Learning Forum

[martin.kopplow]

I am disappointed. 

We are Spaceclaim users for 10+ years now, since I introduced it ty my company in 2007, with up to 10 seats, slightly varying over time. I  recommended Spaceclaim to a few other companies in our area and some bought it. We even made the change to discovery live (for some) about half a year ago. 

When introduced, Spaceclaim represented new thinking in CAD, was evolving fast and both the development and user interface team did an amazing job. The user interface was intuitive, with a good learning curve, modeling was fast, tools were added and enhanced on a regular basis, and even though they apperaed overly simple at first glace, tools were in fact quite capable. The overall speed was improved gradually with each update, though faster is never fast enough, of course. Updates were usually very well tested and not risky to install. We were quite satisfied and Spaceclaim has been a valuable asset in our workflow all the time, allowing us to tackle quite a few time critical projects. 

That changed quite some when ANSYS acquired Spaceclaim. 

The rate on which modeling tools were improved dropped. Some inconsistencies within the UI were introduced. More work was apparently done under the hood, for examplae the scripting interface and such. That did not help the average user very much, even if it was celebrated as a big step ahead, since no adeqate interface was delivered with it, and poor documentation, not to mention a non-existing ecosystem to handle the scripts. It could have been powerful, but wasn't done right. Still, the overall package worked reasonably well until this year. 

Now, the system formerly known as Spaceclaim is obviously moving in the direction of becoming a mere pre-processor for simulation. Overall speed is going down. Maintanance appears to be on low priority, improvement of modeling tools has obviously been slashed and instead of improving tools or processes that were unclear regarding UI, they are simply dropped. Example: Mirrored components. They were important for product developers, now they are gone, without notice. That was a stab in our back. Assembly handling has not been improved, instead in lager models, users need to hover the mouse pointer over a face for up to 30 seconds until it highlights for selection. That is a no-go (and we use powerful PCs). 

Now we move to Discovery Live. The new interface is a mess. Stability is poor. Speed decreasing. Updates less stable. The interface is going back to old thinking, pre-millennium, with lots of redundancies and clutter. Oh, and the black icons on dark grey background, almost invisible. Who approved that? They were perfect before. 
How can we talk interactive changes in the simulation model when half of the tools have been stripped off the interface? Open a component in it's own tab for modification? Gone. Replace components in the tree? Gone. Interaction between a Spaceclaim model and the simulation? De facto nonexistent. Modify a sheet metal part in a fluid simulation? Impossible, there are no sheet tools. What does 'live' mean in this context? Then there is that Keyshot tab in the ribbon. Oh, great! Don't think you could do a nice render of your fluid  or stress simulation, though! There is no 'live' here, too, you only get them ole solids. Instead, when we run both applications, Spaceclaim and Discovery Live, on one PC at the same time, settings of one override the other. That is sure somewhat 'live' but should not happen at all. 

Why can't I no longer move tool pallettes out of the modeling window? Why can't I resize the model tree? Why does scrolling in the solution tree not work properly? Why must parameter studies float over the model window? They clutter the whole place. Why is no simulation I saved one evening working properly anymore the very next morning after I open them? Why are changes I made to the model during simulation in Discovery Live not available when I open it in Spaceclaim next time? 

I could go on like this for hours on end. From my point of view, ANSYS Discovery (live) is heading in a dangerous direction, because it is becoming a bug loaden confused niche product that has nothing to do with predictability any more. That is not what you'd expect of a conceptual CAD and simulation program. Such an app should be on a wide base instead, and remember it is MOST important to a user to foresee how the planned workflow is going to run. Take predictability away, and you put the whole business at risk. Discovery is loosing the target group of conceptual designers, because these are not experimental designers. Don't mess that up! 

I have proposed our management to drop Spaceclaim/Discovery live and switch to another system. 
I have been given approval to research and evaluate, then make a new proposal. 

It's a pity. 

[Naresh Patre]

Hello Martin Kopplow

Thank you for sharing your feedback. It's very unfortunate to hear the experience you had with SpaceClaim and Discovery Live. We would like to work with you to help you get a better experience with our products. So, at first I sincerely request you to reconsider your proposal of switching to other applications. Also, can you please let me know if we can have a meeting early next week as per your convenient timings to discuss this topic further?

