Convergence
Some more comments on convergence of devices. I’ve quoted an old “Ask Don” on the topic and a newer article I ran across recently.
Simple, single purpose devices work better and are easier to learn. Multipurpose devices are more convenient. In the ever-continuing quest for simplicity and convenience, What comes around, goes around. Today simplicity, tomorrow convenience. Tomorrow convenience, the next day simplicity.
I admit, if you’re going to start “converging” somewhere, the cell phone seems like the best place start. It has supplanted any need I have for a land line and it’s just as acceptable sitting on the desk of my office as it is at the gym or home or anywhere else I might be. What could be more admirable for us to create a better user experience than to reduce the amount of devices (DAP, phone, PIM, Pager) in to one single device? So the makers have long since begun shoehorning all this functionality into our phones under the lofty goal of convergence. Software afterthoughts under the guise of simplifying our lives.
Eddie Lopez – Convergence catch-22
There is another direction: Bluetooth. (Simple, wireless interconection of multple separate devices.)
You could still have separate devices, but they would never have to be removed from their carrier (your pocket, briefcase, pocketbook, …). The only visible part would be a universal controller: a Bluetooth device with simple buttons and a screen that connected to and controlled all the separate devices.
I do quite like Don’s way of putting it: It is a tradeoff between convenience and simplicity. We do in fact need forces acting on both sides of the tradeoff to hopefully find our way to the right sorts of devices. Eddie Lopez asks in the end of Don’s proposed Bluetooth solution is in fact the way to go.
I think the unfortunate aspect of Bluetooth is that it still tends to revolve around having a centralized device at some place in the system. I’d like to see future versions of Bluetooth emphasize peer-to-peer more than client-server—each device can learn to get the kind of data it needs from the other device, depending on which profiles are available. Ideally, the network would also be somewhat self-forming. If I connect my phone to my laptop and my PDA to my laptop, my PDA should learn about my phone and be able to interact with it without me having to teach those two devices about one another.
Bluetooth needs to be made simpler before it can attempt to break out of the simplicity/convenience circle.
You’re right. I agree. BT will not get us there as it is now. I believe I was sure to include the word “like” as in: “a bluetooth like world” to emphasis the idea more than the implementation- instead of getting devices to merge to one, focus instead on getting them to share amongst themselves. But you highlight some important details I wish I would have fleshed out- you would need a more P2P implementation than what we have now.
“Bluetooth needs to be made simpler before it can attempt to break out of the simplicity/convenience circle.”
Yet another trade off, security vs simplicity/convenience (not that bluetooth is exactly secure)
Eddie, I wasn’t disagreeing with any of your statements, just adding my perspective on it. I’m glad you posted what you did. I tend to be on the simplicity side of things and give convenience/convergence little credit, even when it is deserving some.
And Justin, yeah, security is an issue too, though Bluetooth isn’t *that* insecure (the large “toothing” problems lie in the fact that you can make a device name be whatever you want—so you make the device name your message). Most modern Bluetooth devices don’t have as many obvious security holes as the older ones.
But as a whole, I view the security issue at a higher level: Security vs. Usability. It’s a much larger problem that’s more widespread. That’s a tradeoff that’s always going on. Within the Usability side of that dichotomy, however, we have these simplicity vs. convenience/convergence issues as well.
Convergence/disconvergence is such a simple god damned problem its disgusting. Converge all processing, distribute the I/O. We demand multiple devices because they have the right i/o rig for the job. A phone fits your ear and your pocket easily well. A PSP or PMP a small pouch, easily in reach. iPod has an audio jack and a dial. We dont value these boxes because of what difference their computational engines have, we value them because the physical I/O is just right.
