Change in House Physics or Viewer Issue?

[Chic Aeon]

I am having issues AGAIN with house physics on the beta grid and also on the main grid IN FIRESTORM.  It isn't the physics model, it is pristine. This is the second time I have had this issue in the last month. I hadn't needed "house physics" as in walking through the doors lately; been making smaller things. 

When the physics are analyzed parts of the doors get closed off in the FS viewer.  I have heard others complain in chat when they had to switch viewers in order to upload :D, but I have never had this problem before.  I know there is a new FS out but I also need to finish some Opensim stuff in a few days so wanted to wait to install the new version. I am HOPING that the change is address in FS. I guess I can install the new one on my notebook and test. LOL.   

So my question is to those in the know -- is this something folks besides me already know has been happening?  My test on the beta grid shows the building works just fine when changed to prim when uploaded with the LINDEN viewer. So it is just me wanted to know what happened. 

This is 5 land impact when changed to prim so happy with that. 

Later:

I installed the new FS viewer on my laptop but alas it doesn't fix the problem. Same exact thing happens. Any insights appreciated.  

 

Edited June 26, 2017 by Chic Aeon

[Aquila Kytori]

Hi Chic

I usually use the Linden viewer but just for you I tried with a Firestom viewer. Not sure how old it is , says August 2016 ..  even gives a warning that Physics isn't going to work with mesh uploads............ but when I rezzed the building and changed to Prim, the collisions surfaces seemed to be working just fine. The Physics mesh I used was not Analyzed.

 

[Chic Aeon]

23 minutes ago, Aquila Kytori said:

Hi Chic

I usually use the Linden viewer but just for you I tried with a Firestom viewer. Not sure how old it is , says August 2016 ..  even gives a warning that Physics isn't going to work with mesh uploads............ but when I rezzed the building and changed to Prim, the collisions surfaces seemed to be working just fine. The Physics mesh I used was not Analyzed.

 

Thank you. I will try that. I have never actually understood when and why  (and I remember you asking me long ago if I did) the difference and since Drongle told me long ago to analyze I almost always do unless I have some odd issue. I did forget to try that.

I didn't get that warning on my current big machine FS Firestorm 5.0.1 (52150) Dec 10 2016 or on the notebook with the newly installed just out of the gate version. IF I would have seen that LOL, I would have had a hint.  I will try the un-analyzed and see how that works compared to the other.

If you have time would you explain "why" you use one over the other (in normal circumstances of course) as hopefully I am smarter now. Hope you waved at Lani if you were on the same sandbox LOL.

 

Thanks very much, she will try momentarily. 

[Chic Aeon]

Update:

So Aquila did find Lani on the beta grid and Lani and "I" and "the typist" all got a lesson again.  

I tried uploading my non-analyzed physics model  and it did work. But the land impact was MUCH higher  - 17 instead of 5.

But, as I understand it (and I am sure she will correct me tomorrow if I got this wrong), while MY cube physics model was just fine to use with ANALYSED physics, I need to use PLANES for my physics model when uploading with non-analyzed. Since planes physics doesn't work in Opensim and I am trying to make tutorials for both (and also -- let's be honest -- because I have been doing it this way for four years LOL) I will just use the Linden viewer for now and hope that FS will get this worked out down the road.

So --- planes (as large as you can make them AND they can overlap) for non-analyzed physics and cubes (not touching) for analyzed.

Here are some photos.This is Aquila's house and my houses and some physics models.  Her house was lower land impact than mine (she has always been smarter  -- no argument there --  and has baled me out of issues for years -TYSM).  My linden viewer house though is only 5 and doesn't increase when resized up. So I am happy with that and will just grit my teeth when I have to upload something with doors in the Linden viewer.  

So set in our ways.

BUT SMARTER NOW. All good. 

Aquila's house, planes physics model and screenshot of physics. 

[Anna Nova]

I just get more confused with all the 'planes/cubes/Firestorm/OfficialViewer/analyzed or not' stuff.   In the end I decided to just try things until they work.  I can't see any logic to it at all.  But it don't half take a while!  Thank God for Beta.....

