Josh Mingus Fungus reminded me that Autodesk's new software for kids, Tinkerbox, is an awful lot like The Incredible Machine. This game, which came out in 1992, was a favorite of mine and my son.
It was a Rube-Goldberg-like puzzle solving game that boasted physics. You had to move objects together, like lights, bowling balls, ropes, inclined planes, and pulleys. Once you got it working, you could go onto the next level.
TIM, as we insiders called it, had gravity and momentum; we could select earth (normal), moon (light), or Jupiter (heavy) gravities. It had object-awareness, so that a rope could attach to pulleys but not to bowling balls. It had snap, so that objects clicked together when we moved one object close to another. It was multi-OS, running on DOS, Windows, Mac, and several gaming systems.
At the time it was popular, I was publishing my paper-based CAD++ newsletter, and in one issue I wondered why CAD software wasn't as intelligent. A few years later, TriSpectives came out, dubbed "the Pro/E killer." It was stunning because it was so interactive. Perhaps too stunning, because the company failed after a few years.
Some of the TriSpectives technology went into IronCAD, and some showing up in Actrix ("the Visio killer") from Autodesk. I wonder if Tinkerbox is based on Actrix code.
I'm not saying my question had anything to do with the development of interactive CAD software. But it is interesting to note how long it takes for user interfaces from games to migrate to CAD.
In the mid '90s when Autodesk introduced ObjectARX and the concept of custom objects, there was great anticipation that a new era of smart objects was dawning. 15 years later we have self-healing walls.
Posted by: Owen Wengerd | Feb 02, 2011 at 11:27 AM
As a serious gamer (who also happens to do CAD on the side), I'm not sure it's so much that CAD is behind as focused on other things.
Games can get away with a lot of approximate answers. Geometry can be close but not exact. The physics engines can push and bump and pull things "kind-of" like in the real world. Collisions can be "close" to where they would be if the objects really touched, etc. In CAD we want it all exact and that takes more time and resources. With games accuracy is secondary to looking cool and being entertaining.
In some games we can "script" objects (like a cube) to do special things. We can say "if the user clicks me, then turn red or become transparent or jump up and down or act like a snap". This is VERY cool but the need just doesn't come up enough in real design. It would be great to be able to script a new car design and drive it around virtually but where would you go? (what I'd like to do is drop that design in a virtual world someplace but thats another topic).
Now UI is a bit different - games come up with some great ideas. Radial and transparent windows, special input devices, subtle use of sounds were quickly adopted in the gaming community. So while guys like me would love to do the same with CAD, as a group us CAD folks are a bit more conservative and the software (mostly) matches that.
Posted by: Mark Burhop | Feb 02, 2011 at 11:51 AM
There's been a lot of discussion of "Gamification" as an emerging method of education on Fora.tv lately. I think this is becoming a popular notion prompted by the latest RockBand games enabling real(ish) musical instruments.
So far I'm all for it, although it is very immature at this stage and I doubt anyone has got it 'right' yet. But would be very interested to hear your direct views on it RG!
Posted by: C Larkin | Feb 03, 2011 at 02:03 AM
Mingus.
Posted by: al Dean | Feb 03, 2011 at 03:17 AM
Incredible Machine was a great game. I played it for many hours and my kids loved it too. I save everything so I must have a copy of it somewhere....
Posted by: Len | Feb 15, 2011 at 04:55 AM