Exactly. And now your "semantic" HTML is full of extra divs that are there entirely to support the layout, along with a whole bunch of float rules that again aren't really there to float content but because it's still the only way to get sensible column layouts in CSS.
div elements have no semantic meaning, and it's perfectly okay to include them to provide style hooks. A page using this trick is still using semantic HTML.
Forgive me for skipping the NSFW link, but in any case, I don't dispute that using CSS for layout is often the right approach. I am merely saying it is not "always the right approach, because".
If you're trying to make a page that is standards compliant and semantically rich, it is the only approach.
OK, let's try a few simple examples. Here's one from a web app I was working on yesterday: I want to display a short, plain text message in some sort of notification or dialog area, with an icon next to it. The icon should be a fixed horizontal distance from the text, and the whole icon+message should be centred within the containing block, with the text lines wrapping if necessary.
There are several ways to solve your example:
Using a float-based lay-out. This is compatible with web browsers going back at least 5 years.
Using display: inline-block on the icon's container and the message's container. This is only compatible with the newest versions of today's web browsers.
Using display: table; and friends to re-create the table using CSS. Sadly, even IE8 hasn't implemented these.
If the icon itself does not have real semantic meaning, you can use it as the message container's background and apply some padding to keep it visible.
Solution 1 and 4 present the additional problem that you can't center something of which the width isn't known. If it's allowed to set a width, no problem. Otherwise two extra containers are needed along with some float rules to fake it.
Now for my example. Years ago I took this page (warning: adult site, may not be work-safe, but when I visited it to get the link it was safe), and re-created it into . It's a good example of a page that makes extensive use of tables for its lay-out, which I successfully converted to show it could be done without hacks.
No, I mean that web developers who are advocating CSS over table layouts frequently justify this on the grounds of accessibility, but I'd bet good money that most of them have never heard a single page read aloud by a screen reader, let alone done actual usability testing with a partially sighted subject.
Do they have to? At least they're making an effort instead of clinging to web design tropes from 1998 that would leave their pages even worse off. I think many developers just need more experience with the craft. I know I'm still learning things despite having a good grasp of it. There are so many things to keep track of while trying to make a page semantically rich.
There are many genuine advantages of CSS and many genuine problems with table-based layout. However, the anti-table crowd still look pretty stupid when you're talking about some trivial page layout and they are advocating 50-line CSS solutions that work on most browsers from the past three years in preference to 5-line table-based solutions that work reliably on every browser since forever.
50-line CSS solutions? What have you been reading?
Websites that make extensive uses of table-based lay-outs take much more than just 5 lines, while the equivalent CSS is lighter by comparison.
They look even more stupid when they "justify" their position based on usability and accessibility concerns that most of them have never experienced, with implications they don't even understand.
You mean web developers are not allowed to argue for accessibility for, say, the blind, because they're not blind themselves? Please tell me you're joking.
This is a flawed analogy. People who never had sex crave it much less than people who have, because they don't know what they're missing. More importantly, without oxygen we die, but we can live without sex.
Honestly, this is common sense. Hardly article worthy.
Sadly, "common sense" isn't common at all. You should know this. Otherwise we wouldn't have most of the population thinking the web's porn domains are the scary part of the web.
They're making a claim involving a ratio. I'm pointing out a flaw in their claim.
There is no flaw in their claim, as there are more porn sites out there than you think. More than a third of the domains consists of porn. Hence if there are 99 regular infected domains for every infected porn domain, there is a lot more malware on regular domains.
And they've "proven" something that was already widely known in the security community and re-packaged it as news.
The question is, though, has it ever been proven before? There are things in life we assume to be correct as part of common knowledge, but they have never been proven. Sometimes those turn out to actually be incorrect, so research like this is useful.
Okay, Avast is not free for business, but I think they deserve less scrutiny because they do give it away for free to everyone else. They're not, say, Symantec. *shudders*
Exactly. 99 regular sites infected to 1 porn site infected is rather telling if there are 1000 times more regular sites, which would mean that you have a ten-fold increase in risk on porn sites.
Way to miss the point, which is that avoiding porn sites doesn't mean you won't get infected. This proves that there are many regular sites out there that are infected, ratio of anything be damned.
As usual, this is a non-story that boils down to nothing more than a press release for Avast: "You're at risk! Buy our crap which will slow down your computer and probably won't detect much anyways!"
I see that happen all the time, where the woman works until birth is close just to get the 3 months extra pay, knowing she is going to not come back. I can't blame her, even if it causes problems for work.
Can't blame her? This is exactly the kind of selfish, opportunist stunt that gives working, pregnant women a bad name. It's abuse of the system.
A good leader with a clear vision and realistic project management will lead to a successful end product. Linus Torvalds has managed to create a superior operating system to Microsoft, who employ thousands and pay very well.
