Hacker News new | past | comments | ask | show | jobs | submit login

So you would still have to deal with the case of someone reading inconsistent data and taking wrong action as a result. If that action is only internal to the system you could go on and do a cleanup when the system becomes consistent again. If the action is external you can not.



Right. If full consistency is a requirement than you can still have it, at the cost of availability. Alternatively, since the dataset is immutable, it contains a history of everything that happened. So you can resolve problematic actions later on (this is similar to what banks do).

It's important to realize that the tradeoff between consistency and availability is a limitation of nature, not of our tooling.


So CAP is still a problem, not beaten.




Join us for AI Startup School this June 16-17 in San Francisco!

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: