‘Pull’ Devices

A few days ago, it occured to me that it would be useful if my alarm clock knew to awake me an hour before my first appointment of the day, as automatically pulled down from my Yahoo! Calendar. All the alarm clock would really need to do is to have a net drop (wifi / ethernet) and every half hour or so attempt to establish a connection to a central server and ask if there were any updates to my schedule. The alarm clock itself wouldn’t need any additional interface itself! When the user first buys the alarm clock, they go to the central website, punch in the serial number on the bottom of the alarm clock and configure their clock on the remote website, such as configuring synchronization with Yahoo! (or other calendaring services). When the alarm clock checks into the server, it self-identifies with its serial number and pulls down your information (as well as the current time, accurate to microseconds, properly adjusted for daylight savings).

As Ethernet/WiFi chips become cheaper and more common, we should expect to see this level of integration in a great number of devices. This differs slightly from previous visions of the “networked home” in which smart devices, such as refrigerators, coffeepots, etc, would have either their own user interfaces or where they would be running their own little HTTP servers responsible for the logic of the device’s operation (similar to how many NATs work these days), since the user interface would not only not be physically on the device (with knobs and buttons and displays), but it wouldn’t even logically be on the device (with a website hosted on the device).

The advantage of this approach is that it reduces to an absolute minimum the intelligence that needs to be on the end device; the server interface could be accessed many different ways in theory (web, email, phone, IM, etc) without any added need for complexity in the end device. As long as there is a simple protocol that exposes to the server the capabilities of the device, the problem is solved from the device’s standpoint and has been future-proofed without needing any dangerous “flash upgrades”.

This methodology works particularly well for simple pieces of equipment that benefit from being synchronized with a digital lifestyle. I’m pretty sure a toaster wouldn’t qualify, but alarm clocks and coffeemakers could plausibly fall under this category. Another variant of this approach could involve first using Rendezvous-like technology to have the alarm clock seek out a local “calendar server source”; failing that, the alarm clock would fall back to attempting to contact a central calendaring server. That way, local resources could be used as appropriate (if the clock is at home), but there is useful operation away from a local calendar server (if the clock is on the road).

On Podcasting

A number of friends have asked me of late what my take is on podcasting. It’s certainly a hot new buzzword; but is it really the “next step for blogging”? It doesn’t seem obvious to me.

Now, from the one perspective, it’s just blogging audio, right? It’s just like blogging, but for your ears, so you can take it with you in the car or on a jog, and it’ll make the blog experience even more immersive. Microphones are cheap and sound editing software isn’t that hard to come by, go the arguments, so with a few more tools and maybe a telephone gateway or two, you should be able to blog your voice!

I’m not sure this is going to fly.

To see why I have my doubts, I’d like to explain why I think blogs work. Call me an optimist, but I think many people have the potential to become passably good writers with but little training – perhaps even a half-day with Strunk & White’s classic “Elements of Style” could suffice for a good many. Text on a screen can be easily scanned for nuggets of interesting insight, indexed and searched, cut-and-pasted for easy quoting, and trivially edited and revised to improve its quality. Comments are left in the same medium as the
content, and an article can be hyperlinked to allow blogs to perform “metareporting” and to cite and draw users to interesting articles and/or discussions going on on other blogs. You can stop reading a piece of text when you get interrupted by something else and can
quickly scan to recover your place, regain context, and continue reading.

Podcasts, aka audioblogs, have none of these advantages. To produce a high-quality audio recording, a quiet sound room is needed with a good quality microphone and, more importantly, a great voice. It is my humble opinion that people can be much more easily trained to be serviceable writers than they can become even approachably pleasant to listen to in monologue. It’s considerably more difficult to edit audio to reword or restructure things; more often than not, a segment needs to be wholly rerecorded in a second, third, or fourth take. The audio itself cannot contain hyperlinks that can let a user dig into a story,
nor can the audio be quickly skimmed – it must be listened to at the rate it was recorded. Comments, if offered, are in a different medium from the podcast itself and are therefore much less likely to be active or seen by those consuming the podcast. The content of the
podcast cannot be searched or indexed in a meaningful fashion (at least until Google supports voice recognition to auto-transcribe podcasts). If you get interrupted while listening to a podcast (likely, especially since hearing words tends to be a much slower
process than reading them, saving the case of a dyslexic listening to an auctioneer), it’s much more difficult to recover your place and recall the context in which your listening was interrupted.

So, with the exception of some professionally-styled productions like Engadget‘s podcast or KenRadio.com, most of what’s out there ends up being poorly-delivered longwinded rants that have very little connection to each other and can’t be skimmed. Think the very worst of AM radio and then some. Not exactly the sort of thing you’d put on over music when going for a jog or driving down the road.

I don’t see it as likely then, even with much-enhanced tools, that “podcasting for the masses” will ever have the same weight, influence, or quality as blogs, given that podcasts lack nearly all of the qualities that made blogs endearing and popular. This is not to say
that we won’t see a handful of professional, targeted podcasts winning out the “morning talk radio” demographic, but I’m baffled at the notion of thousands of listeners tuning in to hear Grandma go on for fifteen minutes about how her hairdrier got stuck on HIGH in the morning. It could be that I’m just being too optimistic about people wanting to spend their time well and that podcasting will actually become tremendously popular, but I wouldn’t bet on it.