At What Point Will SolidWorks Lose Your Confidence? (poll)
Ok, so it seems there are people who don’t see the train coming down the track. There are people who just want to believe whatever they have believed in the past, and for some reason can’t acknowledge that a change is coming, or that that change might not be in their best interest. Which is ok. Whatever.
Here is a list of things that Dassault people have said or done in public, so it’s the closest thing to “facts” that we have:
- Dassault has taken a more obvious role at SolidWorks since about 2010
- “We’re going to drop Parasolid. It’s going to be marginal”
- Dassault is trying to steer their mid-market customers into PLM software based on 3DEXPERIENCE (Catia, V6)
- That change has started with at least one failed product (n!Fuze)
- The change continues with Mechanical Conceptual which is based on Catia
- SolidWorks resellers will be selling Catia based software to their SolidWorks customers by the end of 2013
- There has been an exodus of old SolidWorks employees not thrilled with the changes Dassault has brought
Based on these items, I’m making decisions for the future of my career and my company. I consider everything in this list as fact, based on what people from DS have said in public. I don’t think anyone from DS or SolidWorks would argue these facts (or that some DS decision maker did at least say these things), although I have no doubt they would want to offer some additional spin. If you want to continue to argue about what’s happening or not happening, please refer to this list. If you choose to believe that SolidWorks is going to continue to exist as it has in the past, then that’s your decision.
What I want to know from you, especially those of you who don’t believe your SolidWorks reseller is going to try to sell you a Catia based product in the next year, is at what point will you start to believe that SolidWorks is going to fade away?
[poll id=”13″]
[poll id=”14″]
@Dan Staples
Dan,
I really appreciate your perspective on this issue. I totally agree with you that ST is a step ahead of SW at this time. There is no question that SW employs a solver that wastes time solving geometry that may not be at question relative to the change you just made to the model.
But I think that you are being somewhat self serving in defining CAD as being limited to three dimensions. The CAD that we use today is the platform for many time dependent analysis. If all you are trying to resolve is 3D CNC machining, then it may be difficult to employ multi-core processing. But if you are trying to understand heat transfer on geometrically enhanced surface, with Flow, eight cores may take 48 hours of processing time. If the program does not become unstable in the process.
I care less that that programing for multiple cores will take more time. I can resolve these solutions in Fortran in a non-Windows environment. This is a LOT of work on my part. I think it not unreasonable that a more automated solution may be forthcoming?
Perhaps widgets can be designed effectively with single threaded processing. I totally agree with you on that point. But to take product development to further dimensions, will require multi-core processing. I think there will even be improvement to 3D CAD in that multi core geometry solvers can provide instantaneous rebuilds, error checking of geometry in real time, and maybe even solve long resident bugs created by core kernal issues.
Twenty years from now, I surely hope that our software will support at least 1000 processors. The future will never be as simple as the past.
@David Paulson
On the multi-threaded multi-processor thing, I think there are a lot of misconceptions out there. First of all, why do we have all these multi-core processors? Answer: because they simply can’t make single processors go any faster. If you look up the history of Moore’s Law, you’ll see that around 2005, they hit a wall. So they started adding more processors, simply because they can’t make a single one go faster.
That doesn’t mean its a good thing. To create software that runs across processors requires maybe 5-10X more effort and can result in less stable (crashes more often) software because of the fragility of such an architecture. So is this a good investment for your CAD vendor to make? I would say, not usually. Your CAD vendor should spend YOUR money (your maintenance dollars) on things that provide you the most value. [Note that some things like FEA and Rendering are basically completely parallelizable so it is a no brainer for such apps]
Rather than jumping to the solution, let’s understand the PROBLEM. What is too slow? A lot of people would say “Feature Tree Recalculate” and I would agree. But the solution is not multi-core -in fact, such a thing is not really possible, as each state by definition depends on the prior. The solution is an architecture that does NOT compute things that don’t need to be computed. Why does the 1000th feature depend on the 2nd feature when they have absolutely no design relationships? Answer: It should NOT!
That is in fact the point of Synchronous Technology. We worked closely with Parasolid on all sorts of schemes to speed up the feature tree before we decided on a new architecture that doesn’t require this unnecesary recalculation. In one conversation a very bright guy said “The best way to speed up a calculation is to not do it!” and he was right on point. You should be able to edit one feature without having to pay for another (unless they specifically depend on each other). And that means fundamentally changing the architecture, allowing complete precision and dimension driven features, but without unnecessary dependencies.
I am sorry if someone was offended by my comments! It was not my intend to badmouth or condescend on anyone.
