This is a nonexhaustive summary of the DDevs Q&A stage. Most of the points below are paraphrased for brevity. Ephasis is my own unless explicitly stated, and my opinions on certain things may shine through. Please refer to a VOD of the stage for the exact wording from the Discord Developers.
- Application Commands Permissions V2
- "Super super super close"
- Will not release before the holidays because fixing things on christmas is just not cool
- Closed beta testing will occur after the holidays
- Will not be wide-scale tests due to possibly breaking permissions and that could be bad.
- Desktop frontend is "done"
- woot!!!
- Attachment option types
- iOS Demo video
- Android Demo video
- Desktop Demo Video
- Will be supported on all platforms on release
- Supports video, images, and other file types
- Supports multiple attachments, (1 file per option, multiple attachments per command)
- Attachment is received as an attachment object
- MOBILE SPOILERS ON FILE UPLOAD~!!!!!!!!!
- Localization for Slash commands
- Video
- Still WIP, not available in any public client code
- Backend code is done
- Will start being added to clients "shortly"
- User locale will be received in the interaction payload. Guild locale is also included!!!!!! YESS!!
- Timeout - Should be launching this Monday (December 20)
- Notion Docs + Demo
- A middle ground between kick/ban.
communication_disabled_untilISO 8601 timestamp- Bots and moderators will be able to timeout users.
- If timed out: disconnected from VC, cannot chat in channels
- Custom timeouts may be added in the future for the client, currently its only 60s/5m/10m/1h/1d/1w. API side might be able to set custom timeouts
- Gone are the days of mute roles lol
The following points are answers to questions asked by the community. Questions can be found in the VOD and will not be listed here. All points are listed in order they came up in.
- All of the features for slash commands (perms, attachment opts, etc) are all nearly done and will most likely be shipped well before the message intent deadline at the end of April.
- Other types of input types are planned for message components. "quite a bit more". They were prioritizing the absolutely needed types like buttons. They have a lot of ideas about message components being a "set of building blocks" for displaying "rich data"
- There is a beta testing group for API features, which is available only to a closed set of people (very large bot devs + lib devs). They are hiring for positions for someone that would likely be the person running a public beta program like requested.
- The member roles endpoint is not performant enough to make it widely available, so they have to make time in order to fix them. In the future, they want something like this to be available and performant by default with devs in mind at the get-go rather than an after thought.
- About a sandbox mode: this was asked in the previous Q&A and no further updates about it are given. Mason and the team understand the desire for a sandbox mode and that its a painpoint, and the closed/open betas are leveraged to make sure features are easier to implement. No promises are being made.
- Member object may be getting updated to support guild banner/accent. They have to talk to the team that worked on it to see about getting it implemented
- They want to look into more robust bot profiles. A per-server bio for a bot is "interesting", but having consistent information is also beneficial for the end user. They don't know if they will/won't implement guild bios for bots.
- Modals will get selects and more (radios/checkboxes/etc). They purpose built it so that adding more types would be easy. They however wanted to only launch it with text input so they could make sure things work well
- Bots cannot modify member "interest" for scheduled events at the current time. It doesn't make sense for a bot to modify this as the only thing that happens is the user gets a notification for the upcoming event. However, they will bring it up with the team that worked on scheduled events. Its not their decision to make :(
- They will likely add more action rows for modals, but only once they add more types of components. Multi-page/multi-step forms are definitly something they are aware is requested
- Re: editor upgrades - Voltana has been working on it with Slate. Due to how out of date they were with it, its hard, see this commit image. It has to be done slowly due to how central it is and its size.
- An open text box in a message as chat scrolls by is just not great, a modal is currently one of the better ways to handle needing followup information and there likely won't be an alternative to ask for small followup data.
- Premium memberships are still very very early testing (literally like 10 servers). It likely won't be available in the API until it matures as they don't want to make tons of breaking changes
- Pressing submit on a modal will always close the modal, but multi-step modals would handle the use case of wanting to edit the modal/form "in-place"
- re: variadic arguments (repeating args) - They are still considering it. It gets challenging UX-wise for multiple strings vs multiple users (auto complete things are easier). Its lower priority due to the UX challenges and other important changes being made
- Membership screening is making steady progress, unknown ETA. "it'll be out when its out"
- They understand the value of some sort of in-client bot dashboard, but its something that's very open ended due to all the many use cases. Its not something they'll do anytime soon, if at all. As anyone that needs a complicated dashboard would likely be better off with their own dashboard, but for simple use cases it'll be easier to move something like that into a modal form. It just isn't super feasible right now
- Markdown syntax like tables, anchors, headers, etc in messages isn't something this team (Mason and Ian) would be responsible for. With the new slate upgrades it may be easier, but its surprisingly hard to do, especially rendering a table on mobile. Mason would love this, but its not something he can comment on.
- HTTP-only bots (without bot scope) not needing verification was an oversight and as such, will be able to participate in things for verified bots (eg, app discovery) once they are verified (when they give them support for verification). They also do support "Add to Server", but its broken and they are working on fixing it.
- Date option types are confirmed as being added. After a lot of the more important things. This also might need the new slate upgrade first tho.
- Bots could eventually maybe play video. They need to talk with the native code teams.
- Modals currently don't support mobile (android). They are done on iOS and desktop. They are waiting on several refactorings for the android app first. The estimated (rough and not promised) eta is early 2022.
- Nova's note: Don't go hounding the devs on this please, they're working on it, don't rush them or hold them to this estimated ETA.
As always, big thank you to all the discord developers for your hard work!
Thanks mason
Thank you so much for putting this together! <3