DevSrf for Rhino 6

Will DevSrf (or equivalent) be available for Rhino 6?

Hi David, I do not know what the plan is- maybe @lowell knows…?


At least there will be a plugin equivalent to DevSrf available for v6.
I’d like to do some more work in that area, and its just dependent on workload at this point.

1 Like

I second the request for Devsrf for Rhino 6, Its a great plugin and I really miss in in Rhino 6

Have you tried DevLoft which is the V6 replacement for DevLoft? The -DevLoft version provides some options.

Yer I have tried the new Devloft command or more specifically the -devloft command in Rhino 6, I find its a little dangerous as it will loft a developable surface between any set of curves. even ones i know that will be impossible to manufacture. The issue is there is no way to numerically limit the amount of twist tolerance. you kind of have to turn on curvature display and look for the areas that are not green and adjust your input curves accordingly.
I find the Dloft plugin the best with devsrf plugin a close second.

I developed a method for determining “exact” ruling lines for developable surfaces between two arbitrary curves, and use ruling lines found with that method in conjunction with DevLoft.

I use that method to trim a set of input curves to DevLoft and then usually find the results to be very good. Sometimes I’ll add several internal ruling lines using the AddRuling option of the interactive -DevLoft version.

I may start a thread on improving results using DevLoft.

ok cool, very neat workflow, i like it. Well done!
Question… Correct me if im wrong but using your workflow sets up a set of “optional” ruling lines so the surface created is essentially as “developable” as possible given the input curves used?
There is still no way to limit or measure the amount of twist rhino has put into the surface. again you can loft a “developable” surface between any set of input curves, even ones you know to be impossible to manufacture?

I design boats so when making a “developable” surface I have some practical limits i place on the amount of twist i allow in any surface. this limit differs depending on what material is being used. aluminium will twist far less than foam for example.
As far as i can see using your workflow combined with the Devloft command still dosnt allow the user to specify twist tolerances correct ?

I made a py to measure the twist on a ruling line - I think it works - you can give that a try, it might be of some help.

To use the Python script use RunPythonScript, or a macro:

_-RunPythonScript "Full path to py file inside double-quotes" (1.4 KB)


thanks Pascal, that is indeed usefull.

Thread with method I use for checking twist along ruling line with edge curves. This is the thread which lead to Pascal developing the script he posted above. Ruling line from edge curves twist check

ok thanks David

A true developable surface has zero twist everywhere. The ruling line method I linked to above finds ruling lines between edge curves with zero twist along the ruling lines. I use ruling lines from that method in -DevLoft to obtain surfaces which are very close to a true developable surface. An alternative method is to determine a set of ruling lines and then use Sweep2 with the ruling lines as the sections to create the surface. I use the methods I discussed in Verifying developable surfaces to analyze how close the surface is to a true “no twist” developable surface.

Important - there may not be a developable surface which runs the entire lengths of two edge curves no twist ruling lines between the corresponding ends of the edge curves. In fact for an arbitrary pair of edge curve is very likely that there will not be a developable surface with no twist ruling lines between the corresponding ends of the curves. Note that a true developable surface between such a pair of edge curves will have some ruling lines which do not end on one of the original edge curves.

A major shortcoming of DevLoft is that it always produces a surface which extends the entire length of both edge curves with ruling lines between the corresponding ends of the edge curves. The result is usually a surface which is not a true developable surface.

There are several methods to create a true developable surface between an arbitrary pair of edge curves. They are best described with graphics and would probably be more appropriate in a separate thread.

If the ruling lines fall short on one edge curve, what does DevLoft do to extend the surface the full length?

Can it be stated that the rulings between a point (end of fully ruled edge curve) and an arbitrary curve (remaining curve that has no ruling lines) always form a conic section, hence are purely developable?

DevLoft does not extend curves - hence the inherent problem I mentioned in my previous post. When I use DevLoft I either extend the edge curves manually as needed, or more usually find the portions of the edge curves with no twist between them, and then create developable surfaces at either end which are not conic using a method I developed. I’ll try to find time to post that method in a separate thread.

Yes, it will be a conic surface which is developable. However the radius of curvature will usually approach zero(infinite curvature) at the vertex. The high curvature may cause manufacturing problems. The Developable option in Loft in previous versions of Rhino frequently created conic surfaces at the ends. In my opinion other methods should usually be used for filling in the ends.

There’s a Rhino 6 version of DevSrf here:

This is what I see at that page:

Access Denied

You are not authorized to access this page.

I just put it there this afternoon. I think they have to look at it and authorize it when they get to it. I should have waited to say anything. Or maybe I’m completely confused.

You should just poke @fran for some attention :slight_smile:

I successfully downloaded and installed DevSrf for Rhino 6.