We just need a ton of peripherial I/o devices. My favorite right now, I/O 0.9, is the Griffin PowerMate; the original I/O, the huge fraking knob. USB. Bluetooth is a large first step towards 1.0; already we’re seeing Bluetooth mp3 players, bluetooth speakers, bluetooth speakerphones. With OLED we’ll finally have very small reasonably durable pocket-able displays. If Zigbee werent a bunch of capitalist dickheads we’d have basic input too some day soon. Lots of ski jackets are coming with builtin iPod controls embedded in the clothing themselves. Its not hard to imagine another ten iterations down the line with a do all box and various small i/o devices you can pick up and drop. Converge and unify functional capabilities (moore’s law states nothing more than everything worth having a brain is worth making a full featured brain), diverge the i/o to give greater flexibility and choice.
The only I see which truly superceeds all else, which demands its own real form, is the digital camera. Picture taking demands a sort of I/O functionality you cannot escape from. On the other hand, something like the Casio S500 but with the screen wirelessly accessible from my pocket computer— hells fucking yeah.
On the other hand, I feel so bad for the HCI people. Whole. New. Ballgame.
Yes, Rektide, it would be rather simple if people actually agreed with you. The problem is that since the processing can converge, there are a *lot* of people out there who want the interface to converge (what you’re calling I/O). See: Cameraphones.
It’s only going to get much worse before it gets better. The mobile phone manufacturers still have not figured out that usability can be the greatest feature. They’re really focused on figuring out which features to add next. Nokia already tried a gaming cameraphone (N-Gage) and was met with great failure. You know what they learned from it? They now believe that it just means that it wasn’t the right time to introduce the full-on gaming feature—so they’re going to focus on adding music-playing capabilities to their cameraphones. After they feel successful with that they’re going to try re-adding gaming (like, gameboy-type playing, not just the simple games mobile phones have now)… just to all their phones instead of a specific line.
And then you have the fact that the majority of HCI and usability people I speak to seem to believe in convergence of interface as well. It is more convenient to be carrying fewer devices.
But I’m still of the belief that separate interfaces are necessary, as you argue.
The phone is the wrong place to move. After I was done filling up your comment-space (I hope you do not mind), I went on to fill up Mr. Lopez’s Catch-22 article with further ravings on this particular edge. The phone’s ultimate goal should be to disappear completely; dominate the verbal domain. Its only because we have to carry something at all that we think of making the phone that nexus. Think bluetooth, think of those people who dont touch their own cellphone. The phone is overburdened with requiring a dedicated number pad, a physical form suited to both in hand use and ear use. Requisit lowest common denominator; depotentiated because of what it must fulfil.
The camera is actually the ideal physical nexus. It is the only mandatorily (is that a word?) visual I/o device. I’m skeptical of putting the big CPU & processor on the camera, I want my camera half an inch thick with a massive touchscreen a couple buttons and a big honking battery, not a myraid of convergnce functionalities. But the camera should be the chief visual and physical I/o device, there’s no way you can remove that aspect of the camera. You cant possibly disconverge it, its part of the innate form, so you might as well go omega-factor on the bastard. But really now, who is going to be dumb enough to converge on a camera right? [Anyone dumb enough and with enough money to try, please contact yours truly immediately!] Read more on the catch-22 blog, i had a lot of fun with it and i hope you do to.
Thats why the cellphone is so laughable. Convergence is trying to make it the ultimate invisible un-touched verbal device, while at the same time we’re throwing mpeg4 accelerators, 802.11, primitive-audio-capabilitied java programs and vga screens on them. Sub-optimal optimization. Pardon me here, but… you cant converge on infinity.
Everything in its right place. Disconverge the I/o, converge the processing.