Edited June 27, 2017 by anna2358
you can't abbreviate 'analyzed'.

[Aquila Kytori]

 

@Chic Aeon

Why use planes (not Analyzed) type Physics?

1: WYSIWYG.   Because you are not asking the Physics mesh to be Analyzed, apart from converting it to triangles, the SL mesh Uploader leaves it as it found it. The result is you get collision surfaces exactly where you designed them to be.

2: No problems when it comes to rezzing prims or mesh objects on the surfaces.

3: If made correctly, avoiding small or thin Tris/Quads then generally the Physics cost will be a lot lower than had you used the box method. Less than 1 for the “Chic style” house in the images above.

4: Its quick. Especially for this particular example,

    Starting with a copy of the visual mesh :

    Delete the walls with the windows,    delete the floor, (the visual mesh floor is 6 tris, it will be replaced by a single quad.) Delete the small faces of the step and delete the small thin  quads around the door opening.

    Fill in the floor, fill in the top of the step and finally add faces for the side walls, inside and outside.

 

etc   etc .................

 

 

 

 

Edited June 27, 2017 by Aquila Kytori

[Chic Aeon]

1 hour ago, Aquila Kytori said:

2: No problems when it comes to rezzing prims or mesh objects on the surfaces.

3: If made correctly, avoiding small or thin Tris/Quads then generally the Physics cost will be a lot lower than had you used the box method. Less than 1 for the “Chic style” house in the images above.

Great explanation and I really do get it now (taken awhile). Since you need to use cube physics on some Opensim Opensim grids and since I have been making cube physics for four years LOL, I of course used that in my (currently editing) tutorial. I DID however give an explanation of the other flavor *wink* and even have a screenshot on the beta grid. 

By the time I am finished the land impact will likely be determined by the download and not the physics but I realize the server would likely be happier with a lower "cost". 

Meanwhile I don't have any issues at all with rezzing prims on the floor or roof or whatever and the cube physics model doesn't change in land impact based on the physics (or very minimally) so that is something to consider depending on the build.  No?

Anyway thanks for the lesson!!!!   I may try planes on the next build. Always good to experiment and learn. 

 

[Beq Janus]

Hi @Chic Aeon

There's a whole bunch of little details in this thread that are worth explaining....

TLDR; version

FS 5.0.1 has broken crappy physics. FS 5.0.7 FTW

Havok support is important in very rare cases (analysed mesh only)

a bunch of other stuff.

Verbose version.....

Quote

When the physics are analyzed parts of the doors get closed off in the FS viewer. 

This is (probably) expected behaviour but it's impossible to tell for sure without seeing the meshes and the settings. The analyse function, tries to create a nice physics collision shape and will do it's best to match the shape of the mesh in the minimum number of "hulls". Typically, there are a range of settings available (which you can see in the lab viewer), but in the 64 bit viewer, and in the OS viewer regardless of 32/64 bit, the libraries used to perform the analysis are not present. As a result, only the most basic analyse function is provided and the results are pretty limited in anything but straight forward cases. In you case it may well just throw a hissy fit and fill in gaps.

Quote

  I have heard others complain in chat when they had to switch viewers in order to upload :D, but I have never had this problem before. I know there is a new FS out but I also need to finish some Opensim stuff in a few days so wanted to wait to install the new version. I am HOPING that the change is address in FS.

 

This is quite possibly a part of your problem. When we released Firestorm 5.0.1 (the bento release) just before Xmas, a "minor" bug slipped in from a Linden Lab code change that was merged into the final build. The bug completely messed up the normal way that bounding boxes and thus physics shapes were positioned. It was rapidly backed out by the Lab in one of their frequent updates, however, because we only release Firestorm periodically, mesh physics upload, particularly multi-part mesh, was problematic and the advice we gave was to use the Lab Viewer for uploads until the fix was available. The good news is that version 5.0.7 DOES fix the issue and uploading Mesh with Firestorm is once again back to normal.

