On Thu, Dec 19, 2013 at 11:46 AM, Kevin Ballard <[email protected]> wrote:
> For people who don't want to think about failure cases, I don't see how > this is any better than a genuinely unbounded queue. > As was mentioned earlier, there's no such thing as a "genuinely unbounded queue". It's a system resource-bounded queue. People who "don't want to think about failure cases" are simply accepting whatever failure mode happens by virtue of not thinking about it. As soon as a failure happens, whatever it may be, they will become much more concerned. I'd suggest not giving too much worry to these people. They'll either have everything work swimmingly if they don't process a lot of messages, and one way or another will have everything explode once they hit a resource cap, whether it's bounded by the language or system resources. So I think that entire line of reasoning is a red herring. People writing toy programs that never have their channels fill beyond a small number of messages won't care either way. However, overloaded programs + queues bounded by system resources are a production outage waiting to happen. What's really important here is providing a means of backpressure so overloaded Rust programs don't grow until they consume system resources and OOM. -- Tony Arcieri
_______________________________________________ Rust-dev mailing list [email protected] https://mail.mozilla.org/listinfo/rust-dev
