How do you decide if a feature/API is safe to use in production?

Do you feature detect everything? Wait for some x% of the browsers?

Comments

[Kim Arnett ]

I'm not in the JS world - But for native frameworks, the following questions must be answered yes, or have a good reason why not

• Are regular updates happening?
• Was the last commit within the last month or two? (IE: Isn't abandoned yet)
• Are enough people using it for the contributors to WANT to keep up with it?
• Is it going to save you work?

If going to production - we're a little more picky on the frameworks we use. We'll use more streamline stuff.

[Frederik 👨‍💻➡️🌐 Creemers]

Since you're talking about feature detection, I'm going to assume you're talking about browser features.

First of all, I'd take a look at your web analytics. There, you can get an idea of how common different browsers and versions of these browsers are. Then, you look at a site like caniuse.com to see which browsers/versions have the feature, and see what percentage of your users won't have it.

If a significant percentage of your users will be impacted, and you have no fallback, then don't use the API in production yet.

If you don't have users yet to check your analytics, you can often make predictions. For example, a n audience of developers is likely to use modern browsers, while an audience of bank employees might use older versions of IE.