I found myself musing on the future of the display screen recently, which affects Web and application design, among other things.
Back in the 1970’s, RAM was too slow and expensive to use for a framebuffer, so computer graphics used long-persistence phosphor, then storage tubes (Tektronix 4014, still emulated today). The arrival of the IBM PC with CGA at 320×200 pixels put graphics on most desktops, and resolution has been inching up ever since. My workplace now has a projector wall with 4200×1050 pixels, and a control room with 5120×2048 using eight tiled monitors. Meanwhile, displays on pocket devices have grown too; my phone has 128×128, the iPhone has 320×480, and my Nokia N810 has 800×480. Where will it all end?
I suggest that the pocket display has gone about as far as it can. The Nokia tablets, like the generic men’s wallet, fit in the proverbial vest pocket. If they were any bigger, they would not. The acuity of the human eye (around one arc minute) puts a limit on how small pixels can usefully get, while its focal length puts a limit on how close a user can hold something and still read it. I’m getting on a bit (storage tubes, remember) and need three diopter reading glasses for the N810, but even a 20-year old might need glasses were the pixels even smaller. Pushing the screen to the very edge of the device would give about 1100×570, which I see as the useful limit for a pocket computer.
Desktops, on the other hand, I can see growing to about 6000×2000 by getting physically larger. This represents two distinct target groups for general purpose applications and Web sites (if we forget about wristwatches and legacy cell phones). It will be interesting to see how content evolves to cope with this divergence.
The newer browsers such as Firefox 3 are able to preserve Web layout by scaling graphics and text together, while touch screens and newer mice ease 2-D panning. However, this can still be frustrating for the user, like peering through a mailslot. The use of alternate style sheets on the Web was supposed to address these issues, but if the basic problem is that there is just too much content on a page for a particular display, reformatting it in the browser will not help. There is also theissue, particularly in Canada, of high data charges on cell networks, which make it extremely expensive to view mainstream content on a device such as an iPhone.
One solution may be to generate Web content from a database, applying different layout styles depending on the target device and market. A story might appear on one page on a wall display, three pages on a pocket PC, and the highlights on four pages of cellphone display, each with banners and advertisements sized appropriately. This is already happening to some extent with region-specific advertisements and the .mobi domain Web sites.
