Beta Feedback (v.5.540.6362.0)

So I started using the Parametric Modeling Beta in December and have used it Daily for the past 6 weeks or so and today I’m switching back to non parametric stable release.

Here is my feedback from the past few months:

Bugs (Parametric has been useable but very buggy)

  1. Keyboard responsiveness - There was an issue earlier this year when Mac OS Sonoma was released that caused the keyboard to stop responding in the app until you restarted, this seems to have creeped into the parametric beta as well (Note, I am on windows now). It does not seem to be quite as bad as the Sonoma bug was but I still find the shift key not being recognized or some shortcut not being recognized the first time constantly. For example I will go to select multiple bodies by holding shift and it just doesn’t recognize shift and wont select multiple bodies. I will do control + U to perform a union and the program won’t recognize it. This is only a minor hassle though as bugs go. Just annoying (in a similar vain, please add a keyboard shortcut for Mirror)

  2. Autobackup Generated - Not sure what’s going on here. It seems like every time I return to a file parametric will load in in an auto generated backup folder. There is no explanation as to why it did this, if there was an error or a crash and this was it’s recovery. Half the time its nothing and I can delete the folder and be on my way. Other times the original bodies will have corrupted and at least this will have saved me a good copy. So I can’t complain too much but it would be nice to be able to understand what’s going on it the background that’s causing it to generate and why my files are randomly corrupting.

  3. Ghost Sketches - I honestly hate Parametric. I prefer the 100% direct modeling that the OG Shapr3d was built on. I hate the fact that I can’t delete sketches without it breaking the history and braking my bodies. This is compounded with ghost sketches. I find the parametric beta will constantly generate empty sketch files. I will look at my files list on the left pane to see 6 or 7 open sketches. I will go to move them to my “archived sketches” folder only to notice 1-2 of them have actual sketch data in the workspace and the rest are empty. When i select them nothing appears or highlights in the workspace but because of parametric I don’t want to delete them in case they are somehow linked to something that breaks some chain which then bloats my archived sketches folder which then in turn slows down the entire program and modeling experience with 70-80 sketches being loaded in the background unnecessarily.

  4. Parametric Performance - I really am not impressed with the parametric performance on any level, It only really runs well on a fresh sketch with 1 sketch and 1 body. Once things start getting more complicated I start to see performance hits across all platforms (Windows, iPad, Mac) I am no programmer or systems engineer so I can only guess but my assumption is creating this parametric history/chain forces the program to constantly refer back through the entire chain anytime an update is made or added to it. So as the history gets longer the harder and more power it takes to propagate updates through the hole history. This could again be completely off base but it’s how I imagine it’s working and frankly I don’t see the benefit. The history chain is hidden 99% of the time for me and on the few instances I did try to use it to go back and update an early step It took forever to find in the 100’s of steps that are all named the same thing having to click through each one to find the correct step to modify.

  5. Object Persistence - I’m not exactly sure how to describe this but I’ve found object persistence in the workspace to be spotty. What I mean by this is lets say we have a square in the center of the workspace. We rotate this square 45 degrees. The square rotates and appears as it should in the workspace. I close the workspace, it backups to the cloud sync successfully, everything looks as it should. I then reopen the workspace and the square is back in the orientation is was before being rotated. This happens to me semi frequently. Not enough to be an huge issue but still frequent enough that I find myself reopening files after completion just to double check to make sure the bodies are still oriented as I wanted them to be.

I’m sure there are other issues I am forgetting but these are the main ones that come to mind that all having happened frequently enough and stacking on top of each other have me moving back to the non parametric stable version.

