What are the disadvantages of fixed-width design?
"The key to good web design is in not putting barriers in the way of peoples discovery and use of the information." -- Alexander Johannesen
Lets look at the arguments that keep cropping up that point to fixed-width design as being good. The points in this page are derived from posts of Alan Flavell, Alexander Johannesen, Eric Jarvis and a number of notable regulars in ciwah during an excellent thread in comp.infosystems.www.authoring.html during the second half of 2002.
"Web design should focus more on what designing for the web really is, and not so much about how a design made for paper is going to be done on the net. Re-educate the designers; the web is not paper." -- Alexander Johannesen
Its the numbers!
"The main strength of the Web is that users of any browser and any browsing situation, anywhere, can access the same information rendered in the best way possible for that browsing situation; the content is independent from the presentation, and as such the presentation can be adjusted to the needs of individual readers. No other medium until now has allowed this, and it's really sad seeing people work against the strength of this new medium rather than with it. See Abigail's dream"
Argument: "The reason why 800x600 has been adopted as the standard Web page size is clear: The average screen resolutions of Web surfers is 800x600"
Browsers don't send this information to the server. These numbers are obtained using client-side scripting of one sort or another. However, that makes the statistical variation of the sampled data dependent on the statistical variation of another set of sampled data: the amount of people running with Javascript unavailable, turned off, stripped, or non-functional.
Even the method of collecting the information isn't without bias. Using client-side scripting excludes browsers, and the biggest culprit is the self-fulfilling prophecy - by designing a site that only works in one resolution, visitors not using "The-One-True-Setup" don't visit thus unbalancing the numbers.
There is no need to estimate the stats, we already know the likely range. It goes from mobile phone to 1600px plus screen projection. We know the drawbacks at each extreme, so simply look for ways to control the way the page elements react to very narrow or wide settings. Control the way the extremes of text size are dealt with. Trying to predict settings for some mythical average user is a waste of time and effort.
Fixed design is more readable and has a better graphical profile
"A design that breaks down in the fluid and flexible environment which is the web may not be a good design FOR the web. Any design that takes advantage of the reality in which it exist is better than one which do not."
It is a matter of attitude towards creating designs that cater to the information given instead of the assumptions we have about who wants the information. What exactly justifies a fixed design? Supposedly, readability and graphic profile. The problem here is that there is no readability breach; it is a breach of our assumption on what the user finds readable. The whole thing about graphical profile is something that graphical designers just don't get (at least, not most of them) that the web is not a piece of paper, no matter how much they'd love to think so.
Future screen sizes are bigger!
"I find myself more and more in the need to dump documents and pages to my PDA, and use them as referance material in meetings, especially going to customers where taking the laptop is a bit cumbersome. Have you ever tried dumping your average webpage to an average PDA, like Palm III or Handspring Plus? Most of the time it is a *waste* of time because people don't think that this need is out there. Sure, if enough people shout about it, they might create a "PDA friendly" version, and we're back to poor original designs wasting money and time." -- Alexander Johannesen
Quite the opposite is actually the case. The trend is clearly towards more diversity. The people I see in the street or on the train are looking at rather small displays. They sure aren't lugging 21 inch monitors around with them!
All browsers are maximised!
The stats only look at the monitor sizes, but many surfers do not maximize their browsers. Studies indicate 20% to nearly 40% of the visitors have set their browsers to widths less than their monitors.
The flexible approach is the right way to go since you can't tell what width the browser is.
One site doesn't fit all! Mobile phones will have a different website!
"Any company with a little sense in spending money understands perfectly well that "one site fits all" is better economy that "one page per who-knows-how-many interface devices" -- Alexander Johannesen
There is actually no evidence (or justification) that a person using a mobile phone to access the internet is more likely to want access to a different type of information than another user.
Web users have access to a wide range of browsing technologies, which they'll use as and when it suits them. If they get the impression that a site is making it unreasonably hard for them just because they're presently using their palmtop/cellphone-combo rather than their multimedia station, they're going to be annoyed, and unlikely to revisit later. If the only issue is that the site has some additional content that is technologically inaccessible to them (Java applets on a text-mode browser, let's say), then they'll be more understanding of the situation and might (as long as the author isn't deliberately abusive to them for having the wrong equipment) very well decide to re-visit later to see what the extra stuff can do.
It works in print!
"Any site that uses an organisations pre-existing print designers for the initial site design is liable to end up with a fixed sized design exceedingly similar to their print material. sometimes they move on, sometimes they don't." -- Eric Jarvis
Designing for the web is not like designing for other mediums - you have little ultimate control over how your page will appear and in many repects that is a strength not a weakness of the medium. It's all about making content available to all kinds of different types of devices.
A good analogy is that of a composer. He writes a piece of music, and it has a general sound that may be suggested by annotations in the score, but it will sound different depending on the style of the conductor, the orchestra, and even the venue in which it is performed. But people who appreciate music acknowledge, and even rejoice in these differences, often to the point of collecting recordings of the same piece performed by different orchestras. Similar observations can be made about those who write plays and ballets.
All our target audience use X!
"A target audience consists of people - not of specific computers and (as Arjun says it) wowsers." -- Alan Flavell
Start designing for the Web instead of Print
Website design is where you need to forget absolutely everything you have ever learned about print design and try looking at the web. Ignore all sites you've ever seen - don't compare it with other media.
What do you have to do? Make the site work and look good for as many users as possible.
OK, so how do you do that for the web? You look at the range of browsing situations that the site may encounter and set the parameters you have to operate within. This is on a par with paper is two dimensional. It's a pain in the proverbials if you want to make a sculpture, but there is no point complaining, it is the nature of the medium.
Start by the brutal reality of a text to speech or plain text browser. Make the content make sense in its raw state. Treat that as the starting point, it is part of web design and it will give a solid basis to work a presentation from that will work unless you do something seriously stupid (except now and again when Netscape 4 does something daft above and beyond the call of duty, this is surprisingly rare if the content is marked up to work well on its own).
Once that is set you can get visual - but don't go straight at the overall layout, you'll just box yourself in. The next stage is to get each element of the page so that it works independently of how it ends up compared with the others, because somebody somewhere will end up breaking any layout you can possibly make - but you can set individual elements such as a navigation menu such that they are damn hard to break.
Then you put it all together; letting each element have a place to go when the browsing environment changes, look at what will stack alongside or below an element, look at whether it is better reached by vertical or horizontal scrolling if required (which will be true of just about everything that isn't in the top left corner (unless you are working in Arabic - want a challenge?).
Now you look at specifics; it should be working overall, but you need to play around a bit to get a nice set up for a PDA user, a few twiddles with the css to make the bits that look nice in Opera look nice, but slightly different in IE5 - and not forgetting a completely different set of styles in the imported and linked stylesheets.
It won't be the same in any two situations, but you can make something that looks pretty good in pretty much all of them. But you must stop trying to make it look consistent. It won't be, so don't try. Just get the "branding" across and make it work to suit the situation it finds.