I agree with @Dave Ault
on: “unless you have tried both History based like SW AND direct editing like SE…”
One should at least give both worlds a try before calling the earth flat…
Don’t know about taking up flower arranging! :DDDDDDD
CAD Munkie,
There is no doubt in my mind that VARS from every software company do this including SW which still clutters my inbox and they have been told I will never buy but send anyway. So, are you mad at the message or the messenger? (Remember there have been SW VARS showing up here to badmouth Matt to.) Irregardless of which it is there is still truth here. One of the points Jununee made is the principle reason I bought SE over SW five years ago and that is that I don’t care where the file came from and how it was done. I just import it and go quickly and easily go to work on a kernel that will not be swapped out in my lifetime.
I am a practical guy and I look for utilitarian technological value that best fits my large design build company of one where I do principally milling and turning and fabrication after design. I made a choice years ago to adopt best technology over who had largest market share because I am responsible for how efficiently my company works. I hate to say this but unless you have tried both History based like SW AND direct editing like SE you have no logical basis for such staunch defense of SW.
Your comment above talking about “They do haunt me 😮 ) Most every time I go into work. I guess there’s a distinction to be made between a good feature tree and a bad feature tree.
I’m working on a project with about a dozen other Design Engineers… And the office practice is divergent. Some you can understand what and why the designer has done what they have.. and it follows what is critical on the model…. Others its a long and convoluted feature tree which is disorganized and helps no one understand they mind of the Designer other than they either need SERIOUS professional psychological help or take up FLOWER ARRANGING.”
I don’t have any of these concerns not because I don’t work in an environment like yours but rather because SE by it’s very nature frees me from all these problems you have mentioned. Without intending to do so I am sure the very problems you mention having day to day are great advertising lines for Solid Edge. It is hard for we Solid Edge users to convey to people like you who have never tried or seriously tried, just guessing here of course but I bet I am right, something like the Synchronous Tech in SE. The parts creation problems you describe I used to have too and I will never willingly inflict such painful inefficiencies on myself again. And painful inefficiencies is exactly where straight history based modelers find themselves. As a matter of fact I am looking for files that give SW users problems. If you have a few you would care to send Matt can tell you how to get ahold of me. I am considering making some videos of working on problem imported parts from SW and I bet you have some.
@jununee
Even though you show an obvious bias, I want to thank you for posting your opinion. As has been pointed out, you may have a dog in the hunt, and that certainly moderates any input that you might have.
But please realize that while SE has evolved the Parasolid kernal to provide direct modeling, there are some of us that think that direct modelling is no panacea either. Ultimately, most of us need very accurate geometry that can be generated with 0 defects from a machining viewpoint or from a meshing standpoint in FEA and CFD applications.
Neither Parasolid or CGM kernals harness the available power of multi-core processors. Whether SW, SE or Catia, we are now driving a Prius with a 700 HP engine. Oddly, the engine is 20% the cost of the software…………
I worry that Windows is the bottleneck, an issue that seems to be assumed rather than questioned. Perhaps SW and SE can only do so much with multiple processors because of the OS????
@Jununee again with the rhetoric? It is an active link, I wasn’t implying you were hiding anything… I was pointing out that you are overtly promoting the website of your employer, when you are talking about the failings of a rival CAD product.
Busted huh. 😉
@CAD Munkie
For clarification, I’m only one person/employee, not the reseller company!
Even still, it doesn’t change the sad little fact, that Dassault is not listening to their customers for their needs and wishes, but just blindly pursue on the vision they have for SW.
I used to be a SW user myself, but thank god I don’t have to start converting my CAD-material to something else now just because Dassault decided so (by adapting the cloud or swapping the kernel). Don’t get me wrong, I used to think SW was a great CAD program. Turned out, I just did not know better! SW might be the market leader, but I’m sure it has seen it’s peak. Not only because Solid Edge and Synchronous Technology are a game changer, but also much thanks to “Dassault way of doing business.”
Jununee included a link to his employer, so he wasn’t exactly hiding it. If you remember the late 1990’s and how PTC went down because of a little upstart CAD company, the same thing is going to happen again. It’s just the way it is. Small company evolves into a big company, forgets how to connect to its customers, and drops out of favor. Typical business lifecycle.
Looks like @jununee are a SolidEdge / NX reseller
http://www.ideal.fi/fi/tuotteet/computer-aided-design/
Makes sense that they would say that about Dassult? You’re hardly going to be warm and welcoming about the market leader and leading competitor and sing their praises. If he wasn’t disgruntled about Dassaults way of doing business, there would be something seriously wrong. If he was openly advocating a cometitors product and not slighting them at every opportunity whilst pointing you at a sales page I’d be deeply concerned about his entire business model?
@jununee
When resellers post comments like that it doesn’t help… 😉
Where is the option for the upper poll: six months ago?
I don’t understand why you guys keep on investing on a software which is produced by a company that is not interested on your needs/wishes..?
