XNurbs releases a ground-breaking NURBS software

XNurbs · November 9, 2017, 4:11pm

Hello Rhino users,

By playing the two-minute demo video ( http://www.xnurbs.com/download/XNurbs.mp4 ), you can judge if XNurbs is the most powerful and easiest-to-use NURBS software ever developed. More info is available at http://www.xnurbs.com . XNurbs can solve virtually any NURBS surfaces on the fly regardless of the complexity of constraints.

There are two issues we need some feedback from Rhino community:

We would like also to develop a plugin for Rhino, and need estimate the reasonable pricing for Rhino. (Since the proposed Rhino plugin has basically the same capacity as our SolidWorks plugin, their price should be the same.) Before we start marketing XNurbs, it would be great if you could give us some valuable feedbacks and we may make some adjustment to our arrangement.
A small number of users may get a permanent XNurbs license for free: We will apply for SolidWorks Golden Partner Status, and need a small number of SolidWorks users as our references. If you are a genuine and experienced SolidWorks user and you satisfy XNurbs after your evaluation, please contact us.

(Before posting this message, we have discussed with McNeel. We appreciate the feedback from Rhino community.)

Many thanks.

Kevin
XNurbs

DanBayn · November 9, 2017, 4:25pm

Will there be a demo version we can try?

XNurbs · November 9, 2017, 4:29pm

Yes. Please go to www.xnurbs.com to download the trial or demo version, which is fully functioned.

ec2638 · November 9, 2017, 5:43pm

Looks interesting! We’d use it if it performs in most scenarios.

As deep and as flexible as Rhino is, it needs improved complex modeling tools going forward, IMO.

As for pricing, the more aggressive XNurbs is priced the the easier to swallow. $195 goes down a lot easier than $595.

What I can’t tell you is - whether you’ll sell 4x more seats at at $195 than $595, but obviously there is a relationship.

Edit: and considering the acquisition cost of SW vs Rhino…

XNurbs · November 10, 2017, 12:48pm

@ec2638

Many thanks for your suggestion.

Surprisingly, XNurbs should perform in all scenarios! In fact, both XNurbs and Catia are powered by a technique called variational surfacing. The variational tech powering XNurbs is way ahead of others, and no other system can perform what XNurbs can do.

“As deep and as flexible as Rhino is, it needs improved complex modeling tools…”
As I indicated above, Catia is powered by variational surfacing tech, which delivers some obvious advantages over other methods, e.g., traditional surfacing tech has little chance to compete with the tech in terms of surface quality and the capacity of solving complex model. Without a powerful variational surfacing tech, it would be very difficult for anyone to match Catia’s capacity in surfacing.

“As for pricing, the more aggressive XNurbs is priced the the easier to swallow. $195 goes down a lot easier than $595.”
Since no high-end system can perform what XNurbs can do, resellers actually suggested that “several thousands” is a reasonable price for XNurbs (I know such a price will shut down any possibility with Rhino). We don’t mind to lower the price a bit, but we don’t want to give users an impression that XNurbs is a low-end software. (Many cheap surfacing software is powered by low-end tech, and their price is really very low).

JKayten · November 10, 2017, 4:33pm

I’m very impressed with the demo. I’m also grateful for your attitude towards small time designer sorts like me. I’ve been trying to make use of AD-Fusion application as it includes T-Splines (I’m on a mac) but I keep returning to Rhino to get anything done!

gustojunk · November 10, 2017, 4:43pm

Hi Kevin, all

Welcome to the Rhino community. There’s another surfacing plugin coming out called Xirus, did you know that? Xirus.XNurbs.Xirus.XNurbs. Very very confusing. One of you guys should seriously consider changing the name of your Rhino plugin.

A lot of the feedback I would give you is the same we gave to those guys, link here: Xirus for Rhino: Subdivision, NURBS, polygons & augmented CAD

Also useful feedback to another of your competitors, specifically regarding trials: Clayoo - Organic Modeling

What’s working against all of you is the fact that we users know that most plugins are not financially viable, so you might not be around for too long, or after some initial push in development your continuation or progress of the product, or how you support it, might stall. This has happened with many plugins like PowerSolids, Tsplines, VSR surfacing, …and in Clayoo where the jury is still out if it will make it, especially now that they have a new owner that demands bigger a ROI. Peopel will want to know what’s your funding model" do you have research grants? self-funded? VC money?

Regarding price anywhere is the hundreds of dollars makes sense. Your job is to figure out if it’s $200 or $900. For me the mental block happens past the $600 price for 1 license, or $400/license for 4 licences. It depends if we decide to try it on one user or if a few of us get on it.

