Category: SL User Group Meetings

The following notes are taken from the Content Creation User Group meeting, held on  Thursday, November 16th, 2017 at 13:00 SLT. For the purposes of Animesh testing, the meetings have relocated to the Animesh4 region on Aditi, the beta grid – look for the seating area towards the middle of the region. The meeting is chaired by Vir Linden, and agenda notes, etc, are usually available on the Content Creation User Group wiki page.

Medhue Simoni live streamed the meeting, and his video is embedded at the end of this article – thanks to Medhue, as always, for the recording. Time stamps in the body text will open the video in a separate tab for ease of reference to the relevant parts of the text. However as these notes present the meeting in terms of topics discussed, rather than a chronological breakdown of the meeting, so some time stamps may appear to be out of sequence.

Animesh (Animated Mesh)

“I like the name ‘animated objects’ because I think it’s unambiguous, but it takes a long time to type!” – Vir Linden joking about the name “Animesh”.

Project Summary

The goal of this project is to provide a means of animating rigged mesh objects using the avatar skeleton, in whole or in part, to provide things like independently moveable pets / creatures, and animated scenery features via scripted animation. It involves both viewer and server-side changes.

In short, an Animesh object:

• Can be any object (generally rigged / skinned mesh) which and contains the necessary animations and controlling scripts in its own inventory  (Contents tab of the Build floater) required for it to animate itself.
• Can be a single mesh object or a linkset of objects (link them first, then set them to Animated Mesh via the Build floater > Features).
• Has been flagged as and Animesh object in the project viewer, and so has an avatar skeleton associated with it.
• Can use many existing animations.

However Animated objects will not (initially):

• Have an avatar shape associated with them
• Make use of an avatar-like inventory (although individual parts can contain their own inventory such as animations and scripts)
• Make use of the server-side locomotion graph for walking, etc., and so will not use an AO
• Use the avatar baking service
• Will not support its own attachments in the initial release.

These are considered options for follow-on work, possibly starting with the notion of a body shape (to help with more fully-formed NPCs).

Resources

• Animesh User Guide
• Animesh test content
• Animesh LSL methods:
  • llStartObjectAnimation
  • llStopObjectAnimation
  • llGetObjectAnimationNames
• Animesh feedback thread
• JIRA filter for Animesh

Mod Keys Proposal

For background information on this, please refer to the forum posts by Piscine Mackenzie and Medhue Simoni, together with feature request BUG-139168, together with my notes from the more recent CCUG meetings.

[0:46-2:53] The proposed scripted mod key capability to allow attachments to be made to No Mo Animesh characters has now been determined to be a much more extensive piece of work than originally thought, and so will not be implemented as a part of the initial Animesh project.

However, the Lab will be tackling it as a project in its own right “some time in the future”.

Remaining Bugs / Remaining Items

[3:32-4:55] The list of remaining bugs/ items to be looked at is summarised as:

• A couple of crash bugs in the viewer to be fixed. These are related to switching between the animated and non-animated states with an Animesh object.
• Code optimizations need to be completed and the Animesh code in general requires some clean-up.
• LOD rendering issues on Animesh objects needs to be fixed.
• Animesh imposting needs to be fixed.
• Left-clicking on rigged mesh doesn’t work well, and this applies to Animesh objects as well (right-click Touch should work), so needs to be addressed.
• Performance analysis needs to be carried out to determine necessary / best adjustments to costs / limitations (LI, tri count, complexity).

[4:56-5:45] Baking service updates: there is a couple of bugs in the baking service Anchor Linden is currently investigating, and which need to be resolved for Animesh. One of these can cause incorrect adjustments to be made to an avatar’s height calculation when an Animesh object is attached.

Any other issues people have noted not already filed via the JIRA (see the link to the JIRA filter, above), should be filed as bug reports ASAP.

Costs and Limits

[7:30-8:25] One of the reasons limits need to be examined is that currently, is that LI calculations are based on an object’s scale, whereas the LOD is effectively based on the rigged scale, potentially allowing the cost of the objects to be gamed between the two. This is in part the reason why the Lab has opted for tri counts being a limit with Animesh. However, Vir would like to get a fix in to prevent this kind of gaming; the problem here being the potential for breaking existing content.

[8:25-9:19] Testing will hopefully allows the cost of Animesh objects to be more defined by their complexity than on a fixed land cost, which will be a much smaller component in the overall limits than currently the case 200 LI), and will likely be a “base” LI based on the impact of the avatar skeleton.

