SolidWorks V6 Eviscerated
It looks like the plan to replace the SolidWorks software with a cloud based fiasco have been gutted. They are backing off, and the V6 delivery in 2013 will be scaled back to just a concept modeler. This all jives with what we’ve seen in the 2013 software, and a sort of renewed interest in their old product.
Over the last month or so, I’ve talked with several individuals you wouldn’t expect who have said vaguely encouraging things about this blog and its effect on the software. With this turn around, I’m now seeing why. When you see 2013 beta, I think you’ll agree that this blog has had some effect, or at least been part of whatever caused the effect. The fact that SolidWorks has been willing to engage me at all shows some change in heart somewhere. Users might still have a voice in the tools they use. Of course Autodesk getting their ass handed to them in the stock market because of customer reactions to cloud announcements might also have had something to do with it.
No matter what spin comes out of this later, regardless of the revisionist history that you’re bound to hear next year, this is a reversal. Make no mistake. This is what they should have been developing to begin with. Start with a small cloud tool to test its viability. Something other than n!Fuze. Years ago with Mark Biasotti we all talked about concept modelers. Mark was on the V6 skunkworks team, and this is something useful that has come out of it. Finally. Maybe we should all take a month off and congratulate ourselves on helping SW avoid a Really Big Mistake.
Well, the conic in 2013 is implemented without the ability to make anything curvature continuous to it. Lots of excuses given, but in the end, tangency really isn’t good enough, and why would you implement a conic without c2? Ok, so simultaneously a bit of excitement and a bit of disappointment. I guess if we have to stop complaining about the cloud, we’re going to have to go back to complaining about bugs and half-assed features. Might as well start here. The cycle goes like this:
-implement half-assed feature
-give bad excuses why it can’t be done correctly (none of which include the truth – time)
-couple years later it shows up fixed
-no mention of finding Merlin’s Wand
Yes, this is a prediction. Conic will wind up with the ability to have c2 connections to other sketch entities without rewriting the mathematics of euclidean geometry. The argument, so you know, is that the curvature of the conic is controlled by point placement and a rho value. If you have a conic c2 to items on both ends, it is overconstrained. Well, yes, if you assume that the other items have different curvature at their ends and that they are driving the curvature of the conic. But you could use the conic to drive the other items, which from a user point of view would be perfectly acceptable, as long as the interface isn’t wonky. Someone on the forum may have suggested that it’s “mathematically impossible”, which I’m quite sure its not. In fairness, partial ellipse and parabola also cannot use c2 (in the SolidWorks interface), but this is another fault, not a reason to prevent a conic from using c2. But the arc can have c2 to different items on either end, and it has the same limitations as the conic, ellipse and parabola: curvature at one end is linked to curvature at other end. Because the arc can do it, so (theoretically) can the ellipse, parabola, and ultimately conic. Now it could be that this is a limitation in the Dcubed solver, which is a different issue, and far, far from making it a mathematical impossibility. If people want a conic to make smooth shapes, and the c2 constraint is the way to get the smoothest shapes, not allowing c2 on a conic is ridiculous. This was literally the first thing I tested when I installed 2013 beta. It set a negative tone right off the bat that took a while to overcome.
Reverse compatibility of file swill only bring in the dead geometry?? That is just as compatible as save as STEP. Totally useless. I want all of the sketches, relations, dimensions, all features of all kinds brought in and have them live for changes.
@Jeff Mowry
Jeff it sounds like a Macro that coverts the file into a dumbsolid upon Open. I wonder how Configurations are handled?
From the Newton article:
“Users of SolidWorks 2012 (Service Pack 5) will be able to directly open SolidWorks 2013 files. 2012 users won’t be able to edit dimension-driven features of 2013 models. They will be able to use 2013 models in assemblies and as reference for creating models of molds, fixtures, and other manufacturing tools. SolidWorks plans to maintain this backward –compatibility for one release in the future.”
Sounds like this “backward compatibility” is really turning the SolidWorks-formatted geometry into a dumb solid to use as reference—but not editing/tweaking—uses. One could use a parasolid for the same purpose. I guess the slight advantage here is that if you have someone exactly one version ahead of what you use—and you’re already on SP 5 of the previous version—then you don’t have to go back to the source to ask for the file in parasolid format. (Hardly true “backwards compatibility” folks.)
From Al:
“Speculation amongst those same highly connected users that built the company took this relatively simple message, span it, misconstrued and built up a resistance to a future product that hasn’t even been seen since – and has since been rethought greatly from that original form. The SolidWorks community, wherever it lives (blogs, forums etc) now has a swath of criticism, posturing, denigration and frankly often disturbing commentary on something that simple didn’t exist outside of a development lab. Basically, it went nuts.”