I see a plugin as disposable, because of the uncertainty that I stated earlier. So if you release something that’s stable. and works in Rhino 6 we have basically a few years of subscription if you decide to stop developing or get acquired by Autodesk or another large software company (which has proven to be about the only viable biz. models for past peers of yours).

I really think that you guys (same goes for @Daniel_Schmitter of Xirus and @Xavi_Marrodan of Clayoo ) should consider giving your software for unrestricted use for 3-6 months. Your biggest threat right now is non-consumption. Your second biggest threat is justifying your price (whatever that price is in each case) reflecting your value. Only usage in production will prove/disprove your value. Look no further than Fusion360: They are giving it out for a song ‘as a trial’ . We all know the $30/month ($15 if you get a special) is not sustainable and prices WILL go up, but that will happen once they have enough users finding true value at using it. So it’s all good… except for the communists here that will crucify me for making such statements, but they are lousy customers anyway.

One more thought: Resellers. What they think is irrelevant for the market. What prices THEY need to exist is also irrelevant for the market. We are in 2017, all software purchases are a few clicks and minutes away. If your stuff is good people will know. If it isn’t resellers won’t help you no matter how hard they push sales channels. In other words if you believe them that your products is worth thousands of dollars you are already screwed.

People here buy on value, reliability, and trust. Technology is not that important.

Last thought: I’ll tell you the same thing I tell every group of math geeks trying to make a product for designers, go hire a designers and let her/him make the examples of what your technology can really do. This particular example you posted looks terrible, even if it’s curvature continuous mathematically, visually is abrupt and very ugly, most likely you don’t see it, because that’s not your expertise:

I hope this helps,

G

PS: bonus points for whoever embraces the Rhino Accounts licensing model. For me at least that wins half of the sale. I’m not sure I will continue to use any other plugins that don’t use their common licensing. Talk to @aj1 about it. He can tell you more.

Toshiaki_Takano · November 10, 2017, 4:56pm

Love to try it! Multiblends just takes up time to patch up.

Is there controls for how the blend goes? Like sharper transition with some edges and others softer?

As for the price, if it can do above and patch up also edges which are next to each other without naked edges, I’d go for $300〜$400. Need to also consider the timing as rhino v6 is coming and many might need to upgrade to that before plugins…

Also depends if you can resell, and the kind of license lifecycle it has. I’d like to use it couple of years before any upgraded versions…

schultzeworks · November 10, 2017, 5:28pm

@gustojunk

I think you’re being too harsh. I have projects where I need a surface continuous to 6 floating planes every day of the week. And, then I hit my head on the desk.

JKayten · November 10, 2017, 5:39pm

That’s exactly the game plan as far as I can see. Fusion is still a clunky but promising application suite and if/when it’s ready for prime time there’ll be an user base with years invested in the program. It’s brilliant, but AD has the financial wherewithal to put out a loss-leader program.

gustojunk · November 10, 2017, 5:58pm

I would never release a model to a client with a blend looking like that. we all have different standards. Yeah mine do make my life a bit more difficult some times. Maybe I’m not their target customer.

G

kiwi · November 10, 2017, 6:53pm

Hi Gusto, What If that blend was a 0.4 mm fillet? Would it be an acceptable surface in you opinion? I’m curious. Thx.

gustojunk · November 10, 2017, 6:58pm

Absolutely. At that scale the Patch tool will be fine too. Even a flat bevel that toolmaker can polish.

OSTexo · November 11, 2017, 11:31pm

Hello,

@XNurbs, you may want to consider developing your own effective UX to showcase your work. If the plugin crashes and burns on Rhino you can always sell on the strength of your usability with other platforms.

Stratosfear · November 12, 2017, 7:55pm

I’m impressed. It does a better job of handling a n-side patch than Fusion 360. It was 0.1163 max off G1.

Lagom · November 12, 2017, 11:31pm

Still a lot. How 'bout G2 deviation?

XNurbs · November 13, 2017, 11:18am

@Toshiaki_Takano

Hello Toshiaki_Takano, all
“Is there controls for how the blend goes? Like sharper transition with some edges and others softer?”

