Michael,

>> It would be FAR, FAR better if they were wide and low instead of tall and narrow
>> (as they are now).

>How would you expect the wide and low buttons to look? It is difficult to make wide and >low, and also incorporate both an icon and text.

I really don't see the point in having an icon AND text. You could have text OR an icon and save substantial space. The attached image shows one possible way the MoI icons could be modified. It maximizes the most useful area of the viewport, which is the center. (When people draw on a sheet of paper, they tend to draw in the middle of the paper -- not in the corners!)

In truth, my version of the icons might be a few pixels too small -- but they give you a basic idea of an alternate design. You could, of course, also just use icons and no text at all (this might, in fact, be the best way to go).

I've also included a screencap of a Lightwave 3D viewport. It uses tiny icons tucked into a title bar that GETS THEM OUT OF THE WAY OF THE USER. The icons are very small, but most people tend to use the keyboard for those functions, anyway. However, since part of your goal is to minimize keyboard usage, your icons might need to be a bit bigger -- but not as HUGE as they are now.


>re: Adobe Illustrator

>I'm always amazed that they don't overhaul Illustrator to a greater extent.

>In future versions I do want to explore adding more 2D illustration functionality to MoI, >such as strokes and fills. But that is going to be a ways off.

I really do think MoI is great for 2D drawing. The reason I mention it is that programming 2D graphics -- as opposed to 3D -- HAS to be a whole lot easier and faster. If you added a few essential features (like background image, color fills and color blends, curve blends, and line weights), you could quickly have a commercial application that is ready to sell and PRODUCE AN INCOME STREAM for you.

I think there's a place in the market for a vector drawing program that is BOTH less expensive and more useful than Illustrator. With MoI 3D, it appears that you may still be looking at a fairly substantial development period before it's ready to ship. It's just something to consider... :>)

Regards,

Johnny

> I really don't see the point in having an icon AND text.

They each serve a slightly different function. A Text only button has a problem that it is short. It is easier to hit a target with the mouse if it is square. When one dimension becomes smaller (in the case of text, the height) it increases the difficulty of acquiring it.

Text also requires a bit more effort to read, text by its nature is similar to other text, and when you have a bunch of text together individual words do not stand out as much as icons do.

So being square and having greater differentiation are pros for the icons.

But just differentiating is not quite enough, you not only want to be able to differentiate between buttons, you also want to know what each individual button does as well - text is best for this particular area.

So the combination of both delivers particular things that each one on its own does not.


> The attached image shows one possible way the MoI icons could be modified.
> It maximizes the most useful area of the viewport, which is the center.

It's an interesting idea, but it creates a new problem where the view controls are not clustered together anymore.

If you are having problems running into the view controls while you are modeling, you might want to switch to the single-view mode instead of working in the split view too much.


> I've also included a screencap of a Lightwave 3D viewport. It uses tiny icons
> tucked into a title bar that GETS THEM OUT OF THE WAY OF THE USER.

Personally I find that's exactly the problem - they are so far out of the way and so tiny it is difficult to hit them. I have to really concentrate when I want to use those types of controls, it becomes a disruption in workflow. Being tiny and close together, you have to concentrate more just to pick out the one you want, and then you have to move with great precision to hit it. Also, having them on the top like that means that your hand may be covering up part of the view you are manipulating if you are using something like a TabletPC.

MoI is designed to solve some of these common problems.


If the icons seem too large for you, you can shrink them to your taste using Options / UI size, that will allow you to shrink or enlarge the entire UI including those view control buttons.


> With MoI 3D, it appears that you may still be looking at a fairly substantial
> development period before it's ready to ship.

Well, I'm certainly hoping that it will not be too substantial, I'm targeting something like 4 or 5 months down the road from now.

- Michael

I'm new here, but I found a few things as well...

No back/left/bottom views. Perhaps clicking on the front/right/top views a second time to get the opposite side?

3D could probably have a corresponding ortho mode, for doing planar stuff that doesn't align with other standard views.

Export of .obj a little messy. Seems that each surface is a separate part which is grouped/combined to create the mesh. This means that in some sections vertices don't match, etc. Is it possible to create an .obj where edges match on joining surfaces and vertices are welded? This also explains why Wings3D has strange stuff happen upon import. Considering the way some objects behave, it's suprising they can import to Wings. (Wings will produce "temporary" edges, which suggests something weird going on with vertice order.) If this can be figured out, the program would work better alongside subdiv modelers. As it is now, it's only seems good for straight to render with no other tweaking.

