Requests &...from Pilou:)
 1-20  21-40  41-60  61-80

Previous
Next
 From:  Michael Gibson
207.21 In reply to 207.20 
> Is it the same tangent than a point who will be near the extremity and
> the extremity but "on" the curve?
> Seems yes, but I ask the question :)

Yes, the way NURBS curves work is the end tangent direction of the curve is defined by the direction of the last 2 control points.

But just to be clear - a point on the curve that is near the end but a small distance away from it will not have this tangent quite yet - it is only the very end of the curve that has this tangent. But I think that you have shown that in your graphic. As you move closer and closer along the curve towards the end point, the curve's tangent at that point will get closer and closer to the line between those 2 points.

- Michael
  Reply Reply More Options
Post Options
Reply as PM Reply as PM
Print Print
Mark as unread Mark as unread
Relationship Relationship
IP Logged

Previous
Next
 From:  Frenchy Pilou (PILOU)
207.22 In reply to 207.21 

Yes very very closer but not the same :)
In the file linked...
There are one big curve "curve", a little curve "straight" near the extremity and "on" it, a big curve "straight" tangente to the little curve
I had no preconceived ideas of that, now I am fixed :)
That maybe good to know that in particular extrem conditions :)
So it's better to draw out a curve "curve" by the function "Extend" than by an tangent aligment to the control's points :)

So MoI is a very cool tool for verify geometry concepts :)

Ps Is it possible to have somewhere the Zoom status? (with some decimal numbers :)
I have made that : a curve extend to a line, and an extension by the alignment of the control point (see divergence jpg)
I made some mouse's wheel actions before see a "divergence" :)
What is the "clickety clack zoom dimension" of a wheel action?
I made as an approximation of 125 "single" clikety clack :)

PSs Have you plan an "intersection Snap On" with a surface when you start a line ?

EDITED: 8 Nov 2006 by PILOU


  Reply Reply More Options
Post Options
Reply as PM Reply as PM
Print Print
Mark as unread Mark as unread
Relationship Relationship
IP Logged

Previous
Next
 From:  Michael Gibson
207.23 In reply to 207.22 
> So it's better to draw out a curve "curve" by the function "Extend"
> than by an tangent aligment to the control's points :)

I'm not sure about "better", these do slightly different things. Extend will not change the shape of the existing curve, it will add a line segment to the end of it (except if it is an arc, it will create a bigger arc in that case). This is the equivalent of drawing a line off the end of the curve that is tangent to it, and then joining the line and the curve together. Except if your starting curve is a line it will make a single new line in that case.

When you move a control point, the shape of the curve will change, but if you keep the points aligned in the same direction the end tangent of the curve will be the same.


> Ps Is it possible to have somewhere the Zoom status? (with some decimal numbers :)

It's a little bit hard to add this in quickly because that's a piece of new UI that has to go somewhere. Eventually I think there needs to be some view properties dialog that will let you see and change a few different things about the view, including zoom factor.


> I made some mouse's wheel actions before see a "divergence" :)

A very small divergence is normal in many cases. Some operations such as intersections are calculated to the modeling tolerance value, which by default is 0.01, so it is not unusual that there will be gaps and things when you zoom in tighter than that. Also just mathematics in general is limited on the computer since there is only 8 bytes of data to hold each number. So the range of numbers that can be expressed is limited and this inherently creates small errors just from adding and multiplying numbers together. If you zoom in tight enough you can see these kinds of very small divergences as well.

That's why many things work with a tolerance value, like for instance join will join together two curves if their ends are within the modeling tolerance of each other, they don't have to be exactly touching with 0 divergence.


> What is the "clickety clack zoom dimension" of a wheel action?

One clickety clack will adjust the zoom factor by 0.8 in the top/front/right view. In the 3D view it is slightly different, it doesn't change a zoom factor there instead it moves you by a percentage of the distance between the eye point and the pivot point.


> PSs Have you plan an "intersection Snap On" with a surface when you start a line ?

Probably eventually. But this is a somewhat difficult area since the intersection with a surface is a fairly intensive calculation, so performance is a concern there. That's why I haven't attempted to do it yet.

