Gal Raz on why Dassault had to kill Solidworks

Gal Raz is a blogger in Israel who has written insightfully on why Dassault Systemes had to kill of Solidworks. For 15 years he was a sales rep for Solidworks Solutions and a CSWP, and is now a sales manager for a local Autodesk VAR.

Here, I've condensed his reasons, which he published in May; read the full explanation in two parts at his site displaystatesen.blogspot.com:

• Part I: SolidWorks's Dilemma - How to kill a successful product?
  
  
• Part II: No shortage of challenges to SolidWorks for the coming year
  
  

Solidworks's Dilemma: How to Kill a Successful Product

1. As a new startup in 1995, SolidWorks derailed PTC Pro/E by being the first MCAD system to run on Windows only. Today, SolidWorks doesn't want to be derailed by someone else.

2a. To grow quickly, SolidWorks licensed technology, rather than build in-house. Today, however, it is the only major MCAD system to license a kernel. (Inventor, ProE/Creo, Catia, and Solid Edge/NX all use in-house kernels.)

2b. Because Siemens PLM blocked SynchTech to Parasolid licensees, Solidworks does not have easy access to direct editing technology.

3. Dassault worries about its position, with PTC successful in PLM, and Autodesk successful at everything else.

4. To resolve the file (and other) conflicts with Catia, Solidworks needs to deploy on Dassault's CGM kernel. Existing Solidworks can't do this, and so a new Solidworks based on CGM needs to be written.

5. Solidworks is today the oldest MCAD package. After 17 years, it needs to be rewritten from scratch to use today's programming techniques.

6a. The personnel changes at Solidworks are to ensure that Dassault's dictates will be followed.

6b. To focus on V6, development of Solidworks effectively stopped with 2007.

7. Solidworks is the only major MCAD company failing to invest in new external technology through significant acquisitions (cf. T-splines), because Dassault feels it can provide all needed technology. 

In his concluding remarks, Mr Raz asks:

[Customers] are expected to get new software with a new interface, different file types and more. What will happen to all the applications, macro, patterns, and engineering information [with the switch to V6]? What happens to customers who invest now in Epdm [Enterprise PDM] for their engineering information management [with the switch to Enovia]?

My Conclusion

PTC last year had bumps in the switchover from Pro/E to Creo, mainly because of overhyping a replacement CAD package that won't be up-to-snuff feature-wise for three releases. Solidworks has the same problem, but for the opposite reason: underhyping by failing to make clear a roadmap it owes its one-million-plus users.

 

Comments

[Normand C.]

"Inventor, ProE/Creo, Catia, and Solid Edge/NX all use in-house kernels."

Wait - doesn't Inventor use ACIS which is from Spatial?

Interesting read.

[Jon Banquer]

SolidWorks Corp. should have dumped Siemens Parasolid and the Siemens D-Cubed 2D and 3D DCM constraint managers many years ago. If the old guard resisted then Dassault should have acted to remove them much sooner. By not acting sooner, Dassault has caused themselves a ton of problems that aren't going to go away anytime soon.

Jon Banquer
San Diego, CA
http://cadcamtechnologyleaders.blogspot.com/

[Jon Banquer]

"Wait - doesn't Inventor use ACIS which is from Spatial?"

Autodesk purchased the source code to the ACIS kernel years go and now call it Shape Manager. Back then Autodesk's Buzz Kross made a ton of promises about better hybrid surface and solid tools for Inventor that have never come to fruition. Sadly, no one in the CADCAM press remembers all the promises of better hybrid modeling tools that Buzz Kross claimed were going to happen because of Autodesk exercising their option to purchase the source code to ACIS.

[Jeroen Buring]

Interesting post,I know some of you won't like this comment, but should we not start to judge products once they are on the market? Just a though, what if the new product no longer needs those macro's on the longterm since it has the functionalty as standard with the performance that every wants....I personally would prefer a bit more conversations about the needs and trends from users for the future. If invention is the way apple started to change the world in mobile devices, why should a very financially stable company with many smart people not invest in something on top of what they have, to address needs of the next decade. That is the inner DNA of SolidWorks.... and by the way also from DS. After all SWX and DS where family since years. I like your blog a lot Ralph, nice and critical, mostly on products that exist, and you always seem to take the bla bla out of the PPT's from Vendors. I like that, please stay on that track... Hope you have no plans to go and retire yet!

[Ralph Grabowski]

We cannot judge a product not yet on the market, and in this you are half-right. For we can learn from history what v1.0 of new CAD software will be like -- incomplete. And so there is v1.1, v2.0, and so on. Indeed, Dassault recently shipped V6 R2013.

