Hacker News Reader

Comment by twblalock | Hacker News Reader

One notable thing about the dominance of Chrome (and the decline of Firefox) is that WebKit is now a de facto standard. The desktop and mobile versions of Chrome and Safari are based on it, and according to Microsoft, "any Edge-WebKit differences are bugs that we’re interested in fixing." (https://en.wikipedia.org/wiki/Microsoft_Edge). That covers every major desktop and smartphone browser except Firefox.

At some point, developers are going to target their stylesheets for WebKit only, because Firefox rendering differences are going to be seen as nuisances that aren't worth overcoming in order to reach a tiny minority of users. Firefox will have to work toward WebKit compatibility as Microsoft Edge does.

WebKit is doing pretty well for something that was originally part of KDE.

View on HN

It was true a few years ago, but Chrome no longer uses WebKit, they forked it to their own Blink engine, so now there are differences between them and you have to write stylesheets for both (example: -apple-system font vs BlinkMacSystemFont).

Although some code is still being ported between them (e.g. Safari recently "Enabled support for a modern CSS parser, ported from Blink, that improves performance, specification compliance, and compatibility with other browsers, while also adding support for scientific notation in all CSS numbers" https://webkit.org/blog/7120/release-notes-for-safari-techno...).

I manage 134 branded sites for a university and a handful of other sites. We target IE11, Edge, Safari, FireFox and Chrome compatibility with desktop and mobile (ios, android). Once Microsoft announced sunsetting support for earlier browser versions our job got a lot easier. We all said "thanks MS, took you long enough".

In the last 30 days for the branded sites we have roughly these percentages of sessions by browser:

  chrome  46%
  safari  32% 
  ie      11%
  firefox 8%
  edge    2%

This for about 370,000 sessions.

We will continue to support the same browsers for the next two to three years and then adjust as needed.

Thank you, I'm always interested in seeing browser metrics in other industries. I'm assuming safari is so high due to your higher-ed audience.

We have a lot of Mac users in the student body. More than 60%. Also iOS users.

I bet the majority of your Safari numbers come from iOS users. A lot of Mac users ignore Safari and use Chrome.

Yes, in abstract Webkit only support sounds nice.

In reality, each mobile device and consumer computer has a different Webkit version.

Except Google and the Chrome team have forked WebKit as Blink, and are running quite a different development path.

There's now Blink, WebKit, Gecko, Trident in mainstream and a few others lurking.

Firefox already does that, started long ago. They even added support for -webkit prefixes for the same reasons.

As I can tell you from experience, they don't work all of the time, and it is a source of frustration for front-end developers I know who need to work with it.

You're not supposed to use the webkit prefixes. The only reason firefox has them is because some older websites will stop working.

Most of them have equivalent standardized features, or are in a state where you really shouldn't be using them. Don't expect every webkit prefix to work in Firefox.

I'm not sure we're supposed to think about prefixes at all, but instead just let https://github.com/postcss/autoprefixer do its thing.

So yeah, using prefixes to make props work on older browsers is fine. Using only the prefixed version of a property or a property that doesn't have an unprefixed version at all is not.

Reply on news.ycombinator.com