Replies: 1 comment
|
Manual groups make more sense than automatic ones here. I have around 25 projects in t3code, arranged on disk under a few parent folders. Roughly half have no git remote, so the existing repository grouping produces nothing for them. Anything derived from git identity keeps missing that shape. A manual group doesn't care. Same gap one level up, for the project list itself (#6754). If a group were just a named, collapsible, user-ordered container, one implementation could cover both. |
0 replies
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Uh oh!
There was an error while loading. Please reload this page.
Problem
A project is the right execution boundary, but it is not always the right organizational boundary.
For example, within one project I may have several related threads:
They belong to the same project, but represent different roles or workstreams. As the number of threads grows, a flat project list becomes hard to scan and it is easy to lose the relationship between them.
Proposal
Add user-created thread groups inside a project.
A group would be a named, collapsible section in the project’s thread list:
Users could:
Ungroupedsection.A small per-thread color accent could be a useful follow-up, but group colors are the important first step. Color should be supplementary rather than the only status signal, so the UI remains accessible.
Why manual groups
This is deliberately not a request for automatic grouping, AI-generated tags, or grouping solely by worktree.
Those approaches can be useful, but the person managing the work understands the meaningful relationship between threads: they may be different models reviewing the same task, separate stages of a rollout, or parallel roles in one initiative.
Manual grouping gives that person a lightweight way to express the structure without changing repository, project, branch, or worktree ownership.
Smallest useful scope
Out of scope for a first pass:
Related ideas, but not duplicates
This proposal is a manual, user-defined layer inside the existing project/thread hierarchy.
Question for maintainers
Would this fit the Sidebar V2 direction? In particular, should group membership be durable server-side user metadata, while presentation choices such as collapse state and color remain device-local?
I am looking for product direction first and will not open an implementation PR without it.
All reactions