Macros and other forms of programmable customization let users fill the holes left by CAD vendors. Sure, it is possible that SolidWorks V6 might have features that eliminate the functions provided by some macros -- but this is unlikely.

You know, back in in 1986 WordPerfect had a wonderful macro facility that let me easily create simple functions that are still not available in Word 2010, trivial functions that I as a writer find useful, like switching the order of two words with a single keystroke.

While CAD vendors like to say they want to hear the needs of users, this is a fantasy driven by marketing departments. Reality is that CAD vendors will always be limited by their self-imposed hierarchy of what's possible, given the real-world constraints of time, programming manpower, and size of code. Even crowd-sourcing cannot solve a problem so massive; only individuals can through macros.

[Gsxr1srad]

1&3 are the same,"SolidWorks doesn't want to be derailed by someone else." "Dassault worries about its position" This means they are a proactive company, not really a negative.

2a. In house kernel? they have to do it at some point, why not now? tehy seem to be taking their time and giving a huge amount of heads up to their customer of a change even if its not as much info as they would like.

2b. Your probably correct about direct modeling but nobody has really cornered the market yet and it is not really part of SWX core constituency anyway yet.

3. PTC's "Success" is debatable and Autodesk is good at everything else accept advanced stable solid modeling.

4. to make leaps and bounds improvements in modeling, you are correct, "a new SWX based on CGM needs to be written." I would go futher that they need a V7 with full parallel processing support.

5. almost all CAD package programming needs to be rewritten to fully embrace parallel processing. I hope SWX takes advantage of this opportunity. That will be my test of their innovation.

6a. There is nothing wrong or concerning about having a common vision throughout a company. compare this to all the different divisions and packages at Autodesk where nothing really works well together except for their vault which is anemic compared to any PLM system.

6b. This isnt a reason why they killed SWX and isnt really true. their have been huge improvements since 2k7, many of which their customer are very happy with.

7. other have sufficiently commented on this.

I don't mean to knock your article and you aren't blatantly negative towards SWX so i appreciate your relatively unbiased opinion. These are just some counterpoints that i thought your readers could use.

[Gal Raz]

I would like to thank Ralf for the exposure and to thank you all for the comments.
Gsxr1srad and Jeroen Buring, I read again my posts just to see if I am misunderstood or, is it my poor English.
To make it clear, My post is not about how good our bad will be SWX V6 (I don't have a clue) , It is all about WHY they are developing new product and WHAT can get wrong for the users.
Gal

[Gsxr1srad]

FYI, an interesting, more positive article on what could go right with the future of SWX.

http://www.3dcadtips.com/hitting-the-reset-button-building-the-next-generation-of-solidworks/

[Jon Banquer]

I don't think that most users of history based modelers have a good understanding of how powerful parametric direct modeling can be because most direct modelers are marketed very poorly. Like history based modelers, parametric direct modelers also need much more powerful constraint solvers.

I think that no matter what happens if more powerful constraint solvers can be developed (a good deal of this works seems to be done in Russia) then the total market dominance that SolidWorks now has is going to lesson. This video really gives someone an idea of just how powerful parametric direct modeling can be:

http://www.youtube.com/watch?v=1lCI8EWwM48

Someone commented below the video that it's one of PTC's best kept secrets. I'd certainly agree with that comment.

[Randy Tolbert]

One question no one is asking or answering is will Solidworks V6 finally read Catia files that have been hinted at but posponed for 10 years or more.

[Ralph Grabowski]

My understanding is that Solidworks V6 will read Catia new-V5 and V6 files, because it will use the same CGM kernel, and because data will reside in the same Envoia V6 database.

[Ryan]

I'm not sure what the advantage of reading CATIA V4-6 files is for the majority of SolidWorks users are? And more importantly, from DS point of view, why would you create a software product that would canibailze your sales of higher margin product (CATIA)?

How many of you work with those files, anyway?

[Ralph Grabowski]

The advantage for Dassault might be to sell more copies of SolidWorks into its corporate accounts, as a kind of Catia Lite.

[Gal]

Ralph, SW to/ from Catia is not a technical case . Inventor, SE and Creo can do it for years.
That topic is on the top enhancement request list for several years.

[Gal]

...and the majority CAD users will prefer SolidWorks on steroids rather than Catia Light :-)

[Jon Banquer]

The sad fact is the majority of SolidWorks users are still happy and would prefer as few changes as possible in SolidWorks. It's advanced SolidWorks users that aren't happy. They have every right not to be happy.

Jon Banquer
CADCAM Technology Leaders group on LinkedIn