A little perspective on prefixes

Happy February! No announcements today, but I just wanted to write up something quick about an interesting realization I had. I just answered this Stack Overflow question about CSS vendor prefixes. Here’s an excerpt:

The idea was that authors who wanted to “try” out these experimental features locally could do so with the prefixed properties on shipping browsers, rather than having to deal with nightly builds or even downloading and compiling the source manually, because shipping buggy behavior unprefixed was akin to shipping beta software as production-quality.

What the vendors didn’t see coming was that authors would take the prefixed properties, put them on production sites, and encourage other authors to do the same. As a result the prefixes spread like wildfire to the point where vendors were too afraid to break sites by removing the prefixes after shipping stable, unprefix implementations as originally intended. I mean, just look at what happened when Mozilla dropped support for -moz-opacity, a prefixed property that was relatively little used compared to today’s -webkit- properties, less than 4 years ago. For perspective, -moz-opacity was unprefixed in Firefox 0.9, almost 12 years ago.

(there’s a little more, but these first two paragraphs are the relevant bits)

When writing that bit about Mozilla, I realized this: -moz-opacity was unprefixed before I entered high school, and support for the prefix was dropped after I graduated college.¹ I was 12, and 20, respectively.

In other words, in the time it took for me to go through two whole educational milestones in my life, the fellow who submitted the Bugzilla ticket linked above, and who knows how many other CSS authors the world over, didn’t bother to change -moz-opacity to opacity on their websites.

So… yeah. That’s a bit of perspective for you.