Elegant Web Typography
slideshare.net
slideshare.net
The 62.5% trick doesn't work as well as the designers claim, because there are a lot of people where the default font size isn't 16px tall (e.g. in high DPI mode or when the browser's font size setting is not "normal"). K.I.S.S. and keep everything font-size: 100%, except fine print which can be >= 80%, and headings which can be as large as you want. Basically, any use of "px" and "pt" for text in your stylesheet is just going to cause headaches.
It is true that sIFR works better than what a skeptical person would expect. But, I still have problems with it all the time, copying/pasting stuff into my notes. Don't use it if you can avoid it.
Finally, Jeff completely misses the point of "vertical rhythm" as mentioned in "The Elements of Typographic Style." This is probably the #1 comprehension failure amongst readers of the book. Briefly, on the web you don't need to adhere to the vertical rhythm guidelines as strictly as you need to when you are working with text PRINTED with facing pages and/or double-sided. Unless you have multiple columns of text on your page, you shouldn't be adding extra vertical padding just to maintain a vertical rhythm, because that extra padding just leads to more scrolling, which is horrible for readability.
I already commented elsewhere on the page about the problems with the leading advice.
Regarding quotes...wow, this is already getting too long. I should just write an article and submit it to ALA.
sIFR is designed to a) be frosting on top of a standards compliant HTML cake and b) fail very gracefully in every combination of javascript or flash availability. When you design using sIFR you are heavily encouraged to build the design with common web fonts first and then simply overlay the sIFR to be optionally rendered when possible. It even preserves behaviors like anchoring and copy/paste.
The idea does, at first, sound extremely convoluted and toxic, but should be instead thought of as a really clever hack to make interesting typographic design possible despite constraints of font distribution online.
Edit: Another criticism of this presentation is that it implies Internet Explorer doesn't support @font-face, which is does - in fact, it invented the property which was later standardised in CSS2. The difference is that while other browsers generally support TTF, OpenType or both, Internet Explorer 5-7 only accepts the proprietary Embedded OpenType but a free converter from TTF is available. It also misses one big negative of using sIFR - Flash renders fonts differently from the rest of the page (possibly due to Flash using odd aliasing) making it look weird and 'stand out' compared to the rest of the site - I have no idea why he author is so quick to embrace something that looks so distinctively out of place when trying to encourage coherent design.
Your claim that it is logically harder for the eye to find the start of the next line is easily refuted. Think of the extreme case: each line is half a page apart. Clearly, the eye has no problem finding the next line.
I'd like to hear other people's opinions too.
I don't see how that's a refutation at all - if anything I think it supports my case. If each line is half-a-page apart, when you reach the end of the line you have to look down to find the start of the next one because you can't see where the next line is within your natural eye-line, preventing the sentence from flowing naturally.
Whatsmore, people have no problems reading books or magazines which almost always have a relatively narrower leading than sites like A List Apart.
I think the middle-ground is optimum for readability - if it's too narrow it's easy to mix up which line you're reading, if it's too wide it's difficult for the eye to find the next line so reading doesn't flow naturally. I can't help but conclude that someone read "leading is good" somewhere and took it to its logical conclusion which is making it excessive and this meme has caught on, much like how Fitts's Law has been abused as an excuse to make things unnecessarily and annoyingly big for little gain.
For a typical web page, line spacing as large as 1.4 or 1.5 does more harm than good. First of all, it doesn't do any good (see http://blogs.msdn.com/fontblog/archive/2006/05/17/600507.asp...). Secondly, it makes the user scroll more to read the content. Scrolling is way worse for readability than tight line spacing.
A lot of people use Verdana, and Verdana is a very legible typeface. But, because it is so open, and because of its enormous x-height, it tends to look wrong without excessive line spacing. Also, its normal weight is basically semi-bold at small sizes on LCD screens with Cleartype enabled, and that also encourages excessive line height. I've not seen any evidence that says that these huge line heights actually improve reading performance; the primary benefit of huge line spacing is an improvement in the "color" of the page when using Verdana.
Microsoft's new Cleartype font collection (the C* fonts) are slightly less open, don't have such an enormous x-height, and are less bold at small sizes. Consequently, a line height of 1.2 or 1.25 is plenty when with less than 75 characters per line.
For instance, I'd bet you'd be hard pressed to find many books printed professionally which don't have at least half a line of leading.
How leading affects someone's ability to scan lines, I don't know so much about. In my experience, medium text color is more pleasing to read in general -- I feel less eye strain, less cramped -- but I don't experience much difficulty following lines regardless of leading.
Serif'd fonts are harder to get to look right, absolutely. But they can be used well, and they look incredible when they are.