(Edit: Reading the comments Matt said his reason: “I wouldn’t throw all that accumulated knowledge away just because some company made some bad decisions.”
Personally I’m just disgruntled on Dassaults way of doing business.
@Dan Staples
Dan, that is very exciting news! Can’t wait to see what ST6 brings.
Roger
@Dan,
Sounds really good. June 24 is not very far away either. I’ll expect Matt to do a good write up and honest review. Matt and I often seem to have the same point of view re SW functionality so if Matt has had some input into the new stuff in ST6 that’s encouraging to me and its probably all quite well sorted/ useful. Keep up the good work! I’m keen to see what you have for us. Been saving my SW subs in anticipation 🙂
@Neil
On ST6 surfacing, we can’t spill all the beans yet, but a few things I can say:
1. We won’t have a direct Rhino convertor, but I know quite a few folks that use Rhino and SE together today, so I think the pathway is pretty good already.
2. What you couldn’t do in ST5, but can in ST6 is modify a surface from Rhino that was imported to SE. ST6 has some very nifty technology in this area.
3. In consultation with Matt (who knows any holes we had in this area relative to Solidworks) and the company SEB (makers of Tefal, Moulinex, Krups, All Clad, and more) I believe we have filled any holes we had relative to Solidworks, added a new twist or two, and should be in very good shape to support your needs. (A former Pro/E surfacing user from SEB was just at our beta last week and had pretty glowing things to say about ST6).
I would be happy to arrange a discussion with you or really anyone else in this forum under NDA if you can’t way for June (we will announce/show at Solid Edge University June 24th and ship in July).
On this same note, I might suggest than any “curious onlookers” consider coming to Solid Edge University. It will be in Cincinnati and will be a great forum to learn all about SE from real SE users and my development team. I might even be able to waive conference fees for a limited number of interested parties if you’d be willing to participate in some usability testing we are doing for Solidworks users transitioning to Solid Edge.
@Dan Staples
Yeah, I like that.
@CAD Munkie
So the root of the misconception is totally understandable. Some people think Synchronous=Direct=Non-History and this just flat out wrong. Indeed most non-history is featureless and non parametric and thus of limited use. That is not all what synchronous is. Synchronous is: feature-based, parametric, dimension driven, but is also direct and efficient. The screenshot below is from a PPT we did back in 2008 that describes Sync pretty well I think. Basically trying to take the best of the history-based world and the best of the non-history world and leave the warts of both behind. Voila — Synchronous.
http://screencast.com/t/9FZdZycfy
If this topic remains of interest, I can post a little vignette or two that gives examples of dimension driven/parametric/whatever.
@solidworm
Yeah, and what is the cost of an installation of Enovia? It might not make sense for fewer than say 100 seats. I’m not worried about it. I’ll be using something else before I have to worry about Enovia or Catia Lite.
In many cases the order of history does not make a change in geometry. Solidworks does not seem to realize that merging a to b to c will be the same is merging c to b to a. I am bugged by some of these artificial history artifacts. I am also bugged by absorbed sketches, especially when the sketch is a common edge of two surfaces. Solidworks should let me arrange my sketches in a way that preserves more design intent.
I will drop Solidworks immediately when they go to an unscrutable database for models. Solidworks corrupts assemblies often enough that I make backup copies under different names, sometimes hidden in different subdirectories. Solidworks models are pigs; 1200 times larger thant the information required for the geometry. Imagine what could happen with those kind of bad ideas in a database.
I think Solidworks is heading for a dead end. I do not know when I will bail but the parachute is packed and ready.
@solidworm
I think that’s why we’re seeing such outcry. Small businesses with one or two seats probably represent ~50% (or more) of all business with SolidWorks! To alienate such a large portion of one’s customer base is business suicide. I can think of better business models.
@Matt
If solidworks v6 or whatever, is based on ENOVIA V6 technology, then it means, it can operate within a local company network. this is no longer a guess, its based on information from DS.the drawback is that even for a single user shop, ENOVIA V6 must be installed on a local server for data management purposes and the V6 CAD client must connect to that server to save (propagate in DS terms) into the database.
Just the fact that they are heading for a cloud based model is enough to put me off.
Even apart from the security concerns, there are still places in the world where fast, stable internet connections do not yet exist. How is a cloud based service supposed to work in such places?
@CAD Munkie
Indeed. Interestingly your comments in general and the second last paragraph reinforces my point “what suits one does not suit another”. Therefore statements like “best CAD available, no good alternatives” must be seen only to apply in the specific context.
As a contractor also I am expected to be thorough. Thoroughness comes at a price: if a shaft’s diameter was selected correctly by the “engineer/designer” there exists little reason for a machinist to know why.
