Reviews and prices of digital cameras at Digital Camera HQ

Browser Woes

March 15, 1996
In my designer role, I want to present Sight in a visually stimulating, yet simple and easy to navigate way. One of the frustrations in designing for the Web is the lack of control you have over what the final product will look like on the viewer's computer. You can specify only the type sizes relative to each other; the actual font is controlled by the viewer's browser; and combining text and photographs in a pleasing way is very difficult. To exacerbate these challenges is the fact that there are many different standards for what particular Web browsers will support. It's as if ten people each speaking only their native language and part of another one were all dumped in a lifeboat together. Eventually, everyone would get the essence of a message, but only the speaker would know fully what he or she was trying to say.

There is an HTML standards committee that is working to come up with a universal set of tags that all browsers should support. HTML 3.0 is the current standard, and it contains many powerful features. However, because the two leading makers of browsers, Netscape and Microsoft, have seemingly chosen to support only the tags they want, as well as inventing a number of ones unique to their browsers, the idea of a true standard seems unlikely in the short run.

Despite current appearances, there is always the possibility that Netscape would fail. Because most Web designers are done with Netscape in mind, this would obviously be a problem. It is a dilemma: Netscape currently has a grip on the browser market - I hear figures of between 70 - 80 percent of the market is theirs - and so most sites are designed to take advantage of Netscape tags. Certainly, there are times when tables seem to be the only way to produce a desired layout - yet more rudimentary browsers might not support tables. Other, perhaps more insidious problems can arise from something as simple as the center tag. While most HTML authors would probably assume that the standard is the word "center" surrounded by two brackets, a friend who's looked at the HTML 3.0 standard in more detail says this is actually only the way Netscape has chosen to write this tag.

I have run into a few problems because of browser incompatibility in the design of Sight. In a couple of instance, once with a Web map and another time with a link, I actually created a problem because Netscape was forgiving of an error I'd made in the HTML coding. In other words, I'd put up a page and look at it in Netscape. I didn't think anything of it because the page would work in Netscape. For a couple of months, I was frustrated because people using the America Online browser kept telling me they couldn't access the categories in Sight, because they would get an error message when they clicked on the map I was using. Eventually, I redesigned it so that I wasn't using a map, though my server administrator informed me that there had been an error in the coding that Netscape was able to work through, but AOL's browser could not.

Although I know Sight looks terrible on the AOL browser, I have decided to wait them out, instead of bringing the design of the whole site down to accommodate them. Recent news is that AOL will start being bundled with Microsoft's Internet Explorer, which will be a great boon to AOL users. In the meantime, the AOL browser doesn't reproduce JPEG images reliably, and doesn't support the tables -- it stacks all the elements of a table into one long pile -- and is extremely slow on top of everything else. I know that with the Internet Explorer most of the problems will go away -- with the exception, most likely, of the bandwidth situation, which will get worse before it gets better. This is the most scary part of it to me. I get a fair number of AOL users who may be frustrated because Sight is already so demanding of a good connection.

Ironically, I have received a number of complements on Sight's design from AOL users, which I find amazing. I suppose viewed through the filter of that browser, everything is brought down equally.

The next major feature decision I must make is with frames. I have heard arguments both ways on frames, and for the moment, I am leaning in the direction of not using them. In this case it is not simply the question of whether a particular browser can support frames -- in fact, it is possible to write alternative code for people whose browsers can't support them -- but it is also a matter of how well they work on a standard size monitor. While I like the idea of laying out an index of the site in a left-hand frame -- I think this makes good navigational sense -- I don't like the idea of squeezing the rest of the content into whatever room is left over. What with the menu bar and all the icons, browsers already take up a lot of the vertical space of the screen, so it would seem better to leave some horizontal elbow room.

Building a Web site is rapidly becoming a more complex, technical matter. Serious sites have CGI programmers and will soon need Java programmers to offer the kind of interactivity viewers will expect. While I want to have the best site possible, part of the appeal of it is that I can do most of it by myself-- it is a personal magazine. It doesn't make any money, but it doesn't cost much. If the costs maintaining it were higher, it would change the formula.

I intend to fumble along, picking up the new technologies the best I can and making Sight as appealing as I can to as many people as possible. If I start taking this too seriously, it won't be any fun anymore.

[Sight | Journal]