Auto-connection request for ttop (system monitor)

store-requests privileged-interfaces
1
  1. name: ttop
  2. description: A lightweight TUI system monitor written in C++ that provides real-time insights into CPU, memory, and process activity.
  3. snapcraft: https://github.com/nabhay099/ttop/blob/main/snapcraft.yaml
  4. upstream: https://github.com/nabhay099/ttop
  5. upstream-relation: I am the lead developer and maintainer of the upstream project.
  6. interfaces:

    • system-observe: auto-connection

      • Reason: This is the core requirement for ttop. It allows the app to read the /proc filesystem to display the process list, CPU usage, and memory statistics.

    • process-control: auto-connection

      • Reason: ttop includes a process management feature. This interface is required so users can send signals (like SIGTERM and SIGKILL) to terminate or manage processes directly from the TUI.

    • hardware-observe: auto-connection

      • Reason: Used to detect hardware details and topology to provide accurate performance data and core counts to the user.

    • mount-observe: auto-connection

      • Reason: Allows ttop to monitor disk usage and filesystem partitions, showing the user how much storage is available.

    • network-observe: auto-connection

      • Reason: Required to monitor network interfaces and display real-time upload and download speeds.

Note: Note: I am keeping the confinement strict for safety and to make sure that if there is an error it stays within the confinements and doesn’t run amok.

2

This request has been added to the queue for review by the @reviewers team.

3

Hey @Nabhay , I am fine with all the requested interfaces except cpu-control. For metrics collection, wouldn’t hardware-observe already cover this use case (e.g, per-core frequency and scaling data)?

(#askForInfo)

4

That’s a fair point. I had initially planned to add a CPU frequency scaling/overclocking feature, which is why I included cpu-control. However, since that feature isn’t implemented yet and I’m focusing on the core monitoring features for now, I agree that hardware-observe is sufficient.

I will remove cpu-control from my snapcraft.yaml and stick with the more restricted interfaces. I’ll submit a new request in the future if I ever implement the control features.

Thank you!

5

thanks @Nabhay , so from my end +1 to auto-connection of system-observe, process-control, hardware-observe, mount-observe and network-observe interfaces given the functionality of the snap (#voteFor)

6

Hello @Nabhay!

This is a +1 (#voteFor) from me as well to grant auto-connection to the system-observe, hardware-observe, mount-observe, network-observe, and process-control interfaces for the ttop snap. These are all expected interfaces for a system monitor. Thanks!

7

Voting period has ended. This request is approved with 2 votes for and 0 votes against.

8

+2 votes for, 0 votes against, granting auto-connect of the interfaces system-observe, hardware-observe, mount-observe, network-observe, and process-control to snap ttop. This is now live.