2024 week #20: SL CCUG summary

The Butterfly Effect, May 2024 – blog post

The following notes were taken from my audio recording and chat log transcript of the Content Creation User Group (CCUG) meeting held on Thursday, May 16th, 2024.

Meeting Purpose

  • The CCUG meeting is for discussion of work related to content creation in Second Life, including current and upcoming LL projects, and encompasses requests or comments from the community, together with related viewer development work. This meeting is held on alternate Thursdays at Hippotropolis.
  • In regards to meetings:
    • Dates and times are recorded in the SL Public Calendar.
    • Commence at 13:00 SLT on their respective dates.
    • Are conducted in a mix of Voice and text chat.
    • Are open to all with an interest in content creation.
  • The notes herein are a summary of topics discussed and are not intended to be a full transcript of the meeting.

Official Viewers Status

  • Release viewer: Maintenance X RC (usability improvements), version 7.1.7.8974243247, dated May 8and  promoted May 13 – NEW
  • Release channel cohorts (please see my notes on manually installing RC viewer versions if you wish to install any release candidate(s) yourself).
    • Maintenance C RC (reset skeleton in all viewers), version 7.1.7.8820704257, May 6.
    • Materials Featurettes RC viewer, version 7.1.7.8883017948, May 2.
    • Maintenance B RC (usability updates / imposter changes), version 7.1.7.8820696922, April 29.
  • Project viewers:

Graphics / glTF

Featurettes Viewer – Recap

  • Support for:
    • Viewer-side setting of PBR materials for terrain.
    • 2K texture upload support.
    • Mirrors: using reflection probes to generate local static or dynamic (incl. avatar) reflections in real-time.
  • Both Geenz Linden (mirrors) and Cosmic Linden (PBR terrain) are engaged in follow-up work for both of these projects, which will hopefully appear in a future graphics featurette viewer – not the current version (e.g. allowing PBR materials on terrain to have custom repeats, offsets, rotation).
  • It terms of PBR terrain, it was noted:
    • Emissive is supported on terrain.
    • There is a issue with Mac systems (either emissive or normal may “fall off”).
    • As running all four PBR materials on terrain can put a load on the viewer, there are some automatic fallbacks for systems unable to handle the full load which mean some of the materials may not be applied
    • Those opting to use PBR for terrain, as the viewer capability gains adoption (and region owners enable it), there is not requirement to also provide textures for terrain (to allow for those on non-PBR viewers).

General Notes on glTF / PBR

  • As an extension to the last point above, it was further noted that as PBR gains adoption, there is no enforced requirement for creators to continue to provide fallbacks to handle “non-PBR” viewers.
    • This lead to a further statement (first made at the May 10th TPV Developer meeting) that creators building for SL21B are not required to provide Blinn-Phong materials fallbacks on their builds when using PBR.
    • The above can be taken as an informal view that viewers need to support PBR either fully or with a “beta” release by the time of SL21B (commencing Friday, June 21st, 2024).
  • One of the reasons Linden Water reflections are not as good under PBR is due to attempts to optimise the rendering load so that any potential frame rate hit is minimised for more users.
    • Geenz Linden is looking at improving them once more as a result of the work done for mirrors. This work, if successful won’t bring water reflections up to the standard of pre-PBR water reflections (updated every frame), but will offer an improvement over what is currently seen with PBR.
    • Water as a material has been requested and is something LL “is interested in”, but this is not part of the current glTF / PBR work.
  • LL believe that most users should not see a dramatic fall-off in viewer FPS as a result of PBR, such that their experience is heavily impacted.
  • There remains no current plan to support displacement maps.
  • As well as working on mirrors, Geenz Linden is now working on Index of Refraction (IOR) and transmission, both in line with the glTF 2.0 specification (e.g. allowing transparent surfaces that distort what is behind them, as can sometimes be seen in the physical world).

glTF Scene Import

  • Recap:
    • Runitai Linden is continuing to work on glTF scene import. This has reached a point where (on test viewers) it is now possible to preview a scene (tied to an in-world object) in-world.
    • The initial aim is to get to a point where scenes can be imported and seen, and nodes within them updated with both tools in the viewer and / or using LSL, and ensuring they stay in synch with the rest of the scene.
    • Scenes are liable to use the MSFT glTF extension for Level of Detail (LOD), as this allows LODs to be set per node within a scene, providing more intuitive / consistent LOD switching management (based on screen coverage).
    • There will be constraints placed on scene imports (e.g. will not be able to have a scene which exceeds the capacity of a region; scenes will not be able to span more than one region (so as to avoid issues with physics, etc.); and so on).
  • In terms of scene import / export, while it is still earlier days, Runitai noted again that:
    • The idea is to make the import / export a two-way street for creators so they can modify their scenes with relative ease by taking it back to Blender to fix issues that cannot be fixed in-world.
    • This will be subject to the permissions system to prevent the wholesale export of content added to a scene.
    • Scenes uploaded will have attributions appended by the asset service (we owns the asset, when it was uploaded, etc).

