To understand my point, I have to draw on an analogy from my previous work which includes creating robotics for theme parks and museums. Each technician on the shop floor had his own personal toolbox, aside from the main shop tools, like mills, lathes, chop saws, etc. And every now and then a new guy would come along who had a toolbox the size of a stakebed truck. Whereas I had a drill set of fractional sizes from 1/16 to 1/2 and the tap and drill set for the most common threads (4-40 to 10-24), this guy had a full fractional drill set, a full number set (1 to 60), a full letter set (A to Z), taps for every thread size in the civilized world, and 5/10,000 over- and undersized reamers for every drill.

So, who built the better animatronics: the guy with the gazillion tools, or a guy like me with the basics? Realistically, we did about the same, because 99 percent of the work required holes in fractional sizes and one of the four most common thread sizes. In theory, this guy was five times better equipped than I to do the work, but in reality, most of the work was done with basic tools, and we were equally well equipped.

Another thing that is so often ignored is the quality of the tool. In another analogy, imagine a guy carving a wood duck decoy with a simple two-blade pen knife, as compared to a guy carving a duck decoy with a 38-function Swiss Army knife. Which guy has the better tool? Actually, it's the one with the better quality steel for a blade, and a better sharpener to hone it to peak performance. It's the quality of the tool, not the number of configurations that determines how useful it is to the carver.

So, getting back to computer graphics software, a sophisticated tool repertoire is theoretically of greater value, but 90 percent of the work is done with the basics every application has. And the quality of those basic applications is critical to a great software. And the reviewers often fail miserably to understand the real quality of the tools they describe. So what we get, what most of the industry is brainwashed into believing dogmatically, is that if you count or list the number of tools each application has, if you list the number of variables theoretically adjustable, then you can determine which application is "superior."

What this "tool inventory" approach fails to consider is the following things that I personally consider essential to a great tool:
 

1.  The "artist friendliness" of the tool. Remember, we are talking graphics applications, where the results of their use are ultimately judged by artistic standard, not mechanical ones. So, it's the art that really matters and so how "artist friendly" an application is becomes a critical factor in its quality.

2.  Structured to learn as easily as use. Some applications almost pride themselves in being hard to learn, which is just a stupid exercise in misplaced self-esteem (on the part of the software maker). A great application, regardless of how simple or complex, is one that is designed to be as easy as possible to learn. If you can't learn it, if you struggle to make sense of its tutorials, then the software was badly designed (as were the tutorials).

3.  Get the basics right. Remember, you use the basics most of the time. Forget the "super-whoopie gadgets" and "cutting edge" tool features. If the basics aren't there, it's a lousy tool. (Example of a basic: Being able to position your camera as you look through it. Several applications I tried couldn't even do that!)

4.  Multiple ways to accomplish a task. People are different, and not all of them have their brain wired the same way as the software designers of the application. What's intuitive or logical to them (when they know every line of code in the application) is not what's intuitive to a new user. A good application has two, three or even four ways to accomplish any given task, so each user can find the process that suits him or her.

5.  A very short path to a test render. Fools think the render is something you do when you're done with the scene or file. Artists know the render is actually part of the creative process, and they test render constantly to see things in scene context. I would guess that on an average scene of mine, I will do several hundred test renders, some full screen and some plop-renders of select areas. A quality tool is one that makes it simple and quick to test render any portion of your image that you are concerned with, so you can see changes, additions and alternatives in true scene context. The more procedural garbage you have to struggle with to get a test render, the worse the tool is.

These are my opinions. I welcome your replies, and I'll be posting other opinions soon.