On 2017-6-26 at 9:47 PM, Chic Aeon said:

Thank you. I will try that. I have never actually understood when and why  (and I remember you asking me long ago if I did) the difference and since Drongle told me long ago to analyze I almost always do unless I have some odd issue. I did forget to try that.

The root of this problem almost certainly lies in a combination of analysis and OS/non-havok versions of the viewer.

I'll explain why mesh physics is "different" in a bit. Analysed Physics is, in theory, more reliable and "better" for SecondLife, though in practice that is questionable.

So what does analyse do and why? 

Second Life uses a physics engine called "Havok" to do all of its physics calculations. The Havok engine uses special physics shapes, called "hulls" to optimise the calculation of collisions. The cheapest is the sphere (because collisions can be reduced to "is anything within 'radius' metres from my object centre"), Sphere and cylinders equally have a highly optimised physics collision test. The analyse function is used to convert a triangle based mesh into a set of "hulls", making a robust and efficient physics collision shape. Havok is a commercial engine, it is not available on OS and neither is it (currently) provided for 64-bit viewers. This is why the old viewer that @Aquila Kytori was using had that (misleading) popup. In practice, almost all Havok usage is server-side and hidden away but there are some functions that need viewer-side support. One such function is that physics analysis. In order to enable this, the Lab kindly makes a library available under a special sub-license but because they do not have a 64-bit viewer release they do not provide the Havok library. This is about to change because the lab is on the cusp of launching the 64-bit Alex Ivy viewer. We will then be able to provide a 64-bit Firestorm with FULL Havok support. Without the Havok support, Firestorm performs a "poor persons" analyse which is really not very good and a problem that was further exacerbated I suspect by using the bugged 5.0.1 release.

So what is the difference between non-analysed mesh and analysed mesh?

In non-analysed mesh, the physics engine has to use the triangle in the physics model to perform collision detection. Big triangles are nice and friendly and relatively low cost. long thin triangles are however a nightmare as the physics engine has to work really hard to work out whether things are intersecting. It is this reason, more than any other single reason, that explains why you should make you physics shape as simple as possible. It also explains one of the least well-understood issues with non-analysed mesh. The reason it gets more expensive as you scale it down. Smaller triangles, require far more accurate checks and thus cost more. As you shrink an object with non-analysed physics it's cost progressively grows. until it goes crazy high.....An example taken from my blog is this 9000LI tortured torus.

One of the primary advantages on analysed physics is that it is constant no matter what the scale (sadly in most cases it is "constantly" higher than a sensibly built triangle mesh).

@Aquila Kytori gives an excellent example of how a mesh physics can be made, very efficiently. It has one slight drawback which you may never notice in practice, which is that as a single planar physics surface the collision detection can sometimes "miss" allowing for things to pass through the wall. (it's not normally a real issue)

There is one other mistake that people make, and I have seen it so many times that it features in a blog post about physics issues.

It happens when people make a mesh wall, with a cutout window or door, and don't analyse it. If the wall is less that 0.5m thick then a "safety feature" of second life kicks in and forces the physics shape to behave as if it were a solid hull. The reason for this is simple and explained above if the LI were to continue to grow unchecked as the object got smaller then it would go asymptotic as its scale approached zero, the legendary infinite LI prim :-)

So in summary

Listen to Aquila, she talks sense

But, sometimes analyse IS the right choice, but not that often.

Ditch Firestorm 5.0.1 ASAP it is broken and buggy and we have given you a nice new shiny one. :-)

Look out for Havok enabled 64 bit firestorm in the future and it will give you many more analysis options for those times when you really need it.

 

And finally....

I am working on another blog post all about physics in the very near future. 

I hope this epic first ever post in the forums is vaguely useful, in answering a bunch of the unknowns.

