Hello Nick and Tian, thanks for hosting this AMA!
I was wondering what the plans are for DynamicImage and DynamicMesh, is there any estimation for a release date or for some kind of announcement surrounding those two? Iām really looking forward to it, Iāve had a lot of fun testing them out but I would love to be able to use them in a real game!
Any plans for actually expanding the engine past the current limitations and issues it has right now? Right now itās very hard to see a future in witch ROBLOX can allow the creation of large, detailed and complex experiences without running into heavy performance issues and engine limitations. While you guys will often praise showcase experiences (like most of roblox), itās pretty clear that those experiences somehow suffer from performance issues while not really offering a visual fidelity more advanced than what could be seen on something like Crysis 1 (2007 AAA game) while even STILL having an area of exploration 50 by 50 studs at BEST. Other games that do try to offer higher visual fidelities, larger worlds, etcā¦ often have to rely on hacky optimization tricks that can easily chip away at the games quality in an attempt to make the unreasonably low performance engine take away resources. Another issue i have seen with the ROBLOX engine (and that will continue to be accentuated) would be the fact that the engine itself keeps getting more and more bloated with new features that are forced to have baked into their games regardless of not those features will ever be used. I do not believe simply having a āDisableā button is enough as even THEN those features will be present in memory and still take away some valuable resources regardless.
With all of this, i personally do not believe ROBLOX itself can go too far without tripping over most obstacles due to the lack of a proper game engine. Would there be any chance of the ROBLOX engine getting a major ārewriteā or something similar (in order to alleviate the issues it has right now)? Would there be any chance for ROBLOX themselves to take a more āmodularā and āopenā approach with the engine so developers have way more access to everything? While i do understand that opening the engine up would leave the chance for many bad actors to create many relatively āharmfulā experiences, i personally believe that these sorts of risks have to be taken in order to elevate the engine AND the developer space to higher grounds for a better future.
Hey darkpixlz,
Thanks for sharing your questions. To ensure weāre engaging with as many creators as possible we ask that everyone keep to one question per message. Can you reduce or consolidate? Thanks so much!
Any more plans with highlights? I asked last AMA and didnāt get a response. Thereās a limit of 31 highlights and they donāt work on mobile. We really want to use this feature more broadly.
Will we get more control over terrain?
Examples:
Hello again! Thanks for doing another AMA 
Since I just found out about this AMA right now, Iāll post an abridged version of my primary question, which is the same thing I asked dave in-person at RDC (which did not result in a response with any meaningful substance) : When will transparency and communications about the status of programs improve for the better, and what is Roblox actively doing right now to achieve those? Right now, creators have no way of knowing whatās going on with a given program, where itās heading, or why itās being shut down, nor can they give any feedback on them. The accelerator, stars, devawards and verified badge programs have all either randomly halted or been axed, typically with radio silence from Roblox on the matter. This leads to a lot of problems and frustration towards Roblox as a company!
Iām super happy to chat about this problem further if itās desired, my DMs / email are always open. I have a lot of feedback on this matter that simply would not fit in this reply:
Are there any plans for us to get more UI features such as sliders, UI blur, etc?
I realise I was meant to keep to one question per post. Unfortunately I canāt edit my post (why canāt you edit posts in slow mode - the purported reasoning doesnāt make sense if you only present the limitation after the post is written), so I guess I posted three questions. Whoops.
Due to the slowmode I was unable to edit, hereās the post again;
First off, thanks for trying to improve clarity and community communication. I think itās essential to try and get feedback from everybody, and Iām glad to see if starting to happen more. As for the questions, I only have a few basic ones (mostly relating to UI design):
Thatās all I have for now, thanks for doing one of these once again.
Iām interested in learning more about this upcoming update in particular:
- Community-led Marketplace - Anyone can publish avatar items to the Marketplace (Late 2023)
Have the conditions that must be met in order to publish avatar items to the marketplace been determined and if so, what are they (ex. āAny fee to pay (ex. 750R$) per UGC upload?ā, āActive Premium subscription requirement?ā, āAsset geometry requirements (ex. vert. tri. etc. limit)?ā etc.). I was thinking perhaps the current conditions would change for the public release by tightening or loosening the current upload requirements. Maybe itāll just stay the same. Would appreciate a clarification.
Iām interested in learning this information so that I am prepared for the update as a Roblox user interested in investing into UGC.
Will there be any new rendering capabilities?
Examples:
Hello hello! I know there are plans to rework groups āfrom the ground upā and was wondering if you all could provide any more information on what this entails?
I currently am the owner of a group that holds all the experiences Iāve created. Iād like to have more granular control on which developers can interact with the places under the group. Right now I have to give developers access to all the places under the group if for example they want to upload an animation to it. Thereās also the issue with distributing group funds, is there plans to add the ability to distribute group funds based on an experience and not the entire group funds?
The overall idea would be to treat groups like organizations, and within these organizations could have āgroupsā that have finer control over experiences. This prevents having to create a new group per project which in return then allows you to grow a centralized fan base in one organization rather than having them spread apart against a lot of different organizations.
This is an area I have always felt was lacking and am excited to hear if any of these changes are planned or are being implemented!
Not directly related to the announcements made at RDC, but are there any plans to improve on the universal client? Since itās release 80% of the client is still just a browser pop-up & the performance leaves a lot to be desired.
Hello again,
As of now, the Export Selection in Studio only allows you to export as a .obj file. Will there ever be support to export as other formats, such as .fbx and .glTF?
Thank you.
Will there ever be better tools to debug memory leaks? What we have currently is semi-decent, but Iād love more in-depth tools such as origins or objects that take up a lot of memory.
Will we ever get official support for control over the avatar editor backgrounds and VR home enviroment? Currently itās unoffically possible by tinkering and swapping the mobile.rbxl and the file responsible for the home enviroment. I gave it a try and it looks pretty cool, the only downside that the skybox outside is theme-colored alongside with a thick fog.
Iām ranking my questions based on priority now, so hereās another:
How much of a priority is it to improve Hyperion to patch more exploits which lead to executors/injectors continuing to work? Right now, even after the 64-bit client, anti cheat is still a high priority because of people using injectors, mainly via the UWP app.
Iāve seen that Roblox is adding a new feature for webcam upper body tracking and has recently launched on the Meta Quest. Will there be any future plans for Roblox to support full body tracking for VR?
Actually, the UWP application does have Hyperion inside of it (quite a long time ago). Iām not going into the specifics but currently no exploit executors/injectors are working (from my own knowledge).
Will we ever have API access to AI?
This one is more out there, but Iād figure Iād ask anyway.
Examples:
