Re: [ServiceWorker] Race when registering multiple service workers for the same scope (#783)

Sounds good. It's basically what we discussed in terms of putting register jobs into a job synchronization mechanism. For this, I think the algorithm thread queue should be defined as a per-scope queue instead of a per-registration queue as there's no registration for first time registrations. I'll check out if we have any further problem to solve to spec it this way.

Regarding maintaining multiple queues, do you agree to keep them? Having blocking waits during the register/update job steps (e.g. fetching the script, waiting for oninstall execution, etc.), I think it makes sense to allow the next request to have a chance to replace the current job.

---
Reply to this email directly or view it on GitHub:
https://github.com/slightlyoff/ServiceWorker/issues/783#issuecomment-162732317