In Brief

  • PBR and Bakes on Mesh:
    • LL “keenly aware” this needs to be done, and is noting support for it coming via the Feedback Portal.
    • However – it is not currently a live project.
    • The discussion also again carried the caveat that updating the Bake Service to support 2K textures is only part of the issue: all SL wearable definitions all also need to be revised to be able to support 2K textures, as these are also currently limited in their resolution.
  • There was an extended discussion on content ripping from Second Life and copybotting (with the latter covering a number of flavours). In short:
    • The is the hoary old (but nevertheless true) acknowledgement that Second Life relies on local computer rendering; ergo, regardless of whether it is via a”copybot viewer” or not, content can be ripped from SL and used elsewhere by those determined to do so.
    • Insofar as content being ripped and re-sold in Second Life (or elsewhere) there is the DMCA process – and in terms of other platforms, it does not matter if they have a DMCA Portal or not – a formal identification of content being misused (with proof of ownership) can still be sent to them in order to initiate a take-down process.
    • There was discussion on why “obvious” copyright-infringing content in SL was not taken down (with the example of Chanel products being used). As I understand it (and IANAL), this is bound up with the Safe Harbor provisions of the DMCA, design to provide immunity to ISPs and website operators for copyright infringement committed by their users by providing them with a framework for taking action. Pro-actively removing content potentially circumvents this process, leaving the company open to possible legal action if mistakes are made (and / or for simply not pre-actively addressing all infringements, large and small).
  • There was also a discussion on animations under glTF. However, as this will be a future element of the glTF scene support project, so I’ll leave specifics until such time as the work starts and LL talk more directly as to what they are doing and when.
  • The latter part of the meeting included (an occasionally heated) discussion on the new Modesty Layers being proposed by Linden Lab (see: 2024 SL Governance meeting week #19: Child Avatar Policy). However, as this was somewhat entangled with matters of policy, and those directly involved in making the changes to said policy / overseeing changes were not available  / present at the meeting, ad what was suggested as to possible technical solutions was somewhat speculative, I would prefer to leave further reference out of this summary, and leave matters until those at Linden Lab responsible for the policy have more fully engaged directly with content creators.

Next Meeting

† The header images included in these summaries are not intended to represent anything discussed at the meetings; they are simply here to avoid a repeated image of a gathering of people every week. They are taken from my list of region visits, with a link to the post for those interested.

2024 week #18: SL CCUG summary

Venus by the Water, April 2024 – blog post

The following notes were taken from my audio recording and chat log transcript of the Content Creation User Group (CCUG) meeting held on Thursday, May 2nd, 2024.

Meeting Purpose

  • The CCUG meeting is for discussion of work related to content creation in Second Life, including current and upcoming LL projects, and encompasses requests or comments from the community, together with related viewer development work. This meeting is held on alternate Thursdays at Hippotropolis.
  • In regards to meetings:
    • Dates and times are recorded in the SL Public Calendar.
    • Commence at 13:00 SLT on their respective dates.
    • Are conducted in a mix of Voice and text chat.
    • Are open to all with an interest in content creation.
  • The notes herein are a summary of topics discussed and are not intended to be a full transcript of the meeting.

Official Viewers Status

  • On Thursday, May 2nd, the Materials Featurette RC viewer updated to version 7.1.7.8883017948.