- Michael
  Reply Reply More Options
Post Options
Reply as PM Reply as PM
Print Print
Mark as unread Mark as unread
Relationship Relationship
IP Logged

Previous
Next
 From:  Frenchy Pilou (PILOU)
207.24 

Does exist some limitations or precautions with these sorts of auto Intersect surfaces?
MoI manage them without restrictions?
The "auto trim" don't work is that normal ? (I suppose yes :)
The trim between surface and its generator's curve works fine :)

PS Snap grid disable
Some 2 D curves drawing with intersections a,b,c
When you move a circle by one unordinary point to the intersections a,b,c there is no snaping, but if you move this same circle by its "Quadran" there is snaping on "warm point" a,b,c !
Is that normal? :)
I am agree that is the same but for an ellipse? :)
Aaaaaaaaaah Sorry I have fogotten another time the function "Move" ! Sorry

EDITED: 9 Nov 2006 by PILOU


  Reply Reply More Options
Post Options
Reply as PM Reply as PM
Print Print
Mark as unread Mark as unread
Relationship Relationship
IP Logged

Previous
Next
 From:  tyglik
207.25 
Hi, I hope I may join you, Pilou, to attack Michael with our requests or questions.....


In rhino, there is a way how to define points both in construction plane coordinates
and in world coordinates for each viewport; i can type "x,y,[z]" or "wx,y,[z]".
It seems that it should work identically in MoI, but it doesn't.

-switch into the Front view
-start point command and type 5,5 via keyboard
-repeat it for w5,5
-both points lie at the same location, hmmm?

Is this bug, feature or my mistake?

Petr
  Reply Reply More Options
Post Options
Reply as PM Reply as PM
Print Print
Mark as unread Mark as unread
Relationship Relationship
IP Logged

Previous
Next
 From:  Michael Gibson
207.26 In reply to 207.24 
> Does exist some limitations or precautions with these sorts of auto Intersect surfaces?
> MoI manage them without restrictions?

These are usually called "self-intersections". These types of surfaces will cause problems in several areas, especially things involving surface intersections (which includes trim, booleans, fillets). It creates a problem in the topology of an object, where one 3D point location belongs to multiple different spots on the surface - this can mess up several algorithms.

So it is best to try and avoid making those types of surfaces if you want to do any more advanced operations to them later on.


> When you move a circle by one unordinary point to the intersections
> a,b,c there is no snaping, but if you move this same circle by its "Quadran"
> there is snaping on "warm point" a,b,c !
> Is that normal? :)

It's normal - when you drag an object, it will snap the base point of the drag to either the end, mid, quadrant, or (for solids) the center points just of that object. Intersections are not included because those are not points that sort of "only belong" to the individual object. It would look kind of strange for instance while you were dragging the object around and the "Int" label showed up while you were dragging it. However, having "end, mid, quadrant, or center" show up while you are dragging makes more sense I thought.

Just as you realized, you can use Transform/Move to lock the base point on the intersection if you need to.

- Michael
  Reply Reply More Options
Post Options
Reply as PM Reply as PM
Print Print
Mark as unread Mark as unread
Relationship Relationship
IP Logged

Previous
Next
 From:  Michael Gibson
207.27 In reply to 207.25 
Hi Petr,

> Is this bug, feature or my mistake?

I consider it to be a feature. :)

This particular thing tends to be an area of confusion in Rhino, that typing 5,5 produces a different 3D point location depending on what viewport is active. It is particularly bad in the split view where it is so easy to accidentally change the active viewport just by moving your mouse around without even clicking somewhere.

In MoI this is different - simple coordinate input with no decorations (like 5,5) will always use world coordinates. I think this is really what most people would expect (unless you have become accustomed to other software that does it differently of course).

That just helps to eliminate one area of confusion. The "w" syntax is supported to do the same thing just for people who are accustomed to it from Rhino.

There is currently no way to type in construction-plane relative coordinates. I suppose I could add this in using a new "c" syntax, like "c5,5". What do you think about that? There is not exactly the same concept of an "active viewport" in MoI as there is in Rhino though. But internally there is a concept of the last clicked viewport, and there is also the concept of a full-screen viewport. So the way I would figure it would work is if you have a full screen viewport, it would use that viewport's construction plane, otherwise for the split view it would use the last viewport that was clicked in.