Wrong. Linus Torvalds created a superior operating system kernel. There is no one Linux OS to refer to, because people using Linux as their kernel can't agree on anything else.
For fuck's sake, when will people understand that the listings on w3schools.com ARE ONLY FOR THAT SITE AND NOT REPRESENTATIVE OF THE GLOBAL MARKET SHARE?!
Theora is good. It just doesn't have as good quality as H.264, and a lot of people seem to be hung up on having a codec with equivalent quality, even though most web video doesn't require it at all. Screw them. A codec that runs great on pretty much any computer is preferable H.264 resource-intensiveness.
Yeah, I have the same experience while taking the bus. Almost always the next thing they say after taking the phone is "I'm on the bus.". I think people should buy a portable GPS or something, because that seems to be what most are using their mobile phones as.
Slashdot Mirror
The Nintendo DS doesn't have upgradable firmware and has a thriving homebrew scene, so I wonder what it is you're avoiding by not buying one.
div elements have no semantic meaning, and it's perfectly okay to include them to provide style hooks. A page using this trick is still using semantic HTML.
If you're trying to make a page that is standards compliant and semantically rich, it is the only approach.
Whoops, I made a HTML syntax error and the link to my page got munched.
There are several ways to solve your example:
Solution 1 and 4 present the additional problem that you can't center something of which the width isn't known. If it's allowed to set a width, no problem. Otherwise two extra containers are needed along with some float rules to fake it.
Now for my example. Years ago I took this page (warning: adult site, may not be work-safe, but when I visited it to get the link it was safe), and re-created it into . It's a good example of a page that makes extensive use of tables for its lay-out, which I successfully converted to show it could be done without hacks.
Do they have to? At least they're making an effort instead of clinging to web design tropes from 1998 that would leave their pages even worse off. I think many developers just need more experience with the craft. I know I'm still learning things despite having a good grasp of it. There are so many things to keep track of while trying to make a page semantically rich.
Clearly you don't know what a "niche market" is.
Game Boy Advances and DSs all use ARM processors, and they weren't niche at all.
50-line CSS solutions? What have you been reading?
Websites that make extensive uses of table-based lay-outs take much more than just 5 lines, while the equivalent CSS is lighter by comparison.
You mean web developers are not allowed to argue for accessibility for, say, the blind, because they're not blind themselves? Please tell me you're joking.
This is a flawed analogy. People who never had sex crave it much less than people who have, because they don't know what they're missing. More importantly, without oxygen we die, but we can live without sex.
Sadly, "common sense" isn't common at all. You should know this. Otherwise we wouldn't have most of the population thinking the web's porn domains are the scary part of the web.
There is no flaw in their claim, as there are more porn sites out there than you think. More than a third of the domains consists of porn. Hence if there are 99 regular infected domains for every infected porn domain, there is a lot more malware on regular domains.
The question is, though, has it ever been proven before? There are things in life we assume to be correct as part of common knowledge, but they have never been proven. Sometimes those turn out to actually be incorrect, so research like this is useful.
Okay, Avast is not free for business, but I think they deserve less scrutiny because they do give it away for free to everyone else. They're not, say, Symantec. *shudders*
Wrong. You can't trust any site.
Way to miss the point, which is that avoiding porn sites doesn't mean you won't get infected. This proves that there are many regular sites out there that are infected, ratio of anything be damned.
Avast is free. It even says so in the summary.
It really doesn't. Dating someone like that is hell. You'd be rather be alone.
You did make it past the first sentence. You just quoted two of them.
Can't blame her? This is exactly the kind of selfish, opportunist stunt that gives working, pregnant women a bad name. It's abuse of the system.
Wrong. Linus Torvalds created a superior operating system kernel. There is no one Linux OS to refer to, because people using Linux as their kernel can't agree on anything else.
This isn't the first time I've read someone making a website for a church, so I have to ask: why does a church need/want a website? What's the point?
Definitely. DEFINITELY!
http://eloquentscience.com/wp-content/uploads/2010/02/bobsqu.gif
Please stop being an idiot.
For fuck's sake, when will people understand that the listings on w3schools.com ARE ONLY FOR THAT SITE AND NOT REPRESENTATIVE OF THE GLOBAL MARKET SHARE?!
Theora is good. It just doesn't have as good quality as H.264, and a lot of people seem to be hung up on having a codec with equivalent quality, even though most web video doesn't require it at all. Screw them. A codec that runs great on pretty much any computer is preferable H.264 resource-intensiveness.
It is based on VP3, but a lot has happened since VP3 was open-sourced.
Fortunately, they can never stop the people who really want to make websites from finding another free service or paying for a hosting package.
Yeah, I have the same experience while taking the bus. Almost always the next thing they say after taking the phone is "I'm on the bus.". I think people should buy a portable GPS or something, because that seems to be what most are using their mobile phones as.