The rest of the official viewers remain as:

  • Release viewer: 7.1.6.8745209917, formerly the Maintenance Y/Z RC (My Outfits folder improvements; ability to remove entries from landmark history), dated April 19 and promoted April 23.
  • Release channel cohorts (please see my notes on manually installing RC viewer versions if you wish to install any release candidate(s) yourself).
    • Maintenance B RC (usability updates / imposter changes), version 7.1.7.8820696922, April 29.
    • Maintenance X RC (usability improvements), version 7.1.6.8758996787, April 23.
  • Project viewers:

Graphics / glTF

  • Mirrors: Geenz Linden continues work on bug fixing (notably the fact that mirrors under Linden Water seem to break) and additional updates to improve usability.
  • PBR Terrain:
    • Cosmic has fixed an issue wherein 2K textures were not rendering on the mini-map.
    • In addition, she is looking into the ability to customise repeats in PBR terrain.
  • It is known that screen space reflections (SSR) do not work well on Linden Water, producing moiré effects. There has been talk within the Graphics team of “ripping out” SSR and replacing it with “something” – but no determination has been made as to what should be used, and it is pending work elsewhere within the graphics / rendering system.

glTF Scene Import

  • Runitai Linden is continuing to work on glTF scene import. This has reached a point where (on test viewers) it is now possible to preview a scene (tied to an in-world object) in-world.
  • Work is now being done to set-up a couple of test region on Aditi (the Beta grid) where this can be more widely tested.
  • The overall status of the glTF scene import work is described as “prototyping with much brokenness”, and the project is liable to continue through summer in order to get it into something of a more productive state.
  • The initial aim is to get to a point where scenes can be imported and seen, and nodes within them updated with both tools in the viewer and / or using LSL, and ensuring they stay in synch with the rest of the scene.
  • Once the above has been reached, then the plan is to start looking at performance, working out a Land Impact accounting schema, etc.
    • In as much as Land Impact is concerned, it is likely that nodes within a scene will have their own Land Impact, with those also involved in skeletal animation having an additional Land Impact.
  • Scenes are liable to use the MSFT glTF extension for Level of Detail (LOD), as this allows LODs to be set per node within a scene, providing more intuitive / consistent LOD switching management (based on screen coverage).
    • For automatic LOD generation, the first pass of work might only use the automatic LODs generated by Blender; if a home-ground automatic LOD generation system is used, it will “almost certainly” be Mesh Optimiser.
    • Overall, glTF scene support should leverage Blender’s LOD generation, as Blender is “way better” at this than LL are “any day of the week”.
    • Managing LODs (e.g. setting a default LOD Factor in the viewer is liable to become more firmly set and managed as glTF scene import becomes available (e.g. all viewers working to a fixed LOD of 3.00 or something).
  • The will be constraints placed on scene imports (e.g. will not be able to have a scene which exceeds the capacity of a region; scenes will not be able to span more than one region (so as to avoid issues with physics, etc.); and so on).
  • It was noted that glTF scene support will require a lot of work in areas such as physics management, update message handling (the object-based messaging currently used will not work with glTF scenes). Not of these aspects are seen as potential blockers – but solving all of them will take time.
  • glTF scenes will present more of a two-way street for creators, moving more seamlessly between a design tool such as Blender and SL when building and testing a scene and, following the upload of a scene, the ability to save it back locally again (subject to the SL permissions system where content has been subsequently added to a scene after its initial upload).

In Brief

  • Animations: it was noted that under glTF there is support for at least eight morph channels (with one vertex attribute being considered a channel, so it could be that 2 morphs change 3 attributes), making animation far more flexible in terms of use.
  • Some using the SL Mobile app have apparently been reporting that PBR terrain textures are not showing in that app, as it currently does not have PBR rendering support – this has been passed to the Mobile team.
  • It was noted that one of the reasons for LL adopting the glTF specification is that it is mobile friendly, so the Mobile team is going to be adding PBR, etc., support.
  • A lot of the meeting was taken up with a general discussion on matters relating to glTF scene import – LOD management, collaborative building, licensing matters (e.g. if an asset is available on a third-party asset store with a specifically licensed use, and importing it to SL potential break that license – should LL proactively block the asset from being imported?), etc., which have yet to be fully determined internally at LL, and so fall outside of this summary.

Next Meeting

† The header images included in these summaries are not intended to represent anything discussed at the meetings; they are simply here to avoid a repeated image of a gathering of people every week. They are taken from my list of region visits, with a link to the post for those interested.

2024 week #16: SL CCUG summary

Selen’s Gallery, April 2024: Gem Preiz – Do You Like Gold? blog post

The following notes were taken from my audio recording and chat log transcript of the Content Creation User Group (CCUG) meeting held on Thursday,  April 18th, 2024.

Meeting Purpose

  • The CCUG meeting is for discussion of work related to content creation in Second Life, including current and upcoming LL projects, and encompasses requests or comments from the community, together with related viewer development work. This meeting is held on alternate Thursdays at Hippotropolis.
  • In regards to meetings:
    • Dates and times are recorded in the SL Public Calendar.
    • Commence at 13:00 SLT on their respective dates.
    • Are conducted in a mix of Voice and text chat.
    • Are open to all with an interest in content creation.
  • The notes herein are a summary of topics discussed and are not intended to be a full transcript of the meeting.

