Normalize v2.0.1 - optimized for modern browsers | Hacker News Reader

Normalize v2.0.1 - optimized for modern browsers

necolas.github.com

28 pointsnecolas13 comments

I believe you had a sample/demo page with v1? I thought that highlighted what the stylesheet achieves quite nicely.

Also, I reckon test case examples for each browser quirk would be cool, so people who wanna see what each thing changes/causes, can go through the commented code and see what removing certain declarations would change/effect.

Keep developing cool stuff necolas!

The test page still exists, but isn't currently surfaced on the project page - http://necolas.github.com/normalize.css/2.0.1/test.html. Thanks!

Changelog: https://github.com/necolas/normalize.css/blob/master/CHANGEL...

I don't get it, is everybody ready to drop IE7 support?

Maybe not everybody, but it’s a reasonable default. IE8 has been out for more than three years, and works on Windows versions as far back as XP, which in turn has been out for more than a decade and is nearing the end of support from Microsoft.

There is accommodating customers, usually large organisations, who for whatever reason have chosen stability in their system software. And then there is not taking advantage of new technologies that almost everyone in most markets can use, and that almost everyone else could upgrade to if they needed to, just to support a tiny and shrinking minority.

If you don't need to support legacy browsers, then you can use 2.x.

If you still need legacy browser support, then use 1.x.

Sorry, i did not make myself clear. It was more a general question. I was under the impression most webdevelopers still try to support IE7.

There are different values of "support". I "support" IE6 and 7 (I make sites for small businesses in the UK), as in, the site works, and the content is accessible. Those browsers make up 0.8% and 4.1% of the traffic last month respectively. (Why the hell is IE7 still used? How do people even have it installed???). For those, I'm happy to strip out sections if needed to get things to work, though this is rarely needed to be honest.

IE8 makes up 11.9%, so we do proper testing there, and the content should look as close as possible to modern browsers, but I'm happy to include polyfills etc where needed.

For us, Chrome is 21.1%, Mobile Safari is 17.5%, IE9 is 17.1%, and Firefox is 15.6%, so for us over 70% of users have a modern browser. These 4 are our target browsers, where everything should work.

Nice to see some actual data!!

While I do at least take a peek at IE7, I usually insert a messages on IE6 & 7 users simply telling them to upgrade their browser rather than debug. The only people who are unable to upgrade their browsers are generally gov't types. Everyone else needs to stop being so lazy and devs need to stop being so accommodating. I still support 8, which let's be honest, is basically as bad as 6 or 7 in a lot of ways but it still the highest version of XP. I know people think XP is enough and they stay there, but they can still go download Firefox or Chrome.

As a concession to old browser users, I will use XHTML or an older spec on a site that has large IE<9 usage. But I'm sure as hell not going to spend a lot of time debugging and fluffing the pages anymore. These people don't care enough about the Internet to not use IE7 or 8, then I don't care if their experience is sub par.

Personally I already have - I'm still supporting IE8, because there's no upgrade path from that if you're still on Windows XP, but there's really no reason to be stuck on IE7 these days. (If your workplace has any legacy apps with weird requirements, it's more likely you'll be stuck on IE6 than 7)

In certain segments: yes. If you're doing something aimed at developers or gamers, supporting IE (any version) isn't all that necessary.

I "dropped" IE8 support in 2010: http://impactjs.com/

v2 strips down to 2k and v1 strips down nicely to 2.5k via css minifiers

Hard to tell which mini-comments are needed for hacks or not.

Reply on news.ycombinator.com