[9:39-10:14] The eventual cost might be a combination of a base land impact and a complexity multiplier, although there are issues here with rigged objects being treated as unrigged when calculating rendering costs, which needs to be addressed.

[10:21-10:52] Vir is wary of making too many changes on impact / cost calculations just for Animesh, as there is a broader project to improve complexity cost calculations in general, which is currently on hold pending the introduction of Animesh.

[41:31-43:28] tri count vs Land Impact: in line with the above, Vir is not convinced that “just letting the Land Impact” determine if an Animesh can be rezzed is viable, as has been suggested. There are sufficient issues with the current avatar complexity calculations that, when fixed, could mean that even if a tri count is discounted as a direct Animesh constraint, avatar models converted to Animesh would still be prohibitively expensive from a complexity standpoint.

[43:47-45:16] As with mesh currently, Animesh will not bypass the accounting system if made temp-on-rez.

[45:23-54:50] More discussion on limits / costs, the need for testing, and on the Lab’s approach to preferring to apply costs over limits and the ability for users to determine what they see (as with avatar complexity / Jellydolls), balanced by a desire to get people to think more in terms of optimising their content. Part of the tension with limits / costs seems to be creators are unwilling to develop content, push at the current (arbitrary limits), etc., out of concern that the effort spent building and testing could be wasted / the Lab is hoping people will build, test (and even break) things (including the viewer) so a more accurate assessment of limits can be made prior to release, with the option of making further adjustments further down the road, if necessary.

[54:53-56:11] Would it be possible to have different tri count limits for worn Animesh vs rezzed Animesh – higher for the former compared to the latter? Possibly, but tricky.

Animesh as a Build Floater Option Discussion

[19:28-40:00] A lengthy discussion on having Animesh a toggle feature in the Build floater, rather than something set on upload, and how that might discourage makers of modifiable avatars from producing “Animesh versions” because users could in theory get a set of suitable scripts and animations and attempt to use them with the versions they already have.

The discussion is lengthy, and involves numerous issues, including the extent to which users trying to convert things themselves over buying a “dedicated” Animesh creation (would scripts and animations intended to animate an elephant really work that well in a tiger, for example?); creators’ rights vs. users’ rights with mod items; the decision process behind making the Animesh option a Build floater tool (e.g. it is in keeping with the permissions system – if an item if modifiable, it should include converting it to Animesh, if appropriate; it could help speed the adoption of Animesh – creators could release kits to convert their existing products, rather than entire new products, etc). The overall consensus appears to be having the option in the Build floater is better than trying to restrict it.

Animesh Sitting

[56:28-56:58] As discussed in past meetings, Animesh bodies will not be able to sit in the same manner as an avatar (by being made a child of the object on which it is sitting). Animesh should be able to adopt a sitting animation it contains, like a ground sit, however). The latter obviously won’t give the same flexibility as the former.

Bakes on Mesh

Extending the current avatar baking service to allow wearable textures (skins, tattoos, clothing) to be applied directly to mesh bodies as well as system avatars. This involves server-side changes, including updating the baking service to support 1024×1024 textures, and may in time lead to a reduction in the complexity of mesh avatar bodies and heads. The project is in two phases:

• The current work to update the baking service to support 1024×1024 textures.
• An intended follow-on project to actually support baking textures onto avatar mesh surfaces (and potentially other mesh objects as well). This has yet to fully defined in terms of implementation and when it might be slotted into SL development time frames.

This work does not include normal or specular map support, as these are not part of the existing baking service.

[11:47-12:55] Again, this is a separate project to Animesh, and neither is contingent upon the other. However, using bakes on mesh with Animesh objects will likely come after any follow-on Animesh project to the current work, as Animesh objects will require some notion of an avatar shape to work with the baking service.

[12:56-14:22] There are no plans at present to offer LSL functions or a script API for bakes on mesh, due to the complexities of the baking service. If such a capability is seen as needed, a feature request JIRA explaining why and the benefits should be submitted.

[16:35-19:12] The goal is to have the baking service work in such a way that creators can use the Appearance floater to assign the use of baked textures to object faces which correctly correspond to the different avatar elements (so for example, an upper body composite texture is correctly applied to the upper body faces on a mesh avatar). Once this has been set, users selecting the clothing layer would then see it correctly applied to their avatar in the same way as clothing layers are applied to the system avatar.

Next Meeting

The next CCUG meeting will be on Thursday, November 30th (week #48), due to week #47 being US Thanksgiving week, and the Lab taking a holiday break on Thursday, November 23rd / Friday November 24th.