Official Viewers Status

  • On Friday, April 19th, the Maintenance YZ RC ( My Outfits folder improvements; ability to remove entries from landmark history updated to version 7.1.6.8745209917.
  • On Wednesday, April 17th, Maintenance X RC viewer (usability improvements) updated to version 7.1.6.8709279750.

The rest of the official viewers remain as:

  • Release viewer: version7.1.5.8443591509, formerly the Maintenance-W RC (bug and crash fixes), version and issued March 29th, promoted April 9th –
  • Release channel cohorts:
  • Project viewers:

It is possible the glTF Maintenance-2 RC viewer might be promoted to de facto release status early in week ’13 (commencing Monday, March 25th, 2024).

Graphics / glTF

  • Work continues on the GLTF Featurettes viewer + support.
    • The viewer is in RC with on-going bug fixes.
    • The simulator support is now on a Preflight channel on Agni (the Main grid), which includes the regions Rumpus Room 2048; Rumpus Room 2049; Rumpus Room 2050 and Rumpus Room 2051.
  • 2K Textures:
    • Are not currently supported for upload onto the Main grid.
    • The fees for 2K textures have been set as follows:
      • Basic / Plus account: L$50 per texture.
      • Premium account: L$40 per texture.
      • Premium Plus: L$10 per texture.
    • 2K textures for Bakes on Mesh requires a substantial update to the Bake service (as is frequently mentioned at CCUG meetings), and currently this does not appear to be in the works.
  • Mirrors:
    • Now leverage the Dynamic flag. When set, this tells the mirror to reflect everything in the rendering pipeline including avatars and particles). if the flag is not set, avatars are particles are not rendered, just the environment (e.g. for a reflective floor which reflects the room, but rendering the avatars would cause a massive performance hit).
    • Mirror probes can now be set using LSL.
    • The updated UX for Mirrors is still not done, and it is acknowledged that setting up a mirror is far from intuitive. A good 10+ minutes of the meeting was devoted to “how to make a mirror” – rather proving the point!
  • Work has started on prototyping glTF scene handling (as part of the overall glTF scene import work). Initial work is to enable the showing of a scene locally within the viewer using a debug setting. More on this as the option becomes available in a viewer. But in short:
    • Once imported, glTF scenes will be an inventory asset type.
    • Each scene with by associated with an in-world object (e.g. a prim), which becomes its “handle”.

In Brief

  • Custom skeletons: this is an ongoing debate within LL. glTF allows for custom skeletons and animations. Second Life has a defined avatar skeleton and recognised animation formats for use against it.
    • As such, there is a disparity between SL and glTF, as in theory, adherence to the glTF specification means recognising the use of custom skeletons with their own animations – which would then be incompatible with the existing SL format(s), and lead to potential confusion  / clashes between established avatar types using the SL skeleton and custom skeletons.
    • Thus, some navigation is required to determine if / how custom skeletons might be handled without confusing the existing consumer market in SL in terms of what accessories work with which type of avatar.
  • Emissive strength is seen as “feeble” by some, so a request was made for a numeric box in which a desired emissive strength might be entered. As the glTF specification allows for this, it is something that could be added to a future viewer iteration (but not necessarily would be – Feature Request?).
  • Much of the meeting was about what extensions LL might consider, how things like emissive maps and lighting might be handled in the future, IOR, transmission and the potential to add iridescence – some of which are being worked on (IOR and transmission) and will be reported on in summaries as they come along, others are seen as possibly / maybe.

Next Meeting

† The header images included in these summaries are not intended to represent anything discussed at the meetings; they are simply here to avoid a repeated image of a gathering of people every week. They are taken from my list of region visits, with a link to the post for those interested.

2024 week #14: SL CCUG summary

Memories of Dreams, April 2024 – blog post

The following notes were taken from my audio recording and chat log transcript of the Content Creation User Group (CCUG) meeting held on Thursday, April 4th, 2024.

Meeting Purpose

  • The CCUG meeting is for discussion of work related to content creation in Second Life, including current and upcoming LL projects, and encompasses requests or comments from the community, together with related viewer development work. This meeting is held on alternate Thursdays at Hippotropolis.
  • In regards to meetings:
    • Dates and times are recorded in the SL Public Calendar.
    • Commence at 13:00 SLT on their respective dates.
    • Are conducted in a mix of Voice and text chat.
    • Are open to all with an interest in content creation.
  • The notes herein are a summary of topics discussed and are not intended to be a full transcript of the meeting.