Mea culpa. Maybe I should stop contributing to such catastrophes. Little things like kernel changes have always been taken in stride by users, requiring no more than the average Wednesday to bring up legacy data, so why worry?
http://gfxspeak.com/2012/09/11/full-steam-ahead-for-solidworks/
http://develop3d.com/blog/2012/09/solidworks-2013-what-constitutes-a-good-release
http://www.axiscadsolutions.com/2012/09/solidworks-2013-has-arrived/
So what do we take away from that??
1) v6 is delayed for at least 2 years while they have another go at it? and so we are getting some small harmless easy-to-prepare fingerfood/entree/appetiser next May and apparently even though its a ‘lightweight’ its going to motivate customers to both keep up SW subs indefinitely and buy into this system as well…hmmmm…
2) they are rebranding it as conceptual and complementary which sounds a lot less threatening than saying it is cloud based Catia data to rearrange/permutate and share with *everyone* else and an outright replacement as soon as users are accustomed to it. No killing or pain or insecurity mentioned.. but why would you spend so much effort on a ‘conceptual’ tool and have two kernals in progress indefinitely?…are they really expecting sales to double from the same user base so they can afford twice the staff and in times of a global depression?
3) they are trying hard to make it look like Solidworks is going to be around for a while and also buy some more transition time and credibility by giving 2013 more useful content and feel good hype than recent releases. How good is this release? Man its the best! Roll up, roll up its all cool, no worries…. OK so there is more in it and it might tempt some people who have been sitting out subs to buy again but basically it has to be a necessary sacrifice or a con seen in full context doesn’t it? If SW is going to become so good in itself why would I be interested in buying a second CAD program that only does concept stuff? If Dassault are confident there is still a potential market for SW of another million or two and the brand is strong why set out to replace it or carelessly spook/upset everyone?
I still don’t see how they can honestly reconcile all these intents…it just doesn’t make any sense no matter how its put. WHERE THE HELL ARE WE GOING?? Why are we still going???
Despite everything we are no further ahead in our understanding than we were when the bombshell first landed at SWW. It doesn’t have a name, we dont know what it does, when it will be here, how or if it works or why we need it and yet apparently we luv it. Go figure.
This is still a major management, marketing, PR and technical flop IMHO
@Neil
Neil, who is saying that? Can you give some links?
Looking at some of the news snippets appearing recently in fact they don’t say a concept modellor is coming. They refer to it as a ‘conceptual mechanical design system’. So I think we may well be counting our chickens in thinking the mission has changed. In fact that doesn’t necessarily point to anything ID related. Probably the rewording/reinterpretation is just a way to sidestep referring to the cloud and the kernal change and all the baggage that evokes.
I see various articles actually quite brazenly make out Dassault to be a misunderstood victum in this affair and paints their silence as being avoidance of competitors FUD. I even read they are pressing ahead with ‘the vision they have laid out'(?) and that theirs has been a ‘classy strategy’….
Well I never…I suppose if you have no healthy sense of shame its worth trying it on.
I must have been hallucinating these past 3 years…..
So I am confused now. Is this a genuine revival of Solidworks or a Trojan Horse?
ok,.. let’s simplify this,… her name is “Rho” and she gives the best “G2”! 8^)
Conics are wonderful. This family of second order curves has enough parameters to fit the slope at end points and the rho control sets the squareness of the shoulder. Conics never wiggle as they cannoth have an inflection point. Boundary surfaces work very well with a couple of conic edges. These swoopy curves are controllable and dimension able.
I use the GW3D add in that has conic curves, and an excellent conic surface. The conic surface has many types of control. I usually use Conics for an aircraft fuselage. These sections follow the curve of widest point and the upper profile curve, the initial and final slopes are perpendicular to the right and top planes, the rho or squareness parameter usually has a linear variation, the section is everywhere a conic that is oriented by a spine curve. The surfaces are wonderfully fair. Singular points do something sensible. The fuselage shape is only C1 ath the guide curves, these curves are visually defining and the shapes look good. Since the guide curves are mostly aligned with the flow the aerodynamic properties are excellent. The GW3D conic surfaces never fail so there is less frustration. I now have shapes without tits, curls, butt cracks, hog backs, wrinkles or ripples.
I think that SW2013 Conics will help provide most of these good shape properties. Conics Rule.
What the heck is a conic!? 8^)
I’m no expert but it seems 2013 is one of the best releases (new modeling and design feature wise) since I started using SW in 07. Dare we expect a Conic fillet tool any time before the rapture?
If anyone is interested SW posted a Youtube clip trumpeting the new conic feature here:
http://www.youtube.com/watch?v=0gD_i7k4HYo&feature=em-uploademail
@matt
This is how SpaceClaim began, and now they want to replace 3D CAD systems.
