Originally posted on Google+, for lack of a better place
Most of you wan't care about this post. Heck, a large portion of you won't likely know what it's about.
What it is about is CSS vendor prefixes. There has been a lot said over the past week on the subject, and while I don't think there has been a perfect solution, I also don't think a perfect solution is possible. Basically, we're shooting for least bad here. There have been several proposed.
Stop it.
Specifically, developers should stop using prefixed properties on production sites. These properties are prefixed for a reason. They're not yet a standard. They're still in flux. Which means that when they do change, your site will break. Is that something you want? I know I don't.
Now, when your site does break, due to a vendor changing their browser's implementation, you're probably going to get mad.
Stop it.
Seriously. If you didn't know that these features were experimental, then you haven't been paying attention. It's this developer backlash that vendors are afraid of. It's why they won't change their supposedly-experimental features, and why those features will take a longer time becoming standard. Nobody wants to break the web, but developers have made it easy to break the web by using experimental features as if they were standard and locked-in.
Maybe vendors should break the web. Change your implementations, and see what happens. When developers get angry (and they will), tell them - straight up - that these features were experimental and subject to change. Because it's the truth.
I'm guilty of using prefixed properties as though they were production-ready. I'm aware of this. I've used gradients, transforms, and border-radii. But, for now, I will stop. Until these properties become standard (or, at least, stable and un-prefixed), I will not use them on production sites that are expected to work across browsers.