 

The following notes are taken from the Content Creation User Group meeting, held on  Thursday, November 9th, 2017 at 13:00 SLT. For the purposes of Animesh testing, the meetings have relocated to the Animesh4 region on Aditi, the beta grid – look for the seating area towards the middle of the region. The meeting is chaired by Vir Linden, and agenda notes, etc, are usually available on the Content Creation User Group wiki page.

Medhue Simoni live streamed the meeting, and his video is embedded at the end of this article. Time stamps in the body text will open the video in a separate tab for ease of reference to the relevant parts of the text. However as these notes present the meeting in terms of topics discussed, rather than a chronological breakdown of the meeting, so some time stamps may appear to be out of sequence.

Animesh (Animated Mesh)

“I like the name ‘animated objects’ because I think it’s unambiguous, but it takes a long time to type!” – Vir Linden joking about the name “Animesh”.

Project Summary

The goal of this project is to provide a means of animating rigged mesh objects using the avatar skeleton, in whole or in part, to provide things like independently moveable pets / creatures, and animated scenery features via scripted animation. It involves both viewer and server-side changes.

In short, an Animesh object:

• Can be any object (generally rigged / skinned mesh) which and contains the necessary animations and controlling scripts in its own inventory  (Contents tab of the Build floater) required for it to animate itself.
• Can be a single mesh object or a linkset of objects (link them first, then set them to Animated Mesh via the Build floater > Features).
• Has been flagged as and Animesh object in the project viewer, and so has an avatar skeleton associated with it.
• Can use many existing animations.

However Animated objects will not (initially):

• Have an avatar shape associated with them
• Make use of an avatar-like inventory (although individual parts can contain their own inventory such as animations and scripts)
• Make use of the server-side locomotion graph for walking, etc., and so will not use an AO
• Use the avatar baking service
• Will not support its own attachments in the initial release.

These are considered options for follow-on work, possibly starting with the notion of a body shape (to help with more fully-formed NPCs).

Resources

• Animesh User Guide
• Animesh test content
• Animesh LSL methods:
  • llStartObjectAnimation
  • llStopObjectAnimation
  • llGetObjectAnimationNames
• Animesh feedback thread

Viewer Progress

[3:51-5:33] The project viewer updated to version 5.0.9.329815 on Thursday, November 9th.

• The major change is that Animesh linksets can now include non-rigged items (e.g. prims), and allow the root objects to be a prim rather than a rigged mesh.
• Animesh objects should now render the same way as mesh attachments and generate shadows.
• The viewer also includes a number of bug fixes.

[42:25-44:50 + text chat] However, while the root of an Animesh object can now be a prim, this doesn’t mean mesh objects such as pets can be dropped in-world (e.g. so someone could “pick up” their Animesh pet, carry it, then put it down again somewhere else), because the code required for “dropping” mesh is not provided on the simulator on the ground that dropping meshes can cause physics issues.

Animesh Attachments

Note: there is an extensive discussion in text chat around this subject from the 13:00 minute mark which continues throughout the meeting.

[12:59-18:06] One of the major subjects of discussion with Animesh is enabling attachments to be added to existing Animesh objects (see here and here). In short, many pets and characters are created as No Modify objects, to help protect their capabilities. Thus, as No Mod objects, Animesh characters cannot be accessorized (the owner of an Animesh horse cannot add  / remove a saddle from an Animesh horse, for example). As Piscine Mackenzie explains in the forum, with highlights by Medhue Simoni, making Animesh pets  / characters Modify opens them to the risk of exploitation.

One proposal has been put forward via feature request BUG-139168. Vir Linden has also put forward a number of ideas. Of these, the mod key proposal (2nd bullet point in Vir’s notes and somewhat aligned to BUG-139168) is seen as a potential solution, with LlAllowInventoryDrop used to avoid the need for the root Animesh object having to be modify (and leaving it exposed to the exploits outlined by Piscine and Medhue). Vir is currently seeking further feedback from people on this being a possible approach, or if there are other potential alternatives.