Taken further; I do a mix of things probably similar to that which you do. However I also do other type work one of which is ceramic basins. These take on many shapes from the simple to the quite complex. Over the years I have done ALL these in AutoCAD form (much to surprise of many) and the reason is simple. The companies I do them for require the shape to accurately represent what is required and when done those files are sent around the world, to a number of manufacturers, for competitive quotations and manufacture. That said, each of those manufacturers have their own methods/systems and as such need to use the base data I produce as a starting point should they be successful in winning the job.
Having a dumb solid(s) makes that process the most cost effective for two main reasons. Doing the initial model parametrically would take considerably longer than it does now and, two, it means the manufacturer chosen has accurate geometry (without encumbrances) which can be immediately applied and or used to re-create what is preferred at that site.
Having outlined that approach I would not, personally, consider or presume it is the only way or the manner others should take. However, as it has proven cost effective I have no problem with others competing with more highly structured/complex models and much higher costs.
I just want controllable reliable geometry. Features should never fail, they might do smething a little funny. In the case of an extrude to a surface when the surface does not cover the entire extruce should do something reasonable such as go the the bounding box of the surface. Trims should never flip, relations should be reliable, trims and knits should not vanish. Surfaces should all be trimmed nurbs surfaces with edges that are shared. The surfaces should be controllable interms of the order of the shape. Stamp out wrinkles, ripples, tits, curls butt cracks and hogbacks. Feature tree is not the problem. Geometry must be reilable smooth and accurate.
Btw, just before I disappear from the CAD scene again I would say this. SE only have to convince me they are prepared to deviate a little from their present direction to include more surfacing tools that I need, whereas Dassault are going to have to sell me an entirely new program in spite of all the issues that readers here have identified with the cloud model, and also demonstrate that the company isn’t run by complete idiots. 😉
Matt,
SE is certainly very flexible once you overcome your feature impediment. How far away is ST6? Are we able to know or get an indication of what surfacing tools definitely made it? Will I be able to import Rhino files for example?? Shortly I will be faced with large decisions. Dassault have been unable to tell me anything definitive about what they are doing for years now. Their time is running out and by default so are their prospects. I would suggest to them the time to be starting publicity for whatever winning technology or capability they have is now.
@CAD Munkie
Yeah, the best practice stuff is a real problem in history. Having different ways of working is both a strength and a weakness. None of this is a problem in direct edit. That’s not entirely true, but its much less of a problem. In SE wST, I would build most of a model using synchronous, and then at the end, apply draft and fillets using history-based features. You can do that, or work with just historyfeatures (with a reorderable tree and everything), or work just direct.
In the end, all that matters is the geometry. It takes a while to “unlearn” history. There are some old SE users who stick to just history modeling. It’s ok, but they aren’t seeing the whole benefit.
And to me, this is the other half of the advantage of using Solid Edge over SolidWorks. Solid Edge will integrate new technologies in a way that will allow you to keep using your old workflow. The way Dassault makes changes (say from V4 to V5 to V6) is so entirely disruptive, that to me it would be hard to justify working with them. I don’t have direct experience with the transition between V’s, but google it. Lots of stories out there that aren’t at all complimentary.
The SE guys pay more attention to real world needs of customers. To me, this is an important distinction.
@CAD Munkie
Well, it’s true you can’t go back and change the parameters of an extrude feature, but the parameters don’t matter at all. What matters is the final geometry. In Synchronous Technology, you can apply dimensions and relations just like on sketches or in assemblies, but you do it to the faces within a part.
Here are two videos to help you understand:
Changes in design intent are so much easier.
They do haunt me :o) Most every time I go into work. I guess there’s a distinction to be made between a good feature tree and a bad feature tree.
I’m working on a project with about a dozen other Design Engineers… And the office practice is divergent. Some you can understand what and why the designer has done what they have.. and it follows what is critical on the model…. Others its a long and convoluted feature tree which is disorganized and helps no one understand they mind of the Designer other than they either need SERIOUS professional psychological help or take up FLOWER ARRANGING.