Not much in the way of simple documentation beyond a descriptor. Not everyone comes from a nurbs background with an understanding of how the tools should behave. Some seem simple enough to figure out such as booleans or lofting, but others such as sweep or blend might need more explanation to newbs. (Having a lot of "whaaa?" moments with sweep until I figured out I was doing it in backwards order. I'm still having surfaces balloon out though. Blend doesn't seem to do anything at all.) Even setting up an "official" wiki somewhere where the well initiated could explain things would be nice.

Future pricing? (No nurbs parallel to Wings3d yet, but Silo maybe...) What is the plan? Is it going to be around $100(US) for absolute beginners? Or is it going to be more expensive and cater to professionals who may be able to throw budget money at stuff for a "simple" pipeline tool?

I do like the interface, so you should stick with it. It's simple for the most part and pretty clean. Nothing is really hidden in the depths of the program or relies excessively on hotkeys. So it's not overly complex or confusing like Blender or some other "professional" programs.

Hi Paul, thanks for your feedback!

> No back/left/bottom views. Perhaps clicking on the front/right/top views
> a second time to get the opposite side?

That's a great idea for how to handle it. I had messed around a bit with having a little arrow menu appear over those tabs to let you change them, but I like your idea better. I had skipped doing this because the way I had messed with was too awkward, so I had postponed working on it until after V1. Your idea makes it more feasible, but it still might end up postponed until after V1, just because of general lack of time.


> 3D could probably have a corresponding ortho mode, for doing planar
> stuff that doesn't align with other standard views.

This is another thing that I want to handle eventually, but have currently put in the "after V1" bucket.


> Export of .obj a little messy. Seems that each surface is a separate part which is
> grouped/combined to create the mesh. This means that in some sections vertices
> don't match, etc. Is it possible to create an .obj where edges match on joining
> surfaces and vertices are welded?

The edges should match between joined surfaces exactly, just the points are unwelded. If you're seeing a problem where joined surfaces don't produce the same edge structure, can you please post a model that demonstrates it? That would be a bug.

The new beta that I am releasing today has a greatly overhauled .obj export, so you'll probably want to grab that.

I'm still in the process of working on the meshing and I expect to work on the welding quite soon.


> This also explains why Wings3D has strange stuff happen upon import.

Wings3D appears to have bugs in its .obj import - in this (long) thread: http://moi3d.com/forum/index.php?webtag=MOI&msg=162.1 , some sample OBJ files were loaded into several different applications and they worked fine everywhere except for Wings3D. I also replicated problems with exporting models from Rhino into Wings3D without using MoI at all. This very much leads me to believe that Wings3D has problems with its .obj import, it is possibly related to models that have a closed seam on them, like tubes or cylinders.


> Not much in the way of simple documentation beyond a descriptor.

This will get better more towards the end of the beta. If you find something confusing, please don't hesitate to post a question on it here in this forum.


> Blend doesn't seem to do anything at all.

Right now blend only works if you have selected 2 surface edge curves. If you just select 2 curves instead of edges, it won't do anything right now, eventually it will do a curve blend in that case though.


> Even setting up an "official" wiki somewhere where the well initiated could explain things would be nice.

It would be nice. I'm afraid that the time to work on that is rather nonexistent right now though. In the meantime this forum is the next closest thing.


> Future pricing? (No nurbs parallel to Wings3d yet, but Silo maybe...) What is the plan?
> Is it going to be around $100(US) for absolute beginners? Or is it going to be more
> expensive and cater to professionals who may be able to throw budget money at stuff
> for a "simple" pipeline tool?

The pricing is not completely set yet, but the plan is that it will be very affordable compared to other CAD programs, it shouldn't be too far off from Silo2 pricing, probably a bit more than that but not too much. It will be more than $100, though.

However, I feel that I should mention that MoI is absolutely intended to be a professional tool, that is it is intended to be a tool that is highly productive and easy to use at the same time, and that does a lot of useful things. But I expect people to use it conjunction with other software quite a lot, so it will be priced to be accessible to many.

- Michael