fox@fury | |
Monday, Jul 17, 2000
I haven't been putting all the chapters of this saga on the site, but it should be enough to say that in April I purchased office furniture from IKEA that was to be delivered three weeks later. I finally thought that arduous journey was over today when they delivered 22 boxes to my place (I also got bookcases). Well, it turns out the tabletops for my desk were left in the van, and I still don't have my furniture. I swear, I've spent twice as much time trying to get furniture I've paid for from IKEA then I spent applying to college. Sigh... Saturday, Jul 15, 2000
I was talking to a friend of mine who works at Atari, and he casually mentioned that pinball was dead. Pretty much what I'd expect from someone working on games like WAR: final Assault and Gauntlet, so I told him about the new South Park pinball machine and that it was pretty cool and he said "No, I don't mean it's not cool. It's dead." Apparently Williams Midway has stopped designing new pinball games, and they were the last company. I haven't found the info to back this up yet, so if anyone has any firm info, please let me know. Speaking of dying out, almost all the drive-ins in the Bay Area have turned into cineplexes or flea markets. Saturday, Jul 15, 2000
It may have only been a 3.6, but it still woke me up and now I can't sleep. Still, at least it wasn't on the Hayward Fault that I live on top of. Friday, Jul 14, 2000
I've decided I'm against top-level splash pages. If you type in the URL to a site, you probably want information, not a flash intro. Unless, of course, it's pretty, or better yet, functional. Actually, I can't wait until Gabocorp is back up. Still, if a client tells you they must have a flash intro page, please try to confince them to cookie the user so they only have to see it their first time to the site. Friday, Jul 14, 2000
If Moore's law says processor speed doubles every 18 months (actually, as originally stated, it predicted that the density of transistors on a microprocessor would double every 12 months, but it's been twisted mnay times in the last 25 years to fit history), then why don't our machines feel faster? Well, Moore was a co-founder of Intel and, as we all know, the reason Intel has to keep making faster chips is to make their machines not feel so slow. Mind you, it's not Intel's fault that they feel slow, but because Microsoft has to double the complexity of its software every 18 months to even out the equations. You don't believe me? Take a look at a screen shot of Microsoft Word 2001. PS: If anyone has a screen shot of MS-Word 1 or 3, please forward it to me. I'd love to put it up for comparison. Come to think of it, I might be able to find a copy and run it on the vMac Mac Plus emulator. Friday, Jul 14, 2000
Following up on yesterday's piece about the necessity of catching user error, I found this banner ad that illustrates the point that you're not just saving the user from their mistakes, sometimes you're saving them from yours as well... And before you even bother to try it, the URL they gave doesn't go anywhere. Thursday, Jul 13, 2000
It's strange to me that we think better of ourselves if we feed starving children in third-world countries for only pennies a day, yet boycott stores which pay children in third-world countries only pennies a day. Thursday, Jul 13, 2000
I'm a big believer in designing for common mistakes. If you know a user is likely to make a common mistake, and there's little you can do to prevent it, I believe the best way to handle it is to anticipate what the user meant to do and do it. Case in point: In the beginning, nearly all corporate URLs were www.something.com. Almost nobody put their site up at something.com. Then a few companies decided to drop the www for branding reasons. One of the best examples was CNN. Around 1996 they branded themselves as CNN.com, and not only moved their website to cnn.com, but disabled www.cnn.com entirely. Of course, within a month they realized the error of their ways and mapped both domain names to the same machine. In fact, now if you go to 'cnn.com' you're automatically redirected to www.cnn.com. Companies are very concerned with losing viewers due to the viewer's mistake. For example, it's common practice for companies to buy up the .net and .org versions of their domain such as cnn.net and cnn.org. They want to make sure you get to where you want to go, even if you mess up. This even extends into the academic world (www.ucla.com). The interesting thing is, while companies will go through a great deal of expendature to obtain domain names similar to their own, (www.yahooo.com), they don't seem to care about the more common mistakes of putting too many or too few w's in front of the URL. While stats don't exist because almost nobody logs their own primary DNS server, lookups happen all the time for ww.yahoo.com or wwww.ebay.com, and are refused for reasons not immediately obvious to the user. Site architects should realize the frequency of these mistakes and account for them. It costs no more than a few lines of code in a DNS domain file to map w.cnet.com, ww.cnet.com, wwww.cnet.com or even qqq.cnet.com and eee.cnet.com to map to www.cnet.com. It doesn't harm functionality at all, especially if the web server performs a redirect to the proper connonical name, so even though they type in ww.cnet.com they arrive at www.cnet.com, and