 

Love Beq

x

Edited June 27, 2017 by Beq Janus
sprinkling a hint of grammar through the post

[Chic Aeon]

3 hours ago, Beq Janus said:

The good news is that version 5.0.7 DOES fix the issue and uploading Mesh with Firestorm is once again back to normal.

So I basically knew all that you mentioned, thanks. BUT BUT BUT ---

You missed the part where I tried the NEW just out FS viewer yesterday and got the very same result. It should be noted that before said issue -- I guess as you say it was at Christmas as I haven't been uploading a lot of houses -- the last being for Draftsman in January so done early as I am an early gal .

I was very much hoping that the just out viewer WOULD solve this issue but it hasn't for me. 

I have been using this same method and the same 64 bit SL -OS versions for years now with no problems and lovely walk through doors and rezzable floors. I only noticed a problem a few weeks ago on another build. It was fairly complex and I decided the uploader just really didn't want to think that hard :D. Then too the Linden Viewer uploaded as expected.    But THIS is a very simple build.  

So, while it could have just been me, I am not certain that things are really fixed. Since there are not that many folks that are uploading mesh with cube physics and / or that have upgraded to the newest version of FS I guess time will let us all know how well the improvements took. I got the exact same problem with the new version as with the previous (December) one. 

I just gritted my teeth and quickly uploaded the mesh with the Linden viewer. 

And yes, most of us on the forums know about the skinny walls issue. It has been around for maybe four years? I am sure all that very good info on the problem can be found in the archives -- those archives are sometimes very handy!

Hopefully lots of folks will learn some new stuff!  

 

Later edit:

I will go back and try one more time with the new Firestorm as I am a very good "beta tester", the first being Homesite 1.1 if any of you remember Homesite before it became assimilated :D.  

Will post results here. Will be a few days. Much on my agenda. 

 

 

Edited June 28, 2017 by Chic Aeon

[Chic Aeon]

Wanted to add something to this thread

I started a new -- larger, two-story and more complex house today and I used planes physics. I am sure that Aquila's physics model would have been better than mine but mine did work just fine and I am happy with it. I will work on refining my process in the future. The download cost is mostly because this is a very high LOD build as I want it viewable from far far away for most everyone :D.

AND yesterday while doing a tutorial, I uploaded a simple doorway using CUBE physics and the December version of Firestorm and it worked fine. It was very simple though, so it seems that the problems with cube physics and FS come when things get more complex. 

 

[Chic Aeon]

Just one more note on this :D.   Photos are broken on the forums so won't try and add one, but ---

Installed the new Firestorm, Firestorm 5.0.7 (52912) Jun 13 2017 03:57:58 (Firestorm-Releasex64) with OpenSimulator support  on my desk computer and tested again on an opening -- a fairly simple one as it was the opening at the top of the stairs (so think of it like a window). The physics still didn't work (I had tried in the previous FS version also). The physics work perfectly in the SL viewer. 

So for most purposes (at least in my tests), people using cube physics will need to use the Linden Viewer to upload anything that requires "walk through" including some corners until we get a new shiny model with the promised fixes. 

 

[arton Rotaru]

Third party viewers which have Opensim support are lacking the Havok physics decomposition libraries. So I think it's unlikely that your issues will ever be fixed in those viewers.

[Anastasia Horngold]

8 hours ago, Chic Aeon said:

Installed the new Firestorm, Firestorm 5.0.7 (52912) Jun 13 2017 03:57:58 (Firestorm-Releasex64) with OpenSimulator support  on my desk computer and tested again on an opening -- a fairly simple one as it was the opening at the top of the stairs (so think of it like a window). The physics still didn't work (I had tried in the previous FS version also). The physics work perfectly in the SL viewer. 

So for most purposes (at least in my tests), people using cube physics will need to use the Linden Viewer to upload anything that requires "walk through" including some corners until we get a new shiny model with the promised fixes. 

 

For physics comparable to the SL viewer, you have to use the 32-bit Firestorm for SL with Havok.

