On Sat, 08 Jan 2005 09:17:53 +1100, russ - maxdesign <[EMAIL PROTECTED]> wrote: <...> > As pointed out, these are guidelines only, and open to interpretation. For > example, pixels could be interpreted to be relative units, as explained by > Derek Featherstone: > http://www.wats.ca/articles/pixelsarerelative/65 > The main point that seems to be missing in this discussion is that you could > argue that: > A. pixels are relative units and therefore acceptable within the WAI > guidelines
That's my point. And there is nothing to argue about: pixels are relative units. > B. browsers should support scaling pixels (and therefore IE is wrong) IE for Windows. IE5 for Mac was the first browser with text zoom, IIRC. <...> > Accessibility should not be seen as check points, laws, lawsuits or covering > your bum. It should be about people and empathy - putting yourself in others > shoes. Absolutely. > It is also about the real world. A huge percentage of users are on > Windows IE and within the disabled community, you could argue this figure is > even higher as most accessibility tools seem to be run on that platform. That is the problem - the real world. So far I've seen only assumptions. I did not exactly get your point on disabled community (unless you have other problems in mind, not only impaired vision). For screen readers doesn't matter what size your font is, others may have using low resolutions on big screens, so they are in relatively same position as the rest. > I'd put the question back to the group... Rather than ask "why should I not > use pixels, as there is nowhere that forces me not to", why not ask "how can > I make my content as accessible to the widest audience possible". Because of laziness. As simple as that. Pixels are the easiest way to have consistent result (does http://old.alistapart.com/stories/fear4/ matter any more?). <...> > If you ask this question, then right now, with the current browser > situation, this means that pixels are not desirable as they can possibly > (regardless of whether a guideline or not) adversely affect a large number > of users. <...> Non sequitur. That's why I mentioned 'possibility vs. probability'. Maybe that may affect some users. Maybe not. I may be lazy and implement pixels. Or I may put some effort and make text work with other units. This will be used in case: a) Font size I set looks unacceptably too small for particular user on his screen b) Users uses IE/Win c) User knows how to change font-size d) User chooses to change font-size I have no idea about probability of a). We have very varying results on b) (you can have more exact numbers for site in question though). We have no idea about c) and d). So we may as well end up spending time and money to implement something what is never used. What I see as a biggest overestimation is the idea that user _wants_ to control something on your page. No! User wants content of your page. Or functionality of your application. Or whatever. And d) is very important here: users are much more comfortable with hitting "back" button than in setting font-sizes. So if you committed a sin of infamous font-size-too-small and it is small bellow acceptable level, I'd say there will be one visitor going somewhere else, than one visitor adjusting font-size. But that's an assumption too. Don't get me wrong. I am not advocating pixels as best possible way to set font-size. Methods which allow to scale fonts are indeed preferable, but I still think that evilness of pixels is hugely overestimated as is users will to change it. And whats is not desirable: some percentage based schema, which is broken so users get illegible fonts, or reasonably sized fonts in pixels, plus stylesheet switcher allowing to change size with one click? How big is probability of the first case? What is the probability of user being unhappy with default setting in the second case? Who knows, who knows... I'd say we have more prominent and less arguable problems with accessibility than guessing game about font-sizing. Strive to perfection is nice, though. Don't use pixels. Don't yell, if someone uses. Regards, Rimantas -- http://rimantas.com/ ****************************************************** The discussion list for http://webstandardsgroup.org/ See http://webstandardsgroup.org/mail/guidelines.cfm for some hints on posting to the list & getting help ******************************************************
