Oh my goodness SW behavior in the model feature tree... No wonder they all love Synch

[bnemec]

I've been trying to do some work on an injection molded model, better candidates for ST than others I might say, doing a little test and see on a model I'm not familiar with. I gotta say, I don't know how you all live with SW rolling back and suppressing features in the tree all the time. I tell ya, as awful as this behavior is in ordered mode SW should have been the one to come up with ST, they need it way worse IMO.

Concerning the terrible behavior of automatically suppressing failed features down the tree after editing or suppressing a parent feature was addressed in the top ten list this year? Honestly, I don't know how it took that long. Anyway now that I've done my (censored) I have a question, how do you unsuppres them all? They were all automatically suppressed, surely, they provided a way to single clickedly unsuppressed them once the offending change was undone?

Second thing, please tell me there's a way to stop the automatic rollback madness? What the heck is with that?

Bit of background I had gotten pretty comfortable with ordered mode (the way over rated side) of Solid Edge. It has some form of showing one level of feature dependency like SW has. I would use that but then simply suppressing a feature would show you what happens to the tree. SE didn't roll back anything unless I asked it to and it sure as heck didn't automatically suppress failed features. It just failed to build and often times (mostly) there would be no solid. That was fine, it was showing me the path all the way down the tree what was affected. And with some familiarity of what causes things to fail I could usually get an idea of how to proceed fairly quickly.
Is there no way to do that in SW? I've found I have to close the model then reopen because unsurprising and rolling to the end is... I don't know what word to use.

Clearly, I have much to learn about SW. I don't know anything about modeling any more.

[mike miller]

Can you post a model and tell us what you're trying to do? If nothing else I'll open it with SE and edit the "dumb" geometry.....

[bnemec]

Maybe try SHOW FLAT TREE VIEW? Ctrl+T ?

Thanks for the reminder, Toby. I do more sheet metal work and the show flat tree doesn't work on sheet metal features, so I just don't think of it. Unfortunately, suppressing one feature high in the tree still causes a roll back and many other features to be automatically suppressed with show flat tree view option enabled.

[bnemec]

Can you post a model and tell us what you're trying to do? If nothing else I'll open it with SE and edit the "dumb" geometry.....

No sharing models, sorry. Just trying to remove a Boss Extrude from high up the tree, but there's several violations; drafts, fillets and mirror features sprinkled in the tree instead of done at the end with one mirror body at the very end. Just modeling without forethought. Probably a good part for history-less modeling.

concerning opening it in SE to edit, would still need to put back in SW to use the model. Would be nice if was never modeled in SW in the first place.

[DennisD]

If you have a bunch of downstream features that are suppressed you do not have to go down the Feature Tree unsuppressing them. You can just go to the bottom and start unsuppressing from there. Any feature that also has a suppressed parent feature will cause those parent features to unsuppress. You might have to unsuppress several, but by going from the bottom up you will have fewer to usnsuppress.

You can also hold down the Ctrl key and select all the individual suppressed features and then RMB and unsuppress them at once.

[SPerman]

If possible, you can put all of the affected features in a folder, and then suppress/un-suppress the folder.

[bnemec]

Thank you all for the suggestions. I think I just need to change my approach to how I find what downstream features are broken by proposed changes; my old practice of turning things on and off just isn't suited in SW.

I do wish I could just share the model, I'm sure someone has a clever direct editing solution. I fear the tree is so buggered already DE would be better in long run anyway. These are the kinds of models where Matt says ordered is way over-rated.

[zxys001]

To parametric or not to parametric....that is, SW users can also edit with dumb geometry.. so.. it's a choice, no?

As a past Pro/e (creo) user I would work in dumb mode or neutral formats a lot. I didn't need parametrics. There is a time/place.. just choose.

[bradb]

To parametric or not to parametric....that is, SW users can also edit with dumb geometry.. so.. it's a choice, no?

As a past Pro/e (creo) user I would work in dumb mode or neutral formats a lot. I didn't need parametrics. There is a time/place.. just choose.

I design molds, and it is usually easier to work with the dumb solid rather than try and manipulate what somebody did way up the tree that affects ten other things that have no real correlation to what you want to change. The history tree you build doing it this way is a direct record of what it is you are changing. Some may call it hack and whack, I call it fixing what I need to fix and knowing that I am changing that one thing and that one thing only.
Been bit too many times by going back up the tree changing something that affected something else I didn't know changed. Unless you created it you don't know, and who has the time to try and figure it out.

My $.02. Sorry for the slight drift.

[JSculley]

Thank you all for the suggestions. I think I just need to change my approach to how I find what downstream features are broken by proposed changes; my old practice of turning things on and off just isn't suited in SW.

I find Dynamic Reference Visualization very helpful for this sort of thing.

[DanPihlaja]

No sharing models, sorry. Just trying to remove a Boss Extrude from high up the tree, but there's several violations; drafts, fillets and mirror features sprinkled in the tree instead of done at the end with one mirror body at the very end. Just modeling without forethought. Probably a good part for history-less modeling.

concerning opening it in SE to edit, would still need to put back in SW to use the model. Would be nice if was never modeled in SW in the first place.

If you turn on Dynamic reference Visualization (child)

Then hover over the feature you want to remove, you will see the dependencies.

You have to go to each of those features and remove the dependencies before it will allow you to remove it without an error.

[bnemec]

I design molds, and it is usually easier to work with the dumb solid rather than try and manipulate what somebody did way up the tree that affects ten other things that have no real correlation to what you want to change. The history tree you build doing it this way is a direct record of what it is you are changing. Some may call it hack and whack, I call it fixing what I need to fix and knowing that I am changing that one thing and that one thing only.
Been bit too many times by going back up the tree changing something that affected something else I didn't know changed. Unless you created it you don't know, and who has the time to try and figure it out.

My $.02. Sorry for the slight drift.

I'm torn on that. It is possible to model these parts in a way that don't create needless dependencies so things can be changed, and it's not that hard. On the flip side, when the model was whack and hacked from the beginning there's not much point in doing anything else. Problem is with all the radii and draft being mixed in the tree there are way more faces to deal with than if they were all at the end of the tree so I could hack and whack before them.

As for editing as dumb solid (I assume you mean export and import back in?) That's not ideal as it usually messes up a lot of where used assemblies.

I find Dynamic Reference Visualization very helpful for this sort of thing.

@DanPihlaja and @JSculley Oh yeah, that's certainly on. It only goes to the next reference though and the tree is deep with this one.