Hi RD, re: unexpected chamfer result on rotated face, it looks like that is the correct result for how chamfers and
fillets work in CAD.

Chamfers or fillets between surfaces at different angles will have different widths. It's because fillets and chamfers
in CAD are defined by surface offsets. It's not quite the same thing as in a poly modeling program.

They all match up on your unrotated one because all the surfaces are at the same angle to each other there.

I guess that chamfer in CAD inherits this mechanism from fillets. It's normal for fillets to be wider or narrower
depending on the changing angle between 2 curved surfaces:



There is no guarantee that fillets between different surfaces will have the same width where they run into each other,
unless the edges being filleted are smooth to each other.

Chamfers get constructed in the same way as fillets with the same varying width they just get a straight line put in as
the cross section for the fillet surface rather than an arc being the cross section.

- Michael

For some reason, in nodeeditor, in math node, sin[...] not longer functions properly, but sin(...) does function correctly,
in BOTH MoI V4, and MoI V5beta.

Possible culprits?:
a. Windows 11 upgrade did something.
b. MoI V5beta did something? But to both V5beta and V4.
c. ?

Nodeeditor was not changed by myself...except I did replace arraysExt.js from here:
https://moi3d.com/forum/index.php?webtag=MOI&msg=10671.2

The attached nod program has both [...] and (...) versions of SineWave built with nodes.

Perhaps someone who has not upgraded to Windows 11, or MoI V5beta could try the node program, in MoI V4?

- Brian

Bless you Michael!!!! :)

The Group feature was one of my most desired wishes... I almost lost it when I tried it out!

I quickly assigned Ctrl+G and Gtrl+U that I habitually use with CorelDRAW and this will be so very useful to help me organize my objects and speed up my workflow. Before, selecting and moving objects had the feeling of taking one's groceries home without tote bags to carry them in. I like how you can still drill-select to get into the groups.

There is a little quirk when you try to Ungroup sub-groups, or groups that have been grouped together, everything becomes separate. The Ungroup and UngroupAll commands do the same thing as far as I can tell. (?) A Cut and Paste of each sub-group should keep them from becoming soup if I need to separate them again.

I have a client that insists on 3D type for his tee designs (why did I open that can of worms?) and another that always has me tinkering with his patent ideas with way too many needless iterations, and you can imagine me talking up Moi like a missionary while working with them, that is, until they ask for me to move whole type and object clusters. But now... ;)

Still though Michael, I'm very happy about this and I'm grateful!


I also do appreciate the Dark Theme option. I get eye fatigue easy and this will be nice.
I also find it fitting to Moi's unique character to allow its users a little "house decorating," a definite draw to the more artistically minded.

Also, I'm unable to change the Style color or change the visibility on object outlines for objects that are grouped as a whole, not without drill-selecting into the group or Ungrouping the set. A feature?

[ah... you can change the Style if you select the yellow dot for each individual sub-object. When you change the style of the whole group object, the Style swatch up top says that it has the new color, but the objects themselves remain unchanged.]

Hi Michael, congratulations for the first beta of V5!
Groups was my most wanted feature and I'm really glad that its available already.

I was playing with the new themes feature and i noticed that the Axis Labels are not respected when assigning colors to the Viewport y&z axis.
Is this how it supposed to be and i should just correct it in the theme editor or is it a bug?

Dinos

Hi Dinos,

re:
> Is this how it supposed to be and i should just correct it in the theme editor or is it a bug?

I guess it's a bug but the behavior is the same in v4 for the X/Y/Z axis colors in Options > Grid.

I'll fix it up so that it will take relabeled axis settings into account.

It looks like the axis icon colors (for axis icon in lower left corner) is doing it ok.

Thanks,
- Michael

Hi Mike,

For this part:
> Also, I'm unable to change the Style color or change the visibility on object outlines for objects that
> are grouped as a whole, not without drill-selecting into the group or Ungrouping the set. A feature?
>
> [ah... you can change the Style if you select the yellow dot for each individual sub-object.
> When you change the style of the whole group object, the Style swatch up top says that it has
> the new color, but the objects themselves remain unchanged.]

