is a step in the right
> direction.

Reply | Threaded
Open this post in threaded view
|

Re: Missing operators

andy@kirbyand.co.uk
In reply to this post by tbuser
Ummmmm

Have just had a rummage through pygts.

It strikes me more as a mesh modeler and seems to lack extrude and import functionality.

I took some time over the api and have come to the conclusion that it is massively over complex for our needs. It would realy need a bunch of wrapper code to aproximate scad's cleaner approach.

Probably doable with more python bits added in and a wrapper interface. A lot of work to get back to where scad is now.

Frustrating realy when scad is nearly there all but for variables and math within compound statments and loop bodies.

If OpenSource had already spawned a solution we would'nt be having this conversation.

I respect the wishes of those who like it how it is whilst needing more myself. Question at this point then is whether to fork or use something else.






 
--
Like a rolling stone
----- Original message -----

> No what I mean is, as I understand it after a brief look at their
> webpage (I haven't actually tried it), pygts could be used in a
> similar way to how we're using OpenSCAD.   ie. write code -> generate
> STL using most of the same CSG boolean operations we use in OpenSCAD.
> Except, you use a more full featured language. (python, even though I
> don't particularly like python)   Of course there's a lot more to
> OpenSCAD than that, but still...
>
> On Mon, Sep 13, 2010 at 4:51 PM, Brian Korsedal
> <[hidden email]> wrote:
> > I used a python program to generate a .dxf and then imported that .dxf
> > to OpenSCAD.  I'm not the one who wrote the python program, but I can
> > find a reference to how they did it.  If you want to graft python into
> > OpenSCAD it might we worth checking out.
> >
> > On Mon, Sep 13, 2010 at 12:12 PM, Tony Buser <[hidden email]> wrote:
> > > Has anyone seen this? http://pygts.sourceforge.net/ Would this + an
> > > STL output function get us most of the way towards a Python OpenSCAD?
> _______________________________________________
> OpenSCAD mailing list
> [hidden email]
> http://rocklinux.net/mailman/listinfo/openscad


Reply | Threaded
Open this post in threaded view
|

Re: Missing operators

Giles Bathgate-2

Well open source only offers the solution YOU need if you know how to code or you know someone willing to help you. So its unfair to make demands on a specific bit of software and the community that supports it. However I do feel your pain and do feel that openscad should support basic scripting capabilities and not just be a declarative syntax, we have languages like vrml (X3D) for that

On 14 Sep 2010 22:34, "Andy Kirby" <[hidden email]> wrote:

Ummmmm

Have just had a rummage through pygts.

It strikes me more as a mesh modeler and seems to lack extrude and import functionality.

I took some time over the api and have come to the conclusion that it is massively over complex for our needs. It would realy need a bunch of wrapper code to aproximate scad's cleaner approach.

Probably doable with more python bits added in and a wrapper interface. A lot of work to get back to where scad is now.

Frustrating realy when scad is nearly there all but for variables and math within compound statments and loop bodies.

If OpenSource had already spawned a solution we would'nt be having this conversation.

I respect the wishes of those who like it how it is whilst needing more myself. Question at this point then is whether to fork or use something else.








--
Like a rolling stone
----- Original message -----

> No what I mean is, as I understand it after a brief look at their
> webpage (I haven't actually tr...

Reply | Threaded
Open this post in threaded view
|

Re: Missing operators

tbuser
In reply to this post by andy@kirbyand.co.uk
@andy
> It strikes me more as a mesh modeler and seems to lack extrude and import functionality.

There is an STL -> GTS which can be imported
http://gts.sourceforge.net/darcs/gts-devel/tools/stl2gts.c But you're
right, it does lack a lot of things OpenSCAD has.  However, it does
have a much more powerful language...

@owen
> But I think there *is* legitimate use for a division of labor - not
> making OpenSCAD do everything itself. Say you wanted to use OpenSCAD
> to create a model that included some engraved text in Hindi.

