blender As usual Jira is not perfoming its obligations - Blender Precision difficulties emulating a Prim Slice with a mesh object.

[Love Zhaoying]

14 minutes ago, VirtualKitten said:

The shame of it is why Secondlife or Jira cannot provide a work around?

Matter is now on a ticket [ #1388564 ] with Second life

 

Because they don’t help with content creation issues.

[VirtualKitten]

Love Zhaoying Thanks for your comment  but Its clearly not a content issue am now repeating myself, please read the rest of the thread as I need to be provided a work around. Its gone past a content issue and is affecting my income source on secondlife. It is evident this is a 'precision issue' how then can it be a content issue?

Thanks Denise

Edited July 13, 2019 by VirtualKitten

[sirhc DeSantis]

On 7/12/2019 at 9:50 AM, VirtualKitten said:

I think you are now beginning to understand the nature here. However I am really interested in your scaling information and would like to hear more  , However now we are focused on this deviation occurring in my mesh at approx   0.000036776 m  can you please tell me if it is possible to set a mesh within this precision of nine significant digits in Blender?

I

 

 

< 40 microns oh my. Walls come crashing down etc. Well actually in sub atomic scales thats - lots. I feel a blondie coming on

[VirtualKitten]

Lol Sirhc yes blonde is good, the trouble is second life seems to be dealing in world in these sub atomic scales  and a different precision than blender. The error i get from a increase on 50m on z axis is actually at the size of the item an approx size change required of my mesh of    0.000036776 m  this cannot be done in mesh in blender as it does not work to these significant figures . Now clearly if Second Life can do this by way of a slice cut at 50% of any PRIM. Chin informed of the fact that this is because secondlife uses the fact that A prim is a procedural object - it is generated by the software from a simple mathematics and therefore does not know its end points! Should mesh not be dealt with in the same way with regard to its end points ? why cannot we do this in mesh. The agreed method is to increase the bounding box of an item this is god if you have a door which doesn't grow . However for anything else it just wont cut it !

 

Kind Regards
Denise

[Aquila Kytori]

Hi

40 minutes ago, VirtualKitten said:

Lol Sirhc yes blonde is good, the trouble is second life seems to be dealing in world in these sub atomic scales  and a different precision than blender. The error i get from a increase on 50m on z axis is actually at the size of the item an approx size change required of my mesh of    0.000036776 m  this cannot be done in mesh in blender as it does not work to these significant figures . Now clearly if Second Life can do this by way of a slice cut at 50% of any PRIM. Chin informed of the fact that this is because secondlife uses the fact that A prim is a procedural object - it is generated by the software from a simple mathematics and therefore does not know its end points! Should mesh not be dealt with in the same way with regard to its end points ? why cannot we do this in mesh. The agreed method is to increase the bounding box of an item this is god if you have a door which doesn't grow . However for anything else it just wont cut it !

Earlier Wulfie suggested you were wrong and provided a blender file for you to test. Did you try it ?

I have also done some testing :

4 different objects in Blender each using a single vertice to create the larger bounding box.

 

 

In the SL mesh uploader I used the same high LoD model in each of the LoD slots and used the same for the Analyzed Physics.

I rezzed all 4 on a prim cube then logged out and logged back in again using the Firestorm viewer so that I could have the extra precision in the build floater.

The dimensions shown in the Firestorm build floater were the same as in Blender before exporting.

Next I made a copy of each and changed the X dimension to 50m.

The pivot point did not change, it remained perfectly central for all 4 objects.

 

 

 

My question to you is why would you want to even make a mesh object like that with a dimension to the nearest 1000th of a millimeter?

Why in Blender would you deliberately make a box for use in SL with one side  0.019848m when 0.02m would do? Not only "do" but in every case I can think of would be better lol.  

Just curious

Edited to add: If interested you can download the Blender file with the 4 different objects I used for testing, from  http://pasteall.org/blend/index.php?id=52048

 

Note that it is a Blender 2.80 file and when i tried to open with a 2.79 it opens empty.

Edited July 13, 2019 by Aquila Kytori

[VirtualKitten]

Now I am really curious , Thank you for your input Aquila , However i cannot see why I have not got the same outcome as you have . My size in blender and then into Second life is the same as yours is no difference  from the size in blender to Secondlife properties you can see this in my images.

My mesh is actually a laser beam so it is very slim it is a tube can you download the mesh I made for purposes of transparency i am sharing this Mesh2Prim Mesh Manager dae : http://moe.naonao.biz/mesh/MG_9R2A0LP3IJPV5TH3.html Your boxes then may not be giving the same result as you are testing non like for like .Even Mesh Master could not create this what it created was just a prim half the size of the original however it will give you what I am trying to do this with I would be interested if you get the same results with that., Failing this would it not be better to give a scripted control on the cursor position of an object from within SL scripting so that better control of this axial position could be given to all meshes after they have been imported?  You have to understand that the outside tube is not the part of interest but the inner hole face which is much smaller. IF you i'm me your email address in world i will happily send you a copy of this mesh

Kind Regards

Denise

Edited July 13, 2019 by VirtualKitten

[Aquila Kytori]

 

Upload a copy of your Blender file to   http://pasteall.org/blend/ 

then after Uploading copy and paste the new URL here.