It's a bug - so if you select a group and then set the style by clicking on a color swatch in the Styles section
of the scene browser, it will set the style of all objects in the group.

If you set the style by using the Style menu in the upper right properties panel (the area with the Details...
button) that has different behavior and is currently only setting the style assigned to the group parent
object alone and not any sub objects which has no visible effect and then makes visibility behave a little
weird too.

I'll fix this up to make assigning with the style menu behave the same as how assigning by scene browser
swatch click works currently.

Thanks for reporting this, groups are a major structural change so there's quite a lot of things that need to
be tweaked to handle them better. Hopefully it should be in good shape with the next beta.

- Michael

Hi Brian,

I restored my setup with the original arraysExt.js and have the same node error on both V4 and V5 beta.
Running Win10.

Wayne

Thank you for the detailed explanation.

It seems like Plasticity is having an easier time handling Chamfering then. For example, this is the result in MoI Beta version 5 (also occurs with Version 4) of selecting 4 edges and then Chamfering them. Below, is the result inside Plasticity, doing the same steps.




Here's another example, selecting arbitrarily 2 adjacent edges. Top images are the result of this operation in Plasticity and below, MoI3d.

Hi RD, yes the Russian originating C3D library that Plasticity is using is definitely better at chamfering and filleting than MoI's.

That's of little impact since it isn't feasible to use it anymore. There really isn't any point of comparing with it.

It looks like you should be able to get those chamfers done in MoI ok if you do them one at a time:









In general it's easier to get chamfers and fillets done when all edges at a juncture point are being processed instead
of only some. That then allows for the fillet or chamfer surfaces to stop at a common corner juncture patch rather than
needing to intersect against each other, like this:





- Michael

@RD - it looks like the 2 and 4 edge chamfer cases worked ok in MoI v3 so there's a regression there starting in v4.

I'll see if I can track down what caused the regression and possibly revert it back.

- Michael

Hi Wayne

Thank you for your check.

I thought sin[theta] square brackets used to work, but am not certain.
I think that the square bracket enablement is in Max's nodeeditor code.
Javascript is involved.
I know Windows 11 upgrade modified my MoI 4 default background and alt colors, which seems very strange...
Eventually, in a few weeks, I'll try again on another computer without recent mods.
For now I'll stop worrying about it :-)

- Brian

I am unsure if sinewave should be a Macro/Node combination, or a brand new Sinewave.js?

Thanks for the information Michael.

"it looks like the 2 and 4 edge chamfer cases worked ok in MoI v3 so there's a regression there starting in v4."

Interesting. I started using MoI from version 4, and always assumed that it was just a limitation of the Chamfer feature. I hope that you will be able to revert it back for MoI 5.

Thank you!

Brian,

Please try running attached nod file to see if you get the same results

James
https://www.instagram.com/nodeology/

Thank you James.

After recreating your node, I got exactly the same result as you did.

45 degrees = PI/4 radians.
In nodeeditor math node, sin(45) = sin[PI/4]

The conclusion for nodeeditor, is that when using square brackets, sin[theta] , theta is in degrees, not radians.
And, in nodeeditor, when using parenthesis, sin(theta), theta is to be in radians, not degrees.
Somewhere along the way I missed this distinction.
So in adapting javascript Math.sin(), Max's presets for sin[] and cos[] in Math node are set up for degrees, not radians. (?)

A bracket is sometimes called a square bracket. [ or ]
A parenthesis is sometimes called a round bracket. ( or )

So my last few posts have nothing to do with Windows 11, nor MoI V5beta. (And belong in a different thread...)
Also, a week or two ago, I may have posted incorrect or misleading sine information...

- Brian

A totally separate topic, with no relevance to nodeeditor: In general mathematics, brackets or parenthesis are used in finding domain or range.
https://socratic.org/questions/how-do-you-know-when-to-use-brackets-or-parenthesis-in-finding-domain-or-range

@ Mike K4ICY,

re:
> There is a little quirk when you try to Ungroup sub-groups, or groups that have been
> grouped together, everything becomes separate. The Ungroup and UngroupAll
> commands do the same thing as far as I can tell. (?)

This is another bug. I'm pretty sure I had that right at one point but it must have gotten
messed up when working on what to do when doing "Ungroup" with sub objects selected.

