On 12/25/2011 3:30 PM, Mr wrote:
> I just like to choose the best available open source alternatives - feature-wise
> and *try to adapt* to them. I also like to combine them, and feel very sad for
> all the proselytism from each and every community that leads them to ignore the
> benefits from some other; mostly did you know that in the Blender community
> almost nobody knows of the POV-Ray 3.7 version, they all believe that the
> software has been stuck years ago and prefer to invest time into bridging
> software like Sunflow, which has been declared dead by its authors, or Cycles,
> which is declared immature for a long time to come yet by its authors, than POV.
> I am glad that on the contrary you guys did try the latest version of Blender.
> Your reactions show that Blender's main weakness is documentation, (POV, has one
> of the best around, thanks to Jim H and other contributors).
>

Its not just documentation though, its also UI design, which I know is 
being worked on, and logic of features, in some cases. I can open nearly 
any image editor, and have a clue what I am doing. I can open most 3D 
applications, and have some general idea what is going on. Blender does 
a bloody lot of stuff in... odd ways. In terms of features though, I 
biggest issue comes down to being able to color/paint on, faces. Why? 
Because I may still be using an external editor, like Photoshop, to do 
the texture, but I need to be able to, while in 3D, actually do the 
markup. Examples: Which faces are actually part of which things on the 
object? If I plan to put a panel/hatch, etc. on the side, where on the 
side is it, and what is the side, in the first place? If I do place it, 
which side is up, or am I going to have to flip it, and move it around a 
bunch, before I know its in the right place? Its not enough to be able 
to sloppily paint things, you need to have the capacity to say, draw a 
box on one face, or even across several, so it shows up where it is 
supposed to on the unwrapped UV map, and so you know what you are doing 
in the image editor, when you get there.

Mind, this would be a lot easier too if you could have the unwrapped 
mesh show, kind of like you would make a transparency cutout in 
Photoshop, so that you could slide your image around "under it". I.e., 
move another smaller image around on the UV map, where the mesh 
structure itself acts as a mask, to place it, before backing a final result.

For texture editing, something like that would be a lot easier to work 
with, and make a lot more sense, than the sort of gross spray paint 
effects available.

But that isn't the only case where things are just odd. A lot of it 
comes down to things not working like the person who is trying to use 
the feature *expects*, based on the use of other applications, or the 
specific task they are attempting. I am sure, if I had been working with 
it from day one, and the people that developed it, I probably wouldn't 
find a problem with it. But then, people that used Wordstar couldn't 
comprehend why anyone would want to use Wordperfect either, or people 
that use EMACS for coding (and other things) see someone using something 
like Scite. Documentation only sort of fixes the problem.

And, to be sure, it doesn't fix it at all if, for example, between 2.59 
and 2.6, the book that was supposed to cover the new GUI features in 2.6 
doesn't look anything like what the GUI ended up changing to. :p I ran 
into that one while trying to work out how to use some of the texture 
features. The explanation for what its supposed to do is confusing 
enough, but things that are supposed to be there are not, maybe, in some 
case, because they didn't make sense, and if you don't know what its 
*supposed to do*, you can't tell if your doing something wrong, or its 
working right, and just not doing what you thought it was supposed to.

Yeah, better documentation would help, but as someone that tried 
following a relatively simple tutorial 2-3 times (on bending something 
or other), without getting the same results (and this was using a 
version that didn't have GUI changes yet), while thinking, "If I could 
afford Rhino, I could do this in 10 seconds.", more documentation 
wouldn't have been all that encouraging.

I like my learning curves shallow, for the stuff that should be simple. 
I don't mind so much if its something no one else is doing at all, but 
with Blender, all of it tends to be, "Why does this work this way, or am 
I just thinking its supposed to work some other way, because that is how 
everything else I ever used did it?" Makes for some hair pulling. lol

Post a reply to this message