Edited July 13, 2019 by Aquila Kytori

[VirtualKitten]

Aquila , sorry about that i did this a few days back and didnt know the links expired .

I have place the contents of my dae here https://pastebin.com/A520xee1

Please upload it to SL and try the same experiment and let me have the results please?

Kind Regards

Denise

[Aquila Kytori]

For some reason I can't log into the beta grid at the moment.

@VirtualKitten Unfortunately EVERYTHING about the beam mesh is a mess ! ! !  😯

Even the center of model is nowhere near the inner end of the tube.

 

Please take a few minutes to redo your mesh  from scratch in Blender and either Upload the new Blender file to the link i mentioned earlier or do the test uploads yourself and let us know the result. I would be reasonable sure that it would be working as expected this time

 

If you need a step by step walk through on how to do it let me know the dimensions of the beam part (tube inner and outer diameter and length) of the model and  will get back to you later.

Edited July 13, 2019 by Aquila Kytori

[VirtualKitten]

Aquila , Oh the grid  beta , lol I haven't been able to get into that since it all updated . All i get when i log in is to some strange garden that you cant get out of or build . Its hopeless.  I cant tp out either or set my home elsewhere.

I hope the beam got into a dae for you . You say its a mess not sure why. Yes the beam does not go right to the end it was made like that as i said the quality part is not the outside but the inside of the tube which is important as its nano size

[Wulfie Reanimator]

Why the heck is that tiny beam 268 tris...

Edit: Adding more triangles to a model does not make it "more visible from long distances." You could use a triangle-shaped beam of 6 faces for all LODs and it would be just as visible (if not more so, if you're currently using automatic LODs).

Edited July 13, 2019 by Wulfie Reanimator

[VirtualKitten]

Because it needs visibility Wulfie from 50m away . Oh dear

Kind Regards
Denise

 

[Aquila Kytori]

4 minutes ago, VirtualKitten said:

I hope the beam got into a dae for you . You say its a mess not sure why. Yes the beam does not go right to the end it was made like that as i said the quality part is not the outside but the inside of the tube which is important as its nano size

So you really only want to see the inner red part in the final beam? and the surfaces coloured yellow will not be seen but there just to increase the diameter of the nano tube? if so you are doing it wrong . like Wulfie said ........ lots of tris for such a light beam.

Please redo Kitten.

Like I said in the previous post if you need a step by step give the dimensions

[VirtualKitten]

1 hour ago, Aquila Kytori said:

So you really only want to see the inner red part in the final beam? and the surfaces coloured yellow will not be seen but there just to increase the diameter of the nano tube? if so you are doing it wrong . like Wulfie said ........ lots of tris for such a light beam.

Please redo Kitten.

Like I said in the previous post if you need a step by step give the dimensions

 

Yes its just colored so it can be made invisible i cant get it this small in mesh without it being inside a tube same as in SL nothing strange here . Please upload and give me results first i want to see what you replicate by uploading it to SL i will give you the 11L as its already cost me lots of more testing this. I have already answered wolfie . I cant login currently as it says my account is not accessible until 2019-07-14 : 29:12 Pacific Time no reason given  WTF!

[Kyrah Abattoir]

Linden Lab is not responsible for your inability to use Blender properly, I'm sorry.

And you shouldn't abuse LL employees on Jira because they take a decision you didn't like.

All in all I can't help but wonder why such a simplistic shape has to be a mesh...

[Kyrah Abattoir]

I have but it took me a second read.

Anyway this to me looks like a content creation issue.

The model you linked looks like an exported prim, so I assume blender isn't really your strong suit. I can already tell you that with the object rotation being off kilter like this, precision is the least of your issues.

Object scale isn't uniform either, that typically doesn't cause problems on most models, but that's not how you get a "precision" upload. Meshs should have their scaling transformations applied.

As for the off centering, you can't do that, the uploader doesn't care about custom origins, everything gets reset to the center of the mesh bounding box, so you need at least a single triangle to push the origin away.

27 minutes ago, VirtualKitten said:

The only contribution has been criticism of my mesh created by blender a very basic item.  Its part of a bigger mesh  that is a laser beam please read the whole thread if you would be so kind. It not improperly any use of blender as this is a simple copy of a sliced prim in Second life it could not be simpler to create. As no one will upload my mesh  to SL I have to conclude it does what I have said it does  and this matter has not been resolved .The problem is related to this form in mesh not a cube as shown by Aquila who was most helpful in providing what they could.

Because your mesh is the issue.

Edited July 13, 2019 by Kyrah Abattoir

[VirtualKitten]

Kyrie it was a content issue in only so much to address the fault with the tolerances of the precision. As I have already said in this page the mesh uploaded with the same dimension and rotation that it was in blender to second life no rotation as described earlier in this thread and again here could have had any impact as otherwise the item would be in a different size in second life and different rotation . which i confirm while my account was not upended it was not.

