Mini-challenge

jeff hammond

@unknownuser said:

@jason_maranto said:

I get 10 even -- I think you may have mis-measured.

I'm not talking about the angle -- you don't need any math for that at all, what you need to know is how long the board should be.

Best,
Jason.

i upped the precision to another decimal..

I've got mine maxed out too -- maybe SketchUp is converting something upon opening the file?
[attachment=0:p19ezxh7]challenge_altfinameasurel.jpg[/attachment:p19ezxh7]

Best,
Jason.

delete the post.. i think you're measuring to the top of the post instead of the top of the diagonal board.. it's all so very close together in there that the inference gets confused..

jason_maranto

@unknownuser said:

@jason_maranto said:

I get 10 even -- I think you may have mis-measured.

I'm not talking about the angle -- you don't need any math for that at all, what you need to know is how long the board should be.

Best,
Jason.

i upped the precision to another decimal..

I've got mine maxed out too -- maybe SketchUp is converting something upon opening the file?

I'll tell you what would make it possible to do with just the tools (no math at all), would be if the construction point would snap to the guideline as I rotate it.

Best,
Jason.

jason_maranto

Yep, that found it -- crap... that's some serious precision. Any math would have to calculate down to those many decimal places to match it... and even then it may not be right.

Best,
Jason.

andybot

@pbacot said:

all these methods are the same. You are either rotating a guide point or the board, to an insecure target along a line.... It is just that missing "projected arc intersection" function.

Which TIG's tool neatly restores

pbacot

Jason,

How do you know the 1000th of degrees and upping the precision to another decimal etc.? What do you mean?

Peter

jason_maranto

Under Window> Model Info> Units you can set the precision of SketchUp (to its limits) -- generally I never get this precise because it's not practical for what I do, but it is there if you want it.

Best,
Jason.

jeff hammond

@pbacot said:

all these methods are the same. You are either rotating a guide point or the board, to an insecure target along a line. SU should rotate using the projected arc intersection and that would be no issue for SU? I assume the rotate arc IS on a true curve. It is just that missing "projected arc intersection" function. In my CAD program (which does have arcs of course) a downfall is inability of lines and arcs to find (snap) that projected intersection point without using a trim tool. SU does line completion correctly why not rotation.

It should be part of the SU inference tools (and toggling OFF the inference tools should also be a part! )

right. I think the devs could make sketchup inferencing work the way we're expecting it to without changing the essence of sketchup.. a 'smart rotate' tool might be able to be created via ruby but for all I know, a ruby coder may end up facing the same brick wall?

pbacot

all these methods are the same. You are either rotating a guide point or the board, to an unsecured target along a line. SU should rotate using the projected arc intersection and that would be no issue for SU? I assume the rotate arc IS on a true curve. It is just that missing "projected arc intersection" function. In my CAD program (which does have arcs of course) a downfall is inability of lines and arcs to find (snap) that projected intersection point without using a trim tool. SU does line completion correctly why not rotation.

It should be part of the SU inference tools (and toggling OFF the inference tools should also be a part! )

TIG

Pilou

If the 'cut' end of the rail has the same 'height' when it is flat as it does when it is angled, then the sloping rail has been distorted and it is no longer the same 'width' [i.e. the square/vertical dimension] as it was before it was rotated...

pilou

@unknownuser said:

When testing I recommend you put the posts further apart as with near square shape deviances might be so small that you think you have a correct solution.

I don't understand why you don't receive my test?
All seems perfect!
Where is the glitch ? (I am in decimal with maximum precision alowed by SU )

test_jeff.skp

Gaieus

@unknownuser said:

i'm not quite sure they could give us true arcs in sketchup without entirely changing the way sketchup works.. (how would a cylinder be drawn if there were no segments in the arcs? a nurbs surface? )

I do nor care for any "real" geometry. Let it me an approximated, segemented surface model.

What should be here is circular guides. To let us do geometry we learnt in secondary/high school at our teen ages... That's not much but fairly enough.

TIG

@gaieus said:

@unknownuser said:

i'm not quite sure they could give us true arcs in sketchup without entirely changing the way sketchup works.. (how would a cylinder be drawn if there were no segments in the arcs? a nurbs surface? )

I do nor care for any "real" geometry. Let it me an approximated, segemented surface model.
What should be here is circular guides. To let us do geometry we learnt in secondary/high school at our teen ages... That's not much but fairly enough.
My Tangent Tools fill the gap [for now] and allow you to find the real intersection points of tangents to arcs, lines and arcs, arcs and arcs etc etc...
They are not new

jeff hammond

@unknownuser said:

@unknownuser said:

When testing I recommend you put the posts further apart as with near square shape deviances might be so small that you think you have a correct solution.

I don't understand why you don't receive my test?
All seems perfect!
Where is the glitch ? (I am in decimal with maximum precision alowed by SU )

[attachment=0:1h4o0r1i]test_jeff.jpg[/attachment:1h4o0r1i]

I think presenting the problem in 3D could of been confusing. the direction youre measuring 5 in this latest image is insignificant.. try drawing the 2D version (posted on pg 2 or so)

pilou

Aaaaaaaaaaaaahhhhh
Call me Fredo! I was suffering from hallucinations!

pilou

I just avoid to test the height result! A sort of rotating geometric mirage!
2.670000 against 2.470811m

gilles

I'm back!

Mini-challenge2v6.skp

jeff hammond

@tig said:

@gaieus said:

@unknownuser said:

i'm not quite sure they could give us true arcs in sketchup without entirely changing the way sketchup works.. (how would a cylinder be drawn if there were no segments in the arcs? a nurbs surface? )

I do nor care for any "real" geometry. Let it me an approximated, segemented surface model.
What should be here is circular guides. To let us do geometry we learnt in secondary/high school at our teen ages... That's not much but fairly enough.
My Tangent Tools fill the gap [for now] and allow you to find the real intersection points of tangents to arcs, lines and arcs, arcs and arcs etc etc...
They are not new

yeah, true tangents seems to be the only viable ruby solution at this time.

is it possible to make it more interactive? to where it basically acts like the standard rotate tool but it will snap to an arbitrary point along a line segment? or is something like that impossible to code with the current API?

jeff hammond

@gilles said:

I'm back!

and I'm back on my phone now
can't wait to see what you've come up with this time.

(in 2 more hours or so)

andybot

@gilles said:

I'm back!

I realized this is not strictly geometrically correct. The line that you are putting the guide perpendicular to will not be at the same angle once it is adjusted to the correct width. There is a slight shift that occurs once you adjust both ends.

pilou

Thx for the V6 ! The figure remember something N

I will try another idea come back...in...a week...or months...