Recent Updates Page 2 Toggle Comment Threads

  • Urban 19:16 on 1 Oct. 2014 Permalink  

    If you want something new, you have to stop doing something old

    Peter Drucker
  • Urban 08:00 on 18 Jan. 2013 Permalink  

    Smart Watches, Dolphins and Evolution 

    Watches seem to be a bad fit in the modern world where time is all around us. Glance at your phone or your tablet, and there’s a clock. Glance at your computer screen, TV, car dashboard or digital camera, and there’s a clock. Fish your Fitbit out of your pocket, and it has a clock. Today, keeping time is so cheap that even your oven has its own clock.

    It wasn’t always so; 500 years ago, the clocks might have only been on clock towers. 250 years ago, they came to our living rooms. Then, to your grandfather’s pocket. And then, in an endless feat of miniaturization, they came to your wrist.

    However, in today’s world of Apple iPhones and Google Glass, the uni-tasking device on your wrist has been reduced to a fashion statement and/or a status symbol. But you don’t really need it.

    At least, I don’t.

    What would come in handy, though, is something else — an external screen for my phone. Something to display text messages, incoming calls, e-mails, weather info, and all the things the phone can think of, so that I don’t have to reach for it every time it blips.

    That’s why I’ve been testing a smart watch (Metawatch Strata, more below). It pairs to your phone using Bluetooth Low Energy and does all that, with hopefully more to come.


    A quick review

    Here’s some of the features:

    • It shows texts; see below for notification from one of my servers reporting some stats.1
    • Shows events from calendars synced to your iPhone; this includes Gmail, Exchange, and surprisingly, FB birthdays
    • Displays weather forecast, which is of course location-sensitive; the phone already knows your location, so weather can always be local
    • Shows stocks and phone battery level
    • Displays incoming calls, which you can also reject by pushing a button
    • Has media player controls,
    • Features a vibrating motor to alert you to a call, text or a calendar event,
    • as well as a 3 axis accelerometer and an ambient light sensor, both unused in the current firmware.



    All this works by pairing the watch to your iPhone (Android is supported, but currently lacks many features). In fact, you have to pair it as a Bluetooth 2.1 device for displaying texts and incoming calls (as you would your Bluetooth car kit), and as a Bluetooth 4.0 device for everything else.

    Once paired, you run the app (MetaWatch Manager) that manages the widgets on the watch. You get 4 screens and you can place on them anything from 4 small widgets to 1 large widget occupying the entire 2×2 grid.


    The software is supposed to be open source, and there’s plenty of projects doing more with it. Haven’t looked into it yet.

    One of the more peculiar things is its display, which is nothing I’ve ever seen before. It’s a Sharp low-power memory LCD (video) with 96×96 pixel resolution, but there’s no black or white pixels; the only two states are mirror-reflective and white. This takes some getting used to, so be sure to check if you like it before buying. All photos are cleverly set up to reflect only black, so the effect is not obvious unless you check it in a video. On the bright side, it requires almost no energy when it’s not updating (similarly as e-ink), so the battery lasts for about a week.


    So, is this wearable computing?

    MetaWatch is of course not alone; there’s Pebble, which is just about to ship, with similar features. Pebble uses an e-ink display, so the battery life is reportedly similar (around a week). There’s also the Italian i’m Watch, which is entirely Android powered and uses an active display (hence the battery only lasts for about a day).

    It’s quite obvious something is going on in that space, and that a new computing form-factor is emerging.

    And if you think back, the exact same thing has already happened before: computers started out as large mainframes, where all the computing power was centralized. They were accessed only by a dumb terminal, which was little more than a remote keyboard and screen. Yet, gradually, technological progress killed the mainframe and made the dumb terminal the new protagonist. It became smart enough to survive by itself, sitting under your desk, then on your lap, and finally in your pocket. The smart phone in your pocket is indeed more powerful than your desktop computer was 10 years ago.

    And yet again, there are new dumb terminals to take place of the old. Smart watch is nothing but a dumb terminal for the mainframe in your pocket. Right now it can’t do much more than display what the mainframe has to say. But it might not be long before you won’t need the mainframe anymore.


    The meandering path of evolution

    I’ve been thinking lately how all this relates to biology. Watches seem to have had a similar evolutionary path as dolphins.

    You see, first there were fish (bear with me). Gradually, some fish got tired of water, became mammals, grew lungs and legs and came to land. They evolved further, and eventually they became us. But something else also happened: a group of mammals got tired of land and went back to water. Dolphins are a part of that group, and so are whales. They might swim like fish, but they have lungs and have to come to the surface for air. And funnily, today they mostly hang out with other fish (which have always been fish) and get confused with them by almost everybody.

    Compare that to watches. A clock, as an ideal of craftsmanship, the ultimate precision mechanism, gradually evolves into a mechanical Babbage machine that fills an entire room. Later, that becomes a computer, and the computer gets smaller and smaller, until it becomes wearable, and finally migrates to your wrist. There, it hangs out with the old mechanical relics, only to be confused with them for years to come.



    1. I’m using Nexmo to deliver those, costing me 1 cent per message []
  • Urban 02:35 on 1 Dec. 2012 Permalink  

    625 days of walking 

    I’ve had my Fitbit for over 625 days. It records walking activity with 5 minute resolution and I’ve carried it around almost everywhere.

    So here’s my past 1.7 years of walking in one picture (click to enlarge).

    Every column represents five minutes (288 columns from 00:00 to 24:00) and every line represents a single day (625 lines from March 15 2011 to November 28 2012).

    Green means no activity and red means high activity (highest intensity red is 706 steps in a 5 minute interval).

