I know that a very important thing for V2.0 will be object handling....
I think that object history and transforming history is also very important (even more).... I believe that this has to be very much improved....
I am very impressed with what 3dsmax's modifier stack can do... Those of you that have even the slightest experience, I believe you will vote for this.... or something similar.

I believe that almost 30% of our work is to make repeat and correct objects....
Two of the most important things that I would like to see in Moi are....

1. Working history for object transformations, and immediate update... no matter how many commands- transformations have been done since....

2. Copy or make instances or references (this is totally MAX).... Copy = make a new object, instance= make a new object that will change exactly like the prototype and no matter which one I change, everyone will change the same way, reference=make a new object that will change according to the prototype but any new transformations will be applied only to that object and not to all of them.... That means that a new transformation of the prototype won't apply to it.

My two cents: MoI appeals to me because it's so transparant. I discovered the application just two days ago, but I was able to immediately see through the interface and use it. That's a rare and precious thing. But the difference between an "instance" and a "reference" in 3dsMAX is (in my opinion) obscure. The way you describe it "instances" sounds okay but with an additional "reference" feature things could be getting too complex, as seen from my perspective.

Hi JTB, I definitely want to make improvements to the history mechanism, I have some ideas on doing that for v2. It is a tough area of work though so it is going to be a while into v2 before I'll be able to focus on it, I am planning on focusing on the object organization tools first before that.

Re: #2 - instances / references - I would definitely like to add a kind of "block" instancing mechanism in the future but I'm not quite sure when it will happen.

The focus of that mechanism would be to have clones of a base object that could be automatically updated when the base object is changed.

I'm not so sure about the idea of editing the clone instance and having it propagate back to the original, that kind of seems like an unusual arrangement to me, probably at first that would not be a focus but instead be more focused on Base to Clone updating instead of Clone to Base editing... But we'll see - it will be a while before I will be able to focus on that.

- Michael

"I'm not so sure about the idea of editing the clone instance and having it propagate back to the original, that kind of seems like an unusual arrangement to me,"

This happens only for practical reasons.... I mean, if you place 80 clones of an object in your scene, it will sometimes be difficult to find the original.
Also, don't see it as clones of objects... This means 1 object and 79 clones... I think it is just 80 identical objects....

Hi JTB, it's possible that some kind of "edit base instance" or "go to base definition" type button could make that less of a problem (finding the original object I mean).... But we'll see, this is probably a ways out.

- Michael

I hate to sound like an idiot, but I think the object grouping method (components) used in SketchUp is one of the easiest and more powerful ones I've seen. Just my2c

Hi Bruce, that part of SketchUp is definitely very interesting, it is also incorporates a whole kind of library/browser type mechanism.

That's definitely cool but at first I'm going to try and focus more on methods just to help organize pieces within your current model, not really so much on browsing libraries of external models for now.

- Michael

my 2 cents of impressions
Sketchup has an infinite undo + group / components / sort of inference / management of layers/ etc...
but for all that it can't manage a terrific numbers of polygons without slowdown activity of the computer
I am not sure that is a good thing to make the same in Moi ;)

Hi group,
I use history based parasolid design software at work and in the business we're in we need it. It goes somthing like this;

-Customer supplies 3d product.
-We incorperate the customer 3d model in the design life cycle.
-We design and model in whats called 'Top Down' modeling, where the customers model is at the top and everything designed and modeled is below that.
-Now, when the customer has a revision (design change) of their product we would substitute the old with new and our design which is going to manufacture this product would update within a few minutes and we are back on track again, this is the beauty of history based modeling.
-Thats if it was a small change, in the event the customer has reshaped the product dramatically by removing faces on the model and resurfacing this where history based modeling falls over all links and features used start falling off and have to be traced and rebuilt which takes alot of time, sometimes so bad that we inform the customer of additional charges.

That's one part of our business.

Then there is the individual model with the history tree, with simple models yes it's great but when you start modeling 'monster' models with 1823 features and complex shapes you can easily loose yourself if not careful and don't use proper modeling practice or else those error messages become a part of your career.
Then when I have do something quick or conceptual design I find it easier to blow parameters away and have no history.

I know sometimes in MoI I have to be careful in the way I design because it would be nice to go back and change the size of that fillet by just picking it and apply another value.
So my practice on more complex models is to Incrementally save, this is my history tree, have another session of MoI open and copy and paste part of the model that was ceated say 4 saves ago.

But when you consider MoI @ $195 a seat and the easy and quick work flow compared to the $15000 to $25000 a seat 3D giant that I use, I am as happy as a pig in poo!

Cheers
~Danny~

"sometimes in MoI I have to be careful in the way I design because it would be nice to go back and change the size of that fillet by just picking it and apply another value."

This is exactly what 3dsmax has with the modifier stack.... Not easy, very helpful....

Hi JTB, hi Danny - yeah one thing about very history-focused software is it tends to be a major influence in the UI and in a certain sense kind of forces you to plan things more. Like for example in a parametric solid modeler you may often need to explicitly declare a sketch plane first before drawing on it.... Not all are like that but that's just one example.