OK, that makes sense, a temporary incomplete product. Well, that falls in line with the Subscription Maintenance business model, right? They can just throw it out there and then get paid to fix it via Subscription Maintenance.
Regarding paragraph two of your reply above, you & I have had very similar experiences in this specific area unfortunately. New Orleans(or was it San Diego?) & San Antonio come to mind 😉
@DevonSowell
Devon, thinking about it, “concept modeler” is just a half-assed CAD product when it comes down to it. I’d be willing to guess that the overall plan has not changed. “Concept modeler” is just an unplanned stop along the way. It’s a spin on an incomplete and delayed product.
Oh, and what would O’Malley have to gain from blocking conic? This whole product for the last several years has been about ego, and he had the power to do that apparently. How long did it take to get centered rectangles in the software? Why do we still not have stuff like tangent splines (like tangent arcs)? Why is the curvature on the ends of splines still messed up? Why has Undercut Detection been wrong since it was added?
This change to a “Concept Modeler” must be a recent change because as recently as April 2012 I heard the V6 CAD PDM plan was full speed ahead.
Why would O’Malley block the Conics feature?
This has to be an expensive lesson for SolidWorks & DASSAULT SYSTEMES, they’ve been working on all this for many years, musta spent millions I’m guessing.
Any new information about the SolidWorks V6 ENOVIA PLM type PDM software application that was being tested & considered? I heard that a demo of this prototype software was presented at the last SolidWorks World by Kurt Lundstedt. The comment was, paraphrasing, “it looks just like Enterprise PDM!” I believe Lou Gallo posted that comment on twitter.
Great post Matt ! A few weeks ago Bertrand Sicot sent a confidential email to the SolidWorks VAR channel to clarify SolidWorks product strategy for the future. Thanks to your post now I understand much more some of the statements contained in it.
It’s good to know that people ar DS SolidWorks Corp love current product as much as we do and want to invest in it further. Let’s gather some grear ideas for the upcoming Top 10 enhancement list.
“But you could use the conic to drive the other items, which from a user point of view would be perfectly acceptable”
For this part,a workaround could be, sketching the driving conic on a 2d sketch,and then sketching the driven splines on a 3d planar sketch.it should be doable in 2013,but it probably tosses some other bugs into play 🙂
Jeff, as usual corporate cowardice has overcome any favours they might have done themselves by fronting their customers with the truth however I would take it the cloud has been abandonned and so has a move to another kernal in Catia lite. Deceiving customers at this stage, if their grand plan remains the same but seriously delayed by rework, would be entirely fatal.
The announcement of a conceptual modellor would seem to be a face saving substitute but its probably not intended to be much more than say a homegrown Tsplines but with some of that slick 3d concept sketching derived from Catia incorporated, or at least that’s what I take from it. Its a way of providing some direct edit play space that can exchange with SW and yet keep SW going as is and the brand up to date with technology. I mean they have to deliver something and pretend it was what they intended all along dont they? Well they dont of course but as a low life corporation full of empty shirts sadly thats about all they are capable of in the crunch. No doubt all their sycophantic dependants will line up to promote it as the real deal too. I hope someone has worked out a good cover for why it took so many years to produce and doesn’t bear any resemblance to what was shown at SWW….
For sure this SW range extender could be useful but I wonder how they will manage to incorporate some non history based stuff directly into SW now because clearly SE has something useful if not winning in that. I guess they are relying on SE remaining machinery focused and therefore not a threat to their own specialty of ID/surfacing. Perhaps Freeze, move face etc is deemed to be good enough? Perhaps they just dont have an answer and are kicking the innovation can down the road. Really I think there is a lot more serious strategic thinking to be done beyond this rescue. If they dont refine their ideas and develop a clear vision of where they are going it may well turn out they have boxed themselves in again in a few short years. The modellor is nice to have but this doesn’t address a jump in technology that direct edit presents.
Lets hope they make a decent job of raising the new conceptual modellor to be part of the SW family despite it being an adoption taken on by parents disappointed after they lost their own child. This is still a difficult situation all round.
i think it’s quite possible that your blog and the many replies here made SWX think: okay, maybe our customers don’t want this cloud thingy. so i thank you for that. you’re the jon stewart of CAD blogs – cutting through the marketing bullcrap released by SWX and the vendors.
“I know that Austin O’Malley was responsible for blocking the conic tool for all those years.”
I don’t know how you know that, nor is it really relevant. But if it’s true, what sort of gain could possibly be realized by actively blocking a feature that users have been begging to have for over a decade? (Seems psychotic.)
I’m reluctant to hope that Catia Lite is a dead concept (for now). Is there really a commitment by D’Assault to continue active development for SolidWorks—as SolidWorks? It just seems like a sudden slam of inertia to overcome after several years of telling users what they really want. (And the nature of cynicism is self-defense.)