That is very easy, just add some constraints. You can add any constraints (i.e., curves or points) at any places, which control how the blend goes and control the surface shape generated. (With XNurbs, you are free to use any curves and points as constraints, and for each constraint, you can apply Contact, Tangent and Curvature continuity. So you can easily control sharper or softer transition within a specified part of an edge. The surface quality generated by XNurbs is outstanding, and traditional surfacing tech has little chance to compete with it. (Please read my next response to gusto)

You can send me your constraints (i.e., curves and points) in IGES or STEP formats, I can show you how it can be done. (Before you do it, I would suggest that you download the manual from our website so you know how to prepare the constraints. XNurbs is the easiest-to-use software, and its manual only has a couple of pages. There might be a delay for our response: The latest Windows Updates paralyze our computers running Win 10, and we are still trying to clean up the mess.)

Many surface-related discussions in the Forum don’t make sense, and XNurbs may silence many discussions in surfacing.

XNurbs · November 13, 2017, 11:31am

@gustojunk

Hi Gusto,

Xirus and Clayoo are irrelevant to XNurbs. Xirus is based on a new mathematic model (not NURBS) and Clayoo should base on a voxel model, which is unrelated to CAD. Normally, for a start-up company, they don’t have the capacity and experience to work on industrial-proven tech and compete with big CAD companies (For example, Catia spent millions on their variational surfacing, and a start-up company stands no chance to develop something similar to compete with Catia), so a start-up company has to develop something new (i.e., unproven tech or something no one tried before), but most new methods (95%) won’t work or meet designers’ basic requirements. So as you observed, most start-up companies won’t last several years.

You may be surprised to know how XNurbs tech was developed. When I left one of the biggest CAD company in 2007, there is a written agreement regarding to the IP (variational surfacing). In other words, the tech was something great ten years ago and it is not hard for you to figure out how many years we have been worked on it! It is way beyond the period that a start-up company would normally survive! Now for Ph.D-level developers in big companies, our tech is too intimidated to compete with. It solves all kinds of problems for NURBS creation.

"This particular example you posted looks terrible, "
I am afraid that this cannot be true. I think I can explain how Catia dominates the surfacing field: Its variational surfacing tech. As I said above, the surface quality generated by the variational tech is outstanding, and traditional surfacing tech has little chance to compete with it. In plain English, based on energy-minimization method, the tech generates a NURBS surface that is the one among all possible solutions that has the minimum energy, i.e., the smoothest surfaces. For human-being, it is just impossible to find a better solution - This is why the surface quality generated by Catia is good. So I have two questions for your comments: 1. The attached Blending.IGS (21.9 KB)
is the original model. Could you use your method to generate a NURBS surface from the 12 curves and show it to us? (I am just kidding with you, and I know even Catia cannot do it.) 2. As explained above, it is just impossible for human-being to find a better solution generated by variational surfacing, but it appears that you can get a better one. How? You mentioned the “fillet” method, and “ZebraViewCombined.png” compares the surfaces generated by XNurbs and the fillet method (The bottom image, ten fillet surfaces – only high-end CAD systems can handle this kind of fillets.). Do you think the fillet method is good?

Is the attached NewBlending.png better for you? It is pretty much the same NURBS surface showing in your post. I think you might be misled by the basic graphics settings, which has nothing to do with the NURBS quality. Simple changing graphics settings will generate the attached image.

Blending.IGS (21.9 KB)

XNurbs · November 13, 2017, 11:40am

@Stratosfear
@Lagom

Hi Stratosfear,

Thanks. As said in our demo video, a n-side patch is a very simple and trivial application for XNurbs. XNurbs intend to solve all kinds of problems for NURBS creation – more complex than a n-side patch. Please try new surfacing mode: Users do not restrict how a surface can be generated, and are free to specify any curves and points as constraints to generate NURBS surfaces. Once you understand this, then you will see that many surface-related discussions in Rhino Forum don’t make sense.

Need to point two issues:

Performance. The proposed Rhino plugin should be 5x to 10x faster than SolidWorks plugin. (Some SolidWorks API slows down XNurbs a lot - Notice the performance difference between the preview and the final surface.)
The current version is actually powered by our “low-end” kernel, and we will switch to our “high-end” kernel soon, which further improves surface quality, performance and tolerance. (Some picky designers may push us to make the move.)

XNurbs · November 13, 2017, 11:43am

@Lagom

Hi Lagom,

For XNurbs, the position deviation is about 0.001 mm, and the angular tangent (or G1) deviation is set to about 0.1 degree. In this world, there is no acceptable measurement for Curvature deviation. (XNurbs uses surface normal and curvature, which is more advanced than the traditional G1 and G2). So could you tell me what a good surface normal deviation is? We can change one line of code to satisfy your requirements, but others told me that 0.1 degree deviation is good for tangent. Maybe we could just switch to our “high-end” kernel, which will silence any “picky” designers, e.g., Gusto.