Why I Changed Camps: Part 1
I get contacted every now and then by people or companies doing research before buying a new CAD package.They are interested in the story of a prominent user changing camps. Since I took down the Dezignstuff blog, most of the story that took me from SolidWorks posterboy to Solid Edge employee has been lost. I just want to have the real story on the record and available to those who are curious.
Part one of this story is how I came to this decision, and why I made the choice I made looking at the companies involved. Part two will be why I made the choice I made looking at the technical realities. This is of course mainly my opinion, and what I saw from a particular point of view, but I based a huge career change on these opinions.
The first version of SW I saw was SW96. I joined a SW reseller in 1997. As a SW reseller, we competed against Solid Edge from time to time, but were primarily focused on Pro/E and Autocad.
We were winning deals with SolidWorks for several reasons:
- we were energetic and eager to please
- the SW software was affordable, easy to use, relatively powerful, and run by people who understood the engineering process
- The PC revolution was finally ready to run technical engineering applications with the advent of Windows NT
- the Pro/E salesmen had a reputation for being ruthless bullies, and they scared a lot of business our way
This is really just to point out that we won because of a combination of business and technical reasons, plus having the right idea at the right time. In 1997 PTC seemed like an insurmountable giant to slay. And it took some time, but we did it.
The sale of SW to Dassault in 1997 didn’t affect us at all. Eventually there was a new logo, but there was no other change that the customers could see. Competitors tried to make it an issue, but it wasn’t – yet.
The SW company and its employees were likeable. There was a lot of interaction between development and customers. It retained the air of a company lead by engineers for engineers.
After an 8 year stint at resellers I went independent, doing product design and surfacing, and writing the SolidWorks Bible (2007 – 2013). In 2007 I also wrote some training classes for SolidWorks Corp (Advanced Parts, Advanced Surfacing, and Mold Design). I started blogging for SW in 2007, no pay, just for the fun of it.
Starting with the now-infamous demo at SolidWorks World 2010 of “SolidWorks V6”, and the later move from the Concord headquarters to the consolidation with Dassault in the new Waltham campus, it became clear that there was an internal power struggle about the future of the SolidWorks brand. Even more clear was the idea that who ever was now pulling the strings was pretty out of touch with end users.
With the promised introduction of V6, it started to look like the current incarnation of SolidWorks was going to be replaced with a Catia variant. Most of the key players were gone, the company no longer had “that feeling” about it. Officially most industry wonks professed to be confused about what Dassault was planning, but privately they mostly considered it to be obvious: Dassault was rewriting their mid-level CAD offering to incorporate pieces from Catia. This meant that SolidWorks was probably at the end of its run as a product. Die-hard users still meet this idea with disbelief.
The original software was a lightning strike. I didn’t believe that Dassault could duplicate the original magic. They were clearly talking about a new product, and any new product would have to prove itself, and the odds were just against it living up to the original.
The CAD world churns its top tier every 10 -15 years. After 2012, people were looking for what’s next. It just happens that Solid Edge, the old forgotten competitor, had come out with what it was calling a break-through in CAD. Details were thin, and I was a chief skeptic. I had written blog articles about the problems with history-based modeling, or at least as it was implemented in SW. So I knew that history wasn’t perfect, but I wasn’t ready to throw it out for an even older technology in direct editing.
Eventually I was invited to Huntsville to see the Synchronous Technology product. I was the only one of half a dozen SolidWorks bloggers to do so. After all the hype, Synchronous Technology turned out to be some very nice geometry recognition tools on top of direct modeling. Of course, they still had their traditional, or “ordered” mode, but you couldn’t mix the two methods in a single part. After getting the demo, and a bit of training, and asking a lot of questions, I wasn’t impressed. Part of the reason was that I don’t think Solid Edge really knew how to talk about their new baby yet, and part of the reason was that the idea hadn’t fully matured yet.
Two years later, discontent with SolidWorks growing, I took another look at the updated ST3. By this time, I could really identify with what they had created. It was the subtle change of allowing both methods (ordered and Synchronous) within the same document that made all the difference in the world. History-based modeling had some weaknesses, and this new Synchronous method also had some limitations. But when you lined up the two methods, the weaknesses didn’t overlap. This meant that you should be able to devise a way of working that encountered none of the weaknesses of either method.
