PTC Pro/Engineer was awesome when it could reliably build solid models that were automatically used to create accurate drawings at a time when other companies were barely able to get through solid modeling demos without crashing* or required manual interventions** to generate drawing views.
These are no long awesome characteristics. Those companies that didn't die or get absorbed finally caught up.
So what would place Creo ahead of the pack today in a way that one could call awesome?
1) At the top of my list is deformable parts and assemblies. It would act on parts and subassemblies something like the spinal bend does, but at the assembly level and with better control. Unlike spinal bend it wouldn't lop off the portions that aren't being deformed and would only alter selected volumes.
Sometimes this can be handled with flexible features or with Flexible extension, but the applications I'm looking at won't work with rigid geometry motions. For example, widely used electical cables or fabric structures using zippers and Velcro need multiple versions of base items and those items need to be shaped such that building family tables of coordinates or spinal bends that are coordinated amongst all the affected parts is unrealistic. The alternative is creating separate files representing the variants.
2) A terrific feature would be a lowly Sketcher item - the curvature controlled curve. In its least form it is defined by length and curvature. This is great for items that need to bend but may also be exactly straight. A spline can sort of do the job, but it isn't length invariant and doesn't carry a curvature value. An arc can be controlled for length, but it can't be perfectly straight and can't reverse curvature. With the curvature controlled curve, positive values cause a smooth bend one way; zero causes the curve to be straight, and negative values cause curvature the other way. Apply the right hand rule to determine curve direction. Using a curvature controlled curve, it would be easy to build formed pipes that could also be shown as straight and live hinges. Items that are bent on installation, like cardboard boxes and tabs, are also good uses.
3) Next on the list is support for POV-Ray. This is a terrific renderer. It has the ability to generate animation frames, a general purpose programming language, even an extension to do wavelength sensitive ray-tracing. The main reason I like it is that the camera is controlled using parameters so that one can accurately reproduce image setups. It also has parametric textures which can be far more realistic than applied bit maps. See http://hof.povray.org/ for some examples, YouTube*** for more recent examples. It is available under GNU licensing http://www.povray.org/povlegal.html . It used to be accessible using slp2pov, but that utility doesn't run under Win7 and up because Microsoft removed DOS support. It did run under NT and I think XP, so if you have an XP box available you can still try it.
4) Could we please have colors and textures directly in family tables? For goodness sake, it's been almost 20 years of asking. We've been down the path of materials having related colors and textures and placing those in family tables, but those who say it works haven't backed up the claim, and those who asked PTC were told it doesn't work. So, put the items in the family table directly and be done with it. Besides, material colors only work for entire parts, not for indivdual surfaces. From buttons with different colors, lamp covers, and decals/logos it is uselessly time consuming to handle in other ways. It would be awesome for these to be added to the family tables.
5) In-line comments in the model tree and general comments for features and parts and assemblies to record what is supposed to happen and how the models are supposed to work****. Parametric CAD is essentially a software development process, but unlike all other software development, Creo has no sensible commenting storage; nothing that shows when the developer scans the tree. There's no obvious place for "About this model" information.
6) Genuine longshot - actual VBA for Creo. No more Pro/Program, no WebLink, no JLink, no Toolkit, no more relations editor. Just the VBA enviroment. Sure it could be Python, but VBA is fine. Drawing symbols would have VBA links allowing them to adapt their geometry to the application, instead of being the list-driven inclusive/exclusive simple (and simply painful) constructs they are. Features and components could be event triggers and event driven. Parts or assemblies could have code that would query files or databases to store or retrieve information and drive or be driven by other applications.
What else could make Creo awesome?
*I was at at least two separate Computervision invited demos ended by software crash while updating the location of a single hole in a simple plate. A third demo didn't happen because the hard drive enclosure was packed in a shock-absorbing container, but the hard drive inside the enclosure wasn't attached to anything. Shaken like a pea in a can, it didn't survive shipping.
**I'm looking at you View Cells in Bravo! At the time, views on drawings were placed View Cells, which were batch processed from CSG models. They had dimensions manually added to them.
***A little of what POV-Ray can do http://www.youtube.com/watch?v=_ZiATJsISQc ; http://www.youtube.com/watch?v=_XAJJIMUyRI
****They say a man with one watch knows the time, and a man with two is never sure. I say a man who understands how multiple watches are supposed to work to show the same time can build a Global Positioning System that allows locations to be more accurately determined than any other method. If you have a model and a description of what the model is to do, then you know something needs fixing in the event the two don't agree; and the chances are small they are both wrong, but agree accidently. If you only have a model, you are never sure if it's right.
