Some Thoughts and a Discussion

Late last night I was in a discussion with a friend on the current state of computers and their usefulness to the general public. By general public, I mean to say the people we know, such as our families or friends, who are less technically proficient than we are. Computers are our hobby and livelihood; however, to them, they are confusing tools, boxes of the Unknown. The transcript which follows is from that conversation.

Shawn: Personally, I still think the PC has a seriously long way to go before it’s as intuitive as a television.
Clarkk: Indeed it does.
Shawn: Anyone who has done technical support can testify. PREACH IT!
For every “I tried to download my webpage to my cdrom, but the kids can’t play this game,” there is an equal but opposite, “It’s…I…hold on…no, see it’s…let me just do it for you, ok?”
Clarkk: Yeah. :0 VNC is wonderful for that. Esp. over VPN link to offices half-way around the planet.
Shawn: Yeah. I find myself praying that my sister and bro-law never ever get another computer ever again.
Clarkk: Do they not have one right now?
Shawn: Nope. They’ve had two.
Clarkk: Ah.
Shawn: Granted, they have three kids. But if they have to maintain a computer and keep it running, they’re better off having an Xbox. Actually, if my family gets computers, those machines need to be the most simple, most dumbed-down surfing-the-web devices. Web browsing, writing school papers, storing digital pictures, maybe tracking money. And that is all.
The I-Opener had so much potential because of its extreme simplicity. Perhaps down the road I can build a mini-itx system with an lcd screen, small keyboard, and a stripped-down OS. And hand them a set of USB keychains to carry files on. They understand things like that. Like memory cards for game consoles.
When they log in, a window pops up giving them access to their own file bucket. And they can drag files between their keychain and their bucket, or to a shared bucket. I don’t want them to see a single piece of OS filesystem.
Clarkk: Yeah.
Shawn: They don’t understand guts. They understand objects. Things.
Clarkk: Which is why people came up with the file and folder abstractions. Though, when you have enough files and folders, things still get “interesting”.
Shawn: Yeah. Even now, the file and folder idiom is obsolete.
Clarkk: Well…it has been around since the early 80’s.
Shawn: Yeah, it was created to resemble the idioms of the office to better help offices in their transition to using computers. And HFS held some promise, but Apple is letting it die. A folder for programs. A folder for the system. A folder for prefs. A folder for personal files. Clean.
Clarkk: Apple isn’t letting HFS die…HFS+ is still the native install format for OSX.
Shawn: Ah. heh. Still, it relies on folders and files. Say you have a ton of pictures. You have them clumped together in a pictures pile. Some of those pictures are from a visit to Big Bend. You also have video from your visit to Big Bend, and a pdf brochure of the Big Bend area. How do you organize this?
Most people would clump all the files into one big folder. Others, in a naive effort to be more savvy, might create a subfolder called “Big Bend” and dump everything related into it.
Clarkk: Yeah…
Shawn: The problem comes when you want to view all the pictures you have featuring your little brother. You’d have a clump of pictures to search through, but if you’re not so savvy, you’d probably accidentally skip the Big Bend pix. So a cure would be to copy the pictures, make duplicates in each relevant subfolder. When you ultimately run out of disk space, you go on a cleaning jihad, wherein it’s likely that you’ll unwittingly delete all copies of a duplicate, sending the image off to the aether forever.
Clarkk: The major problem is sheer amount of stuff…and wanting to put it multiple places.
Shawn: Yeah, exactly. I’ve observed people behaving in ways that tend towards this while they were using their laptops. Watch people long enough, you see their habits.
Clarkk: Yeah.
Shawn: Everything gets dumped into “My Documents”, and they scroll and scroll until they find their item to manipulate.
Clarkk: Flickr has some interesting ways to work on things…though you end up working with the images as files until you upload them.
Shawn: Exactly.
Clarkk: Yeah… But i’ve been used to organizing my files for a long time…and can keep enough of where stuff is in my head, that it isn’t painful to do.
Shawn: Yeah, I have my files stored in a certain prescribed, yet inconsistent, hierarchy. Anyone who is unfortunate enough to use my computer will be lost.
Clarkk: Yeah…sounds like me. I can find stuff, but other people would have to grep and so forth.
Shawn: A folder for images. A folder for music. A folder for video. Etc. But even within and between those folders there’s relation and inconsistency. So. What is a solution?
Clarkk: Tags?
Shawn: All the files go in a bucket. When you put them there, you can give them tags, quick blurbs of what the items are. If they’re pictures, you can say where it was taken, who’s in the picture, etc. If it’s a song, the ID3 tag adds detail.
Clarkk: Store all the data in a DB, and access it via other interfaces…. :)
Shawn: Then, you can sort and retrieve. It’s possible to have a user interface that handles “things” in a way that’s relational.
Clarkk: True…but we’ve not moved to interfaces that handle things relationally, and in some cases where we have, they’ve occasionally broken, so there need to be ways to fix them.
Shawn: Yeah, as far as administration of a machine goes, yes there can be an expert mode, just as a webserver has both a public-facing and an internal-facing interface.
The Web is giving us a slew of new ways to handle things. It will be a good day when that flexibility reaches the desktop. Google Desktop is a start.
Clarkk: Other than it stores things off your computer, creating privacy issues.
Shawn: That’s why I said it was a start. :)
Clarkk: :)
Shawn: So users can potentially store, retrieve, and manipulate things of various predefined “types”. If a new thingtype comes into existence, they can add that capability through the UI. More advanced users can create thingtypes as well. This is kind of the tack I’m taking with my website engine, as we may have discussed before.
Clarkk: Ahh. :)
Shawn: Internally, all the Things are in one big bucket. Each Thing is of a certain Thingtype. Each Thingtype has a certain set of ways to look at and manipulate it. You can take Things and put them into Collections. A Thing can be in any number of Collections. I assume Flickr is similar in allowing you to assign an image to any number of sets.
Clarkk: Yeah. Any number of sets, any number of tags on the image, any number of pools (shared sets, basically). Search by tag or tag set.
Shawn nods
Shawn: This is the methodology people are getting savvy to. This needs to be on their computers. Windows Search, Macintosh Sherlock (or its modern variants), Google Desktop…those are getting close. But not close enough for the Daily Joe or Sometimes Sue.
Clarkk: Yeah. Spotlight is the later variant of Sherlock on OSX. Though I think Sherlock is still included.
Shawn: Once people are in a web browser, they understand things. Once they close their browser, they are lost in their own front yard. It’s not a condemnation on them, it’s a condemnation on the technology.