Official Viewers Status

  • On Wednesday, April 3rd, the Materials Featurettes RC viewer updated to version 7.1.5.8472515256.
  • On Tuesday, April 2nd, the Maintenance X RC viewer (usability improvements) updated to version 7.1.5.8443777128.

Both of these updates were to bring the above viewers to code parity with the current release viewer (with bug fixes to to the Featurettes viewer as well). The rest of the official viewers remain as:

  • Release viewer: version 7.1.4.8149792635, formerly glTF PBR Materials Maintenance-2 RC viewer, promoted March 26, 2024.
  • Release channel cohorts:
    • Maintenance-W RC (bug and crash fixes), version 7.1.5.8443591509, March 29.
    • Maintenance Y RC ( My Outfits folder improvements; ability to remove entries from landmark history + Maint Z RC integration), version 7.1.5.8448596295, March 29.
  • Project viewers:

It is possible the glTF Maintenance-2 RC viewer might be promoted to de facto release status early in week ’13 (commencing Monday, March 25th, 2024).

Graphics / glTF

  • Work continues on the GLTF Featurettes viewer + support.
    • The view is in RC, but the simulator-side support is still working its way through the system for full deployment on Agni (the Main grid).
    • The RC viewer has been undergoing numerous bug fixes related to both PBR Terrain textures and Mirrors, some of which are in the most recent update, others of which will be included at the viewer continues to be updated.
    • The UI/UX changes required for Mirrors to overcome confusion relating to the placement of things like Mirror reflection probes (with the creation of Mirrors described a “not intuitive”) is still under discussion at the Lab and so yet to be implemented within the Featurettes viewer.
    • The pricing structure for uploading 2K textures is currently being finalised, although it is not ready to be formally announced.
      • As with other services / uploads, the pricing looks like it will be based upon subscription level (e.g. cheaper for the likes of Premium Plus / Premium than for Plus / Basic) and the overall resolution of the texture.
      • Further work will also be required on the back-end services to fully support the upload of 2K textures.
      • An upshot of this is that at the time of writing, there is a communications mismatch between the viewer and the Aditi test regions supporting 2K texture uploads, resulting in such attempts falling over.
  • The HDRI support (replacing the sky with an HDRI file when rendered locally within a viewer to assist creators in checking to confirm their objects render in SL as they are seen within the tools they use to create their content – e.g. Substance Painter – and any variations are not related to the shader / import pipe) is still being tweaked.
    • In particular, now it has been tested internally and seen as offering some potentially interesting uses over and above pure EEP settings, consideration is being given to making HDRI an actual sky setting with simulator-side support.
    • One aspect of this would be the ability to use the lighting effects of HDRI within EEP sky settings, thus providing more true-to-life lighting without losing the use of existing EEP settings. Another might be using HDRI maps inside buildings and structure and not having the sky setting lighting leak into the structure.
    • Again, it’s important to not adding back-end support for HDRI is a future implementation, rather than anything being worked on right now, and would require work on both the simulator and back-end services (e.g. a new asset type to support HDRI would be required).
  • Light sources: there have been requests for the likes of point light shadows (e.g. so a map on a table casts is own shadows); the ability to explicitly state whether or not a light source casts a shadow or not; ability to support shadows from multiple light sources (depending on rendering settings).
    • Some of this might be addressed through glTF scene import, and / or through further adoption of the glTF specification for lighting (e.g. punctual lights / physical lights  – such as spotlight cones, etc.).
  • The work of glTF scene imports officially commenced in week #14. This work will be tackled in a similar manner to the Featurettes viewer – elements added to the viewer but placed behind debug flags until such time as they can be properly tested as the simulator support can be added (initially via Aditi, the Beta grid).
    • The will allow the viewer to be worked on and tested (with user testing) iteratively, rather than the work disappearing into a viewer fork that doesn’t see the light of day for months.
  • It was asked whether or not EXR might be a better option for SL skies rather than EEP + HDRI.
    • Runitai Linden noted he has experimented with EXR alongside of EEP and found that while EXR comes “close” to matching EEP skies, its doesn’t quite have the same range, whereas the HDRI lighting overlap with EEP offers easier flexibility of use.
    • He also noted again that the primary reason for providing HDRI support was to allow creators that means of reference checking their content between SL and their modelling tool to ensure it is appearing as expected after going through the SL upload pipe, etc., when compared to its appearance into their development tool.
  • Does the emissive map “knock out” the normal map if it is bright? No, but it can overpower the normal may and cause bumpiness, etc., to be less distinct.

