Temporary prototype update agent for local testing
Categories
(Toolkit :: Application Update, task, P1)
Tracking
()
People
(Reporter: molly, Assigned: molly)
Details
As a way to validate all of the currently largely disconnected parts of the update agent that we have built and are building, we should implement a prototype skeleton version of the top-level logic and allow manually enabling it to use for limited local testing, either with builds from oak or perhaps official nightly builds (we'd have to enable building and packaging the agent for nightly, but probably not register its task yet by default).
Just to set expectations, since this is a prototype, it wouldn't have much or any error handling or even a lot of normal state management that the final version will need, and we should expect to discard it and start over at least once, because it is the fate of software prototypes to never have been gotten right the first time.
Comment 1•4 years ago
|
||
With the updated approach for the background update agent I don't think this is relevant anymore. Please re-open if I am mistaken.
Description
•