I’ve read others feedback on parametric and have seen the responses from the team that parametric is the future and will be the sole app eventually, but I can’t help but keep telling myself I sure wish it wasn’t. I love Shapr3d, have used it since 2016 and use it daily. It is software I cannot live without. In a sea of CAD competition Shapr3d stood out through it’s beautifully simple interface, it’s cross platform support, it’s amazing value for money and it’s fresh take on what CAD software can be. I can’t help but think that implementing parametric was a massive undertaking in development time and cost and as a team and the idea of throwing that time and cost away is unthinkable but honestly I wonder if it might be a case of sunk cost fallacy. That as a team you are already too deep and have invested too much to turn back. I would highly suggest doing a little introspection to see if Parametric really should be the path forward for Shapr3d. Look at the feedback provided by the community. I don’t recall seeing many asking for this. I see lot’s of people asking for plugin support, hole/thread wizard, I would love to see some slick shapr3d vision of simulation but not really anyone asking for parametric.

Maybe someone can explain why the team decided to go down the path to Parametric? I understand some of the benefits it brings but it also makes Shapr3d just another CAD program in the sea dominated by Solidworks etc. I say keep Shapr3d Unique, add the features users request and keep down your own path than trying to be like the others.

All if this is obviously personal opinion, subjective and open to feedback, criticism. I would be happy to be proved wrong but as of today. This is where I stand as a long term dedicated subscriber.

2 Likes

Thank you for the detailed and candid feedback.

Our top priority around parametric design right now is to increase stability and performance. Many of the issues you’re facing seem to be straight-up bugs. It’d be a huge help for us in verifying if the fixes are indeed on the way if you could export the designs in question as .shapr files and send them to beta@shapr3d.com

Regarding the direction we chose: we definitely don’t want to build yet another Solidworks in term of user experience. But we also believe that it’s possible to have an awesome, fast, stable, easy-to-learn and effective-to-use multi-platform CAD that can do direct modeling – and a lot more. There are countless workflows that are better served by having a parametric engine under the hood, including the ones you mentioned (plugins or holes/threads), but even very simple ones like adjusting or fine-tuning your revolved, swept or lofted shapes by being able to adjust the origin sketches. But if you don’t need those, you can keep the design history closed – as you said you do 99% of the time. The issues you mentioned above are indeed getting in the way right now, but those are not inherent to the principle of parametric design, rather bugs and UX issues we need to work ourselves through.

Thanks again for the time you invested in testing the beta. Please don’t forget to share the problematic designs with us, it really does help.

3 Likes

Thanks for reaching out. I’ve emailed a .shapr file to the beta email address as requested.

I do understand your point and I get the benefit. One of the beautiful things of Shapr3d has always been how smooth it runs on more “entry level” systems. You never needed some massive spec’d desktop to get good performance. Does Parametric have a fatal flaw in it’s technical architecture that having to constantly refer back to the history even if only in the background will constantly cause performance issues and be a constant chase of optimization for the team?

I truly hope the team is able to get it right. If it runs how it always did and I can just hide Parametric in the background I’ll be happy. I am not trying to bash down on the team either. As someone who has no say in the companies direction other than being a paying customer and only wanting the best for Shapr3d as such a passionate fan. I want the best for the future of the software and company and people who work for it.

Thanks, we got the file, we’ll look into the issue.

Parametric modeling does incur some performance penalty, but there’s a ton of room in optimization compared to the current state of the beta. Our goal (which we believe is achievable) to make the parametric version perform roughly the same as the direct modeling version does as long as one keeps only adding new steps and does not go back editing the history.

1 Like

This is partially why I’ve recommended branching within the model so you can put a hard stop on the history chain within the model that then can be reverted back to. In the current state I’ve been exporting the model and opening a new workspace to clean up that history trail.

On a side note, when exporting a model in resolution you can choose from Low, High, Custom. I can’t seem to find a reference for what Angle Tolerance and Deviation Tolerance is set when you choose Low or High.

The functionality to manually discard parts or all of the history (with the exact same result as an export/import cycle that you’ve been doing) is coming very soon, we are working on it as we speak.

I’ll check the tolerances question and get you an answer.

2 Likes

That’s awesome news! I will 100% switch back to the beta and give it a shot when the feature is rolled out.

Thank you again Peter for your response and insight

Happy to hear that, thanks for the team listen user feedback and keep Ameliorate