This is the kind of list that works not to the advancement of standards-based creation, but its detriment. Firstly, the tips are just thrown out there, which will certainly not help the CSS novice understand the reasoning behind them. Second, they're mostly outdated/unnecessary. And third, as already mentioned in this thread, there's a case to be made for less hacks.
Regarding hacks specifically: if you feel like you need to use hacks, at least try to understand how they work and the problem they solve. i.e: don't just throw something out there to solve IE problems if you don't know what they mean. It's amazing how many people use box model hacks in their code but don't even know how the box model actually works. It is certainly one part of the spec that isn't really obvious for most people, but copying/pasting a few lines will certainly not make you a better coder.
I completely agree. When I first learned CSS it was largely through tips and tricks like these, but without any sort of mental model or big picture understanding of what was going on my progress was unimaginably slow. I had to actually read part of the spec and start looking at and modifying stylesheets used on real sites before I became at all comfortable.
When I was learning, I got really quite confused by the intersection of floats, overflow and positioning. Reading the spec and actually understanding all of the rules at the same time, rather than just what I could glean from random hacking and tutorials, was really valuable.
I read all 31 snippets and I am still a terrible coder. Perhaps a better title would be "31 random CSS hacks, copied from other blogs, in no particular logical order"
Personally i believe that you should use the least amount of CSS hacks as humanly possible. The only one i use on a regular basis is clearfix. I usually dont have to use any others to get my layout working in everything except IE. I then use an IE specific style sheet (conditional comments) to break my layout for IE.
I agree. Anytime I use a CSS hack it makes me feel dirty. An IE specific stylesheet is an unfortunate necessity, but there seems to be a cottage industry of people using browser-specific hacks for everything. It feels like it goes against the whole spirit the CSS standard. Maybe it is a case of just getting the job done, and being forced into nasty hacks by a half-dozen browsers implementers who are supporting different parts of the standards on their own schedule.
After struggling a lot with the browser specific hacks to target IE/Webkit/Whatever, I have now switched to CSS browser selector [1] and never looked back. It simply saves a lot of time and effort. If you think about it, the User Agent String is the only accurate information that can be used to sniff the browser name and version, not some random CSS hack.
BTW the article is mostly useless for any practical use. If you really want to target IE6/IE7, this article [2] is a nice start. Learn about what types of bugs IE support and try to remember the name of the bug so that you can google a fix for it when you actually encounter one. As others suggested here, try to understand at least the basics of box model and how it is different in IE, but don't try to get a PhD in any IE bug you encounter. Fix it as fast as you can using whatever easy way out you can see an move on. There are better things to do in life.
Using the User Agent String to do anything in CSS or JS is a giant no-no. IE conditional comments, CSS hacks, and feature detection / polyfills are the industry standards. The UA String can be non-existent, mishandled, or just plain-old fabricated.
I find I rarely need to target any browser other than IE with specific styles, so I usually end up doing something like this, which doesn't rely on JS.
Actually I didn't know about that article, but is nice to see the origin of this technique. Mostly took inspiration from HTML5 Boilerplate (also Paul Irish), and previous articles I read. But yes, Paul Irish is awesome, and deserves lots of credit.
>User Agent String is the only accurate information that can be used to sniff the browser name and version, not some random CSS hack.
Those random CSS hacks should always work as they are relying on the specific functions of the UA which can not be altered (though they might clash with new browsers, * html anyone?), whilst as a sibling comment notes UA can be easily modified.
A word of caution regarding "ultimate font stacks" - most of these are an utopia. Not all fonts have the same height and x-height (read - the same letter size and general visual appearance) at the same pixel size. Most notably, Constantia just does NOT work as a drop-in Georgia replacement. Nor does Calibri work as a Verdana substitute. They needs either a different font-size or line-height.
In other words - stay with a conservative font stack or do on the fly font detection and restyle the document respectively.
Knowing how to use CSS hacks to target specific browsers does not make you a better coder. Knowing how not to need those hacks makes you a better coder.
Regarding hacks specifically: if you feel like you need to use hacks, at least try to understand how they work and the problem they solve. i.e: don't just throw something out there to solve IE problems if you don't know what they mean. It's amazing how many people use box model hacks in their code but don't even know how the box model actually works. It is certainly one part of the spec that isn't really obvious for most people, but copying/pasting a few lines will certainly not make you a better coder.