MoI discussion forum
MoI discussion forum

Full Version: V4 beta May-12-2018 available now

Show messages:  1-19  20-39  40-59  60

From: naxos
15 May 2018   [#20] In reply to [#17]
That would be the Holy Grail for modellers worldwide : precise modeling, booleans that works, so as chamfers, and to be able to open in a mainstream 3D software for animation, and so on...

As a nice quad-meshed polyset, ready for smoothing...

Nice quad meshes like those :
https://i.ytimg.com/vi/bVzfndz-vPQ/maxresdefault.jpg
https://i.ytimg.com/vi/sh-pwCQMQEs/maxresdefault.jpg
https://i.pinimg.com/originals/89/5a/4c/895a4c75ba28d1015efd3f9a70d3bb66.jpg
From: klaudio
15 May 2018   [#21] In reply to [#17]
Great to hear you have reach this milestone and finally going into new features! I am looking forward to it for sure.

Would be cool to see a list of what is planned for features. I know there was a lot of talks about it before but plans were also changing along the road so, as far as i know, there is still nothing set in stone, so to say. Perhaps now at this point there is?

My suggestion is the same one i already mentioned hundred times and see people asking (and complaining about) hundred times more > better and robust fillets. Fusion360 like fillets flexibility would be dream come true for me. I know, though, it's not that easy so just a remainder. ;)

Best
From: schneich
15 May 2018   [#22]
thanks for the update,

I haven't had the chance to test the other features, but the Mac display bug ist still unchanged, still jaggy jumpy...


cheers from cologne


t.
From: Michael Gibson
15 May 2018   [#23] In reply to [#20]
Hi naxos,

re:
> As a nice quad-meshed polyset, ready for smoothing...
>
> Nice quad meshes like those :

Actually I'm kind of confused by these examples because they all look like things where you'd just put fillets on them in NURBS and then I don't know why you would want to apply additional sub-d smoothing in addition to the fillets.

They all seem to be examples that are so much easier to do in NURBS rather than sub-d.

- Michael
From: Michael Gibson
15 May 2018   [#24] In reply to [#22]
Hi torsten,

re:
> I haven't had the chance to test the other features, but the Mac display bug ist
> still unchanged, still jaggy jumpy...

Can you please e-mail me at moi@moi3d.com and also send your system specs?

I'd like to send you a couple of test builds to see about solving this problem.

- Michael
From: Michael Gibson
15 May 2018   [#25] In reply to [#18]
Hi Ralf,

> Which new feature will be the next you working on...
> Just meddlesomely... ;-}

I'll probably be trying a few experimental ideas, but I'm not sure if they will work or not.

- Michael
From: Michael Gibson
15 May 2018   [#26] In reply to [#21]
Hi klaudio,

re:
> Would be cool to see a list of what is planned for features. I know there was a lot
> of talks about it before but plans were also changing along the road so, as far as
> i know, there is still nothing set in stone, so to say. Perhaps now at this point there is?

Sorry no still nothing set in stone. I don't really like to work that way too much. It's probably a good time to try some experimental things which may or may not amount to anything.

- Michael
From: pafurijaz
15 May 2018   [#27]
I wanted to highlight a problem with the import of a STEP file once imported into Moi all the objects are grouped in the center, with FreeCAD there are no problems. this is also present in version 3. Last thing the STEP files are not assigned colors or layers.

https://grabcad.com/library/raspberry-pi-3-b-1







Image Attachments:
fc_rb.png  moi_rb.png 


From: mkdm
15 May 2018   [#28] In reply to [#1]
Ciao Michael!

Thanks a lot again for this latest "crucial" beta!

Well done!

With that said...

Since (soon) you will be able to put your hands on new things, you'll forgive me if I remember you the other two "basement" of which we talked about on this forum, when you first announced V4.

That is, (1) INSTANCING and (2) GROUPING.

Two fundamental things that we can't miss.

I hope that soon you will put your hands on these things along with a new "display and mesh engine" that can leverage on the power of the modern GPU.

Thanks.
From: Michael Gibson
15 May 2018   [#29] In reply to [#28]
Hi Marco, my plans have changed since that time - I'm no longer sure about getting grouping and instancing into v4 at this time. It kind of depends on how difficult they are to implement, there is a possibility that they may require changes to many commands and I'm not too excited about something that requires lots of changes at this time.

Performance work to better use powerful cards with lots of VRAM is probably pretty soon though.

- Michael
From: Michael Gibson
15 May 2018   [#30] In reply to [#27]
Hi pafurijaz, that looks like instancing transformations not being read correctly. There is an updated version available of the library MoI uses for STEP, IGES, and SAT importing. I'll see what's involved in updating to use that it may be fixed there.

