Re: [streams] Do writable streams need separate signals for write-success vs. write-accepted? (#316)

Thanks for the summary. The sort of tasks for which the API with success/failure signal is useful would be one consists of a sequence of works where:
- each work is transaction-ish.
- it makes sense to buffer the works with back-pressure
    - this implies that the user want to issue multiple operations without waiting for completion of previous ones

I guess the domain of problems that satisfy both of these conditions is not so big.

Another use case for which the API is suitable is to represent RPC stub via which we issue mutually independent computation jobs to the cluster. In this case, write() calls may be fulfilled out of order, but backpressure makes sense to tell the developer if the cluster is overloaded (or load is exceeding CPU quota), and the user really want to see the result and we can return the result via the promise returned by write() call.

---
Reply to this email directly or view it on GitHub:
https://github.com/whatwg/streams/issues/316#issuecomment-90846064

Received on Wednesday, 8 April 2015 08:49:43 UTC