Here i will sound like a CAD snob, and in part I’m going to take that.. Will these comments come back to haunt me, yes and no. I would guess that people that don’t know their oppose-able thumb from their unconstrained sketches will not understand, and push back on these comments… Personally I’m tired of peoples BAD CAD, I’m a contractor and spend most of my days fixing others peoples CAD faux pas…
So to me I believe passionately that yes a feature tree is an essential component of good design because it is part of a good and ordered narrative that got you to where your design ended up… Because if you work with others, there needs to be good effective communication and this is ONE of the tools CAD Munkies have of communicating that design intent…
Free form design is fine if that’s what you want to accomplish, moulds, that is fine and well. Most things I design are designed to be made by a machinist working on a lathe or a Mill, on a part which is one component of hundreds in an assembly in other assemblies… without feature trees, as far as i understand we roll back to AutoCAD R14 (R13?) solid modelling, where there was no roll back, or back button… I like the security of a feature tree and the parametric capability of knowing that diameter is that diameter for ‘this’ reason’. Yes I am that anal, yes I do label my feature trees, and yes I do add comments to help other people that I work with, or more often than not myself understand what I was doing on that day, why I made those choices,and follow the design intent of what I was asked to do as I did it… Because I have a bad memory, and I am very dyslexic (I SINCERELY apologise to anyone that that has made it this far in my ramblings), and because if I do it right the first time around I don’t have to waste my time over and again…
It is the simple things that make life better, and maybe this is just where I found my level of ‘niche’ which agrees with my mind set and I am not capable of going on to ‘learn’ new practices.. for me this way of working is a right way of working…
For what its worth
CAD
This is my ‘misconception’ yes :o)
For me if there was a way of building geometry expressed in a discrete way or area, the build of a part expressed as extrude cut revolve dimensions and design intent, then it is called a feature tree?
Am i missing something here? I’ve been using a few modellers, where you just pull and push faces around, to get the shape you want, there’s not back tracking or understanding what the ultimate constraints are on the part, or its relation to other parts which it may be attached to in the wider world? This is what I understand featureless modelling to be.
James
James
@R.Paul Waddington
Totally with you Paul,
Seriously. Who cares about history? I used to, but I can’t care less these days. The files I submit for manufacturing are either IGS, STEP or Parasolid. And guess what, no feature tree, just the geometry that needs to be pulled from a mold somewhere down the line.
And if this geometry is created in SolidWorks, SolidEdge, Inventor, Rhino, Alias, Alibre, you name it. If the resulting geometry is what you were looking for, mission accomplished.
I for one absolutely love, and I mean love that my feature tree in a complex surfacing part never ever breaks in Rhino. Because there is no feature tree. There’s just geometry I always work in 3D, I never have to worry about references that go missing, flip or break. For me Rhino in combination with VSR shape modeling has been great. I will tell you that I struggled immensely during the switch from a decade plus of feature based modeling to Direct modeling. Easy, no. Worth it, it is for me.
I still use feature based modeling a lot for mechanical parts and think it is fantastic. but surface modeling…… No more feature based for me.
@CAD Munkie
“Maybe this is my mis-understanding here, but would any serious Design Engineer, designing for manufacture, seriously want there to be no traceability or narrative to the creation of a part?”
Oh! dear: for an old guy who can remember design before CAD your comments CAD Munkie will haunt you.
Pray tell, as a “serious Design Engineer “- what ever that may be – can you not see alternative? Equally I find it interesting a user of CAD – including a serious Design Engineer – with a history tree would believe the tree captures the intent in any manner other than it was the order the model was created. So much more is done in a designers head, for many designs, which is never, and never will be, captured by a history tree.
@matt
“Best CAD on the market”. Compared to what and in what way?
I agree with Matt here Doni. your comment is pointless without detailed evidence.
The next problem you (will) face is, does your evidence have relevance to others? The reality will be, for some it may and for others, like me, it will not.
I find the primary complaint of straight history based modelers to be that they feel they lose control of their design. In history you start with sketches and then constrain and dimension your sketches and then the waltz through the history tree for any edits and occasional problem solving for attempts to do things that did not work as planed. And what if your part blows up and you can’t recover? Start over.
In direct editing like ST you start with sketches and can also fully constrain and dimension. After that however your changes are not locked to history and it is impossible to lose your part, or at least that has been my experience with five years of ST under my belt. “What about unintended changes since I can’t control my geometry” is the next biggest myth I hear. All dimensions can be locked down or unlocked to make certain edits possible without unintended consequences. All dimensions can be used to drive changes in direct editing with no history tree to go through nor dependencies to break and struggle for hours to fix.
You have all the control in Synchronous to create parts as you would with parametric modeling
except with Synchronous you have complete freedom to alter existing parts, create families of parts, edit immediately geometry from any CAD program and never worry about what some bone head did before you to create the part or indeed where the part came from and arrive at the same quality conclusion as history based but much faster and easier.
I believe this is why SW is going through this whole torturous Catia Lite CGM kernel change because they can’t make direct editing work as well as it needs to and be competitive so they switch the kernel to one they hope will. ( I have been told that yes the Parasold kernel in it’s entirety is sold to customers like Dassault. But I have also been told that the goodies that make direct editing work well with parasolds is not for sale to Siemens competitors.) I say hope because there really is nothing yet Dassault has produced for direct editing for SW. Anyone who actually gives direct editing a serious try will be another lost customer for Dassault’s SW and they know they have to do something.
