What is your policy on flags in browser compatibility data?

(Jmedley) #1

I ask because I’m the person who officially does this for Chrome. We deliberately do not add features that are only behind a flag. Yet, MDN’s instructions for updating compatibility data use Chrome in the examples for flags.

What I really want to know is is MDN enthusiastic about listing flag-supported features on MDN or is it only listed in the instructions because you know people will try to add it anyway?

(Eric Shepherd) #2

Generally, we encourage anything that’s exposed to the open web to be included on MDN, even if it’s behind a flag. It’s just important that flagged items be properly labeled, and ideally include explanations of why they’re behind a flag and whether or not they’re part of an evolving spec. That sort of thing—to help readers recognize whether or not they want to take the “risk” (so to speak) of using the flagged features.

Eric Shepherd
Senior Technical Writer
MDN Web Docs: https://developer.mozilla.org/
Blog: http://www.bitstampede.com/
Twitter: http://twitter.com/sheppy

(Chris Mills) #3

We could’ve picked a Firefox example or a Chrome example really; there was no particular reason for picking Chrome. historically, we have tended to list features that are only supported behind a flag, but we tend to say “feature not supported”, and then “But available behind flag x” in a footnote.

I seem to recall that we are not following this pattern in the new browser compat data schema. If not, we probably should do.

I think we would have not bothered listing features that only exist behind a flag, but users have expressed interest in this information along the way.

(Stephanie Hobson) #4

Developers expect to have information about features that are behind flags, it’s important that they can get that from our data.

Personally, I find it very useful because if a feature is behind a flag I know it is under development.

We can display “no” when it comes to rendering the compat tables but having the version number in the data is more useful than having false in the data :wink:


(Eric Shepherd) #5

Also, if we have the information in the compatibility database, we have the option of deciding whether or not to include it in the actual displayed tables or not. It’s trivial to filter them out if you decide not to include them. Indeed, it’s hypothetically possible that one day the site (or any tool using the database) could let the user decide whether or not to include flagged items, prefixed items, or items based on whether or not they’re obsolete, etc.

That’s my dream, anyway. :slight_smile:

Eric Shepherd

Senior Technical Writer, MDN

MDN: https://developer.mozilla.org/

Blog: https://www.bitstampede.com/