Suddenly to me, this became a big thing. A big part of my books was to devise clever ways to manage or get around the weaknesses of history-based modeling, and some of the non-history problems with SolidWorks. I started to realize that modeling didn’t have to be this constantly painful process of build-change-repair. You really don’t have to be superstitious about in-context modeling or external references – this may be one of the biggest secrets about Synchronous Technology.
So I had this situation with SolidWorks (the company and the software) that was gradually falling apart. The technology was not aging well, the company was not paying attention to its users, and the product was bloated with useless junk and increasingly buggy. The salad days were clearly over. On the other hand, Solid Edge, a name from years gone by, shows signs of not just life, but brilliance with this combination of old technology and new technology. Works was making worse and worse decisions, and Edge was making better and better decisions.
When you evaluate companies to work with as a partner, one of the things you have to look at is how that company deals with change. The word “disruptive” gets thrown around a lot, but in the end, disruption is not usually a good thing, especially in manufacturing business. If you look at the way Siemens handles large change and new technologies (integrating NX and SDRC, integrating Synchronous Technology), the change happens over time in such a way that you can continue working the way you work.
If you look at the change that Dassault is proposing for its customers with Mechanical Conceptual, it’s a “get off the bus, and get on a different bus” sort of dead-end change. Long term, the Siemens approach is much better for customers. The Dassault approach (in addition to being still entirely imaginary) is simply disruptive. In a bad way.
And when Dassault makes their change, where are they going? Likely to the place where Siemens and Solid Edge have been for nearly seven years already – history/direct 
hybrid.
As a CAD industry commentator, I feel that it’s part of my job to at least have an idea where things are headed. SolidWorks is old news, although there are areas of the customer base that don’t know it yet. SW may still be the fashionable “it” software in some circles, but PTC also went through a long period of denial after losing their crown in the early 2000s. All newly developed CAD (including V6) contain some hybrid of history and direct, and that makes Solid Edge the one with the most mature technology (not to mention the most copied modeling paradigm in 30 years), and SolidWorks in dead last, as far as new technology and the big four mainstream CAD vendors.
Don’t let the cloud fool you. The cloud has nothing to do with CAD. It’s just an IT delivery method. Developing for the cloud does nothing for your CAD needs.
If you had to pick one, which would you pick? PTC? They are as distracted by non-CAD stuff as Dassault. Autodesk? I see them as the biggest competitor, as they are building out a complete vision with top-of-the-line packages in Alias, 3dsMax, Mudbox, Delcam, HSMWorks, T-Splines, Moldflow, and others. Still, Autodesk is building its mechanical side on Inventor, which in my opinion is a third rate product compared to the other players.
When you look at it as a question of which of these 4 implementations of history/direct hybrid will come out on top, (Creo, Fusion, Synchronous, and Dassault’s imaginary mid-range V6), Synchronous Technology is far ahead of the rest. It is now in its 6th production release, and NX shares a lot of the technology. Siemens eats its own dog food, they are a huge user of their own CAD. To me, that says a lot.
So I picked Siemens. I really like the direction of the development of the Solid Edge product. Having seen ST7, I can say it’s full of great stuff that you will use. This wasn’t a choice that was made lightly. I didn’t get a job, then change my mind. This was a realization that took years, and a lot of research into comparing CAD products and companies.
Siemens has great technology just when the market is waking back up, and starting to realize that it’s time to shop for CAD software again. SolidWorks may seem like an insurmountable giant to slay in 2014, but if anyone has the current CAD product to do it, it’s Solid Edge.
Hello Matt,
I am glad to hear you have transitioned to a product that is at least ‘part way’ there – meaning one that understands and recognizes the long standing issues with history based modeling, but does not go all the way to a pure direct, vs. a hybrid approach? I am curious as to why/how you think history methods are still necessary to help the design process, or why direct editing tools alone cannot do the job?
In terms of saying direct editing is old school, the new paradigms for direct geometry creation and editing were pioneered by Kubotek USA’s KeyCreator ~ eight years ago are anything but old – in fact Kubotek introduced these concepts for direct editing, and feature recognition tools well ahead of SolidEdge ST, SpaceClaim, etc. To be fair, only Solid Designer [now Creo Direct] had some similar concepts to the modern version of direct modeling way back when PTC was really gaining traction.
It is direct the modeling component that offers the speed, flexibility and intuitiveness to more efficiently create and manage our designs, not the history component… And a pure direct modeling approach allows for better reuse of data, regardless of the source… Just some thoughts for your consideration. Thank you.