Those that stand on the outside looking in without trying Synchronous or direct editing really have no basis for comparison and over time are going to look a little silly to those who do understand and have made the change.
@matt
Very well said Matt, you explained everything pretty darn good to me.
Billy
It would be constructive to compare this time with 1995. I was an autocad user with many files in autocad. When I saw a demonstration of SW95, I was completely blown away and decided to switch then and there. The reason was this new product provided a vast improvement in functionality over autocad.
I apply that same logic every year- has some new CAD program come along with vastly improved CAD functionality? For this reason, I will be open to the evolution of SW but I will evaluate what the improvements are before I switch- to any software.
I can certainly understand the hesitation to exclusively save your files on the cloud- the risk of having everything compromised with a single “hack” is terrifying. At the same time I do back up my hard drive on the cloud, so there is some hypocrisy here on my part….
I can also understand the unsuitability of HAVING to be online to use your CAD software- that is also a non starter….
So there are non- negotiables (above) and there is also the continual search for a vastly improved solution. I remember when SW was rapidly improving- very satisfying to know I bet on the right horse back then. However, I’ve been very disappointed for at least 5 years now at the almost complete lack of improvement in the program!!!
@CAD Munkie
James, yeah, I think there is a lot you’re missing here. It turns out that the history tree is an artificial abstraction made necessary by the hardware of a couple of decades ago, because it couldn’t solve all of the geometry at the same time. So the software developers set it up to solve it in smaller chunks. So for the last couple of decades, we’ve all been learning how to control a part as if it were a live set of programming instructions. We came to believe this was a great thing, because it is difficult at first, we learned to master it as far as it could be mastered.
The thing about a feature tree with a lot of “definition” stored in it is that sometimes you want to change the definition. For some things like changing a number, that’s easy, but with other things, like angling the faces of a block, it becomes more difficult, at times even impossible without deleting and starting over again.
With “direct edit”, stuff like that is easy. It makes possible some things you would never even attempt in SolidWorks. Remember the discussion on modeling faceted parts here? I did a stealth fighter (http://www.dezignstuff.com/blog/?p=6383), and readers did a number of other things. This kind of stuff is far easier in a direct editor than it is in SolidWorks. Especially if you are concerned about “best practice”. In a direct editor, it doesn’t really matter how you get there, as long as the geometry is right. In SolidWorks, the geometry has to be right, and the way you got there also has to conform to whatever you’ve decided best practice is. But you and I might disagree about best practice.
I’ve said this a number of times, but you guys that are coming to the party late have to have everything re-stated for you. The system I believe is best is to combine direct with history. Make the main part of the model with direct edit. Then add cosmetic stuff with history. There are many reasons for this. The main weakness of direct edit is when topology is removed from the BREP. It can’t just be added back. So changes that might do that should be handled by history. Everything else, especially changes to prismatic geometry, should be handled by direct edit.
Direct edit can put dimensions on the 3D model, use equations, have functions similar to SW configurations, handle master model situations, and have tools immensely superior to in-context techniques for assemblies. You can make changes with numbers, don’t let all that handle pulling fool you into thinking this is just a visual looks-like approximation. Instant 3D looks like direct edit, but it is not.
In a specific tool like Solid Edge Synch Tech 5 or later, what is changed depends on what you select, which can include faces from different parts. In SolidWorks, you have to set up the whole scenario before hand. In SE, you set it up when you need it.
So in a SE assembly, you have parts that are completely unrelated. You make a change that picks faces from two different parts. You tell the faces to move 4″ parallel to some other face. When you are done, the parts are still not related, no funky file management relationships, or external links or anything. The next time you make a change, you can move just one of the faces if you want.
In SW, it takes a lot of pre-planning to get a single dimension to update either symmetrically, to the left or to the right. In SE, you tell the dimension how to update when you make the change, and changing how it changes is a one-click sort of affair, not a whole research project like it can be in SW.
I think there is a lot you’re missing. I’ve been going through Solid Edge trying to find weaknesses for a couple of years. And there are specific weaknesses. But overall, the concept of “design intent on the fly” in Solid Edge is much easier to change and much more robust through changes than history.
In SW, you make a change, and you get some features that fail. Features don’t fail in SE wST, unless they are ordered (history-based) features. Direct edit will sometimes not allow you to make a change, but SE wST it tells you what is preventing the change. You can turn that off, and complete the change.
If this is the first you’ve looked at it, I don’t expect you to understand it all right away. Plus, over time, I hope I’ll get better at explaining it. And the first try at Synchronous Technology didn’t get it right. ST1 and ST2 I thought were interesting, but not too exciting. When they got to ST3, they had figured out to combine history and direct rather than separate them. That was the ticket. It made a lot more sense to me, and opened up more possibilities.