[martin.kopplow]

Hello Naresh, 

sorry for falling into the house with the front door. I did not want to shock anyone, I just need clarity.

Sure we can arrange a meeting*, though I must underline we are no longer talking isolated issues or individual bugs. I had too many user complaints from within the company in the recent past to go though them one by one. Also, management is requesting justification for the amount of time spent with certain processes, for example drawing creation. These things have added up, and we need to talk general strategy. 

I am afraid I can no longer hold back the considerations about switching, though no steps will be taken immediately, think in the interval of maybe the two next minor updates. Such changes need to be well prepared, and the outcome is of course still open. 

Within this process, it would be very helpful, if we knew the direction Discovery/Spacelaim is going to move, and if it would be approximately the same direction we go. 

To us, right now, it looks as if ANSYS are taking Spaceclaim down the simulation road, and have either bitten off more than they can chew, or are merely using it as a teaser for the 'real' packages. Either way, the creation of geometry and a designer's 'bread and butter' functions appear to be no longer the focus of budget and development. 

That would not help us at all in the multi-CAD prototyping and product development business and does in fact force every user who is not a simulator in the first place to run another CAD package besides Spaceclaim to satisfy the basic needs. Once that is established, Spaceclaim quickly loses it's justification. 

 

We need: 

Good communication: Import/Export that translates all info available, fast. (even Converter.exe has been dropped ...)

Good documentation: Reliable drawings of parts and large assemblies, fast and well organized. (Drawings need to stick to their components, part properties management must be improved for BOM, Sheet layouts improved)

Good models: Improved modelling tools, especially when it comes to surfaces. Spline editing is still a mess! What about tangency control? Patterning needs more control and must be unified. Elliptical rounds, ... 

Advanced tools: The hole tool needs a fastener tool integrated, beams need to have a 'solidify' option, ... 

Mechanical tools: What about gear creation? (I know about the add-in, but that is very basic, what about chain/belt drives? ..) The add-in ecosystem is not developing as anticipaded, is it?

Interactivity: Simulation needs to run on the very same model as modelling, and have ALL the expected tools available for modifications. 

Live Dynamics: We design moving assemblies, we need tools to represent that. Visualization, constraint management, collision warning, ...  

Configurations: The current solution is unusable. 

Streamlining: Material libraries for BOM and SIM must be unified. 

Simulation: Assemblies, adequate resolution for thin walled objects, RELIABILITY/STABILITY. 

Versioning: Nothing has happened here in years, this should be a standard by now. 

(...) 

 

*) I am in central europe time zone (GMT+1). 

[Naresh Patre]

Martin Kopplow Thanks for your reply. I am discussing this with the team and will get back to you.

[martin.kopplow]

Hi Naresh, 

How are the talks with the team going?

A week has passed, and obviously some people have been in the loop, as I have received emails and phone calls. That at least confirms my message has made it past the firewall. Now what is the outcome? 

I have received offers to help fixing individual issues with tools, geometry and files, one by one. That is all very much appreciated, though no solution or strategy for the long term. Others said it was about time someone speaks up. 

What I have not received is an answer to my strategical question. It appears people much deeper into Spaceclaim/Discovery than I am have made observations not much different from mine, only earlier. That confirms there actually is an issue with the future of Spaceclaim, and everybody knows, while trying to keep the lid on it. 

Apparently, there is still a struggle within ANSYS over the direction to go, with ANSYS being more concerned with their internal processes than product quality, and consequently, resources are not assigned as targeted as should be, strategies are not outlined as clearly as desirable, and subsequently Spaceclaim product development is not on track. They rather administer the status quo, which leads to a level of uncertainty and confusion that can already be recognized on the user level, right though the product. 