Matt,
I don’t know as much about KeyCreator as I should. I’m not here to promote it.
The reason a hybrid approach is better than a straight direct edit approach is that direct editing has some weaknesses tied to consumption of faces in the brep. Some edits going in one direction may consume a face, and editing in the reverse direction would require a different face type to be created out of thin air. This causes a lot of direct edit type edits to fail, and usually the user has no idea why. In my work, I’ve noticed that this happens most with fillets. If the main body of a part is created as a synchronous body, and the fillets are added as history-dependent features, you avoid fillets causing the brep problems. Plus, it make a whole lot of other things easier, like dimensions or relations to sharp edges.
This is why I was unconvinced when Solid Edge used their initial strategy of parts that were Ordered or Synchronous. Many parts benefit from an Ordered and Synchronous strategy. Combining the methods is stronger than either of the methods on their own.
Matt,
Great to read about your CAD history leading up to being part of the Siemens/Solid Edge team. I feel like we’re in the NFL and we just pick you up on waivers from our biggest rival in the league. You’ve really have brought an amazing perspective to the Solid Edge community with your vast experience with Design, Engineering, and the CAD tools needed to get ‘er done.
Funny I had a similar realization of Synchronous Technology, but from the inside as a Solid Edge user for 13 years. As you were looking “into “SE back in the ST3 release time frame, I was looking outside to replace replace it… I really thought Synch was great in concept, but wasn’t ever going to replace the try and true structure of History based modeling. It had been years since I demoed other CAD products, but with the confusion and frustration of learning NON History modeling, I decided an investment in time needed to be made if I were to continue using any CAD software. So I either had to suck it up and learn Synchronous, or move to SolidWorks or Inventor.
What was most frustrating was trying to see through the sales pitch and see if my over negative reaction to Synch was impulsive. Well ST4 showed up in the nick of time, and it really was a breakthrough release for SE and Synchronous Technology. Dan Staples and the Solid Edge Development team really listened to the users and addressed the main reason users were having such a hard time embracing the Synchronous workflow. Granted most of the resistance came from old time History minded dinosaurs like me, who were stuck on stupid. But ST4 broke through the fuzziness of how Synch and it’s powerful Live Rules wanted to work… well at least for me anyway. And with ST7 coming later this summer, it has gotten more powerful and easier to use with every release since ST4… they’ve just never stopped refining Synchronous and Solid Edge.
So beyond the power of using Synchronous Technology, what makes using Solid Edge exciting is the sense that this is not a flash in the pan. Siemens is focused on CAD tools, and the silly Cloud, which you point out is pure IT crap with the Marketing types and bean counters salivating over the elusive new revenue potential it brings.
Very much looking forward to reading more… Keep up the great work.
Bob
Bob,
Yeah, that’s a great story. I think we still have something to learn about how to talk about ST with people unfamiliar with it. It’s not that complicated, and maybe that is perceived as a a weakness. To be powerful, maybe people are thinking that it has to be complicated like history-based modeling.
Thanks again for the comment.
Matt,
I’m curious why you think Inventor is a 3rd rate mechanical product?
I reviewed ST6 in August for CadDigest and was very impressed. Its certainly very powerful & compelling. However, i didn’t see a vast ocean of difference between the two,like you are claiming. If your comment was based on outright surfacing capabilities, then there is a lot of difference there.
Where Solid Edge lets itself down is in the UI. Its dated in a lot of areas, highly inconsistent & in some cases overly complex.
So I’m surprised that a man of your stature & experience is so dismissive of Inventor. To the point that I’m questioning myself, what am I missing? Or are you being deliberately dismissive of a product which in light of SWx’s problems is the only true competitor of Solid Edges’ in many different areas.
I look forward to reading a sensible & matter of fact response.
cheers
Scott
Scott,
About Inventor, every time I used it, it just seemed to have one layer of functionality, and was obviously copying a competitor. Where other software would have layers of options, Inventor stopped at the first level. Sorry I don’t have detailed examples, but that was my impression of it as I worked through a few things. Nice enough interface, but the beauty was skin deep. I don’t think I mentioned surfacing. In the technical side of the discussion, I’m pretty clear that Synchronous Technology’s huge advantage is in prismatic modeling.
The “vast difference” is in Synchronous. Not worrying about parent/child and not caring about feature order.