I agree.  I'd love to work on adding more DXF entity support to
OpenSCAD which would allow easier importing DXF files created outside
OpenSCAD for things like fonts to be extruded.  See my attempts at:
http://tonybuser.com/3d-text-and-openscad-bitmaps However, needing to
use one language to generate code in another language which gets
compiled into something else to generate a model is painful.

Reply | Threaded
Open this post in threaded view
|

Re: Missing operators

andy@kirbyand.co.uk
In reply to this post by Giles Bathgate-2
I agree fully.

I feel very little pain at this point though.

Fortunately the shortfall was discovered before I had reccomended scad's use too widely.

The ability to build smart-parts and give them to people to use who are not parts creators is why I came to scad.

If I make parts for people that they have to come back to me for tweaking every time, then I am responsible for perpetrating what you sugested below.

If however I gift people with parts thatm are smart and can self tweak (within parameters, pun intended) using a simple (doable by someone less able than myself) toolchain with minimal dependancies I have done more for more people with out inflicting intelectual vendor lock-in on them.

Openscad is currently the closest option to acheiving this vision at the moment.

But ultimately it is no hardship to me to wait a while longer.  

Open source and free software are charity of the intelect.

Like all charity some is given with a heavy price tag and there are more charitys asking for contributions than an individual can physicaly donate to.

Maybe sometimes any solution is not the best solution. Giving has to be selective. Open source that takes more to use than the benefit it brings is unlikely to survive.

Personaly (and i suspect the same is true of most folk here) i am up to my lugs in giving to open source and at the moment am so overdrawn that i cannot give to scad as i would like.

I thought to mention and ask if scad was heading the way of my vision to find out if it was worth waiting a while or if there was a minor way i could contribute. (Testing, ideas, whatever etc)

If not I will leave you guys to it. Openscad is good and has a bright future. Because you are making it that way.  
 

--
Like a rolling stone
----- Original message -----

> Well open source only offers the solution YOU need if you know how to
> code or you know someone willing to help you. So its unfair to make
> demands on a specific bit of software and the community that supports
> it. However I do feel your pain and do feel that openscad should support
> basic scripting capabilities and not just be a declarative syntax, we
> have languages like vrml (X3D) for that
>
> On 14 Sep 2010 22:34, "Andy Kirby" <[hidden email]> wrote:
>
> Ummmmm
>
> Have just had a rummage through pygts.
>
> It strikes me more as a mesh modeler and seems to lack extrude and import
> functionality.
>
> I took some time over the api and have come to the conclusion that it is
> massively over complex for our needs. It would realy need a bunch of
> wrapper code to aproximate scad's cleaner approach.
>
> Probably doable with more python bits added in and a wrapper interface. A
> lot of work to get back to where scad is now.
>
> Frustrating realy when scad is nearly there all but for variables and
> math within compound statments and loop bodies.
>
> If OpenSource had already spawned a solution we would'nt be having this
> conversation.
>
> I respect the wishes of those who like it how it is whilst needing more
> myself. Question at this point then is whether to fork or use something
> else.
>
>
>
>
>
>
>
>
> --
> Like a rolling stone
> ----- Original message -----
>
> > No what I mean is, as I understand it after a brief look at their
> > webpage (I haven't actually tr...


Reply | Threaded
Open this post in threaded view
|

Re: Missing operators

Steven Dick
In reply to this post by andy@kirbyand.co.uk
// Here's how you make an array of hex objects
// I'm too lazy to do the trig to get them touching perfectly, but this is close enough
// to show it is possible and easy

rad = 10;
seg = sin(60)*rad;
xstep = rad*2;
ystep = seg + rad; // this is almost right

for (y=[0:5]){
  for (x=[0:5]){
    translate([x*xstep+y%2*xstep/2,y*ystep,0])
      for (i= [0:6]){
    rotate([0,0,i*60]) translate([rad,0,0]) cube([1,seg,1],center=true);
      }
  }
}


Reply | Threaded
Open this post in threaded view
|

Re: Missing operators

andy@kirbyand.co.uk
Ok where do the 5 and 6 come from.