What I find most condemning is that no major software or hardware manufacturer has stepped forward publicly to reddress the idioms that they’ve established their businesses upon. A lot of us understand the folder/file/desktop idiom. A lot of us comprehend the guts of our computer systems. But, even with the current state of User Interface design, most of the general public is completely lost and dazed when placed in front of a computer and asked to do a simple search for anything. It becomes increasingly complicated once you have a lot of naive users with a large install base to support (“naive” here is not intended as an insult).

Most operations a typical user needs to do are through a web browser; however, on owning and controlling their own machine, they are left with a heap of confusion. Popups convince them their system needs “cleaning software”. Unscrupulous vendors offer software at premium price to allow users to sort and manage their files, a feature natively offered by their operating system for free. Vendors of prebuilt systems offer “desktop launcher apps” that allow users to click on graphical pictures representing various functions like word processing, music playback, email, etcetera, but once clicked and the main applications launch, users are left to contend with the operating system and its filesystem underneath on their own.

The core of a public-class operating system does not need to be simple; underneath can exist a bulk of maintenance apps, firewalling, networking suites, and so on, things to make the computer work. The user interface, however, does need to be this simplified, object-relative idiom if computers are ever going to be as intuitive as any other piece of consumer electronics.

Read the Anti-Mac Interface, a paper published in 1996 by Don Gentner and Jakob Nielsen, which got me to thinking about the state of our User Interface idioms and where they could be greatly improved upon.

Published by Shawn

He's just this guy, you know?