Inventor was always a product trying to catch up, and never quite made it. The new stuff with Tsplines is much more interesting, and admittedly deserves a new look.
But here’s the other thing. While you seem like you’re focused on Inventor, I’m really not. I could easily write 20 blog posts without referring to Inventor once. I’m not going to give a detailed critique of Inventor, because I don’t have any need to use it. I may make an exception just out of curiosity about Tsplines. This is really a story about a transition from Works to Edge, and Inventor doesn’t play a role in that, except as one of the available options. Part of my prejudice against Inventor has to do with my history with Mechanical Desktop, which was an awful product by any measure. I used it for a couple of years, and never forgave Autodesk for that experience.
About Solid Edge’s UI, I hear these comments frequently enough they can’t be dismissed, but here’s my take on it. Out of the box, the Solid Edge UI is very optimized for expert users. It’s very sparse in explanatory text, and mostly icon driven. It’s totally efficient. Unless of course you’re not an expert. There are other ways of displaying the UI so that it has more text and is easier for beginners to use. There has been a bit of development around this part of the interface (personas) the last couple of releases because of feedback like yours. Another part of the UI is that Edge is to some extent assuming you are familiar with the workflow. Like anything, the UI does take a little getting familiar, and it does make some different assumptions than some other CAD software I’ve used. Once you’re familiar with it, it’s extremely efficient. I’ll grant you that there are some dialog boxes that have a dated look, and development addresses these as they work on various areas of the software. Function over fashion.
With other software, they go through fads. First, they make the argument that they want to clean up the graphics area, so they move all the dialogs to an uncomfortably narrow area on one side of the screen. Then they call it “heads up display”, and start putting dialog boxes back in the graphics area. Solid Edge is more consistent than that. So if it looks “dated” to you because they aren’t following fickle trends, then so be it.
Matt,
First of all, thanks for approving my comment and your subsequent reply. Nevertheless, I’m disappointed you mainly focussed on my dated comment regarding ST6’s UI, especially by devaluing my polite initial comment by ending your response with it. Obviously just because it’s dated doesn’t make it inefficient. Regarding inconsistency & complexity of the UI, a key driver to the success of a product in the market place is how quickly new users get up to speed. Your response reads of prejudice towards inexperienced users. One of the reason SolidWorks was so successful in the beginning was do to the ease with which new users were able to become familiar with the UI. Solid Edge needs to drastically improve it’s UI to attract a new & larger user base.
Your partial justification for why you have a prejudice against Inventor, just because of your experience with a completely different product is extremely odd & frankly irrational. That would be like me saying I have a prejudice against Solid Edge, because while studying engineering at University I had to use SRDC Ideas. I disliked that product & found a large number of bugs, to the point that my lecturer marked me down on a project because I found a workaround for a bug he acknowledged, it’s just my model browser/history didn’t match the example anymore…. Ridiculous indeed…. but I don’t let it cloud my judgement of products since derived from it. Inventor wasn’t in anyway derived from Mechanical Desktop.
Also your comment about me being primarily focussed on Inventor is a tad insulting, I haven’t disrespected your skills at any point, I don’t know you from a bar of soap. Just like you don’t know me from one either beyond a quick Google search? You were primarily focused on SolidWorks for years & now Solid Edge. Which is fine, that is how you earn your living. I don’t hold that against you and I don’t write off competitors products just because I know Inventor in more detail than I do the others.
Yes, in the past Autodesk have been playing catch up with Inventor. They still are with surfacing. I don’t think they are in a lot of other areas anymore. What was the last version of Inventor you reviewed? I’m glad to see you will have a look at Inventor once T-Splines have been integrated. How long would you typically spend with a product before you are satisfied you know it well enough to credit it or write it off? I believe each of those end results would take different amounts of time, such that I would want to spend more time with a product before I write it off fairly.
Just to end, my initial comment wasn’t posted with the intention of this being an Inventor Vs Solid Edge pi**ing match. This comment isn’t either, it’s more about respecting the opposition for what they are good at, recognising their strengths & weaknesses so you are best placed to win. Just like any top flight athlete does. The impression I got from this post, is you don’t see it that way, hence why I initiated this conversation.
Incidentally I think your reasoning for switching from SWx to Solid Edge is sound. I would like have felt the same way in your situation.
Respectfully
Scott Moyse