• [21:11-22:51] Two concerns on this approach are that a) any scripted mod key capability could be withdrawn at any time by the creator, potentially leaving users with broken content. However, there is no obvious means of safeguarding against this (unless it is made some kind of one-time function operation by the Lab, if possible); b) there is a risk the mod key could be accidentally exposed / leaked in being shared between those using it.
• [25:52-26:12] The most “obvious” means of handling this issue would be to implement a change to the permission system to cater for the specific user-case of allowing attachments to be added to Animesh objects. However, the Lab is hesitant to consider such changes due to the risk of unintended impact, risk of content breakage, etc. As such a precise proposal on how to update the permissions system and the use cases it would meet would have to be properly defined for consideration.
• [28:15-30:40] A suggestion is made to have a new flag “only allow scripts compiled by the object creator run on this object”. However, this could limit cases where models are built by one person, animated by another and scripted by someone else, or situations where a creator is using full permission scripts sold by another creator, etc., and so seen as potentially less than ideal.
• [56:07-1:01:56] Determining whether Animesh attachments capabilities (e.g. using the mod keys idea) can or cannot be implemented relatively easily is seen by the Lab as a critical aspect to moving Animesh towards a release. If doing so proves problematic, the project will likely move ahead without such a capability being implemented. In the meantime, as there isn’t currently an available solution, creators are advised to handle Animesh the same as any other product they are developing for the time being, and focus on Animesh at the functional level.
  • [1:01:57-1:05:05] An example of complexity is defining how a mod keys capability would work. If it is limited to one specific group of people creating a poduct line, it is relatively easy to control. However, if a creator wants too support an ecosystem whereby other creators can all offer attachments for a product, then it becomes far more complicated in terms of keeping the mod key secure.

Given the complexity of this subject, continued discussion through the Animesh forum thread is encouraged.

Animation Playback Speed

[51:47-52:37] It has been noted that the speed of the animation depends on the distance from the viewer’s camera to the animated object – see BUG-134259 and this MP4 (note particularly the two elephants). This appears to be the result of the debug UseAnimationTimeSteps being set to TRUE (default viewer behaviour). Setting it to FALSE fixes the issue, and doesn’t appear to cause any other performance impacts.

Land Impact and Other Limits

[54:53-56:05] It’s unlikely that the current baseline limits (LI, tri count, rendering), will be altered  until there is confidence that the viewer functionality is relatively stable in terms of features, bugs, etc., at which point reliable performance testing can begin. Once this point is reached, then the Lab will be able to gather reliable data in order to start adjusting / potentially relaxing the limits.

Environment Enhancement Project (EEP)

Project Summary

A set of environmental enhancements, including the ability to define the environment (sky, sun, moon, clouds, water settings) at the parcel level; a new environment asset type that can be stored in inventory and traded through the Marketplace / exchanged with others; scripted, experience-based environment functions, an extended day cycle and extended environmental parameters. This work involves both a viewer updates (with a project viewer coming soon) and server-side updates.

Current Status

[8:25-10:32] Rider Linden is “very close” to removing the old windlight environment code from his EEP version of the viewer, allowing it to use all the new EEP assets, allowing the local environment to be set via windlight inventory assets.

Once this can be done, test regions capable of supporting the assets will be established on Aditi (the beta grid), and a project viewer will be made available for more general testing.

Bakes on Mesh

Project Summary

Extending the current avatar baking service to allow wearable textures (skins, tattoos, clothing) to be applied directly to mesh bodies as well as system avatars. This involves server-side changes, including updating the baking service to support 1024×1024 textures, and may in time lead to a reduction in the complexity of mesh avatar bodies and heads. The project is in two phases:

• The current work to update the baking service to support 1024×1024 textures.
• An intended follow-on project to actually support baking textures onto avatar mesh surfaces (and potentially other mesh objects as well). This has yet to fully defined in terms of implementation and when it might be slotted into SL development time frames.

This work does not include normal or specular map support, as these are not part of the existing baking service.

Current Progress

[35:41-36:33] The performance testing has been carried out, and the results point to the need for additional bake host servers to account for the additional load in handling the higher resolution textures.

Other Items – In Brief

• [36:30-37:29] Region crossing loads: the question is asked about what is the biggest load / cause of lag on region crossings (direct or TP). The short answer is scripts. That is, suspending them, saving their state, packaging them, handing them off to the next simulator, unpacking them alongside everything else (e.g. the avatar and / or vehicle they are running against), setting them back to running from their saved state. Obviously, the more individual scripts an avatar (+ vehicle, where used) is running (regardless of individual script size), the bigger the load placed on the hand-off process / receiving region.
• [47:03-47:19] Materials system update: this is frequently requested, but currently the Lab do not have any plans to revisit the materials system.
• [49:32-49:45] Cannot pathfind through a hollow object – intentional? Short answer – yes; physics simulations going into concave objects is expensive
• [49:45-50:18] Will RenderVolumeLOD be revisited? – Short answer – yes, as a part of the work in re-evaluating rendering costs, land impact, etc.