- Michael
  Reply Reply More Options
Post Options
Reply as PM Reply as PM
Print Print
Mark as unread Mark as unread
Relationship Relationship
IP Logged

Previous
Next
 From:  Frenchy Pilou (PILOU)
207.28 In reply to 207.26 
So reserve 'auto intersect' for exotic topologic forms like Boy's Surface or Klein' s Bottle or Big Bang / Big Crunch and Black's Hole :)
  Reply Reply More Options
Post Options
Reply as PM Reply as PM
Print Print
Mark as unread Mark as unread
Relationship Relationship
IP Logged

Previous
Next
 From:  tyglik
207.29 In reply to 207.27 
Hi,

>>This particular thing tends to be an area of confusion in Rhino, that typing 5,5 produces a different 3D point location depending on what viewport is active.

...not to be confusing for me. When I work in 2D view I am just drawing and don't care about 3D.

>>It is particularly bad in the split view where it is so easy to accidentally change the active viewport just by moving your mouse around without even clicking somewhere

yes, I must agree with you in that, it sometimes happens to me, haha...

>> ...simple coordinate input with no decorations (like 5,5) will always use world coordinates. I think this is really what most people would expect...

It is hard not only for me to place a points in 3D space while looking at "non-corresponding" 2D view, so I might value something like "c5,5" or it seems to be better idea to add world axis icon to non-3D viewport, so I wouldn't be confused enough. On the other hand, MoI's construction lines is so powerful (it just groups Rhino's ortho mode, elevator mode, planar mode, tab constraint and AshlarVellum's DraftAssistant together) that users are encouraged to work more in the 3D view where "w5,5" or "5,5" is the same. Besides, it can always be done with drawing in Top viewport and then rotating the drawing.


>>There is currently no way to type in construction-plane relative coordinates. I suppose I could add this in using a new "c" syntax, like "c5,5". What do you think about that?

I think that the rhino's concept world coordinate system 'wx,y,z' versus local one is essential because there is a way how to change the direction and origin of the construction plane. For instance, if I change the origin in Rhino's Perspective view and type "5,5,5" vs. "w5,5,5" there will be absolutely different location of the points. Thus, if you decided not to add the command handling a construction plane, you could give up new 'c' idea perhaps


By the way, it makes no sense to initiate an input field after typing "w" or "r" without running command, so this is a misfeature I guess


Petr
  Reply Reply More Options
Post Options
Reply as PM Reply as PM
Print Print
Mark as unread Mark as unread
Relationship Relationship
IP Logged

Previous
Next
 From:  Michael Gibson
207.30 In reply to 207.29 
> On the other hand, MoI's construction lines is so powerful (it just groups
> Rhino's ortho mode, elevator mode, planar mode, tab constraint and
> AshlarVellum's DraftAssistant together)

Cool, I'm glad that you're finding it useful! I think the other important factor is that it stays out of your way until you call for it with the drag gesture. I've seen many other systems that get triggered too easily and then sometimes end up getting in the way.


> Thus, if you decided not to add the command handling a construction
> plane, you could give up new 'c' idea perhaps

I'm not planning on adding construction plane handling in for V1, but I haven't given up on it forever. I think it will come in future versions.

Once you have set a custom construction plane, I figure it will probably affect all views and then at that point it would probably make sense for typed-in coordinates to be in that new coordinate system instead of world. I don't think that would be surprising anymore, because you would have to do something in particular to activate it.

But just having the coordinate system shift by having a particular view active is more of a problem.


> By the way, it makes no sense to initiate an input field after typing "w" or
> "r" without running command, so this is a misfeature I guess

I've fixed this for the next beta so that focus will only go to the xyz coordinate input if there is a point currently being picked.

- Michael
  Reply Reply More Options
Post Options
Reply as PM Reply as PM
Print Print
Mark as unread Mark as unread
Relationship Relationship
IP Logged

Previous
Next
 From:  Frenchy Pilou (PILOU)
207.31 

Does exist a "full screen" without the UI? (just keeping Zoom Pan Rotate...)

