Merovingien wrote: ↑Thu Sep 02, 2021 9:24 am
...
i understand you matt, but you will not be able to talk about that and have a mature discussion here.
I'm glad you understand. Thanks for saying that. I think that a mature discussion is exactly what's happening. I state my case, and people hopefully understand it. I'm not the police, so they keep doing what they want. I hope what I say at least makes people think. And I'm totally cool with people fixing bad practice workflows, by the way.
Plus, I think I've at least offered some situations where it's not such a bad idea, like really simple sheet metal, or one person responsible for their own stuff. (Ad hoc hack and whack modeling)
I have to admit that I too have a certain perspective. For sheet metal, I rarely do that in my every day work, but in consulting work I deal with it all the time, and with companies that have very different means of getting things done. I'm usually brought in to do best practice consulting, specialized training, process improvement, file management implementation, and so on. So I'm trying to get companies set up with the best possible process. If you're just one person, and you're responsible for your own stuff, I'd still recommend better practice, but it's less of a disaster. The more people you have and the wider the range of skill levels you have, the more of a problem something like this becomes.
My perspective is modeling practice that creates the least number of problems when you look at the whole scenario. That almost always means there's a hierarchy of practice, and there's a definite orthodoxy. You should understand how it all works - the correct way and the backdoor workarounds, because there are times when you need those workarounds. My point of view is that workarounds should not become your primary techniques.
I wrote one time about the "Rings of Fire" or something like that. It was a set of concentric circles moving from stuff that is pretty safe to stuff that is less safe. Most of my modeling work is in the WARNING SEVERE INSTABILITY range, with surfaces and trims, and whatnot.
And since I'm kind of on the Solid Edge kick lately, this is less of a problem in synchronous because of the lack of the history problem. Why would you introduce history-based limitations between parts? You have to have the features of one part before or after the features of another part - mixing them together would be another level of disaster. Do you really have the discipline in an already undisciplined method to keep the features separate?
Synchronous assemblies would allow you to match features in the parts - going either direction or both directions (A=>B or B=>A)- without incurring in-context or circular references. You can actually also do that in Solidworks, but for some reason people keep forgetting it.
This is just like the mania that surrounds zero-thickness errors. You've got a whole class of people who claim they NEED to design with zero thickness conditions in their models. But they don't really need it, obviously. Same exact deal. People claim there's some problem with assemblies, and they keep trying to make parts into assemblies. But there are so many things you give up when you do that. I can't say I understand it, but people accept a lot of limitations just to feel like they're getting away with something.
I've said it all multiple times now, so for those who feel the need, it's no skin off my nose what they choose to do, really.