Re: [fetch] Aborting a fetch (#27)

My concern is that `fetch` has been promoted as some sort of eldorado, nice and simple, but also powerful and better replacement for XHR. I've personally talked at QCon about it and suggested it's not bad but **only** if used for small interactions or everything that does not require a progress bar.

I also empathized that WHATWG page should explicitly, and inside a *strong* , communicate that `.abort()` is also missing in its core so at least developers will be aware of such limit and chose the right API for the right job.

My only extra fear leaving fetch this way thought, is that most developers **will** think it's good for everything and the one that will pay the consequence is every user surfing the Web.
That, plus the fact that inevitably Streams will be implemented x-browser later because of priority, but I see them more important than fetch for the simple reason that similar API can be polyfilled to provide its most common and basic use case. Streams? Not a chance to meaningfully polyfill them.

I'll keep watching hoping for better solutions or proposals, I don't want to distract too much who's still trying to solve this problem with my thoughts.

Best Regards

---
Reply to this email directly or view it on GitHub:
https://github.com/whatwg/fetch/issues/27#issuecomment-86722668