Tracking changes in Discord Canary updatesCurrent build hash: 4f0263eeecb48e5f7238d60c2946a9d27eda0250
4f0263eeecb48e5f7238d60c2946a9d27eda0250
Tests a Windows voice-call setting that reduces echo when speakers are used instead of headphones.
Speaker calls can create echo when the other person’s audio is picked up by the mic. Splitting mix and auto lets Discord tune echo reduction with less impact on voice quality.
Shows device-type icons in voice/video device lists for microphones, speakers, and cameras.
Device icons can reduce wrong microphone, speaker, or camera selection before a call starts.
Syncs the desktop DESKTOP_TTI_HTTP_CDT native setting from the experiment value for startup measurement.
It is not a visible UI change, but controlling the TTI measurement path helps Discord compare desktop startup improvements more accurately.
Tests a redesigned Clips settings surface and its interaction with autoclipping conditions.
Clips depends on users understanding capture and automatic clipping options, so a clearer settings surface can affect feature adoption.
Makes servers previewed before joining easier to keep and find again through the lurker preview flow.
stickyLurk is the key value for preview persistence, connecting the experiment to server discovery and join conversion after previewing a guild.
Changes the invite modal title when adding people to an existing one-to-one DM.
The title at the moment a one-to-one DM becomes a group DM can clarify whether the user is adding a friend or expanding the conversation.
Tests group-DM customization and cancel-button behavior before adding people to a one-to-one DM.
canCustomizeGDM controls pre-invite editing, while addCancelButton controls exit handling, both of which can reduce accidental group DM changes and improve invite conversion.
Exposes dice rolling as a Discord built-in slash command.
It turns lightweight chat play into a default Discord capability that servers can use without installing a bot.