Ps What is the utility when you make a loft of the "Clic on curve for change seam"?
Ok that permit to "move" the seam, works fine but what is the use of that :)

EDITED: 19 Nov 2006 by PILOU

  Reply Reply More Options
Post Options
Reply as PM Reply as PM
Print Print
Mark as unread Mark as unread
Relationship Relationship
IP Logged

Previous
Next
 From:  Michael Gibson
207.32 In reply to 207.31 
> Does exist a "full screen" without the UI? (just keeping Zoom Pan Rotate...)

No not right now. I've thought a little bit about making some way to collapse/expand the side pane and bottom toolbar to make this possible, but I don't think that this will be available for V1.


> Ps What is the utility when you make a loft of the "Clic on curve for change seam"?
> Ok that permit to "move" the seam, works fine but what is the use of that :)

Most of the time there isn't any use in it because the automatic calculation for how to align things works pretty well. But that method is provided in case the automatic alignment didn't work very well.

- Michael
  Reply Reply More Options
Post Options
Reply as PM Reply as PM
Print Print
Mark as unread Mark as unread
Relationship Relationship
IP Logged

Previous
Next
 From:  Frenchy Pilou (PILOU)
207.33 

Filet in this case is pemited?

Edit Very curious I have reload the file and all was Ok!!!
So Does exist a "refresh" function? (reinialize of Pan Rotate don't refresh that)

ps What is exactly history?
Can make this?
Object filet, modif, modif....
Replace at the end the first filet by a chamfer

EDITED: 20 Nov 2006 by PILOU

Attachments:

  Reply Reply More Options
Post Options
Reply as PM Reply as PM
Print Print
Mark as unread Mark as unread
Relationship Relationship
IP Logged

Previous
Next
 From:  Michael Gibson
207.34 In reply to 207.33 
Hi Pilou, this is a bug in the display. The object itself is fine (all the surfaces and trim curves, etc..), and when you reload it, the thing that triggers the bug will be gone and so it will work then. This display bug is fixed for the next beta.

> ps What is exactly history?

History is a mechanism that saves the parameters that you used to create an object. But right now the only parameters that you can change are the input objects. For instance if you do an extrude from a curve, if you change the curve the extrude will update.


> Can make this?
> Object filet, modif, modif....
> Replace at the end the first filet by a chamfer

Not yet, because of a couple of reasons. First, it is not changing an input object, it would be changing different parameters such as a radius or the type of operation. Eventually there will be some kind of UI mechanism that allows you to see and modify these types of values, but that is going to be quite complicated to finish so it won't be ready anytime soon.

The other thing is that when you do a fillet, the original object (before the fillet) is deleted. The history mechanism needs to be able to find the original object in order to apply changed parameters to it. I have some ideas on how to make this possible, but again it is fairly difficult so it will be a while.

There is a different style of CAD software called a "Feature-based parametric solid modeler" which is very much designed to operate like you're asking for. Some examples of this are SolidWorks, Alibre, Inventor, Pro/E, etc... MoI is not really set up to do exactly the same thing as these.

- Michael
  Reply Reply More Options
Post Options
Reply as PM Reply as PM
Print Print
Mark as unread Mark as unread
Relationship Relationship
IP Logged

Previous
Next
 From:  Frenchy Pilou (PILOU)
207.35 In reply to 207.34 
Thx!
All is clear :)
  Reply Reply More Options
Post Options
Reply as PM Reply as PM
Print Print
Mark as unread Mark as unread
Relationship Relationship
IP Logged

Previous
Next
 From:  Frenchy Pilou (PILOU)
207.36 

What about the "cracks" in general?
Is just an visual effect due the screen definition & zoom
or model of Breps(Boundary Representation) volumes result are always "watertighness" in Moi?
Is there an option for regulate that ?

Whish list :)
A "plan work-grid (or -view)" aligned by 3points (or by view) ?

Attachments:

  Reply Reply More Options
Post Options
Reply as PM Reply as PM
Print Print
Mark as unread Mark as unread
Relationship Relationship
IP Logged

Previous
Next
 From:  Michael Gibson
207.37 In reply to 207.36 
> What about the "cracks" in general?
> Is just an visual effect due the screen definition & zoom

Cracks like you show there are usually a visual effect due to slightly different mesh structures being created.