Edited July 2, 2017 by Anastasia Horngold

[mikka Luik]

The 32 bit TPV I use for OS seems to handle physics well enough - was playing with some doorframes/windows this evening and no problem walking through. To be on the safe side, all thats destined for SL is uploaded with the SL viewer of course.

[Chic Aeon]

3 hours ago, mikka Luik said:

The 32 bit TPV I use for OS seems to handle physics well enough - was playing with some doorframes/windows this evening and no problem walking through. To be on the safe side, all thats destined for SL is uploaded with the SL viewer of course.

Agreed that some things still work :D, but the minute it starts getting complex, it does not. And for OS if it is for someone besides you, best to test on the Sandbox III I am pretty sure on OS grid for UbOde.    LOTS of items uploaded in OS no longer work for those using that type of physics now. Much chatter likely a year ago so not really new news.

AND while I am typing on this thread.  I DID make some plane physics models and I do more or less understand it now, but (BUT) all my tests (and granted I may still be doing it wrong) had the physics with planes sort of "spongy" in that while there was a barrier, testing showed that part of your body went into the wall (that single plane collision surface I am guessing). I really don't like that.

I even had one wall that had a solid plane as the physics barrier and you could fall though one section which make absolutely NO sense to me at all. Again, might just be me. Still LOL. 

It seems messy and kind of "not right" to me. Since I am making things with long LODs now (and VERY long for my full sim LEA build, my physics cost is never the highest of my build. So I will likely be sticking with cube physics. I will also likely be "working around" things as much as possible so that I don't have to use the Linden viewer. 

 

 

[Beq Janus]

@arton Rotaru Havok being the issue only applies if the physics is analysed. If you upload the triangle mesh physics shape then that will not be the case. That said we are expecting to have the full Havok decomposition restored to the 64 bit FS in the next release....watch this space.

@Chic Aeon I'd love to work with you on resolving or at least understanding this if we can't resolve it. If it is an issue in the viewer then I can hopefully fix it, if it is a workflow thing then we can work that out.

 

 

[Chic Aeon]

30 minutes ago, Beq Janus said:

I'd love to work with you on resolving or at least understanding this if we can't resolve it. If it is an issue in the viewer then I can hopefully fix it, if it is a workflow thing then we can work that out.

Chatted at length in world folks. No real issues except my need for "all in one" house physics of the analyzed version.  Working around that some so that I don't have to delve into Linden Viewer land all that often. I never uploaded anything with the previous viewer in the problematic areas so for ME there was no difference between the new release and the previous one. 

 Looking forward to 64 bit Havok in the Fall!  In the meantime I am designing with as little need for Havok as possible. That works too. 

Just so the thread is a bit more complete. 

[arton Rotaru]

4 hours ago, Beq Janus said:

Havok being the issue only applies if the physics is analysed. If you upload the triangle mesh physics shape then that will not be the case. That said we are expecting to have the full Havok decomposition restored to the 64 bit FS in the next release....watch this space.

That's what I said.

I doubt that Linden Lab will allow Havok to be included in 64 Bit TPV with opensim support still.

[Whirly Fizzle]

14 hours ago, arton Rotaru said:

I doubt that Linden Lab will allow Havok to be included in 64 Bit TPV with opensim support still.

They won't.  There will be a two 64bit Firestorm viewer versions, one with Havok that can only connect to SL grids & one without Havok that can connect to any grid..

[arton Rotaru]

I know that indeed. Though, from what I understood of Chics posts, her problem revoles around convex decomposition which will be reqired for her opensim projects. And she was hoping for a viewer fix for that. Hence my reply that it's unlikely that this fix will happen.

[Chic Aeon]

Just wanted to clarify what I do and do not know. That a 64 bit FS viewer will be available (soon - in the Fall - September? all answers from reputable sources so I am considering that a "good" bit of info. As to timing who knows when the next viewer will be out ? -- lots of variables.  No one said anything to me about it being in the Opensim viewer.  One way or another. 

Since there are three physics versions in OS at the moment - Bullet, ODE and UbOde I can't see that it would be all that helpful and since many folks uploading in OS just let the default uploader do the work with NO physics model probably few folks care. 