In Brief

  • There was a general discussion on reflection probes and why they are limited in terms of shape (sphere or cube) and supporting other shapes+ illumination issues where probes overlap.
    • The shapes are necessarily limited largely as a result of performance stability.
    • It was acknowledged that blending lighting between probes is more a matter of improving / altering SL’s default ambient lighting model – which may well be looked at down the road.
    • The current advice is to – if necessary – “bomb” a scene with reflection probes  – and if that doesn’t work, disable probe ambience in favour of the global ambience – it will still allow the probe radiance and so generate reflections.
  • Occlusion volumes: the question was asked if it will ever be possible to have occlusion volumes in SL such that a room and its contents are never rendered in a viewer until the user actually enters the room in question.
    • It was acknowledged that this is the kind of thing the work with reflection probes might lead to, in terms of the basic mechanics being similar from a user’s perspective.
    • However, there are no current plans for developing such occlusion volumes at present, and a feature request on how user might want to use them / see them as working has been requested so the idea can be more formally reviewed / considered.

Next Meeting

† The header images included in these summaries are not intended to represent anything discussed at the meetings; they are simply here to avoid a repeated image of a gathering of people every week. They are taken from my list of region visits, with a link to the post for those interested.

2024 week #12: SL CCUG summary

Serene Retreat, February 2024 – blog post

The following notes were taken from my audio recording and chat log transcript of the Content Creation User Group (CCUG) meeting held on Thursday, March 21st, 2024.

Again, apologies for the lateness of this appearing; a major RL home project is getting finished-up, but it is still taking a good deal of my time.

Meeting Purpose

  • The CCUG meeting is for discussion of work related to content creation in Second Life, including current and upcoming LL projects, and encompasses requests or comments from the community, together with related viewer development work. This meeting is held on alternate Thursdays at Hippotropolis.
  • In regards to meetings:
    • Dates and times are recorded in the SL Public Calendar.
    • Commence at 13:00 SLT on their respective dates.
    • Are conducted in a mix of Voice and text chat.
    • Are open to all with an interest in content creation.
  • The notes herein are a summary of topics discussed and are not intended to be a full transcript of the meeting.

Official Viewers Status

No changes to any of the official viewers through the week, leaving the list as:

  • Release viewer: version 7.1.3.7878383867, the Emoji Viewer, issued February 15, promoted March 1st, 2024 – NoChange.
  • Release channel cohorts (please see my notes on manually installing RC viewer versions if you wish to install any release candidate(s) yourself).
    • glTF PBR Materials Maintenance-2 RC viewer, version 7.1.4.8149792635, March 11, 2024.
    • Maintenance X RC (usability improvements), version 7.1.4.8148263040 , March 11, 2024.
    • Maintenance-W RC (bug and crash fixes), version 7.1.4.8113624779, March 6, 2024.
    • Maintenance Y RC ( My Outfits folder improvements; ability to remove entries from landmark history + Maint Z RC integration) updated to version 7.1.4.8114240508, March 6, 2024.
  • Project viewers:

It is possible the glTF Maintenance-2 RC viewer might be promoted to de facto release status early in week ’13 (commencing Monday, March 25th, 2024).

Graphics / glTF

  • Work continues on the GLTF Featurettes viewer.
    • There have been refinements to the applying glTF materials to terrain, notably the UV density was halved in the interest of 2K textures, and further work put into making sure repeats are less noticeable.
    • It will now be possible to have alpha under PBR terrain “under limited circumstances” (e.g. single-sided and alpha masked)
    • Discussions have started at the Lab on the fees applicable for 2K texture uploads, but no decision on what this is likely to be.
    • Mirrors have also bee going through further refinements (e.g. a Hero reflection probe for a mirror can now have similar – but not the same – fall-off as other reflection probes, so it is no longer necessary to have a “really massive” probe to get a reflection; also, a mirror reflection probe now has the ability to go between a box probe or a sphere probe, influencing how it gets masked on a reflective surface).
    • It has been noticed that people trying the pre-release versions of the Featurettes viewer have been very confused as to the placement of things like mirror probes. Therefore, the viewer is going through a further UI/UX review in order to try to make things less confusing.
    • There are also some optimisation issues (again with mirrors) which may have to be revisited.
    • Brad Linden has been working on the server-side support for the Featurettes viewer through QA. If this is successful, it will allow the graphics team to work with the simulator team to get the code into an Agni (Main grid) channel (either as Preflight or as an RC channel).
  • Index of Reflection (IOR):
    • Initially will be a numerical value which can be adjusted, and which will be “bundled” with transmission
    • LL will attempt to get volume and dispersion included at some future iteration.
    • In this, it was pointed out that within the glTF specification, IOR requires both transmission and volume: transmission gives the basics for refraction (e.g. glossy refraction); whilst volume allows for things like colours within the refraction.
    • It’s not clear if attenuation will be used with respect to IOR.