And what when I want to change the tile size from say 3 hex cells per side to 11 hex cells per side??

Dont get me wrong a solution would be great.

I take it that you are using square instead of a hex cell. I used circle drawn with 6 segments.


Btw you might need to re-synchronise terms. Your definition of easy is questionable.



--
Like a rolling stone
----- Original message -----

> // Here's how you make an array of hex objects
> // I'm too lazy to do the trig to get them touching perfectly, but this
> is close enough
> // to show it is possible and easy
>
> rad = 10;
> seg = sin(60)*rad;
> xstep = rad*2;
> ystep = seg + rad; // this is almost right
>
> for (y=[0:5]){
>     for (x=[0:5]){
>         translate([x*xstep+y%2*xstep/2,y*ystep,0])
>             for (i= [0:6]){
>         rotate([0,0,i*60]) translate([rad,0,0]) cube([1,seg,1],center=true);
>             }
>     }
> }


Reply | Threaded
Open this post in threaded view
|

Re: Missing operators

andy@kirbyand.co.uk
In reply to this post by Steven Dick
Ok

run all that what it makes is a square matrix by drawing the sides as oblons aranged in a hex patter.

Clever. I like it. But definitely not what is needed.

What is needed is:-

a hexagonal tile (it must tesselate) to be used to make larger sheets.

The hexagonal tile is made of hexagonal units or cells that also tesselate.

Each hexagonal unit is made of a top and bottom rim which is hexagonal of characteristic radis rad (we got that bit). The hexagonal tube between the rims must be h long and be hexagonaly biconical ie have a waste in the middle of characteristic radius w

The tile will be of a module or cell size of m units per side.

That is why it is anything but easy.

Making the cells is easy. Auto calculating the teselatable (it has to be a hexagon of hexagons, draw it on hex graph paper if you dont believe me) variable m tile in scad without variables is not.

Please prove me wrong.





 



--
Like a rolling stone
----- Original message -----

> // Here's how you make an array of hex objects
> // I'm too lazy to do the trig to get them touching perfectly, but this
> is close enough
> // to show it is possible and easy
>
> rad = 10;
> seg = sin(60)*rad;
> xstep = rad*2;
> ystep = seg + rad; // this is almost right
>
> for (y=[0:5]){
>     for (x=[0:5]){
>         translate([x*xstep+y%2*xstep/2,y*ystep,0])
>             for (i= [0:6]){
>         rotate([0,0,i*60]) translate([rad,0,0]) cube([1,seg,1],center=true);
>             }
>     }
> }


Reply | Threaded
Open this post in threaded view
|

Re: Missing operators

clothbot
Lost me on the hexagonaly biconical bit, but here's how I would approach parametric generation of a simple hexagon-of-hexagons:

http://github.com/clothbot/makerbot/blob/master/tests/hexagons/hex_fill.scad

Andrew.

On 2010-09-14, at 11:09 PM, Andy Kirby wrote:

> Ok
>
> run all that what it makes is a square matrix by drawing the sides as oblons aranged in a hex patter.
>
> Clever. I like it. But definitely not what is needed.
>
> What is needed is:-
>
> a hexagonal tile (it must tesselate) to be used to make larger sheets.
>
> The hexagonal tile is made of hexagonal units or cells that also tesselate.
>
> Each hexagonal unit is made of a top and bottom rim which is hexagonal of characteristic radis rad (we got that bit). The hexagonal tube between the rims must be h long and be hexagonaly biconical ie have a waste in the middle of characteristic radius w
>
> The tile will be of a module or cell size of m units per side.
>
> That is why it is anything but easy.
>
> Making the cells is easy. Auto calculating the teselatable (it has to be a hexagon of hexagons, draw it on hex graph paper if you dont believe me) variable m tile in scad without variables is not.
>
> Please prove me wrong.
>
>
>
>
>
>
>
>
>
> --
> Like a rolling stone
> ----- Original message -----
>> // Here's how you make an array of hex objects
>> // I'm too lazy to do the trig to get them touching perfectly, but this
>> is close enough
>> // to show it is possible and easy
>>
>> rad = 10;
>> seg = sin(60)*rad;
>> xstep = rad*2;
>> ystep = seg + rad; // this is almost right
>>
>> for (y=[0:5]){
>>      for (x=[0:5]){
>>          translate([x*xstep+y%2*xstep/2,y*ystep,0])
>>              for (i= [0:6]){
>>          rotate([0,0,i*60]) translate([rad,0,0]) cube([1,seg,1],center=true);
>>              }
>>      }
>> }
>
> _______________________________________________
> OpenSCAD mailing list
> [hidden email]
> http://rocklinux.net/mailman/listinfo/openscad