I decided today (one of my ten second life changing decisions) not to build any more for OS. There are some great people there but the blatant, bragging, in your face theft is getting a bit hard to swallow. I have done lots of tutorials to help the folks there make their own goods. I have been supportive for almost three years, but my time there is pretty much over I think. All my close friends there have disappeared and I never made a ton of money -- again, just trying to give folks choices.

So, for ME, September or whenever will be a wonderful change.  Meanwhile I learned a bit about planes physics (good) and will just grit my teeth when I need Havok.  It's funny really as I didn't need it for hardly anything these last years but with bigger buildings going up in my LEA sim I "DO" need it (still not liking those spongy planes physics walls  ) unless I do some workarounds which I have been doing to a point.

That's all I know. Like so much (Sansar for one which has been in the news again this week), we will know when we see it happen. 

Well duh! Rereading I see that Whirly confirmed what I thought LOL. Late to the party am I. 

Edited July 8, 2017 by Chic Aeon

[Aquila Kytori]

Hi Chic

 

14 hours ago, Chic Aeon said:

It's funny really as I didn't need it for hardly anything these last years but with bigger buildings going up in my LEA sim I "DO" need it (still not liking those spongy planes physics walls  ) unless I do some workarounds which I have been doing to a point.

By "spongy" , do you mean that you can put your arm, leg or half of your avatar through the wall if you aproach to close?

If this is happening make sure all floors and ceilings have two collision surfaces.

A ground room floor should have the upper collision surface to walk around on and also an underside plane below it, with the normals facing in the opposite direction. Just as an second storey floor should have a ceiling plane and above that the floor plane.

 

 

I have sent you the two examples to test on the Beta grid; Pink is the single plane, "spongy" walls and the green one double plane, solid walls.

 

[Chic Aeon]

4 hours ago, Aquila Kytori said:

Hi Chic

 

By "spongy" , do you mean that you can put your arm, leg or half of your avatar through the wall if you aproach to close?

If this is happening make sure all floors and ceilings have two collision surfaces.

A ground room floor should have the upper collision surface to walk around on and also an underside plane below it, with the normals facing in the opposite direction. Just as an second storey floor should have a ceiling plane and above that the floor plane.

 

 

I have sent you the two examples to test on the Beta grid; Pink is the single plane, "spongy" walls and the green one double plane, solid walls.

 

Thank you once again and yes, that is what I meant as "spongy". I have trying an experiment with two planes "on my list" :D. But list is long LOL -- oh so long.  So it seems like I could use my "cube physics" and just take out the thin edges?  I KNOW that isn't as elegant as your example but I am such a creature of habit and for me, the cube physics models are so much faster --- LOL.   BUT, as always I am happy to LEARN and I will TRY .

 

One more time, "Thank you". 

 

[Chic Aeon]

Just and update for my TUTOR :D.   Started a little fantasy-esce house today and it had that problematic look AND I could actually "see" what I needed to do, so I did a planes physics test and it worked perfectly (in Firestorm). 

Just wanted Aquila to know that I was paying attention. 

2 land impact at 10 meters for the main room only. 

 

PS. I forgot an analyzed this the first upload (habit, habit). WOW, was THAT a MESS!   LOL. 

[Aquila Kytori]

Hmmmm........ !

Unfortunately Chic I can’t mark you higher than a 12/20* for this one. 

Firstly your screenshot, I’m guessing that the mesh on the right is the Physics mesh.  Why do I have to guess? It should have been clearly labeled as the Physics mesh.

 

Secondly, and this is really where you’re losing  points,  the first rule for creating  a planes type physics  mesh is, the planes /triangles should be as large as possible.  Optimize !

Just as your visual mesh should be optimized as much as possible so should the Physics mesh.

Vertices not contributing to the shape should be merged to ones that are.

 

* but you do get a 18/20 for design.

 

Edited July 11, 2017 by Aquila Kytori