    What’s immediately visible is when I wind down and go to bed. Then there’s some displacements where I was obviously traveling in a different time zone.

    There’s also a trend that I’m not fond of: I seem to be rising earlier and earlier, since the active (yellow/red) band is moving slightly to the left as time progresses.

    A quick note on how this was compiled: the data was extracted from the Fitbit website, using the undocumented API that fuels the flash charts (the cookie that serves as auth was copied from the browser’s inspector into the script). I looped through the entire date range and saved every day into a separate XML file. Another quick and dirty script combed through the files with a regex and concatenated all 5-minute intervals in a day into a single comma separated line. The resulting CSV file was then imported into Excel where simple normalization was done, as well as some color coding using conditional formatting. Yes, this is extremely lame, but it was all finished before I could even decide what other tool to use.



  • Urban 11:00 on 17 Nov. 2012 Permalink  

    Least popular 192.168 networks 

    Having a home VPN server on the default subnet is a pain due to address collisions. Indeed, if I haven’t bothered to change it before, why should I expect a random cyber cafe/hotel/company or anyone at all to use a different default subnet?

    So I’ve decided to renumber my home network. But first, I wanted to find a 192.168.x.0/24 subnet with the lowest popularity, so I could minimize the potential collisions.

    I asked Google1, querying all the default gateways2 of the form “192.168.x.1”, and got the following result (x axis is the number of Google results, in log scale). Click to enlarge.

    Keep in mind that the Google AJAX search API used here returns somewhat strange numbers, which are by a factor of 10 lower than in plain old desktop Google search, but at least that factor appears to be consistent (I checked the first 10 IPs).

    In fact, the data here is not intuitively comprehensible. The pie chart below is better, showing that anything except the 7 subnets in the legend should give you a sufficiently low probability of collision (lower than 1%). Click to enlarge.

    Note: all data as of Nov 14 2012. Here’s the raw dataset.

    But renumbering the network is a lot of work, really, so maybe I should just pick something from the 10 or 172 subnets. These seem to be much rarer (especially 172). Of course I’d have to retrain my muscle memory.

    There’s other options as well, but I don’t find them very practical. For example, using a 1:1 NAT to provide additional address mapping to VPN users just complicates the network, not to mention firewall rules. And using static client side routes is another non-option, since it can’t be used on locked down devices, such as  iPhones and iPads. So renumbering it is.



    1. I used Google AJAX Search API, stuck the URL into a Ruby script and iterated from 0 to 254 []
    2. I made two hopefully reasonable assumptions, namely that the network popularity is proportional to the amount of people talking about its default gateway, and that the default gateway is has the .1 address []
  • Urban 09:00 on 14 Nov. 2012 Permalink  

    Smart bulbs (and other musings) 

    As a gadget enthusiast I instinctively clicked “Back this” when I saw the Lifx project on Kickstarter. I was torn, however, when I saw the public outcry regarding the founder and his alleged incapability to ship a cardboard box. I hesitated until the last day, not sure whether to keep the pledge or cancel it.

    Meanwhile, I happened to stumble upon another similar project at the Mini Maker faire at Strataconf NY–the Visualight. It instantly caught my eye and one of the founders explained to me how he was just finishing the writeup when the Lifx project came online.

    So I said, “convince me that you’re better and I’ll cancel the Lifx pledge and back yours instead.”

    He did give me a pitch with plenty of differentiation, saying that Visualight is a great data visualization tool which sports open APIs for all the communication. Who needs disco effects and music visualization, when you can have the light change color according to weather, stocks or subway service (kind of like the Nabaztag / Karotz).  He also showed me a working prototype.

    But right there, I couldn’t decide which one had a better premise. It all boiled down to the “smart bulb, stupid network” (Lifx) vs. “smart network1, stupid bulb” (Visualight) dilemma, and I got an instant case of analysis paralysis.

    I started thinking that I’ve seen the story many times before.

    For example, in computers.

    In computing we started off with a centralized design (mainframes) and dumb terminals. Then the brain moved to the local box (PC), and now, finally, it’s moving back to the network (cloud), with the clients getting more and more stupid once again (just take a look at Chromebook).

    Something similar seems to be happening in mobile phones, with the brain first moving from the network to your iPhone, and now slowly creeping back into the datacenter (Siri, anyone? Or maps with server-computed turn-by-turn?)

    But right now the infrastructure is not quite there yet. It’s not infallible and 99.999% robust, and it pisses us off when Siri can’t take a simple note. Imagine you can’t turn on your light at 2AM because your server’s down.

    So that’s what I was thinking while standing there, staring blankly into empty space. I decided that (at least my) world might not be ready for a remote controlled stupid bulb.. yet.

    And a couple of days later, Philips announced the Hue. It’s severely limited (iOS only, and the bulbs are not self-sufficient; it uses an additional ethernet-connected gateway which communicates with bulbs via ZigBee). However, with its market cap, lighting expertise, reputation and virtually the same price point, Philips might have just eaten the lunch of every other lighting startup.

    Then there’s another issue where Philips wins: safety. A product like that, done wrong, can easily burn down your house. I’ve already seen the remains of an exploding Chinese USB charger, and this is indeed a great concern. Compared with cheap Chinese LED bulbs that I bought en masse years ago, such a smart bulb has to be always on to benefit from its embedded computer. If you switch it off, it’s dead.

    So we’ll have to wait and see who’s going to be the winner here. The race is long. In fact, it’s never-ending.

    And no, I haven’t cancelled the pledge.



    1. and here, by “smart network” I mean “a server” []
Compose new post
Next post/Next comment
Previous post/Previous comment
Show/Hide comments
Go to top
Go to login
Show/Hide help
shift + esc