Topic: Alternative CAD Interfaces. Looking for feedback please.

[gmyroup]

Great question Dan...

how do you initially specify what type of object it would create, a line, a circle, a camel etc.

A camel???  I don't think we've covered that... seriously, thats the beauty of the system... you don't need to specify what type of object you wish to create. How you use the mouse while in the CREATE mode tells the system what you are wanting to create.

As I noted in my previous post... to create a circle you simply double-click the center point and pick another point to define the diameter or enter the diameter and press ENTER.  If you're already working in the CREATE mode... you only need to perform 2 interactions. If you working in another mode you would need to press the Ctrl key first... so in that case it would be 3 interactions.

So at the worst the number of interactions is no greater than the 3 you must perform to draw a circle using your marco button.  And in most cases... you're creating 1 or more entities.

[CADaver]

Okay, show of hands,
"How many of us still draw lines arcs and circles on a regular basis?" 
"How many of those that still do, plan on doing so three years from now?"

Speaking for me and mine, we draw darned few lines, and I haven't drawn a plain circle in years or an arc in decades, nor do I plan on doing so any time soon.  Even back when we did draw 2D lines, arcs, and circles, it was only a fraction of what we drew (that wasn't blocks or parametrically automated), and then we had it customized to control layers at the same time.

I have some questions:
1.)  How does this new "feature" help me in the 3D world, using solids, blocks, dynamic blocks, push-pull, etc.?
2.)  Where is this feature going in the next five years? How will it grow to meet new releases?
3.)  What is the learning curve for power users? What's the ROI?
4.)  How customizable is the interface?  Will I be able to bend it to how we work?
5.)  How well does the interface play with already heavily customized systems?
6.)  What's your target market, and just how much do you think that is of the total market?

[gmyroup]

Okay, show of hands,
"How many of us still draw lines arcs and circles on a regular basis?"

If you look at numbers supplied by AutoDesk... you see that migration to their 3D products is not nearly as great as they would like it to be.  There are still manyof us out there that DO have need to create things using basic primitives.

"How many of those that still do, plan on doing so three years from now?"

As I cited above.. migration is not going well... its getting better but had you asked someone like yourself 5 years ago if Autodesk would still have 85% of their users still using 2D... 5 years from then... they probably would have 'No Way'

Speaking for me and mine, we draw darned few lines, and I haven't drawn a plain circle in years or an arc in decades, nor do I plan on doing so any time soon.  Even back when we did draw 2D lines, arcs, and circles, it was only a fraction of what we drew (that wasn't blocks or parametrically automated), and then we had it customized to control layers at the same time.

I have some questions:
1.)  How does this new "feature" help me in the 3D world, using solids, blocks, dynamic blocks, push-pull, etc.?

It doesn't help in the 3D world right now.  Please try to keep in mind that we are only looking for a few dedicated users to try out the concept...

2.)  Where is this feature going in the next five years? How will it grow to meet new releases?
3.)  What is the learning curve for power users? What's the ROI?
4.)  How customizable is the interface?  Will I be able to bend it to how we work?
5.)  How well does the interface play with already heavily customized systems?
6.)  What's your target market, and just how much do you think that is of the total market?

On Item#2... where will CAD be in 5 years? It could be entirely different in theory but you will always have many users who don't always update just because it a 'new' version.  Since the interface is actually no interface at all?? as long as there is need to create entities in whatever interface ACAD may have 5 years up the road... it will still work just fine

On Item#3... The learning curve is typical to any new way of doing something that is pretty night and day different from how they do it now... if someone sees the benefit of being able to create, edit, and transform entites nearly as fast as they think... it may take them a dedicated week to get the hang of it.  The ROI... well you figure it out.  If the interface improves... I mean TRUELY improves their productivity by even 10%... that means in return for the 40 hours spent learning the interface.. they would save 4 hours per week... times 51 weeks for the first year... 52 for each subsequent year. By saving 4 hours MINIMUM per week it would take 10 weeks for them to break even... everything beyond that is REAL savings.

On Items 4 & 5... like I said... since the interface is really not an interface at all... at least one you can see... so you can't customize it.  It works using a technology called 'Task Interaction Pattern Recognition'

The beauty of this whole concept is that it'll work nicely with your current system without requiring any modifications... and if we port the interface over to other CAD systems... you can be equally productive with any CAD system that supports add-ins.

[daron]

I've been looking into all sorts of software lately and have known about a 3d program, similar to MAX, Maya, etc., called Blender. I decided the other day to load it and see what it can do. Well, it is very difficult to learn how to do things without the tutorials, but one thing I learned by toying with some controls is that a selected object or objects can be moved, rotated, scaled, linked, etc. by using certain mouse gestures. For instance, if I pick the screen and move in a circular pattern, the object(s) will start to rotate. If I move in a linear pattern, they'll move or scale. This sounds like what gmyroup is looking at doing for 2d, but really, if it is similar to the blender controls, it sounds like it could work for 3d. Now, I'm not backing gmyroup's concept here, but the blender interface made me think of his concept. Good luck man.