You’re definately right about worse before better. N-Gage example isnt fair though, there was no OpenGL ES spec at that point! But yes, in a nut shell, we’re devolving at an extremely fast pace. Even by my own metrics, I’m stepping out on a limb here, but I’d wager that evolution towards convergence is not possible. Look at the cellphone, how hard they’ve tried. But J2me is too cumbersome for hte phone hackers, and phones, ironically enough, lack the communication mechanisms for ubiquity. They lack the refined multi-threading to run background services, at least from every J2ME app I’ve ever seen (grabbing the entire I/o on the phone). All our attempts are to graft some other device onto another device. To capture obvious convergence niches (eg, TiVo). Thats why we need the box, the computing core we carry. Something general purpose and all purpose. Convergence is tyrannied by applications, by still being tied to the box it originally came in. Put hte damned thing in a sleek black box with two blinkenlights no buttons and call it a day. Revolution. Force the change; even if we’re going backwards in the process. Then we can start evolving.
Its going to be the hackable hardware which pushes the entire world towards something better, which shows the possibility. I’m dead drop dazzled at the gp2x; a handheld meant for hacking. SDL powered! Brilliant! They botched a number of things; the USB host is still up in the air, there’s no linux SDIO support (not their fault), but it is, in all essences, the first coming. It has no intent. Its a generic media box, something built for possibility. If you build it, they will hack.
Once we get that hackability, the plasticity to build and experiment, it seems to me only obvious that seperate I/o devices are the way to go. But they will be naturally freed from the original constraints upon which they were birthed. I/o for the sake of I/o. No longer purposed as “a cellphone screen”, “a walkman remote”.
Btw, seems you might have some knowledge on this, can bluetooth dongles communicate with more than one bluetooth device? My plan has been to route all comm through my laptop but I’m no longer so sure if its that easy to patch my cellphone through my computer. I may have to buy two dongles. :-/
Yes, the entire point of Bluetooth is multi-device communication. I have several Bluetooth devices connected to my laptop. Most cell phones don’t support very many Bluetooth profiles though, so depending on what you mean by “patch my cellphone through my computer” you may not be able to do it.
The problem with Bluetooth (as I got into above) is that it’s more centered around having hubs. Right now my laptop is the hub for my mobile phone, Palm, and mouse. If I want to browse the web on my Palm through my mobile phone’s connection, I must set up a separate connection between my Palm and mobile phone, making the phone the hub. My Palm should already know about my mobile phone since both are connected to my laptop—it should simply use the mobile phone for GPRS access when needed.
I’ve read each of your posts Rektide, and I’m still not exactly sure what’s going on. I agree with you, because I think we are ultimately saying the same thing just in different ways.
I think I argue that the cell phone is not the right interface as you say. I also imply that it’s a good place to start since you always have it with you. So if we are talking about convergence, we are both talking about a small device with an adaptable UI. How is that different than what you are saying? The camera seems just as laughable to me to start at converging. I find any camera without a decent optical zoom (greater or equal to 3X optical) to be useless… I haven’t shopped in a year or so, but every camera with optical zoom does not seem like something I would want to carry around in my pocket (vs… say the RAZR). Also, if you’re going to put a touch screen on a camera, how is that any different than on the phone? I think both are laughable as they are for convergence, but either has just as much potential as we march on developing: voice driven interfaces, or touch screens, or “glove” like typing or any other type of I/O, UI.
So even if we do use a camera and converge telephony onto it so that there is no “phone” interface. What are you likely to use more during the day, the telephony or the camera? Fill in the blank here, when you leave the house in a hurry, what do you tell your wife: “I have to run back in, I forgot to grab my _____.” Can you go a day without a camera? How about a phone?
You keep saying the cell phone is the wrong place to move. But I think you are focusing too much on the capabilities of these devices. The softwre, hardware and user interaction. I agree, that is the wrong software, hardware etc. But it’s the moving part that we need to focus on. The phone is the best place to start because it is most needed function. If you want a camera type device that has no phone interface but you can place calls on it, I would imagine people will call it phone in the long run. If you want it to be invisible, fine I agree there too. It’s voice medium, there’s work to be done, but it’s possible. But when you voice prompt that box in your jacket pocket to call the office, I would bet those around you will say: “hey, that’s a neat phone”