Rocket.Chat vs Slack
Side-by-side trajectory, velocity, and editorial themes.
Rocket.Chat's 8.6 RC line adds self-hostable translation and a unified presence engine
This feed tracks Rocket.Chat GitHub release-candidate tags, and the top of the window is dominated by empty 8.6.0-rc.x and 8.5.0-rc.x 'Bump meteor version' cuts with the real content concentrated in the 8.6.0-rc.0 minor release. Note: this appears to be a duplicate product row of the other Rocket.Chat entry in the catalog (same RocketChat/Rocket.Chat repo, same releases, different slug/UUID); it is being classified independently off its own entries. Because these are RCs, capabilities are staged into a pre-release train rather than GA.
The 8.6 cycle leans into self-hosted and privacy-controlled deployments: LibreTranslate for fully on-premise message auto-translation, Virtru as an external ABAC attribute store, and a unified presence engine with priority-based claims. In parallel there is a broad, deliberate migration of legacy DDP methods to REST endpoints (settings, spotlight, im.blockUser, e2e key requests, rooms.join), signaling an API-surface modernization ahead of a 9.0.0 removal.
The rc.x cadence points to an 8.6.0 GA cut once the release candidates settle. Expect the DDP-to-REST migration to continue toward the flagged 9.0.0 removal.
Slack is quietly rebuilding itself as a runtime for third-party agents.
Slack's developer platform has shifted its center of gravity from bots-that-reply to agents-that-act. The last month is dominated by agent primitives: apps can now receive the context a user is looking at, Slackbot can call external tools over MCP, and a dedicated agent messaging surface ships alongside steady CLI and Block Kit work.
Each release fills in a piece of an agent platform — context in, tools out, and a native place for agents to converse. Block Kit is gaining richer primitives (containers, data visualization) that read as the display layer for agent output. Three CLI releases in a month show the tooling keeping pace with the expanding surface.
Expect the next moves to connect these pieces: agent context feeding MCP tool calls, and Block Kit's new blocks becoming the standard way agents render results in-channel.
See more alternatives to Rocket.Chat →
See more alternatives to Slack →