--

"The future is already here.  It's just not very evenly distributed" -- William Gibson

Me: http://clothbot.com/wiki/




Reply | Threaded
Open this post in threaded view
|

Re: Missing operators

andy@kirbyand.co.uk
In reply to this post by andy@kirbyand.co.uk
Funnily enough the square grid may do for the friend I am writing this for. I will ping him and see.

If he can use a squared grid of hexagons cells that tesselates. Or trim it down to make a hexagon grid of cells it might be good enough.

The problem is that photon multiplier tubes are round. So a hexagon tile is a good match.

--
Like a rolling stone
----- Original message -----

> Ok
>
> run all that what it makes is a square matrix by drawing the sides as
> oblons aranged in a hex patter.
>
> Clever. I like it. But definitely not what is needed.
>
> What is needed is:-
>
> a hexagonal tile (it must tesselate) to be used to make larger sheets.
>
> The hexagonal tile is made of hexagonal units or cells that also
> tesselate.
>
> Each hexagonal unit is made of a top and bottom rim which is hexagonal
> of characteristic radis rad (we got that bit). The hexagonal tube
> between the rims must be h long and be hexagonaly biconical ie have a
> waste in the middle of characteristic radius w
>
> The tile will be of a module or cell size of m units per side.
>
> That is why it is anything but easy.
>
> Making the cells is easy. Auto calculating the teselatable (it has to be
> a hexagon of hexagons, draw it on hex graph paper if you dont believe
> me) variable m tile in scad without variables is not.
>
> Please prove me wrong.
>
>
>
>
>
>     
>
>
>
> --
> Like a rolling stone
> ----- Original message -----
> > // Here's how you make an array of hex objects
> > // I'm too lazy to do the trig to get them touching perfectly, but this
> > is close enough
> > // to show it is possible and easy
> >
> > rad = 10;
> > seg = sin(60)*rad;
> > xstep = rad*2;
> > ystep = seg + rad; // this is almost right
> >
> > for (y=[0:5]){
> >       for (x=[0:5]){
> >           translate([x*xstep+y%2*xstep/2,y*ystep,0])
> >               for (i= [0:6]){
> >           rotate([0,0,i*60]) translate([rad,0,0])
> > cube([1,seg,1],center=true);              }
> >       }
> > }
>
> _______________________________________________
> OpenSCAD mailing list
> [hidden email]
> http://rocklinux.net/mailman/listinfo/openscad


Reply | Threaded
Open this post in threaded view
|

Re: Missing operators

andy@kirbyand.co.uk
In reply to this post by clothbot
Ta muchly

I will have a crack at that one next. It looks very promising though.

If you were to make an hour glass using two cones it would be bi conical. With the waist being the narrow bit in the middle. If the cones were hexagonal in shape instead of round you would pretty much have it.

I will pinch anything that works but will add attribution for contributors to the header. I am a big believer in recognising contribution.

Whey hey might not nead variables after all.

Cheers guys.

