Public: anyone can read this; the crew can change it.
Asks and crew answers fold into the core: what waits for a person is a state on an issue or pull, shown where that issue or pull already is #13787
Landed in pull #13838 · 2 runs · run
the crew's run was interrupted: restart, resumes on the next pass — cancelled
Sign in to join this conversation.
new above
offline
The owner, 2026-11-16: "the whole asks thing and the 'crew answers' things are just flaky abstractions of the core we're trying to get at … they take up a lot of UI space throughout the forge as well and I wonder if they contribute to slowness." His asks page took about 20 s to open in two browsers at once.
Measured on 0.7.613.
/askstakes 300 ms on the server with nothing else running (render 220 ms; the page budget is 100 ms), and it subscribes to eight live event kinds.held10.8 ms,stopped8.3 ms) are 19 of the page's 41 ms: nearly half of home's server time computes rows he asked to have off the home./forges/:name/stopand engine jobs), which every page shares; asks did not cause that, but a 300 ms page feels it most.The core. A repository, its issues and pulls, the crew, an app, a forge. "Something waits for a person" is a STATE of an issue or a pull (who is next and why: the
nextthe forge already computes), not a second kind of object with its own page, blocks, counters and vocabulary. A crew member's answer is a comment on that issue or pull.Ask. Design first (one short note in the pull or a docs page), then remove more than is added:
next.who. No separate asks inbox page;/asksredirects to that filter./asks) move to the apps list as each app's state and action.next.Budgets. Home loses the
heldandstoppedphases; the "waits for me" view answers under 100 ms warm with 500 items.Done when. There is no page, block or counter named asks; a person finds what waits for them on the kanban; the home's server time is under 25 ms.