I hope this clarifies this point again

Kind Regards Denise

[Aquila Kytori]

46 minutes ago, VirtualKitten said:

Aquila, Please upload the dae i sent you

When rezzed in world it is obvious that the offset is in the incorrect position and of course when stretched to 50m the offset is over 2 meters from the inner end of the beam.

This incorrect offset can be seen in the .dae file when opened in Blender.

The whole of this thread could have been avoided if someone had been more careful when creating her original mesh in Blender.

To keep your inner beam nano mesh diameter (5mm) you just need to add an extra quad as explained in the images below. Make this plane a separate material then when rezzed inworld  set it to 100% transparent.

 

This new beam .blend file can be downloaded from  http://pasteall.org/blend/index.php?id=52049

 For me and I do hope for you now that the problem has been solved.

Your mesh was a mess lol

The origin of any precision errors was all in the original mesh.

Best is next time post your .blend file here first before posting Jiras 

 

 

Edited July 13, 2019 by Aquila Kytori

[Kyrah Abattoir]

2 minutes ago, VirtualKitten said:

Kyrie it was a content issue in only so much to address the fault with the tolerances of the precision. As I have already said in this page the mesh uploaded with the same dimension and rotation that it was in blender to second life no rotation as described earlier in this thread and again here could have had any impact as otherwise the item would be in a different size in second life and different rotation . which i confirm while my account was not upended it was not.

I hope this clarifies this point again

Kind Regards Denise

Do you know what floating point numbers are?

[Ghost Menjou]

I am confused, is this mad man thinking Jira is a seperate group or company?

It's a software packet for issue/project management. Jira is used by LL, it's not something separate. 

Also you have no idea how GDPR works. 

[VirtualKitten]

 

1 hour ago, Kyrah Abattoir said:

Do you know what floating point numbers are? 

Yes thank you. You cannot alter a item in mesh in blender with a precision of 0.000036776 m  which is a required adjustment approximately to my mesh to change the position of the little part that increases the boundary box  as when the mesh was increased to 50m on z axis the error was:

I asked someone could be so kind an upload this mesh and repeat my findings and verify if you get the same problem its here https://pastebin.com/A520xee1

just save it in a note card to a .dae file

I hope this clarifies my points

1) to replicate go to link https://pastebin.com/A520xee1

2) copy the content into notepad and save as beam.dae

3) upload to LSL 11L$ and rez in world duplicate the re-zed item and move so its has its end the same as other but separated

4) click one of the items and increase z to 50 m

5) Please inform me if the other end moves like mine has by .0100m?

Edited July 13, 2019 by VirtualKitten

[Wulfie Reanimator]

4 minutes ago, VirtualKitten said:

5) Please inform me if the other end moves like mine has by .0100m?

Yes, it does. Because you have created the model wrong. The center of the object needs to be in the correct place, but it isn't.

This has all been explained to you so many times, by so many people, in so many ways, that I have no idea what more anybody could possibly say to help you.

[VirtualKitten]

1 hour ago, Aquila Kytori said:

When rezzed in world it is obvious that the offset is in the incorrect position and of course when stretched to 50m the offset is over 2 meters from the inner end of the beam.

This incorrect offset can be seen in the .dae file when opened in Blender.

The whole of this thread could have been avoided if someone had been more careful when creating her original mesh in Blender.

Thank you So you are detecting the same fault as i did but did not show the extension of two meters you encountered in pictures why would you not show this

Now i have already informed this forum the size of my objects .0914 and the result of it doubled as being .018388  the same as imported into SL

I do not understand why your figures are different did you kindly repeat my steps 1 - 5?

 

I hope this clarifies my points

1) to replicate go to link https://pastebin.com/A520xee1

2) copy the content into notepad and save as beam.dae

3) upload to LSL 11L$ and rez in world duplicate the re-zed item and move so its has its end the same as other but separated

4) click one of the items and increase z to 50 m

5) Please inform me if the other end moves like mine has by .0100m?

Kind Regards
Denise

 

Edited July 13, 2019 by VirtualKitten
Adding media; Readed steps to reproduce

[VirtualKitten]

1 hour ago, Wulfie Reanimator said:

Yes, it does. Because you have created the model wrong. The center of the object needs to be in the correct place, but it isn't.

This has all been explained to you so many times, by so many people, in so many ways, that I have no idea what more anybody could possibly say to help you.

Wolfie, I would expect the same results to be repeated from my dae as was in my original bug report the items scale and size is significantly different than what my import is

Therefore why would i listen to you please desist from commenting it is not required and just is presenting to me as ambiguity.

[ChinRey]

1 hour ago, Kyrah Abattoir said:

I can already tell you that with the object rotation being off kilter like this, precision is the least of your issues.

The two are directly related because that rotation itself adds a huge margin of error, not only in SL but in Blender and all other 3D modelling programs I know of too.

 

43 minutes ago, VirtualKitten said:

Yes thank you. You cannot alter a item in mesh in blender with a precision of 0.000036776 m  which is a required adjustment approximately to my mesh to change the position of the little part that increases the boundary box  as when the mesh was increased to 50m on z axis the error was:

But you still haven't given any reason why you scale it up in-world. Why don't you make it at full size in Blender?