 

--
Like a rolling stone
----- Original message -----

> Lost me on the hexagonaly biconical bit, but here's how I would approach
> parametric generation of a simple hexagon-of-hexagons:
>
> http://github.com/clothbot/makerbot/blob/master/tests/hexagons/hex_fill.scad
>
> Andrew.
>
> On 2010-09-14, at 11:09 PM, Andy Kirby wrote:
>
> > Ok
> >
> > run all that what it makes is a square matrix by drawing the sides as
> > oblons aranged in a hex patter.
> >
> > Clever. I like it. But definitely not what is needed.
> >
> > What is needed is:-
> >
> > a hexagonal tile (it must tesselate) to be used to make larger sheets.
> >
> > The hexagonal tile is made of hexagonal units or cells that also
> > tesselate.
> >
> > Each hexagonal unit is made of a top and bottom rim which is hexagonal
> > of characteristic radis rad (we got that bit). The hexagonal tube
> > between the rims must be h long and be hexagonaly biconical ie have a
> > waste in the middle of characteristic radius w
> >
> > The tile will be of a module or cell size of m units per side.
> >
> > That is why it is anything but easy.
> >
> > Making the cells is easy. Auto calculating the teselatable (it has to
> > be a hexagon of hexagons, draw it on hex graph paper if you dont
> > believe me) variable m tile in scad without variables is not.
> >
> > Please prove me wrong.
> >
> >
> >
> >
> >
> >
> >
> >
> >
> > --
> > Like a rolling stone
> > ----- Original message -----
> > > // Here's how you make an array of hex objects
> > > // I'm too lazy to do the trig to get them touching perfectly, but
> > > this is close enough
> > > // to show it is possible and easy
> > >
> > > rad = 10;
> > > seg = sin(60)*rad;
> > > xstep = rad*2;
> > > ystep = seg + rad; // this is almost right
> > >
> > > for (y=[0:5]){
> > > for (x=[0:5]){
> > > translate([x*xstep+y%2*xstep/2,y*ystep,0])
> > > for (i= [0:6]){
> > > rotate([0,0,i*60]) translate([rad,0,0]) cube([1,seg,1],center=true);
> > > }
> > > }
> > > }
> >
> > _______________________________________________
> > OpenSCAD mailing list
> > [hidden email]
> > http://rocklinux.net/mailman/listinfo/openscad
>
> --
>
> "The future is already here.   It's just not very evenly distributed" --
> William Gibson
>
> Me: http://clothbot.com/wiki/
>
>
>
> _______________________________________________
> OpenSCAD mailing list
> [hidden email]
> http://rocklinux.net/mailman/listinfo/openscad


Reply | Threaded
Open this post in threaded view
|

Re: Missing operators

Steven Dick
rad = 10;
seg = sin(60)*rad;
xstep = rad*2;
ystep = seg + rad; // this is almost right

hexmap(5);

module hexmap(side)
{
  m = side-1;
  for (y=[0:m]){
    row (y,m,m-(y+y%2)/2);
  }
  for (y=[m+1:m*2]){
    row (y,m,(y-y%2)/2);
  }
}

module row(y,m,start)
{
  width = (m-start)*2 - y%2 + m;
  echo (start,width,m,y);
  for (x=[start : start+width]){
    translate([x*xstep+y%2*xstep/2,y*ystep,0]) hexcell(rad);
  }
}

module hexcell(rad)
{
      for (i= [0:5]){
    rotate([0,0,i*60]) translate([rad,0,0]) cube([1,seg,1],center=true);
      }
}

module cylcell(rad)
{
  cylinder(r=rad,h=1);
}

Reply | Threaded
Open this post in threaded view
|

Re: Missing operators

Steven Dick
Not sure if I made some silly error and OpenSCAD really needs fixing, or if it already supports it, but...

After writing this code, OpenSCAD's grammar may need adjusting so that every *block* supports new variables, not just every module...  Even if we do keep it purely functional without side effects.

Not that I care either way if real variables get added or not.
I see advantages both ways, and can't decide.