It looks like you have two separate surfaces there - if you join them together into one single brep that has a shared edge, you shouldn't see a crack there anymore. That's because MoI will create the same mesh structure along shared edges.


> or model of Breps(Boundary Representation) volumes result are always "watertighness" in Moi?
> Is there an option for regulate that ?

Breps that have no "naked" edges are watertight - that is when every edge of every surface is joined to some other edge.

This makes it topologically "watertight".

If you have just individual surfaces that are positioned next to each other but not joined, that is not watertight because none of the algorithms in MoI have any knowledge that the surfaces are really supposed to be connected to each other.


> Whish list :)
> A "plan work-grid (or -view)" aligned by 3points (or by view) ?

In the future at some point I want to add some functionality like this for changing the working plane (which I call the "Construction plane"). But I don't think that this will be available for the initial V1 release though.

- Michael
  Reply Reply More Options
Post Options
Reply as PM Reply as PM
Print Print
Mark as unread Mark as unread
Relationship Relationship
IP Logged

Previous
Next
 From:  Frenchy Pilou (PILOU)
207.38 In reply to 207.37 

Thx
Another little question :)
What is the normal use of a volume made in a Polygonal Prog ---> transform in Nurbs --> import in MoI
(so some Nurbs surfaces :)
Select all / Separate / Select all / Boolean Union ?
(file 3dm linked: volume on the left is the "poly" transformed in Nurbs) on the right the operations line above )
Seems all vertical facetts don't weld in only one facett?

Ps Does exist in Moi an automatic function after an auto-trimming of several nurbs surfaces for delete all that not define a volume?

Coons and Gordon surfaces are implicit inside MoI or don't exist?
coons planes http://www.cs.technion.ac.il/~cs234325/Applets/applets/coons/html/index.html

Gordon surface generated by latitude curves and longitudes curves with for each one point commun with each other :)

EDITED: 23 Nov 2006 by PILOU

Attachments:

  Reply Reply More Options
Post Options
Reply as PM Reply as PM
Print Print
Mark as unread Mark as unread
Relationship Relationship
IP Logged

Previous
Next
 From:  Michael Gibson
207.39 In reply to 207.38 
> (file 3dm linked: volume on the left is the "poly" transformed in
> Nurbs) on the right the operations line above )
> Seems all vertical facetts don't weld in only one facett?

The object on the right appears to have a lot of duplicated surfaces in it, I think that is causing the problem there.

One other tip - although boolean union will often work for this, it is really not quite designed for this type of operation, it is more intended for objects that intersect each other (by pushing through each other, not just along edges only). If you have a bunch of surfaces that just touch each other edge-to-edge, then you'll get better results from using Edit/Join to bring them together, it is designed for that type of situation.


> Ps Does exist in Moi an automatic function after an auto-trimming of
> several nurbs surfaces for delete all that not define a volume?

No this does not exist in MoI currently. It would be possible to add this function, but I'm not very sure that it is really that useful, since it only functions in this particular case where you have created several surfaces that are extended past each other on all sides. You can get the same results right now by just doing regular trim followed with a join.

Maybe at some point in the future I can add this in as a special case for one of the boolean operations.


> Coons and Gordon surfaces are implicit inside MoI or don't exist?

They don't exist right now, but I do plan on adding these.

- Michael
  Reply Reply More Options
Post Options
Reply as PM Reply as PM
Print Print
Mark as unread Mark as unread
Relationship Relationship
IP Logged

Previous
Next
 From:  Frenchy Pilou (PILOU)
207.40 In reply to 207.39 

Thx
Something curious
Curve / Revolution / Shell / Booelan Diff with a Sphere
Why the shell resulting is "open" on the Diff internal facetts ?

Attachments:

  Reply Reply More Options
Post Options
Reply as PM Reply as PM
Print Print
Mark as unread Mark as unread
Relationship Relationship
IP Logged
 

Reply to All Reply to All

 

 
Show messages:  1-20  21-40  41-60  61-80