Remember, I’m the guy who wrote the SW bible, so I understand the history-based paradigm pretty well. I wouldn’t throw all that accumulated knowledge away just because some company made some bad decisions. The concept of direct edit used in conjunction with history-based modeling is far better than history alone or direct edit alone. I believe this is where SW/V6 is headed, but can’t say for sure, based on a bunch of Youtube videos and V6 videos on the DS website. Solid Edge has already gone there. Inventor is possibly also going that direction. I’ve heard that Creo is going there as well.
Hybrid direct-history is the future. In fact, it’s the present. You’re just behind, and clinging to an outdated way of doing things. If SolidWorks admits this now, customers may jump ship and just go with Solid Edge, which already has a working system. This is at least part of the reason why DS is dumping parasolid. Because ST is not part of parasolid, and yet they recognize that they need to do what ST does – combine history and direct.
@CAD Munkie
So your misconception is that all non-history modelers cannot fully define a part.
This is certainly not true for all, especially Synchronous Technology in Solid Edge. While a “house of cards” approach to modeling a part is not needed in Synchronous Technology as is with history based CAD like SolidWorks, Driving Dimensions and Geometric Relations between faces are supported and can be applied/changed/removed at will without causing the model to fail, and this not dependent on any pre-planning for anticipated changes.
Contrast this to the “house of cards” approach used in SolidWorks and other history modelers where one small edit in the base features can sometimes cause catastrophic failure of the rest of the model and almost always a failure of a few features… Risk goes up with the complexity of the model history tree, the deviation from anticipated change, and the unfamiliarity of the user with regards to how the model was constructed.
Now, I will not argue the fact that sometimes history makes sense, but those areas where it makes sense are limited. The beauty of Solid Edge is that you can use that history only in those places where it makes sense and you aren’t forced to use it where it doesn’t make sense.
@matt
Only if they are FROSTED!
I might be missing something here?
I am a design Engineer… Or CAD Munkie …
One of the things I feel most strongly is that implicit to GOOD CAD practice is the feature tree.
It is the Darwinian leap between us and the ‘Felt Tip fairies’ – The Graphic Designers … Please read ‘Fairy’ as a mythical entity that is able to magic fantasy shapes as if from nothing ;o)
Fundamentally what use is a part that is designed that has no definition? The WORST button in SolidWorks is Instant 3D… Why or how any human being ever thought that was a good idea should have their thumbs recanted publicly, and awarded the Darwin award for CAD.
Maybe this is my mis-understanding here, but would any serious Design Engineer, designing for manufacture, seriously want there to be no traceability or narrative to the creation of a part?
James
@Doni
Why can’t people learn to give concrete examples? Synchronous Technology enables users to edit parts without need for feature history, and it does it in such a way that enables you to determine the design intent on the fly. You can change design intent to suit the changes you need to make. It’s way easier than using history based sketches and feature definition, and reordering features to do it. You just manipulate the geometry directly. Grab a face or edge, move it. Click a dimension, determine which direction the change will affect, and change the number. You can’t do this in SolidWorks as efficiently. That is something that has been demonstrated many times.
It’s not as clear as it used to be that SolidWorks is “better” than Solid Edge. Can you give some examples that demonstrate your point?
You’re right that there aren’t any great books for Solid Edge yet. They are out there, but not sure of the quality.
There is a reason why the next version of CAD your SolidWorks reseller tries to sell you will not be exclusively history-based. Dassault is moving away from the history-based idea, like most of the rest of the market.
Just give me some reason to take you seriously. Corn Flakes are #1!!! Go Corn Flakes!!!
Solidworks is still the best CAD on the market
Solid Edge is still far away
Not even a good book can be found for Solid Edge
I am drifting away from solidworks, but using solidworks. Most of the surfaces that I use are GW3D surfaces. A bit harder to use but better surface quality and more reliable.
@matt
Perhaps a better way to word it, is that Dassault does not have control over the Parasolid kernel, the way that PTC does with Granite, Autodesk with ShapeManager, and Siemens with Parasolid. These major competitors can tweak their in-house kernels to suit the needs of their MCAD software functions.
Dassault must be frustrated that it owns TWO kernels, but has to rely on the whims of a competitor to add functions to SolidWorks!
@Robert Katsma
???
Angry? Banished from the Garden of Eden?
I’m not clear about what your point is, other than “we’re not in the Garden of Eden”, which I grant without argument.
There are lots of 3D programs out there that license technology. Writing your own kernel is not an option for most developers. Parasolid is a great choice. I’m sure better choices will develop in the future, and Parasolid will also improve, but we’ve got to live in the present.