Wow, Matt, this is the most positive blog about SolidWorks that you have done in a very long time! It’s too bad SolidWorks keeps some of the same old carrots hanging out there a bit longer. Otherwise, they could have moved on to some other areas of improvement. The 2013 release is definitely an improvement. I really hope they can continue (or pick up) that momentum for the 2014 release. There is so much more that can be improved.
Yeah, I can see that. I’m saying that it doesn’t matter where the problem is. It just has to be fixed. Whether in Dcubed or in SW, it still is going to require someone sitting down and fixing it. I don’t believe it’s impossible.
@matt
I don’t get where you are coming from on this Matt. If there is a deficiency in the DCM this affects constraints. So far nobody has confirmed to me you can apply g2 to a rho conic in any DCubed DCM app. Anyone?
OK so they have given up. Hooray for that.
Of course they aren’t big enough to say that but we will take a win for customers and common sense anyway.
Several hundred wrung out Concord employees and a few useless Dassault suits can thank their lucky stars they still have jobs although I think its obvious one person should go regardless. The cloud was a major mistake and a costly one. In times gone by people had some honour but I suspect today they need to be dragged from the building clinging to a golden ripcord…
I suppose there is a constellation of VARS and 3rd party providers also breathing a large sigh of relief that they will have ongoing jobs. The downside is that this back down lands us with the same flawed business model we had before. Dassault can’t rest here long. A return to where we were is not a solution for tomorrow.
A concept modellor sounds like something useful and a bit of a white flag sweetener. Presumably it is lifted from Catia so it ought to be fairly sorted when it arrives.
I still need convincing though that they are dedicated to existing Solidworks. Even if it is embarrassing I think they need to make a public statement setting out exactly what they about now. There are a whole lot of perturbed customers out there who need reassurance. Regardless of any encouraging noises and carrots dangled I won’t be handing over subs money until I am convinced my existing SW investment will continue as the Dassault midrange solution.
There are also a lot of operational issues to sort out regarding existing SW as far as I am concerned.
We aren’t just going to go back to extended search engines and the half assed features and call it good. Before daily life resumes SW folk need to be refocused on their real mission so I think there is still some talking to be done on this blog about where we need to go now in regard of what customers want/need and what the company will do their best to deliver. Although the cloud has been a battle SW had been drifting off course for some years already. Now is the time to fix that and to move forward with real purpose and effectiveness.
@Kevin Quigley
Kevin,
It doesn’t matter where the problem is. It doesn’t matter what happens or doesn’t happen in any other software. The only thing that matters about SolidWorks conics is that it draws a pretty conic shape, but has limited use where it will be needed. It’s like a bucket of blue paint with nothing to put it on. Unless you design a lot of stuff that only requires a single conic sketch all by its lonesome, maybe tangent to a straight line.
Matt regarding the conics, we discussed this on the forums at the time. In Rhino you can apply g2 and 3 relationships to a conic but there is no way to constrain it. I’d like to know if any other app using D Cubed constraints manager allows G2 onto conics. If it can, then I agree with you that there is no reason why you cannot apply G2. If not, then the issue is with the DCM used by SolidWorks.
Presumably you have tried this already with SolidEdge’s conic curve sketch tool?
I personally think 2013 release is a good one. There are a lot of very nice geometry tools in there that I will use and will save me a lot of time. I found them very robust during testing.
@solidworm
The conic propmgr has a value for the curvature at the “shoulder” point. It doesn’t allow you to place curvature controls like splines or min curvature indicators.
About the newly added conic entity, i haven’t used solidworks 2013 yet. but generally you can compute the curvature at each end of a rho controlled curve with a simple formula, and input those values as the curvature dimension for the two splines connecting to the conic , which can be done with solidworks. i’ll look up the formula if u needed that. but i agree, they could automate this process easily.
//sorry, solidworm, not solidworks :p
Veerryy Interesting……
I’d venture that “conceptual modeler” is the easiest kind of CAD package to create from scratch, because by definition the vendor gets to leave out anything, yet manages to look respectable.
Autodesk is doing the same with Inventor Fusion for Mac. A new conceptual modeler? Great! For the CAD vendor, this is much easier than putting out a rigorous solid modeler along the lines of a SolidWorks or Inventor.
In contrast, consider the grief Autodesk got when it first released AutoCAD for Mac, and users found that a third of the commands were missing (including crucial ones, like defining plot styles), yet the price tag was the same as AutoCAD for Windows.
As for n!confuse, it made sense from the vendor point of view. DSSW probably thought, “Let’s put out a Google-like product that lets our customers get their feet wet in cloud and social-media type things.” Great concept, except that it forgot that engineers are loners by temperament.