While this is great news, I’m confused. 
The two most important tools for NURBS modeling, “Match surface” and “Blend surface”, are being neglected for years. Lets start new Serengeti topics named:
“Rhino 9 Feature: MatchSrf re-implemented”
“Rhino 9 Feature: BlendSrf re-implemented”
I was thinking “which one of you numbnuts confuse fillsrf and filletsrf?”
But then it’s better not to delve into it just like we have warning labels on everything there is a reason for it😑
Patch it is… Retains what everyone already knows anyways everyone happy.
Patch has been neglected for so long now that it’s useful it’s gona leave a bad after taste for a while
“-Just p batrgh parfch🤢 …PATCH it and everything will be fine almost automagically~”
Many important surface modeling commands need to be rewritten; those two commands you mentioned are essential.
Let’s hope for Rhino 10, several years from now.
@menno i must highly object.
in its current state i dont see FillSrf to be a replacement for Patch at all not even close. FillSrf does not fulfil any of the initial abilities of patch besides patching holes under certain boundary circumstances.
Patch is much more than just a “fill” tool. the ability to create surfaces from many input data is absolute crucial for many surface creation methodologies, like creating surfaces from points, clouds etc without the need of boundary surfaces. i would rather have patch work as is and the fitting algorithm updated to reflect proper continuity demands and not just eliminating most of what patch actually constitutes instead.
here the screen shot of the Command, none of which the current FillSrf is capable of.
renaming the patch als PatchOld and not even having it autocomplete is not only bothersome but indicates that this command is seen unnecessary and might even be left out completely at some point. that would be a huge deal breaker.
edit:
hmm sorry maybe scratch that, quickly testing out it actually seems to work but the output is just way too different.. i will report back later when i had time to test that.
one small but very key difference to anyone testing this-
oldpatch almost never joined to the surfaces around it unless you went crazy with the span counts.
patch (newpatch) now almost always does in testing here.
simple things like capping a cylinder with a patch and a point (history enabled for iteration) do join and maintain continuity, where in the past using oldpatch you had to crank the isos to crazytown or do a pipe trim/ blend srf transition. I use this technique a lot for buttons and the like, Patch(new) works like a charm and can achieve G2, which was not something oldpatch could do.
see simple example below
also of note, the continuity is much better
Any way to have a command line option “CapCirclesWithHemispheres=Yes”? Similar output like when one builds a pipe with a spherical end cap.
I can see this further expanded with a point near the circle’s/cylinder’s end (just like on your image above), in order to control the flatness of the hemisphere. Should be History-enabled, too.
Of course, this will only work when the input geometry is an exact cylinder or a round hole. It will produce a clean, easy to edit output surface with the minimum possible control points. This way, the tangency deviation should be 0,0000000 degrees.
Patch with a hemisphere cap.3dm (1.2 MB)
What Bobi says.
Anyways, good job on the new, improved patch!
Does it have a “MaintainHeight=Yes” option to keep the top shape parallel to the input edge (useful for buttons or letters)? 
What that shape will looks like when the input geometry includes a horizontal curve instead of a point to define the top of the button?
no, but I’d add a guide curve or two and set them to I to make that happen.
did the old one either just now you need points additionally to (round-)cap non circular boundaries, while in the old patch you didnt.
So this new Patch command will be available in the next build?
already there:
current: _fillSrf (and the old _patch as it is)
next: _patch (the former _fillSrf) _patchOld (as it was)
it looks like the new patch does want a point as a 2nd reference atm for this type of cap-
for fast caps it would be interesting not having to add additionally geometry. i am not going to die over it, certainly not looking at the advantages we have now, but still hoping!
@menno would have to chime in about the tech limitations or requirements for this type of surface, but I agree, it would be nice to have a 1 click solution for this, especially for situations like Bobi mentioned capping text, etc..
Yes please. A ‘patch fast’ command would be great. Where you select object with non-planar close loop holes and just caps them however it can.
Maybe a scripted version that can apply trimmed planar caps to planar holes and patches no non-planar holes would be nice too.
Thx!
G
looks like the special case “vertical walls” is not handled without additional geometry…
as soon as there is some tapered direction it works without…
@menno for the left shape i get a crash or at least a unresponsive rhino for a few minutes on my
test_fillsrf_cap.3dm (481.9 KB)
macbook
How the new Patch command handles filling the open top of vertical extrusions such like the letters A, R, F, B if using some funny rounded font?
@menno here some of my concerns:
Patch as seen in the images below on the left works fast and without messing around, results in a very symmetric surface and UV. the outcome of FillSrf on the right is pulled towards one side resulting in a non symmetric surface and very wild UV, i could not care less about the UV in a patch honestly but it should be symmetric at least, and still the surface symmetry is much more important anyway. i fiddled around for minutes but could not get anyhow close to the left result, rotating the uv made it slightly more symmetric but worse in general bulging out some areas.
here one more, on the left Patch worked without any further guide curves points etc, while we know that and hope that this could change, still the example on the right with FillSrf does not even manage to accomplish it with 2 extra points and one guide curve. with a lot of fiddling i managed to set it to something that could work but it only wraps half of the geometry.
FillSrf:Patch.3dm (287.7 KB)
