Update: In this post, I’m trying to point out that disagreement isn’t a bad thing, as it can lead to discussion. I’m also providing a little history, to show that not everything new is unprecedented. Apparently my sense of parody is a bit obscure, because some folks thought I was picking on Deelip with this post. I think I need to make this even clearer than I thought I made it originally: I like and respect Deelip, and my ribbing him is entirely good natured.
A few days ago, after I poked Deelip Menezes pretty good in a comment to one of his blog articles, he wrote me, saying “people like you and me were designed to disagree. However, I don’t think either of us wants to entertain the world when doing so.”
I responded, “the fact that we disagree is not a bad thing. It gives a basis for discussion. And, if it entertains the world — that’s OK.”
It all comes down to motive. I think Deelip’s motives are generally pretty good. He’s an enthusiastic developer of CAD utility software, who has an interesting, and occasionally fresh, perspective on things. (Not to say that he doesn’t get things wrong sometimes. Still, he’s got a lot of catching up to match my record in that realm.)
Today, I was reading his post, How SpaceClaim Did It Differently. I can’t resist adding some perspective, and disagreement, of my own.
Fourteen years ago, I wrote about Trispectives (a new CAD product) for CADalyst magazine, saying it could end up being “the Pro/E killer.”
The people at Trispectives liked what I said so much that they quoted me in their advertising, and on their product packaging. Years later, I learned that Buzz Kross, the VP of Autodesk’s Mechanical CAD division, saw that quotation, and came to the conclusion that I was nuts. Or an idiot. Or something similar.
Buzz had a good point, since Trispectives ended up killing nothing, especially Pro/E. Over the course of years, as we got a chance to talk every year at COFES, I think Buzz came to understand that I wasn’t an idiot. And, I got to explain that the context of the quote was important: I wasn’t suggesting that Trispectives was going to beat Pro/E in a feature-by-feature faceoff, but that it brought a new approach to CAD that made it usable for people who weren’t gurus.
Rather than using the parametric feature-based modeling process typical of the day, Trispectives let users model by starting with basic building blocks, then push, pull, drag, and drop them into the final form. This process involves directly selecting and manipulating features in a model, and has sometimes been called “direct modeling.” I prefer the term “direct feature modeling”, to distinguish from direct face modeling, a somewhat different technology.
Trispectives was the first mechanical CAD product to implement direct feature modeling. It was rolled out to the CAD industry at the Daratech CAD conference in 1995–the same conference where SolidWorks was unveiled. (Charles Foundyller of Daratech deserves credit for the tremendous impact his conferences had on the growth of the CAD industry in the 1990’s.)
I remember sitting and chatting with Mike Payne, of SolidWorks, at that conference. Mike was quite familiar with Trispectives, but was of the opinion that the market really needed (or wanted) a product that, rather than redefining how CAD worked, simply built on the parametric feature-based modeling method used by the standard of the day: Pro/E. Given that Mike was also one of the founders of PTC, and had been responsible for much of the development of Pro/E, his opinion was worth listening to with care — though, at the time, I thought he was just biased, because of his background.
Turns out, of course, Mike was right. SolidWorks was a big hit. It was the first of a new wave of mainstream CAD products, which included (after a false start) Solid Edge, and (after rushing to avoid missing the wave) Inventor.
Unfortunately, Trispectives was a market failure. It never delivered on the initial spark of promise it showed. Possibly because the time was not yet right, but more likely, because its technology wasn’t yet ready for prime time. Still, just because it failed commercially, doesn’t mean it had no impact on the market.
Today, SpaceClaim is a shining example of what can be done with direct feature modeling. It was born out of a deep understanding of Pro/E and the other mainstream CAD products, and a desire to build a CAD system to overcome the limitations inherent in the parametric feature-based modeling approach which those systems used.
It could be a coincidence, but it turns out that Mike Payne, the same person I chatted with about Trispectives at the Daratech conference 14 years ago, was the founding CEO of SpaceClaim.
SpaceClaim is by no means the only company to recognize that, possibly, Trispectives was onto something way back when.
Siemens PLM has Synchronous Technology, which is the engine that enables direct feature modeling in both their Solid Edge and NX products. CoCreate, now a part of PTC, never jumped on the parametric feature-based modeling bandwagon, but over time added direct feature modeling capabilities to their eponymous CAD product. Kubotek pioneered direct feature modeling of dumb (e.g., IGES or STEP) models through feature inference.
Deelip, in his blog post today, said:
“[T]hey [other CAD vendors] have been offering direct modeling much before SpaceClaim, but what they did not offer (and some still don’t) is the dynamic push-pull user interface that make SpaceClaim so easy and intuitive to use.”
Deelip is right, with respect to some vendors, but, in general, he’s wrong, as the example of Trispectives shows.
(Did you hear that Deelip? You’re Wrong!!!)
It’s OK… I’m only picking on Deelip to provide some entertainment for the CAD world. I’m certain he’ll return the favor.
Since I’m at it, let me pull a couple of more excerpts from Deelip’s article:
[I]f another MCAD vendor whose solution is built upon ACIS (for example Kubotek) wishes to add push-pull direct modeling capabilities to their software, they need to spend some more “quality” time with Spatial and learn some new moves themselves. . . .
I believe what SpaceClaim did was really commendable. Its founders had a vision of an easy to use MCAD software and challenged the kernel developers to do things differently. Spatial took up the challenge and delivered. There is a word for that. Its called innovation and it needs to be appreciated. . . .
SpaceClaim may have got some recognition for its innovation. However, I am not sure Spatial ever got any. For whatever its worth, I dedicate this post to Spatial and its brilliant programmers that helped make SpaceClaim happen.
Deelip has some good points here. The programmers at Spatial do deserve recognition for their work.
While there are many who are due credit, one name comes immediately to mind–Spatial Technical Fellow, principal author of the ACIS Space Warping patent, and the leader of the team that developed the core ACIS technology that helped make SpaceClaim happen: Dr. Paul Stallings.
As to Deelip’s point about MCAD vendors (for example Kubotek) needing to spend some more quality time with Spatial: I think it’s a great idea to talk to the people who are actually responsible for developing a technology.
I’ve had some very interesting conversations with Dr. Stallings at COFES. He has tremendous technical insight into what it takes to do push-pull direct modeling–from his experience developing deformable modeling as a team leader Spatial, to his pioneering work in feature inference as VP of Development at Kubotek.
Oops. Well, maybe Kubotek already took Deelip’s advice. But, instead of just talking to one of the brilliant programmers at Spatial, they hired him.
(Incidentally, Kubotek KeyCreator doesn’t support SpaceClaim-style push-pull direct modeling–but it’s not for lack of knowledge of how to do so. Keycreator uses direct modification of inferred parameters. This method seems to work better for Kubotek’s customer base, who often use KeyCreator for modifying or repairing third-party models.)