In Brief

  • “Flexi mesh” – a long-standing request to have mesh interact with physics to make it flexible in the same manner as flexiprims (i.e. movement but without collisions being calculated). The response to this was:
    • The flexiprim system doesn’t really work for mesh.
    • What would be required is dynamic bones (as seen on other platforms). However, providing this has a number of pre-requisites (e.g. custom skeletons on hair for hair movement). As such, it is not something that is liable to be implemented in the near-term (if doable).
    • In terms of using Havoc physics interacting with mesh, the short answer is: not possible. It would require Havoc physics libraries which cannot (due to licensing) be made available to the viewer, due to the latter’s open-source nature.
  • Bake Service updates:
    • LL have acknowledged that it has been some time since the Bake Service has been updates, and that there are multiple requests for improvements and additions to the service (e.g. PBR materials support).
    • A potential project to update the Bake Service and implement some of the requested updates is currently being considered. However, a decision on where this will fit in the roadmap and what it will include has yet to be taken.
    • There is a fear among some that if the Bake Service is not updated specifically to support 2K textures when that is deployed, it could see Bakes On Mesh deprecated by creators in favour of applier systems which can utilise 2K textures (even if the system downsamples them on the basis of size).
  • With regards to 2K textures:
    • It appears that LL are trying to make heir use system specific: if a user is on a computer with the capacity to handle 2K textures, then they get 2K; if a user is on an older system (e.g. one limited by VRAM), then the largest textures they get are 1K.
    • For PBR terrain using 2K textures, there are some glTF specification-defined fallbacks the Lab hare leveraging to ease the load on lower-end systems.
  • It appears the PBR release has broken the Render Resolution Debug setting (most often used by RLV(/a) for screen blurring). A fix for this might be in the glTF Maintenance-2 RC viewer.

Next Meeting

† The header images included in these summaries are not intended to represent anything discussed at the meetings; they are simply here to avoid a repeated image of a gathering of people every week. They are taken from my list of region visits, with a link to the post for those interested.

2024 week #10: SL CCUG summary

Monkey Island, February 2024 – blog post

The following notes were taken from my audio recording and chat log transcript of the Content Creation User Group (CCUG) meeting held on Thursday, March 7th, 2024.

My apologies for the lateness of this report appearing; right now I have a lot going on in the physical world, and the blog and SL are having to take a back seat through until early April, after which things should start to (hopefully) return to normal.

Meeting Purpose

  • The CCUG meeting is for discussion of work related to content creation in Second Life, including current and upcoming LL projects, and encompasses requests or comments from the community, together with related viewer development work. This meeting is held on alternate Thursdays at Hippotropolis.
  • In regards to meetings:
    • Dates and times are recorded in the SL Public Calendar.
    • Commence at 13:00 SLT on their respective dates.
    • Are conducted in a mix of Voice and text chat.
    • Are open to all with an interest in content creation.
  • The notes herein are a summary of topics discussed and are not intended to be a full transcript of the meeting.

Official Viewers Status

  • The Maintenance-W RC (bug and crash fixes) updated to version 7.1.4.8113624779, on March 6th, 2024.
  • The Maintenance Y RC ( My Outfits folder improvements; ability to remove entries from landmark history + Maint Z RC integration) updated to version 7.1.4.8114240508, March 6, 2024.