I don’t deny SW the right to change. But some changes are better than others. So I’m making a change. You like change, I like change, SW likes change. We’re all one big happy change family. ;0P
Matt,
You sound angry-like you were unfairly tricked into biting the apple and are now being banished from the Garden of Eden. However, you chose a user interface (SolidWorks) that didn’t own a kernel, and that model is unsustainable today. It was probably never a good model. Solid Edge, in my experience, always had superior functionality-not that either one of them were excellent. The point being, we are not in the Garden of Eden. SolidWorks is not so good that we can’t do without it. I like change. At least it provides the hope of improvement.
Regardless of our viewpoints I do and will continue to respect you as someone that speaks his mind. This blog is always on my “read” list.
@TOP
Paul, I don’t know anything about that. The thing is, it might be hard to judge, because they’ve said that there is a fuzzy line between SW and DS development, in terms of developers.
@Mark Landsaat
Hmmm. Glad you found something that works. Never heard of VSR. Have to check it out.
@Jeff Holliday
Jeff, my spam filter is sticky. I have to pull half a dozen messages out of it every day. It’s getting annoying.
I don’t think we disagree about your assessment. In fact, I think even SolidWorks people would agree with the basic premise that I lay out. Where we diverge is that I think the company they have turned into is not one that I want to work with, and they have a very slim chance (given all the hints they’ve dropped along the way) of creating software that I would actually pay money for. This is why I’m expanding my horizons.
I’m using SWX way less than I used to. I started using T-Splines for organic sufacing a while back which gave me exposure to Rhino.
Since then I have added VSR shape modeling as a plugin for Rhino and find that for the work that I do it actually works better. I also find that Rhino and SWX play pretty good together.
I used to do a lot of parametric surfacing, but now that I have VSR shape modeling I can see that the surface quality from parts made in SWX really isn’t that good. Less than optimal surface transitions and very, very dense geometry that is impossilbe to edit. I can get better results with VSR in less time.
Still use SWX and probably will continue to for the foreseeable future, but VSR is my software of choice for surface definition. By the way, the company behind VSR also does work for Alias and a lot of the guys came from ICEM. They are very responsive and actively developing surface modeling tools. I can’t wait until they release 3.0
Matt, what do you know about SW R&D budget? That would seem to be the real indicator and that is one graph they don’t show at SWW.
@CAD Munkie
It’s fine if that’s what you believe. For some companies it’s the right thing. I don’t see why a SW fanboy would buy into DS’s way of doing things, but that’s up to you.
There’s an option you’re missing there?
I believe it IS going to happen…
Another option… I WANT it to happen. :o)
Still a SoildWorks FanBoy,
James
I truly do believe you “want to know” and that you are genuinely hoping/looking to make the best choice for your future. I also believe you are endeavoring to help others with that choice.
There has been a communication problem without a doubt on the part of SWorks. They have made statements that are very confusing whether they are taken out of context or not. The fact that we were lead to believe V6 was on the verge of becoming reality, then it may not be, then it could become something else is confusing. A statement about Parasolid disappearing while being told that SWorks as we know it is going to be continued for the foreseeable future is confusing. “Guesswork” is what happens when any of us try to come up with an answer about a distant future event will entail. The guess could go towards the negative (Dassault is going to leave us high and dry) or positive (we will continue and/or resume being increasingly productive). The fact is that none of us know for sure. Dassault may not know for sure. We have a choice of how to move ahead.
As much as I have been a strong supporter of SWorks (maybe even a fanboy) I could change depending on the future outcomes. If I could see a path toward a future product that serves my needs both productivity-wise and economic-wise, I really don’t care if it is called “Catia-lite” and/or whether my VAR begins adding other products to their business.
As I stated earlier – I will consider changing or staying only after many more details emerge, not based solely on any of your choices in the poll.
..when we learn the french interpretation of DS really means BS!?
@Jeff Holliday
Jeff, can you be explicit? What here is “guesswork”? Be specific. Really, I want to know.
@Charles
“Disingenuous” may be a little strong, but I acknowledge your point. My idea with this question was to see how many people would take whatever DS tells them to take.
I think your second poll is disingenuous. We have no way to rate this new product that doesn’t exist yet, and may not even exist at DS yet. I would be more than willing to consider a new product from DS, but I wouldn’t just take it because I’m told to.
“At What Point Will SolidWorks Lose Your Confidence?”
For me; at a specific meeting I had with SolidWorks employees at the 2011 San Antonio SolidWorks World.
That is very similar in content to many of the polls we have been inundated with over the past many political elections. They ask very pointed questions and/or offer very few “reasonable” tick-boxes. I wish you had included a selection in the first part of your poll along the lines of “When facts have been presented showing the choices you will have ” instead of being based on inuendos, personal guesswork and suppositions.