- Michael
From: JaXXoN
15 May 2018   [#31] In reply to [#1]
Hi Michael,

Thanks for the new release. Unfortunately there's still the issues with Fedora Linux that it's not possible to properly select objects.
(see http://moi3d.com/forum/messages.php?webtag=MOI&msg=8839.1)

This is with "GraphicsAPI=OpenGL". When applying "GraphicsAPI=Direct3D", instead, then it's not possible to select any objects at all.

Please let me know if I can do any further testing.

best regards

Bernhard
From: Michael Gibson
15 May 2018   [#32] In reply to [#27]
Hi pafurijaz, it looks like the .igs file of that same model has all the geometry come through ok.

- Michael
From: Michael Gibson
15 May 2018   [#33] In reply to [#31]
Hi Bernhard, thanks for testing it. I have done a little testing over here with Linux Mint and I am able to reproduce the same selection problems. So that will probably help me figure out what's going wrong there. I believe it to be a problem in one specific area of texture read back from the GPU to CPU, so it may not be too difficult to fix it.

It's a little strange that it works fine on some cards though.

- Michael
From: knob32
15 May 2018   [#34] In reply to [#1]
Hi Michael,
I made a Japanese language file with UTF-16, but characters are not displayed properly.
Thanks.
#This is a machine translation.

Image Attachments:
img.jpg 


From: Michael Gibson
16 May 2018   [#35] In reply to [#34]
Hi knob32 - that means that those glyphs are not present in the default font that MoI uses. You'll need to set the UI font to something that has those glyphs in it. That can be done in moi.ini under:

[UI]
UIFontNameV4=Open Sans

It works ok in the rest of the UI because the HTML part has a font fallback mechanism where it automatically uses a different font if needed. I was thinking of trying to implement that for the 3D viewport text display as well.

- Michael
From: mkdm
16 May 2018   [#36] In reply to [#29]
OK Michael.

Thanks for your reply.

While I'm happy to read that you want to boost the GFX routines of MOI, leveraging of modern GPU (I have a super powerful Gtx 1080 Ti with 11 GB of VRAM), honestly I cannot hide that I'm rather disappointed to learn that you decided to abandon the development of such important things, that is Instancing and Grouping.

Nowadays these are two very important features in almost all CAD systems.

I can imagine the effort needed to code these things but I really hope that you will code at least GROUPING features, because actually Moi is too weak about managing "object hierarchy" and has an outliner that is too simple, too limited, when you have to deal with medium scenes or even little, very little, assembly.

I hope you will understand my point of view.

Have a nice day.
From: Michael Gibson
16 May 2018   [#37] In reply to [#36]
Hi Marco, I definitely agree that those are important features! However there are even more basic CAD things that MoI is still missing such as dimensions.

32-bit memory limits and the fragility of the Wineskin based Mac version were causing a lot of problems so those were even more basic things yet.

If you need to do complex hierarchical assemblies, MoI probably isn't the best suited tool for that area right now but there are many existing CAD programs that can do that type of work.

- Michael
From: mkdm
16 May 2018   [#38] In reply to [#37]
Michael, I understand your point of view and all your efforts.
I do the same job as you (software developing) :)

@You : "...If you need to do complex hierarchical assemblies..."

No Michael, I don't manage complex assembly.

But, actually, with both the current outliner and the absence of true "Parenting and Grouping", it's very hard to manage also very simple assembly.

But...anyway...I hope that one day you will put your hands also on these things.

Have a nice day ;)
From: naxos
16 May 2018   [#39] In reply to [#23]
You're right as i did not found more specific examples, but don't forget that if we need to go to poly softwares (like 3dsmax), it is mostly for rendering and more of that for animation... and for light polysets.

- Polygons are easy to manipulate, deform (for animation for ex.), and a light polymesh with OpenSubdiv / Turbosmooth is very fast and usefull when you need lowpoly for VR / RT outputs.
- Exported polys from Nurbs softwares are eavy in polygons count, and sometimes with ugly triangles around chamfers, booleans etc.
> Moi3D has a better polygon exporter (quads), but as you know for sure, each surface is treated separatelly, so the vertices from sides (of a cube for ex) and those from the chamfer are not the same, so they can't be welded for further treatment (deformation, smoothing, nice UV unfolding...)
- Whenever i don't need to deform, UVmap, lowPoly... or when i'm sure that having lots of polys is not an issue for the project, i can use the Moi3D obj exporter.
- I'm done trying to open Nurbs within 3dsmax : it can handle it, but soooo poorly, slow and not really usable, i prefer getting polys.

3dsmax (and Maya and others) can do booleans, and chamfers... But not really both of them on the same mesh, as polygons booleans are almost always very poor in meshing (long flat triangles, non-aligned / non-weldable vertices...

That are the reasons why i'd love to get a clever Nurbs to Polygons exporter, that would know about chamfers from Nurbs and that would create a OpenSubdiv-ready polymesh.

If interested, i could model some examples with both nurbs and polys to show what does not work and what should.

Regards

Show messages:  1-19  20-39  40-59  60