Both of the above are at parity with the Emoji release viewer. The rest of the current crop of official viewers remains:

  • Release viewer: version 7.1.3.7878383867, the Emoji Viewer, issued February 15, promoted March 1st, 2024 – NEW.
  • Release channel cohorts (please see my notes on manually installing RC viewer versions if you wish to install any release candidate(s) yourself).
    • glTF PBR Materials Maintenance-2 RC viewer, version 7.1.3.7821226606, February 20, 2024.
    • Maintenance X RC (usability improvements), version 7.1.3.7721015131 , February 14, 2024.
  • Project viewers:

Graphics / glTF

  • Work continues on the GLTF Featurettes viewer, and it is hoped the release on an official project viewer is now not far off. This viewer includes:
    • Geenz Linden’s work on Mirrors, including his latest updates to the capability.
    • Cosmic Linden’s work on applying glTF materials to terrain (currently going through UI bug fixing + this is going to possibly reduce the number of texture repeats in light of 2K textures).
    • Updates which will allow the glTF / PBR swatch boxes in the Build / Edit floater display previews of the materials being selected, rather than grey boxes.
    • The first series of additional glTF extensions – Index of Refraction.
    • Initial work on 2K textures support.
    • Local high dynamic range imaging (HDRI) preview. This will allow content creators to view content in-world under the same HRDI settings as they are using in their tool of choice as a sanity check as SL environments / EEP will render them differently, leading to potential confusion and a belief that the HDRI settings are “wrong”.
    • The nightly builds version of the Featurette viewer is available on Github, but is not suitable for use as a primary viewer, and the capabilities are not supported on the Main grid. As such, this viewer should only be used by those wishing to test the capabilities.
    • The Rumpus Room 2048 region (and neighbours) on Aditi provides back-end support for the Featurettes viewer.
  • With reference to 2K textures, it was noted that while the viewer will have the ability to upload them, the actual ability to do so will be under the control of the simulator code; so just having the viewer-side update does not necessarily mean 2K textures can be uploaded grid-wide.
  • glTF terrain texturing: the focus has been on bug fixing and improving usability.

In Brief

  • Materials support for Bakes on Mesh: this has frequently been requested, and while it has a placeholder in the the Lab’s plans for possible implementation, it is is not currently scheduled to be worked on.
  • It was reiterated that blend shapes will be part of the work to support glTF scene import.
    • glTF scene import will allow and entire glTF scene file (scripts, node hierarchy, etc.) to be imported to SL and essentially “hung off” of a single object in-world.
    • Exactly how this will work, what limitations it will have, etc., in order to avoid “breaking” SL or opening griefing vectors, etc., has yet to be decided.
    • Similarly, glTF scene support will have its own Land Impact calculations, which are currently still TBD within the Lab.
    • Currently, glTF is due to start-up as a major project after the glTF featurette viewer has reached release status.
  • Splat Maps have been requested in the past with terrain painting. These are described as “in the early phase of consideration”, and if taken on, might be a “PBR Terrain Phase II” feature.
  • The subject of separating head and body shapes was raised – this has been the subject of a feature request, which has been accepted (Tracked). As always, “accepted” does not necessarily equate to it being quickly implemented.
  • Land Impact:
    • An update to the Land Impact is due to be deployed on the simulator side, in the Hearts and Flowers release (due to start deployment on March 13th, 2024).
    • This does not change the total Land Capacity (“prims”) for a region; but it does reduce the download weight calculated for meshes
    • This equates to a “15% discount” on mesh object Land Impact (all mesh with the exception of Animesh, not just mesh items downloaded after the update is deployed).
  • Reference was made to the upcoming shift of object inventory messaging moving to HTTP.
    • This should result in the contents of objects rezzed in-world loading “many time faster” than is currently the case.
    • The update will be deployed to the simulator code first, and then be subject to a viewer-side update.
  • With the switch to Github / Canny, LL is trying to be more transparent in terms of planning, projects, issues, etc., looking to have as much public as possible (except things like security issues, which need to be limited in who can view them for obvious reasons). However, while this will allow users to see more of what is in the works / being considered, it also means that projects and work can appear to sit apparently dormant for periods even longer than on Jira
  • How (viewer) feature requests are prioritised (in short):
    • Those which are accepted and form small, relatively stand-alone requests will be directed towards “featurette viewers” of the kind like the glTF featurette viewer mentioned above. The hope is that this will allow devs to pick them up and work on theme in between larger projects or when larger projects are stalled for whatever reason, allowing more of this type of request to filter through the the viewer more readily.
      • In addition, if this works and a Canny Board of such requests can be built-up, it might allow LL to open it to open-source developers to pick items they’d like to work on and then contribute the code to LL.
    • Larger-scale feature requests (such as adding materials support to BOM) will be subject to more extensive investigation / evaluation and internal prioritisation against other large-scale / high-value work, by LL.

Next Meeting

† The header images included in these summaries are not intended to represent anything discussed at the meetings; they are simply here to avoid a repeated image of a gathering of people every week. They are taken from my list of region visits, with a link to the post for those interested.