Hey Michael
Yup that looks pretty unchoked to me.
I'll be glad to help any way I can and I promise to try to choke it again.....hehehe.
Thnks for your efforts and promptness

Hi niko, I've just posted a new v2 beta on the download page which has this blend fix in it, please give it a try and let me know if you see any problems.

Thanks,
- Michael

10.1 MB installation file. 9.6 seconds to run the install on a P4.

Simply amazing!

Worked beautifully on the choked file over here!!!

Hi niko, also the bug you reported back here:
http://moi3d.com/forum/index.php?webtag=MOI&msg=3344.1

Where construction line snaps were not being obeyed in all cases when using the Move command should be fixed in this new v2 beta as well.

- Michael

Two thumbs up Michael.
Both of the bugs I reported are fixed, and I haven't been able to brake MOI yet.
I've attached another difficult blend in which MOI proved to be a troop. Awesome work man!!!!! Keep it up..

Users can add as many profiles as they wish, so Rhino has the tools to avoid self intersections. But it doesn't make any guesses for you... Anyway I rarely use Blendsrf if I want a precise blend, then I would use networksrf because results are much more predictable and allows fitting within tolerance... The same I think is true in MoI that Blend is good for quickly sketching but usually a poor choice for modeling to spec within acceptable deviation...

jonah

Hi Jonah,

Network requires a lot more setup work than Blend does, because you have to create all the curves you want to use as a separate step. If you have something where you want the created surface to have a kind of even proportion between the edges then Blend is the right tool to use to get the job done more quickly.

Why would you think that Blend does not fit within tolerance? In MoI the Blend surface is fit to a tolerance of 0.001 units so that it can be joined to the other surfaces you are blending between.

In Rhino it should also be getting fit to the model tolerance as well, if you have run into cases where using a Blend in Rhino did not produce a surface within tolerance, then that's just a fitting bug in Rhino that you should report to them so that it can get fixed.

There are lots of ways that a fitting algorithm can get fooled into thinking things are within tolerance when they actually aren't, I fixed some bugs in this area for MoI v2 a couple of betas ago.

There is no reason to think that you can't use Blend to produce results within an acceptable deviation, it should produce the same accuracy as any mechanism that involves an adaptive fitting process like Sweep for example.

- Michael

Hi Michael. Regarding fitting, I was only referring to Rhino's lack of user defined tolerance within the command. I've never seen it actually produce a result which didn't match at the edges. But it's unlike NetworkSrf which gives you tolerance options and can make a big difference in resulting surface match and density... About deviation, just speaking of matching shapes to required positions and angles according to design specs. If you need to precisely fit the interior portions of the surface to existing geometry, then Blend is not the best tool...

jonah

Hi jonah, well most of the time Network is composed of a lot of bendy curves and is not particularly a "match to dimensioned spec" type command either.

> But it's unlike NetworkSrf which gives you tolerance options
> and can make a big difference in resulting surface match
> and density...

Other commands in Rhino often do not have a separately adjustable rail fitting tolerance as well for example Sweep1 and Sweep2. That does not mean that you should use Network instead of Sweep when Sweep would be easier.

In Rhino it is generally intended that things should use the overall model tolerance instead of each command trying to have its own individual one.

In the case of NetworkSrf, the reason why Rhino has it separate there is not actually to increase accuracy but instead to decrease it. Network by its nature has a tendency to kind of explode with a lot of control points and decreasing accuracy helps to lighten the control point density.

In MoI v3 I'll probably be doing some refinements on MoI's curve fitting process to try and produce some lighter results automatically without needing to tweak tolerances.

It's pretty nice if things can just work well without forcing people to have to fiddle with various semi-arcane tolerance settings.

- Michael

I have to agree with Michael on that one.