I still don't quite understand how divs / CSS make it easier to change a page design when you consider the amount of testing to ensure the changes work on all browsers, especially with all the hacks and so on. Its a hugely frustrating and tedious process which wastes large amounts of time and money. Thankfully the 508 legislation only applies to the USA (land of the free - yeah right) and so being as I am based in the UK, I can develop web sites in whatever style I choose. Contrary to popular (US) belief, legislation made in Washington doesn't apply to the rest of the world - thank goodness for that. It must have been a good day in the world of web design the day the legislation came into force. I think I'm becoming increasingly cynical as I get older with 25 years in the IT industry but I feel sure this kind of legislation is just to protect jobs. In reality anyone can knock together a reasonable web page with a couple of tables. It takes a lot more effort and knowledge to do this with DIVs / CSS. In my experience it can take hours and hours Googling to find solutions to quite simple problems and reading incomprehensible articles in forums full of idealistic zealots all argueing about the 'right' way to do things. You can't just dip your toe in the water and get things to work properly in every case. It also seems to me that the lack of a definitive guide to using DIVS / CSS "out of the box", that applies to all situations, working on browsers, and written using 'normal' language with no geek speak, also smells of a bit of protectionism. I'm an application developer and I would say it takes almost twice as long to figure out layout problems and test against all browsers than it does to create the basic application, design and implement business objects, and create the database back end. My time = money, both for me and my customers alike so I am sorry if I don't reject all the pro DIV / CSS arguments in favour of cutting costs and providing value for money for my customers. Maybe its just the way that developers minds work, but it seems to me far easier to change a complex table structure than it is to modify DIVs / CSS. Thankfully it now appears that a solution to these issues is now available - its called WPF.

根据我对表的了解，如果嵌套了太多的表，在呈现页面时浏览器会有很大的开销。

1 -浏览器必须等待呈现最终视图，直到整个表被加载。

2 - The algorithm to render the table is expensive and is not in a single go. The browser, as and when, gets the contents, will try to render calculating the content width and height. So, if you are having nested tables, say, the browser has received the first row and the 1st cell is having large amount of content and width and height not defined, it will calculate the width and will render the first row, In the mean while it gets the 2nd row will cell#2 having loads of content! It will now calculate the width for 2nd row cells.. What about the first ? It will calculate widths recursively. That's bad at client side. (To site an example) As a programmer, you'll optimize stuffs such as time to fetch data, optimized data structures and etc. You optimize things to complete on server side, say in2 secs, but end user in getting the final view in 8 secs. What is wrong here ? 1. May be network is slow! What if network is fine ? What is network is delivering the contents in next 1 sec ? Where is this extra 5 secs getting consumed ? Thing to worry about-- The browser might be taking lot of time in estimating and rendering the tables!

如何优化表? 如果你使用表格，我建议，总是定义单元格的宽度。这并不能保证浏览器会盲目地只取这个宽度，但会对浏览器决定初始宽度有很大的帮助。

但是，最后，div是CSS可以被浏览器缓存的好方法;而表没有缓存!

超级简单的回答:用表格设计可维护的网站是很困难的，而用标准的方法来做是很简单的。

网站不是一个表格，它是相互交互的组件的集合。把它描述成一个表是没有意义的。

如果你在这方面支持表格角度，找一个有表格的网站，然后给自己买一个屏幕阅读器——关掉屏幕阅读器，关掉你的显示器。

然后尝试一个不错的语义正确的div布局网站。

你会发现其中的不同。

如果表格中的数据是表格而不是为了布局页面，那么表格并不是邪恶的。

我尽量避免使用TABLEs，但是当我们设计复杂的表单，混合了多种控件类型和不同的标题位置，并对分组进行了非常严格的控制时，使用DIVs是不可靠的，甚至几乎是不可能的。

现在，我不会说这些表单不能重新设计以更好地适应基于DIV的布局，但是对于其中的一些表单，我们的客户坚决不改变以前版本(用经典ASP编写的)的现有布局，因为它与用户熟悉的纸质表单相似。

因为表单的表示是动态的(其中某些部分的显示是基于案例的状态或用户的权限)，所以我们使用一组堆叠的div，每个div包含一个由逻辑分组的表单元素组成的TABLE。TABLE的每一列都被分类，这样CSS就可以控制它们。这样，我们就可以关闭表单的不同部分，而不会出现在div中不是表来换行的问题。

1:是的，你的用户很关心。如果他们使用屏幕阅读器，它就会丢失。如果我使用任何其他试图从页面中提取信息的工具，遇到不用于表示表格数据的表是一种误导。

div或span用于分隔内容是可以接受的，因为这正是那些元素的含义。当我，一个搜索引擎，一个屏幕阅读器或其他任何东西，遇到一个表格元素，我们期望这意味着“以下是表格数据，表示在一个表中”。当我们遇到div时，我们期望“这是一个用于将内容划分为单独部分或区域的元素。

2 .可读性:错误。如果所有的表示代码都是css，我可以阅读html，我将理解页面的内容。或者我可以阅读css和理解的表示。如果html中的所有内容都是混杂在一起的，那么在我甚至可以看到什么是内容，什么不是内容之前，我必须在脑海中剔除所有与表示相关的部分。 此外，我害怕遇到一个不懂css的web开发人员，所以我真的不认为这是一个问题。

3:表比较慢:是的，它们比较慢。原因很简单:在呈现表之前，必须对表进行完整的解析，包括其内容。可以在遇到div时呈现它，甚至在它的内容被解析之前。这意味着div会在页面加载完成之前显示出来。

还有一个好处是，表格更加脆弱，在不同的浏览器中呈现的效果并不总是一样的，不同的字体和字体大小以及所有其他可能导致布局变化的因素。表格是一种很好的方法，可以确保你的网站在某些浏览器中会偏离一两个像素，当用户改变字体大小或以任何其他方式改变设置时，它不会很好地缩放。

Of course #1 is the big one. A lot of tools and applications depend on the semantic meaning of a webpage. The usual example is screen-readers for visually impaired users. If you're a web developer, you'll find that many large companies who may otherwise hire you to work on a site, require that the site is accessible even in this case. Which means you have to think about the semantic meaning of your html. With the semantic web, or more relevantly, microformats, rss readers and other tools, your page content is no longer viewed exclusively through a browser.