Paul Noone wrote: > Ric Raftis wrote: > > Paul Noone wrote:
>> >So setting the font size for the html element to 100.01% and then >> >adjusting it in the body (or elsewhere) is no longer recommended >> >I tried to find fault with Owen Briggs' Text Sizing >> ><http://www.thenoodleincident.com/tutorials/box_lesson/font/index.html> >> >article which uses a simple declaration of font-size: 76% in the body. But >> >no amount of nested lists in nested tables could reduce the usual array of >> >inherited sizing that I recall from not so long ago. >> >So now I can cut yet more dead wood from my CSS. Samuel will be so proud. :) >> Where did you get that from in that article? Setting the font size to 100% >> and then setting individual elements to ems is how I do all my pages. As >> far as I know it is the recommended method so users have control of their >> own viewport. > Where I got it from was the supplied stylesheet. The comments within also > explain why 76% was chosen as a figure. > The 100.01% size for html or body elements was/is a much practiced method > which was expounded on this very list not so long ago. 100.01% on body serves multiple purposes. First it's to avoid a serious IE inheritance bug often seen when setting a size in body in ems. Second, some old Opera browsers have a rounding problem with inherited sizes that the fraction fixes. Third, it ratifies the fact that the default size is the user's preference size, a statement of respect for the user. Briggs is really no one deserving the status of example to repeatedly point people to. Early on he says "most browsers default to a text size that I have to back up to the kitchen to read", which he follows shortly with "it's easier to read text that's smaller than default, and a little larger than the toolbar font", but without any indication what he means by "a little". His latter I agree with (see http://members.ij.net/mrmazda/auth/defaultsize.html#note1 ), but then he goes on to elaborately recommend body be set to 76%. First note the impact of 'body {font-size: 76%}'. CSS "sizes" are nominal. Real sizes are multi dimensional. As applied to screen fonts, there are two applicable dimensions, height, and width. Anything you make 76% shorter you also make 76% narrower. The effect then is multiplied. If your default initially (100%) is 16px, your character box should have 16 vertical pixels, and about 8 horizontal pixels, for a total of 128 pixels. Applying the 76% rule, you get roughly 6 horizontal by 12 vertical, for a total of 72 pixels. That's 72/128 -> 9/16 (56.25%) of the original 16px _size_. See http://members.ij.net/mrmazda/auth/area76.html to visualize the area vs nominal size concept. Now let's apply some math to what he says, using my two most used systems as a singular example. Both have the toolbar/menu text set to 10pt, and a default size that equates to 12pt, or 20% larger nominally, which is a bit less than 44% larger in area. Hopefully, this would fit within Briggs' definition of "little larger than the toolbar font". Now apply his 76% to my 12pt default, and guess what happens? 9.12pt (12.16px @ 96 DPI), or _smaller_ page text than toolbar text! Browser makers provide users with a preference adjustment precisely so that they can optimize to the size that best suits them. This personalization is one reason why the machines most use are called personal computers. 76% is totally arbitrary, in spite of Briggs' supposed rationalization, and applies no matter what the default, however larger or small, happens to be. Designers should instead defer to whatever the users prefer, leaving content P text unsized, respecting that personalization, however many or few actually do it. Too small text is the #1 complaint from web users: http://www.useit.com/alertbox/designmistakes.html Make the web accessible. Use your visitor's pref size for most of your content. It's something they have a right to expect you to respect. http://members.ij.net/mrmazda/auth/accessibility.html -- Jesus Christ is the reason for the season. Team OS/2 ** Reg. Linux User #211409 Felix Miata *** http://members.ij.net/mrmazda/auth/ ****************************************************** The discussion list for http://webstandardsgroup.org/ See http://webstandardsgroup.org/mail/guidelines.cfm for some hints on posting to the list & getting help ******************************************************
