Odd timing - I was at a user group two weeks ago and Lloyd Beachy, in missionary mode, handed me a sheet with this exact enhancement request on it hoping I would sign on to it.
The real drag is that THIS WOULD BE EASY TO IMPLEMENT (except for the "caveat", below). But dopn't let the caveat stop it - even if it was half done, it would be better than not doing it.
Paul, you of course remember the bad old days when we couldn't roll back a part at all without it rebuilding from the very top of the tree. This went away circa 2003 (I don't remember the exact release, but it was around then). This is because SolidWorks now saves parasolid snapshots of the model at various points in the feature tree so when we rollback or make changes we only have to rebuild from the last snapshot - not from the top. This is called 'partitioned rollback'
This was one of the reasons (I suspect) Thilo stepped up and improved ecosqueeze - he wanted to remove parasolid info because that was causing mega-bloat in our files. Anyone want to know why your files are getting so huge? Partition rollback is the biggest cause, and that's what ecosqueeze removes when you tell it to remove parasolid data. New folks - check out, experiment with, and cherish ecosqueeze at
formatting link
So I think the path to glory on your request has two parts:
1) when you delete a feature, SolidWorks checks to see if there are dependents. There is already a parent-child list so this info is already available. Any of you macro guys want to chime in, here is your opening - seems like it would be easy to me (except for the "Caveat", below).
2) When you open a part from an earlier version of SolidWorks, we know it's a crapshoot on whether it will rebuild the same (or at all). What gets me is that there is already parasolid snapshots behind the scenes, so why can't we use it? Here's the case - last week I opened a job from 2005 and needed to make a new part to fit it. Of course the 2005 part erupted in bloody errors that prevented anything from rebuilding, and even if it did rebuild there is no guarantee that the features will rebuild the same. Fortunately we had an IGES of the original released part lying around that we could use to for references to make the new part (tips to the newbies - always save a parasolid or IGES of all released products so you can reference a locked version of your geometries if you need to do any other development in a later release!) The thing that gets my goat is that the feature tree of that part from
2005 probably already has the parasolid info locked into its tree. Why can't I get an option to use that instead?
The promised "Caveat" - This is what I bet Solid Work's line will be
- what about equations? For instance, if I have an offset plane in my part, and if I have another dim in a sketch that through an equation is set to be related to the offset-dim of that plane, I have a problem. A parent-child on that plane shows no children, but if I delete the plane the equation is now in error.
There are also in-context ramifications - parent parts don't know about their kids except when child relations are made to the sketch in the parent (a tag is left somehow, and will let you know that there are outside relations when you try to delete a referenced segment, though you don't know if it is local or through the assembly - it's weird, and I bet no one in Concord actually knows this) SolidWorks would have to take on a major project to clean this all up and have it make sense. Without adding more to an already long post, it's a big nut for them and it is far easier to just-make-us rebuild than actually work it out. Maybe around 2015 they will finally clean up their references, get all kids under their parents so we can see everything that parts already know, add new stuff they need to know, and otherwise get all this stuff in line. Sorry to end with a bummer, but I think this is what's going on and why you have to rebuild.
Ed