What comes much clearer now, though nobody confirms it, is that ANSYS does not have the Spaceclaim user base in focus, but are looking for a readily available geometry editor for their simulation products in the first place, be it Spaceclaim or something else. Anyway, ANSYS don't want to put much effort into it, since they like to acquire rather than develop, and so they slash functions important for modeling, if they don't feel they need them for sim prepping, rather than improve or even fix them. That will most probably cripple Spaceclaim as a CAD geometry creation tool in the mid term. 

What would happen to Spaceclaim if decisions were made to use someting else as the geometry frontend to ANSYS simulation products, and when will that be decided upon? Will ANSYS just drop it, or will there be a Spaceclaim after ANSYS? 

Your users need to make strategic plans for their businesses just as much as you do, and they will most probably choose a paved well mapped out way over a nebulous swampy one. Again, it is all about predictability and process stability.

Where is Spaceclaim heading? 

Looking forward, 
Martin 

[Naresh Patre]

Hello Martin Kopplow 

I will check with the concerned team and update you.

[Naresh Patre]

Hello Martin Kopplow

Thank you for your feedback.  I believe that our Product Manager had reached out to you after your initial post to address the points you raised.  Also, our Account Manager had recently discussed with you some of the concerns you brought to our attention.

We strive to include customer-requested features while balancing robustness, so your feedback has been very important and helpful for us.  Please reach out to your Account Manager regarding feedback of this nature, and please continue to post any technical questions to this Discovery Community Forum.

[martin.kopplow]

Hi Naresh, 

it is my understandíng that feedback is an integral part of any development process. Nevertheless, playing feedback ping-pong sometimes takes a lot of time, which is not always available. Anyway, I see some improvements have been made, for example the mirror functionality on component level has made it to a checkbox in the options on the advaced tab.

Sketching has some new features, too, and tracking is kind of blurring the line between direct modeling and feature modeling. One thing I really miss is an option to show constraint conditions in a sketch. Showing them only when touched is not enough. 

There are attempts on 3D sketching, too, though this area needs more thought to be put into. If you want to learn about a good implementation, you might want to take a look at the Ashlar 3d drafting assistant: They probably have an ages old patent on that one, but it is really good. Buy or copy. 

Surface modeling has not improved, as far as I can tell. The shell tool is still weak. Rounds and fillets - in design-driven times like these - are seldom radii, and SC can't deal with 'designer's rounds' which are made up of splines or ellipses, neither can it create nor remove them. All things tangency and spline editing are still a big unsolved issue. 

There are some new reverse engineering tools, which is great, though they would benefit from analysis tools that actually support editing. For example the cavities tool: It can reveal cavities that make patching up a tesselated surface difficult or even impossible, but don't provide a logical workflow to remove these. Wouldn't it be nice if the cavitie's faces already highlighted by the tool could be immediately selected and deleted/filled/smoothed out? 

I still have a feeling that features are preferably added at the periphery, but not much improvements are made at the modeling core, and UI consistency is still at risk. I hope, though, that people at ANSYS have understood, that to lure in designers interested in basic simulation, it is necessary to first provide rock solid bread and butter functionality, and only then add the cream and cherry, because they will certainly not be willing or able to live on cream and cherries alone. ;o) . 

 

Best,

Martin

[Naresh Patre]

Hello Martin Kopplow

Thanks again for this feedback. Our team will continue working with you to address the issues.

[Naresh Patre]

Hello Martin Kopplow

Regarding the sketch constraints visibility issue that you mentioned, there is a tool named “Show Constraint Tips” to display the constraints on the sketch. Below is the option highlighted. 

[martin.kopplow]

Thank you Naresh, I see now. They are so unrecognizeably small ob my screen*, though, so I did not identify them as symbols, just tiny blueish dots ... in which - looking real close - are in fact kind of symbols, but they are brown or grey on blue, which blurs it the more. They need to be bigger, and more contrasting. 

Constraint sketching does only work when no second drawing window is open at any one time, so I cannot use it for prodictive work yet, anyway. It would be welcome, though. 

 

Best, Martin

 

*(2x24"@1920x1200 using 120% zoom)

[Naresh Patre]

Hi Martin Kopplow

I agree that the constraints are difficult to recognize. I will pass on your suggestions to the development team for improving their visibility.