It's been a common occurrence that fixing up one aspect of group handling has broken some other
behavior. After a couple more rounds of tuneups it should get there.

So the way that it is supposed to work is that regular "Ungroup" should just get rid of one level of
grouping at a time while "UngroupAll" is supposed to be the one that blasts things all the way apart
with just one application.

I think that when sub-objects are selected the regular "Ungroup" is working as it is supposed to
right now, it should remove the sub-group with the contents of the sub-group being promoted
up one level to now belong to the next higher up parent group.

The other direction of having a top level group selected with "ungroup" is what seems to have
gotten broken. It's on the fixup list for the next beta now, thanks!

- Michael

Thank you Michael!

I've spent the last few years getting a bit familiar with C++ as far as AVR microcontrollers are concerned. I know what you mean about tweaking one area and another going on the fritz, like changing some prescaler timing registers and screwing up the ADC. But what you do with Moi on your own is God-level stuff and I certainly appreciate the fruits of your countless hours of head scratching. :)

Grouping

Having Issues with groups select objects functionality.

Invert select does not deselect the currently selected group!
Can I isolate selections to items of a current group ?
Can I "select all" from just within a group ?
Can I invert a selection just from within a group ?
Can I iteratively expand or collapse entries listed in a group ?
Can I array a group of objects and have that original group and all the resulting groups wind up in the same parents group?

What is the layers' relationship between groups and their constituent objects ?

Thank you for all your good work !
MindSet

Hi Mindset, yes groups are still a work in progress. Hopefully most loose ends should be tuned
up in the next beta.


> Invert select does not deselect the currently selected group!

It does if you have selection just on the top level group. For example if you draw 2 rectangles,
then group them, you will now have a selected group. If you now do an invert that currently
selected group will get deselected.

The part that needs tuning up is the behavior when there is a sub object selection inside the group
instead of just top level selection. Both invert selection and select all still need to be tuned up for
that case.


> Can I isolate selections to items of a current group ?

I'm not sure of the answer to this because there isn't a concept of a "current group".

But you can isolate a selection to a specific sub item in a group using the scene browser. You
do that by right-clicking on the selection dot. A right click on the selection dot in the scene
browser will isolate the selection to just that item, deselecting other things. Left click toggles
just that item's selection state without altering other items.


> Can I "select all" from just within a group ?

The easiest way to select an entire group is to have no sub object selection in it and then just
click on it inside the viewport. The first click on an entirely unselected group will select the
entire group.

If you have a sub object selection then I do want to make "select all" (and invert) to limit
selection changes to other siblings like it does currently for an edge or face sub object
selection on a solid.


> Can I invert a selection just from within a group ?

Not yet but I'll be working on that shortly.


> Can I iteratively expand or collapse entries listed in a group ?

You mean other than by clicking on the +/- expander box? Like using left and right
arrow keys? Currently that's not possible because it would need the concept of a
scene browser item having keyboard focus.


> Can I array a group of objects and have that original group and all the resulting
> groups wind up in the same parents group?

No, I had it behaving like that at one point earlier on and it seemed weird. I think because
it's out of sync with how sub object selection already works on solids.

If you have an edge or face sub object of a solid selected and do an array on that, the
copies that are generated are independent objects and not connected to the the solid
that had sub object selection in it.

Since MoI already has a type of sub object selection mechanism in it for edges and faces
inside of a parent solid, it's good to make sub object selection in groups behave in
a similar fashion.


> What is the layers' relationship between groups and their constituent objects ?

You mean the style assignment of the parent group object? It's intended for the style
of the parent group to be just disregarded and instead have effectively be an
amalgamation of the styles of the child objects same as the behavior for solids. That's
not working correctly in the current beta but I have got it fixed up for the next one.

- Michael

Hi Michael,

Ref V5 beta:
This model is exported with an OBJ format. Then imported with the Sub-d. The orientation is off by 90 degrees on the X axis. OBJ Import orientation is set at Z-top. Is there a setting on Moi that requires adjusting on my system?

Update: Changed the OBJ setting to Y-Top and it is working okay. Z-Top makes more sense.


Thank you,
Wayne