it smooths the user experience. It's surprising that designers will mandate that no page take longer than 4 seconds to load, but don't have a problem with erroneous DNS lookups that can take 45 seconds before timing out, leaving the user with a blank screen. To take it one step further, a DNS server can be configured so that any unrecognized address can be dynamically mapped to the web server (and similarly can be MXed to the primary mail server) so that mail to me@whoziwhatsis.yahoo.com goes to me@yahoo.com whether whoziwhatsis exists or not. The underlying principle here is that every action a person performs has a base case which occurs if the act is not performed in an expected or acceptable fashion. When walking, the base case is falling down, as you have to keep walking correctly to stay up. In flying, it's hitting the ground. The idea here is to make the base case as pleasant as possible, and whenever possible turning the base case away from a failure case, unlike the examples above. In Part 2: An example of how to handle failure cases on the other half of the URL or: "The user should never see a 404." Wednesday, Jul 12, 2000
Tonight I went to the monthly BayCHI meeting (Bay Area Computer-Human Interaction special interest group) at Xerox PARC. There were two speakers. The first, Niel Scott from Stanford, talked for about 90 minutes on the struggles of adapting (or augmenting) computer systems for use by the disabled, and how this problem is compounded both by the lack of support within modern OSes for overriding of input and output streams, and by the variety of forms a disability, or combination of disabilities, can take. Their solution is to essentailly add a physical hardware layer on top of the computer, which can in turn have its own input and output devices, translating the computer's video into terms the disabled user can understand (by way of actual computer vision techniques, recognizing what windows, buttons, and icons look like) and in turn taking the disabled user's responses and translating them into keyboard and mouse movements that are fed into the computer's keyboard and mouse busses. Basically this system takes the digital of a computer, converts it to a form of analog, understands its salient features, and presents them to the user in a digital fashion again. Then the whole process happens in reverse for input. The second speaker was Dan Russell from IBM's Almaden Research Center. He was a lot more etherial and visionary, talking about pervasive versus ubiquitous computing, and the future where everything with an embedded processor talks with everything else. Very inspiring, well thought out, and interesting. What nobody brought up in the subsequent Q&A, and I didn't realize until I was driving home, was that the two people were almost talking about the same thing. On one hand Niel was talking about having an arbitrary range of input devices controling an arbitrary number of different kinds of computing platforms (he gave the real-world example of controlling a Sun box via a Visor with a bluetooth card and special input software for cursor and text-entry control, as well as voice control of any number of devices) and on the other hand Dan was talking about the coexistance of the myriad of electronic devices, including the imminent problem of standardising data protocols and 'manners' for how these devices will communicate and control each other. The interesting bit is that this is exactly the failure which necessitated Niel's foray into abstracting the computer out of the interface, in order to let people control it. Essentially, in Dan's future world, all of Niel's problems would be solved. One more thing from Dan's presentation stuck in my mind. He said "The poor are always with you." in reference to the need to design not only for 800x600 24-bit color but for the 9" monochrome screen. I remember thinking that this goes both ways. It's not only the poor we must keep in mind, but the ultra-bleeding edge. The techiest of techies are using WAP phones and PDAs, and AvantGo browsers. Bleeding-edge as they may be, they require a simpler interface than any I've made since Mosaic 1.1. Technology is a curvy road that sometimes loops back on itself for another go. Wednesday, Jul 12, 2000
Once upon a time, information architects were handed a load of content and functionality, and it was their job to sort it all into a cohesive, cogent model. As would often happen, there were some items which just didn't fit into the mold, and so these extra pieces were put in a clearinghouse. Two examples would be the 'FAQ' on web sites and the menu bar on the PalmOS. Logically, if you have an informational site, people are coming to your site with questions. By that token, their questions should be answered by the site, not by an additional FAQ. Its existance comprimises the intention of the site, unless you choose to put all your information into the FAQ. This thought isn't fully fleshed out yet. I might add to it later. |
aboutme
Hi, I'm Kevin Fox. I also have a resume. electricimp
I'm co-founder in The Imp is a computer and wi-fi connection smaller and cheaper than a memory card. We're also hiring. followme
I post most frequently on Twitter as @kfury and on Google Plus. pastwork
I've led design at Mozilla Labs, designed Gmail 1.0, Google Reader 2.0, FriendFeed, and a few special projects at Facebook. ©2012 Kevin Fox |