The Max method is similar but kind of evolved for a different purpose which was to be able to animate those different modifications. Again there is some complexity you're not just modeling, you're building a kind of animate-able recipe for your final object.

The focus in MoI is much more on kind of immediate free drawing.

So I don't really want the history to become a very dominating thing in MoI. That's what is nice about MoI's current history mechanism, even though it usually will only apply to something that you recently built, the good part is it doesn't force you to go through rigid steps or setup, or dominate the UI.

I'd like to extend MoI's current history mechanism to have a deeper link through more actions, but at the same time remain a kind of optional thing. Instead of an explicit tree, it will be kind of more implicitly built just by recording your actions.

I have a bunch of ideas on how to move forward on this, but it is a lot of work and it will take like a month or two of dedicated time to really make progress on it. But I think I will be able to make some progress on it later on in v2.

- Michael

I tried Max a few years back and it was so complicated. Seems like the total opposite of where MoI is headed

I can see how in certain situations this kind of functionality is to be desired, but it always comes at the price of complexity. The beauty of MoI is that it is easy, both simple and powerful without complexity.

On the other hand I can see how being able to change fillets, etc later on would be great. After all I can change the size of a cube, why not other things?

I'm sure that Michael will come up with an elegant way to handle this in a MoI way :-)

Please don't get me wrong, I am not a Max user, at least not a good one.... I find it difficult and I don't need it's capabilities more than 3%.
I just like some of the methods very much....

I try to imagine MoI's future as a user.... I would like to be able to do many things even if I don't need them every day.... For example, we have an array command... What if we could make this array attached to a surface? I mean what if we make a 10X10 objects and then make them follow the surface of an object, something like trees on a mountain..... or the very nice array on a curve that we already have in MoI.

Another example.... What if we created this array and now we want to replace the initial object we used (and every other object of the array) with a better object we imported.... Would be nice to have such a tool.... Or, scale one object and then everyone follows... That's instances.

Another great feature of Max is the modifiers stack.... Make a box, make a fillet, open a hole.... No, it has to be a bigger box.... Go back and change box's one dimension, 2.3cm.... and now the fillet can be a bit bigger, change that too....
You get the picture I think....

I am not a programmer, I judge every product from the result I get.... MoI is great but there is a lot of work to do so my only purpose is to help Michael and not make a clone of Max which I don't like very much.

I wish I had some experience with better apps like Maya or Rhino but since I am not a CG artist and I work with buildings, I never had the need.

Michael !
Its not the library aspect of the componants that
I love its the way you can tell that component to snap
into place that is so cool !!!!!!!
Like you can build a window and move it into a building
you then save as a component how that should snap onto the surface
next time .
It works with arrays too !
You build a model up fast that way too .
A building is set up in a matter of minutes where you need a few hours in others !
Peter

Hi JTB,

> I would like to be able to do many things even if
> I don't need them every day....

Sure, I can understand that, and I will be trying to make that happen.

It is just that the #1 priority for me are the simple things that you do need to use more every day. That's why having a full set of "updating" kind of tools was not the highest priority for version 1.0 .

Like for example right now the priority has been on making tools that make it as easy as possible to do the initial drawing, rather than focusing primarily on later updating.

These other programs that are focused on updates have the reverse priority - they make you go through more steps in the initial drawing which becomes less convenient but then later updating is easier.


I want to keep the general focus on doing the initial drawing easily, but I do have some ideas that should make later updating work better than it does now.

- Michael

Hi Peter,

> Its not the library aspect of the componants that
> I love its the way you can tell that component to snap
> into place that is so cool !!!!!!!

That part is definitely cool!

But please keep in mind that it is something very focused on architectural modeling.

SketchUp is a tool that is dedicated for use in architectural modeling.

MoI is a more general purpose tool not so specialized for one particular area so much like SketchUp. This has weaknesses but also different strengths as well, like for instance it is easier to do curved forms in MoI.


That's not to say that it would not be cool to have that type of component snapping in place in MoI as well, it is just probably going to be a while before that happens. In the meantime, if you need to do buildings with a bunch of windows in them, that is really easier to do in an architecture-specific type program, it is just a better tool for that kind of a job than MoI. In a certain sense, that actually makes it a kind of lower priority for me to add that stuff into MoI, after all if there is another tool out there that is better for a particular job I just really expect that you should use that tool instead of MoI.

- Michael

Michael !
In general I´m with you with what you say .... but this time I´m not !!!!
Its not something that is worthwhile only for architectural porposes !
If you make a mechanical object and make a rivet a component you could easily place those on
a surface using this feature !
I see it as part of an updated array tool too !
I understand that its not No 1 priority but its not ONLy usefull for architecture !
It would be a "COOL" feature for the future even for mechanicals !
Peter

Hi Peter, yeah you're right and in fact I guess that mechanical solid modelers like SolidWorks and Pro/E often have a pretty similar kind of "feature library" type thing as well.

But again if you have the need to create such repeated fasteners quite a lot, then again I would suggest using one of those modelers that is focused very much on mechanical part design like SolidWorks, they will do a better job of that type thing in a similar sense to what I was describing for an architecture-oriented modeler being easier to use for architecture.

- Michael

I'm sure that what you have in mind will work OK, I just wanted to share my small experience...!