API Prefixing: Making a Data-Driven Decision – Help Wanted

Background: API prefixing is when we give new DOM functions a name beginning with “moz” to show they are experimental, e.g. mozDrawText on <canvas>. There is some discussion as to whether this is actually a helpful thing to do. We want to get some data to help us decide.

What we need is a list of the finished APIs we’ve added in the past (say) 3 years, the date (A) they went in prefixed, the date (C) they were unprefixed, and then (and this is the trick) the date (B) we could have unprefixed them if we had perfect vision of the future – i.e. when did the last incompatible change go in.

The question then is: is B usually close to A, close to C, or does it range about from API to API?

We’d love someone to gather data on this, to help us make a decision on whether API prefixing is a good idea. To do this work, you just need to be able to use Bugzilla, read English, and understand the concepts in this message. If you have some time, drop me an email and I’ll help you get started.

3 thoughts on “API Prefixing: Making a Data-Driven Decision – Help Wanted

  1. Have we prefixed and unprefixed enough DOM APIs to get a real sample size here? I think CSS properties would be more interesting.

  2. Also worth looking at APIs that evolved unprefixed and seeing what kind of doom and gloom ensued (or didn’t ensue). It would be worthwhile to look at the Canvas 2D API and the APIs associated with the video and audio elements.