Once again Mr Quigley - I agree 100% with you !

well not quite ... actually the M and N's don't have to match exactly

if there is a gap (within an allowed tolerance) the surface is formed half way between

that's how I see it here anyway

I have never understood why Tim will not see that a "Match Points" tool is required for just such intersects. Match point would ensure an intersect and hold a constraint there, so the 2 points cannot fall apart when one line is edited.

Perhaps Tim sees this as part of some possible future work on 3D constraints - I'm not sure

Meanwhile sadly surfacing in Shark is not making the strides forward that solids have in recent years

Thats a good video but it also highlights one of the issues. In all the videos I see for surfacing the spline by points tool is used, yet when designing many designers prefer the spline by control points - it is far easier to get smoother curves this way.

When using the CV spline the only points you can ensure that touch the perpendicular cuves are the end points, yet it is the points in the curves within the surface that you often need to control.

It is great that the tool has this flexibility, but that is really only useful for demos and wild surfaces. Most of the time you are designing to specific requirements - to match existing surfaces or to control dimensions. What we really need is a constraint that applies to two crossing M and N curves to force an intersection - and ideally to allow that intersection point to be locked to a specific plane. Only when we have this will we be able to fully control the resultant surface with precision.

Ernesto - are you talking about the net surface tool - or another surface tool ?

Ernesto, your observations point to you trying to make net surfaces from curves with gaps beyond tolerance, not that there's no tolerance.

I think you're confusing the issues. If the snap doesn't work and your curves' gaps are within tolerance, you should get a result anyway.

Hello Tim,
so you suggest to loosen the precision for surface modeling and subsequent Solid generation? Doesnt this imply problems to? I havent tried but I think not all solid operations are possible when I build an open solid.
Would be really a great benefit if control points of curves could be "nailed" to edges/curves etc......

Here's more info on the net surface.

http://doc.spatial.com/index.php/Net_Surfaces

The only tolerance that we updated is the maximum distance used to govern the interpolation process used to create the NET surface if the curves do not intersect.

Tim