What I do think important, however, is that if real variables are added, they use proper scoping rules.
This might include distinguishing between creating a new variable and assigning to a global one.


Of course, I still think that OpenSCAD should directly support exporting multiple STL files from the same run.
Yes, I know you can do multiple runs on the same file and get the same thing, but that doesn't take advantage of the internal caching that might otherwise occur.
Reply | Threaded
Open this post in threaded view
|

Re: Missing operators

andy@kirbyand.co.uk
Maybe i am being dim but cant you just copy the stl to get multiple stl files.

Sorry it has been a long night shift.

--
Like a rolling stone
----- Original message -----

> Not sure if I made some silly error and OpenSCAD really needs fixing, or
> if it already supports it, but...
>
> After writing this code, OpenSCAD's grammar may need adjusting so that
> every *block* supports new variables, not just every module...   Even if
> we do keep it purely functional without side effects.
>
> Not that I care either way if real variables get added or not.
> I see advantages both ways, and can't decide.
>
> What I do think important, however, is that if real variables are added,
> they use proper scoping rules.
> This might include distinguishing between creating a new variable and
> assigning to a global one.
>
>
> Of course, I still think that OpenSCAD should directly support exporting
> multiple STL files from the same run.
> Yes, I know you can do multiple runs on the same file and get the same
> thing, but that doesn't take advantage of the internal caching that might
> otherwise occur.


Reply | Threaded
Open this post in threaded view
|

Re: Missing operators

Steven Dick


On Wed, Sep 15, 2010 at 12:49 AM, Andy Kirby <[hidden email]> wrote:
Maybe i am being dim but cant you just copy the stl to get multiple stl files.


Sure, if I want the same object.

No, what I'm doing is I'm making related parts and putting them all in the same OpenSCAD file.

In some cases, one part has a complex surface, and the other part has a face that has the first part cut out of it.
Or something like that.  Or maybe they just share geometry in a more obvious way.

Reply | Threaded
Open this post in threaded view
|

Re: Missing operators

clothbot
That's why I structure my code with the "render_part" variable at the top and use the module declarations to manage variable scope.

I'm able to control which part is rendered out from the command-line in a Makefile like this: http://github.com/clothbot/eda/blob/master/openscad/libraries/avago/Makefile

Andrew.

On 2010-09-15, at 12:36 AM, Steven Dick wrote:

>
>
> On Wed, Sep 15, 2010 at 12:49 AM, Andy Kirby <[hidden email]> wrote:
> Maybe i am being dim but cant you just copy the stl to get multiple stl files.
>
>
> Sure, if I want the same object.
>
> No, what I'm doing is I'm making related parts and putting them all in the same OpenSCAD file.
>
> In some cases, one part has a complex surface, and the other part has a face that has the first part cut out of it.
> Or something like that.  Or maybe they just share geometry in a more obvious way.
>
> _______________________________________________
> OpenSCAD mailing list
> [hidden email]
> http://rocklinux.net/mailman/listinfo/openscad

--

"The future is already here.  It's just not very evenly distributed" -- William Gibson

Me: http://clothbot.com/wiki/




Reply | Threaded
Open this post in threaded view
|

Re: Missing operators

Britton Kerin-2
In reply to this post by Owen Taylor
On Tue, Sep 14, 2010 at 2:40 PM, Owen Taylor <[hidden email]> wrote:
> I'm not going to take a position on what features the OpenSCAD
> language should have and should not have - it's clearly a judgement
> call.

I'll go ahead and take a position on this.  OpenSCAD should really be a
library and a separate little program to view the renders.  Ideally with
numerous language bindings available.

Don't get me wrong OpenSCAD almost made my cry with joy when I found it but
there is no good reason for it to have its own language.  The language is *not*
the useful part of this brilliant piece of software.  The non-coders aren't
going to start programming in openscadese they will just use a conventional
CAD anyway.

I know it caches some things but this could be done with a cccache-type
approach.

Adding more language features is effort in the wrong direction IMO.

Britton

123