Mesh/Archive/2011-07-11
< Mesh
Jump to navigation
Jump to search
Revision as of 15:17, 11 July 2011 by Charlar Linden (talk | contribs) (→Transcript for Monday July 11, 2011)
Agenda
- Possibility of 'cheap' meshes that have the same collisions as sculpted prims, in order to reduce their impact on the simulators. --Kyrah Abattoir 16:42, 2 July 2011 (PDT)
- Is this the final prim equivalency accounting as it is right now, or are you still working on it? [Vivienne Daguerre]
- Given that people will work around the increased prim count on linking mixed tool sets by not linking them, will you reconsider taking prim and sculpt complexity into accounting on linking to mesh? I think this also breaks legacy content, sculpt tool kits for builders--windows, doors, arches, etc. These things were made to be linked to builds. If a sculpt or a prim ever counts as more than one prim, linked or not to other things, then in my opinion legacy content is broken.[Vivienne Daguerre]
- Using different DAE files for the various LODs and Physical model gives undesirable results if the bounding box is not the same in all files: each file/object is scaled and centered individually in the upload process, resulting in uneven scale and offsets (see also JIRA CTS-244). Adding a bounding box object in all the files and including it in the exported scene does not help (even multiple objects in a DAE file get scaled and centered independently). How do we deal with optimization for objects that have extruded details only needed for higher LODs ? Can't all the objects involved in the model upload share the same space ? [Bunnie Mills]
- Please provide documentation of the algorithm used for calculating mesh upload fees. This is necessary for sensible optimisation. [Drongle McMahon]
- Which physics shape type is the physics cost shown in the uploader for, and are we going to get the option to change it? [Drongle]
- The current limitations for vehicles (32 linked parts, 32 PE, 32 physical cost) seem too tough in the PE figure to allow good detail when many rounded shapes are involved. Maybe keep the max physics cost low, but allow some overhead in the PE (even 48 would make a difference) ? [Bunnie Mills]
- The uploader requires that each LOD's model has the same number of faces (materials). This means some drastic optimization won't be possible or some workaround should be used, like keeping at least a few vertices for each material in the reduced models, just to preserve the face count in the upload process. Please clarify [Bunnie Mills]
- Can you confirm that a 64M prim used as a regular prim will be a regarded as a legacy prim? [Graham Dartmouth]
- Possible bug in the importer for rigged av's. I've checked my vertex groups over and over. This seems to be an issue with the importer, and subtle too. Read jira for details. Hoping it was error on my part, but I did what I could to rule out possibilities. https://jira.secondlife.com/browse/CTS-689 [Raz Welles]
- Are the upload fees and PE costs finalized once Mesh is rolled out in the main grid RC channels? If not, is there a chance of content loss when using mesh on RC sims? [Maygray Heron]
- What is the LOD-Generator in the mesh-client based on? Any chance on an optional seam-preserving algorithm across all LODs? (usefull for large-scale terrains to prevent gaps between different LOD levels) [Maygray Heron]
- What solution can be achieved to address the problems associated with creating and selling mesh clothing items since you have stated you will not develop any type of parametric deformer or cage system to compensate for different avatar sizes and shapes. Currently, clothing is almost unusable/unmarketable unless is it a complete outfit from neck to feet, or an entire avatar, which is even then very limited in adjustability. Being that clothing is one of the largest markets in SL, this would seem to be a viable issue that needs to be solved. The same problem occurred in bluemars, and was only solved with the parametric deformer system, which is similar to how clothing textures currently "shrink fit" to your shape. This would also alleviate the need for skeletal rigging, which is extremely difficult. [Maxwell Graf]
- What is the primary aim of introducing mesh, i.e. - what do you want to see us use it for? [Maxwell Graf]
- Is the license of the llm file of the Viewer LGPL? Or is it CC BY 3.0? [Yuzuru Jewell]
- Any news on pivot points? Are pivots working again? [Ashasekayi Ra]
- Any documentation/glossary releases for the layman yet? [Eleanora Newell]
- Is anyone currently working on mesh planning to leave Linden Lab immediately after it is released? Draconis Neurocam 12:45, 11 July 2011 (PDT)
- can we get a central location listing of tutorials by beta users? [Eleanora Newell]