Forum Moderators: open
By far the most attractive and professional looking were the fixed width pages. Some of the expandable ones weren't bad, but design compromises are inevitable when you give the browser a lot of latititude. (The ideal from a design standpoint would be one big image map or sliced image table - leave nothing to chance! Not real effective for SEO, though.)
I'm inclined to go with the 800 pixel width that seems to be most popular with major sites. Any feedback?
Many of our sites run around 10% of visitors at 640x480. We want to make sure that we don't make life too difficult for 1 in 10!
This oftens means significant white space, wide margins, etc. All that actually helps a page to make an effective visual communication.
-H
is that through any objective criteria or simply that fixed width pages relate most closely to traditional paper design standards?
As we can notice you get different opinions. What about the audience of your site? What are the products sold? If it is a general audience, can you live with a site too wide for 8 to 10% of users?
If the audience is more tech savy, or computer graphics oriented then you can go wider.
Hmmm, I'm not sure how to answer, Eric. My comments were based on the balance between graphics and text, and the way the viewer's attention was drawn to the focal points of the page. Some pages look like a real graphic designer created them, and others look more like a code-jockey made them. I'm probably heavily influenced by print design criteria, but then so are the majority of visitors who have been using print media for their entire lives.
I think that by and large good design is good design - elements have natural relationships to each other that create a pleasing balance. If that balance changes too much due to browser settings, it's no longer pleasing. Paper catalogs and magazines are very high design content media - people expect a high degree of sophistication, and IMO this translates into expectations for the web sites that correspond to these catalogs.
It's kind of like telling an interior designer, "by the way, the sofa might double in length, but the picture grouping over it won't, and the lamps might shrink, but make sure it looks good no matter what."
I have a high end magazine publisher for a client who insisted that all the copy needed to be knocked out of the images and he further insisted that the images should be "full bleed".
I told him he was just wasting money with his design concepts and that the only person who would wait for his site to download would be his mother, and only if she really loved him a lot and had a T1 connection.
Now he wonders why he is not getting any hits. Maybe one day I will tell him again.
Agreed, rogerd. And I would add: ... that create a pleasing balance in the service of an effective communication."
For many years recently, the print media has developed a kind of "high concept" design that wins awards from other designers and artists, but does not communicate well. Companies have poured large amounts of money into ad campagins that essentially did them no good at all, except the designer may have won a Clio!
When this "design-above-all" mindset hit the web, the companies involved had horrid results. I feel that functionality and usability must come first, and design must serve these ends, not obscure them. No matter how beautiful a page as a work of art, if the business purpose isn't served, then the design is dysfunctional.
A second point: any print designer worth their salary creates an advertising piece in concert with the press that they use - they understand something about the technology of print: the inks, the challenges of registration, etc, etc. Designers who work on the web must gain a working knowledge of the web's technology, or else they will waste their company's resources.
The current technology of the web makes "liquid design" very challenging. It's understandable that fixed page design is so common - it's easier to code and often looks better on most monitors and browsers.
Can we evolve a kind of web design metaphor that works in the liquid fashion? Those who want this must push for it, and show its possibilities. Right now, I'm very unhappy with the limits and the complications of liquid design, and I'm hoolding off until browser technologies and CSS support mature. I'd rather put my clients' resources into refining the message, the targeting, the usability and the functionality of their sites.
Personally, I prefer to create variable width pages. With a few tricks and good knowledge of how different browsers render different html(so you dont get frustrated), this can look really good. You can usually assume that your user is surfing at some minimal screen res, such as 640x480, and design your squishy tables and image widths to work with nothing less. I would suspect that increasingly more people are surfing without giving the individual browser 100% of the screen to work with. This means that any fixed width setup would be upset somewhat. This is especially true for those who run at res's higher than 800x600, and have large screens.
I did a design for a guy who wanted the page to display "nicely" from an 800 to 1280 screen. Text was not a problem and the page wasn't too graphic heavy, but the logo graphic proved to be difficult. It was in the top data cell flush left. It would have looked odd to let it float center. I wrote a screen detection script that simply switched the graphic, but warned him that it wouldn't work without Java being enabled. The solution was to fill the cell with a close color in that event. I did the same trick with the text using three style sheets. Don't know if this helps. Hey, it's like a birthday present, it's the thought that counts :)
There is something nice about being able to resize my browser and watch a page resize with it (try this with NS4 at your own risk !!!). With a little bit of planning and tweaking image aligment with tables IS a viable option. It always amazes me that more big corporate sites don't use percentages but hey what do I know hehehe.
One thing I do almost every time so as not to make a page with only two or three lines of wrapped text at 1600 x 1200 is put all content inside a one cell , centered table with no border, padding, spacing etc. Then I set the width of that table to 80% or something.. whichever looks best on all common resolutions.
I is all due to rounding. :)
?
I would really try to stay away from fixed widths. I really would. The small screen stuff, just doesn't look good at higher resolutions and when you go 800, you leave the 5-10% behind. Who is to say who that 640x480 user is? It could be a ceo on a laptop looking to make it your lucky day.
I find many sites difficult to view in anything wider than 1024 even if it's not fixed with. This is especially true when there's a lot of text even if the font is adjusted. It's like reading one very loooooooong line.
There's a page on my site where I give some tips on CSS. On the page there are examples how images appear in screen widths from 640 to 1920 wide. Anyone interested can drop me a StickyMail and I'll send you the URL.
(edited by: tedster at 5:49 pm (gmt) on Sep. 17, 2001
I rarely use some electronic assistant to add up stuff like that. The point is how different browsers deal with this. There is no magic number. Just some better choice.
I am not in the market to make decisions, just to recommend some client who sell high end video cards and add up 20 pages a day to choose some screen resolution and some other clients who want to reach every one with is 2k e-commerce site selling blue, red or green widgets to get another one.
If 100 means 102, I will repay him his 2, and get 100000% fidelity in return. Funny how fuzzy can numbers be.
If it is necessary to go to a fixed-width design for artistic reasons (like the major news sites and some major catalog outlets), a design width of something less than 800 (say, 770 or so) is probably the best choice. This will be fine for the many users at 600 x 800, and will produce a display of acceptable (if less than ideal) size for those with higher resolution. 5 - 8% of the visitors will be limited to 640, and will end up having to scroll horizontally.
In the catalog biz, we used to do split run tests - say, mail a catalog with two different cover designs to see which generated more orders. I'd be interested to see the test results of a high-design content, fixed width site vs. a less-designed, flexible width site. Maybe it's worth losing some of the 640 users if orders from the rest of the visitors are 10% higher. Or, maybe the 640 users would be so grateful for a site they could browse easily that their additional orders would offset any design-related gains. Who knows? That's why you run tests... Thanks for the comments, everyone!
most of the major news sites couldn't handle the damand last week...CNN dropped down to a plain image plus a few text links format within a couple of hours...don't expect them to stay with the heavy graphics based designs for all that much longer...they've been slimming down year by year
a great looking news site is no use if nobody can reach it
there are big developments happening with mobile web access...that will throw a new set of common screen sizes into the mix...I'd say this is a very risky time to be fixed width unless it is absolutely essential...it rarely is
I'm not sure I agree that the news sites (and others) are slimming down. I'd say they are holding their own or even beefing up as fewer users have slow dial-up connections and more connect via various broadband solutions. Certainly, they are making a lot more rich media available, and even the HTML pages don't look much slimmer to me.
