I shouldn’t have written black-on-white or white-on-black but have gone with dark-on-light and light-on-dark respectively. Sorry for the oversight. I was never planning to use a pure B/W colour scheme. It would probably even hard to find anyone recommending such a high contrast for reading long text on a screen. (Benchmark: #333 on #FFF. ¶ informationarchitects.jp/…)

Thanks for reminding me of that article again Kroc. I’ve read it before but I’ll bookmark it this time. It makes much the same point as Information Architects did above: the default size is 16px for a reason.

However I think you (Torbjörn) are still wrong with your assumption that the overall contrast is “[m]ore important than whether you have black on white or white on black”. Wikipedia for one will tell you that “vision and perception researchers [dispute] about whether it is actually easier or healthier to read text on dark or light background”. (secure.wikimedia.org/…) Renowned designer Mark Boulton went out of his way to mention the changes between typesetting DoL vs LoD in his series to better typography. (markboulton.co.uk/…)

I agree with the fact that the legibility of fonts is not decided by serifs, but when we’re talking about the web, you really just split it up into serif and sans-serif fonts. Either you go with a Lucida/Arial/Helvetica inspired stack, or you’re stuck to a Georgia/Times inspired stack. Since I’m only trying to collect some kind of statistic on what people prefer to read on their screen it really is a question of serifs. I’m not going to say either of them is actually proven to be better (they aren’t ¶ alexpoole.info/… I’m just looking for the one that is most preferred.

Good thing you brought up line measure, the reason I didn’t ask about it is because I wanted to ask quick and simple questions to get a base. Another way of checking your measure is the 2–3 alphabet lengths trick. You put the alphabet (meaning A-Z!) 2 or 3 times on a single line and take that as your measure. This is a good way of taking measure because it takes into account the variable letter width of the font you will be using.

Recap (TL;DR as the in-crowd calls it): thanks for the pointers Torbjörn, but I’m really just trying to collect personal preferences on the asked questions here ;-) I’m fairly up-to-date when it comes to typography.

I take a lot of inspiration from W. W. Norton & Company’s Hunting of the Snark (my all time favourite book). They use a big text size with a small measure. I uploaded some quick pictures of their typesetting here: flickr.com/…

Thanks everyone for the great input so far!

Watch out, the fourth link was parsed including the ), that should go behind it, here is a clickable one: alexpoole.info/…

Thanks for your input Grimmeh, much appreciated.

As far as this forum is concerned, Kroc’s statement from earlier (“If they’ve coded it well, then it won’t be too hard for me to change the colour and size using CSS if I disagree with it.”) obviously stands for this forum as well. If you feel the line-length is bothering you just apply something like the following to make the central column smaller again:

~~~ CSS ~~~>
body > section,
body > header > nav {
max-width: 760px;
}
<~~~

It’s not a bug, it’s a feature.

Readability (or Safari Reader) give points to wrapping elements depending on the content and then takes the wrapping element with the best score. A problem it would be having on this forum is that there is no real wrapper element that contains all content. (It ignores the BODY element because for most sites there is just too much in it.)

For the HTML of this forum it means every single reply is fighting to be the content displayed by Readability.

And yes, the idea with this topic was to get to some kind of baseline for displaying running text. Most readers will not know anything about setting their own style