Update on using the TypeMatrix keyboard

It’s been two months now since the last post, and most of what I said before is still relevent. What has changed is:

  • I’ve got used to using it with the silicon cover. Think I even prefer the feel using the cover now, since it slightly cushions the fingers. That said, I can’t say I really prefer the TypeMatrix keys over my ThinkPad’s keys (nor the other way around).
  • Switching between the layouts of the ThinkPad and TypeMatrix keyboards is less problematic, though I still occaisionally mis-hit keys.

More significantly, I’ve tried using the TypeMatrix while gaming. Here it doesn’t do so well, for two reasons: the layout, and the Alt+Tab button. In games you really need to be able to hit the right keys quickly and without taking your eyes off the screen, so not being able to feel a gap between the number row and the F-keys, or between the bottom letter row and the play/menu/app-switch buttons doesn’t help. And that app-switch button: if you’re playing a competitive game on Windows, the last thing you want to do is accidentally hit Alt+Tab. And that button is right between Alt and C… personally my feeling is that we could do without the Alt+Tab button completely.

Posted in Uncategorized | Leave a comment


I’ve recently bought a TypeMatrix keyboard. Here’s my thoughts.
QWERTY model

There’s two things to comment on really: the physical construction and feel, and the new layout. I’ll start with the former.

Key travel is good, a little harder to press than my Thinkpad keyboard but with better feel. It’s short with relatively smooth pressure feedback; I’ve had to get used to pressing the keys slightly harder, but otherwise I like it. I use the silicon skin; without this, keys require a lighter press and feel like they bottom out more suddenly — a more sudden shock to the finger than with the Thinkpad keyboard, but more precise too. I couldn’t say which of these is best for the hands — to be honest I find all three variants perfectly good enough, with the minor down-side to the TypeMatrix with skin being that I tend to press the keys slightly harder than absolutely necessary to make sure letters get pressed in the correct order.

The physical construction of the device — it feels high quality, if a bit heavy. Size is small, although it’s not especially thin. Some people reportedly carry the keyboard with them but I certainly prefer leaving it on my desk due to weight and the captive cable. It has no extra ports, switches, etc., just the keys on the front and one protruding cable.

Angling the two halves out in a fan might be slightly more ergonomic, but I don’t think it would actually be significant — ergonomically I find the keyboard very good (except that the right-shift key requires a long reach with the pinky).

Moving on to the layout, the most striking thing is that keys are vertically aligned in a grid. I’ve only been using the keyboard 1-2 weeks and am still getting used to the layout so I can only talk about initial impressions, which are that this arrangement of keys feels much more natural and works well with a little adjustment (I am typing this article on the keyboard, and at my normal typing speed). Dropping old habits (or switching back to a “normal” keyboard) takes a little effort though; in particular you have to reach for the keys on the lower row (ZXC…) quite differently, and reaching to where you expect the B key to be tends to result in pressing both B and Enter simultaneously.

Another thing that takes some getting used to is the large grid of keys on the right hand side with little about them you can feel to tell where your fingers are. There’s the usual tactile lump on the index finger keys (F and J on QWERTY) and another on the down arrow key, but getting used to where the arrow keys are and in particular the right control key in relation to the arrow keys takes a bit of getting used to. (Note: I use the silicon cover. Without this, keys feel crisper and the tactile lumps are easier to feel, which does improve the situation.)

Other changes are easier. The tall shift keys and central enter and backspace keys do feel strange at first but aren’t actually hard to get used to. I’ve actually hit the shift keys several times intending to hit either Enter or Capslock (which I use as backspace); this is not so bad since it results in nothing happening instead of what you expected (on the other hand, I’ve hit enter unintentionally a few times, submitting forms with incomplete input).

Other things: the home/end keys by the arrows are easy to get used to. The two keys on the bottom-right corner function as page up/down normally and back/forward when ‘fn’ is pressed, which also works well (though I’d have preferred not to have to use ‘fn’ for back/forward). The three new cut/copy/paste keys (used with ‘fn’) function perfectly (various linux software), but just seem redundant (and are shifted to the left compared to what you’d expect). These keys (without ‘fn’) also work fine, but seem a bit strange (the app-switch key is just redundant as is the desktop key, “right click” is not where you’d expect it and the play key… I sometimes press accidentally).

The number-pad area works fine, but I pretty-much never use those so won’t comment further.

Across the top there are ‘eject’, ‘power’, ‘sleep’ and ‘wake’ keys. For me, ‘power’ shuts-down the computer (which I didn’t want it to, so disabled it), ‘eject’ and ‘wake’ do nothing, and ‘sleep’ works correctly. I’d have preferred a ‘lock screen’/’screensaver’ key.
There are also calculator, mail and browser keys on the right edge, which are useless to me. I might remap them as forward/back or alternate media controls or something.

: if you use them a lot, you might find the TypeMatrix layout a little annoying. They’re separated between F5 and F6 instead of the usual three groups of four, and — as with a lot of special keys on the TypeMatrix — there’s not much tactile feedback to tell you which buttons are which. This is another area where I prefer the (2010/11) ThinkPad layout.

Overall, I like this thing a lot and think it’s well worth the money if like me you spend a large portion of your day using a keyboard. Not having each row of keys shifted by some strange amount makes the layout feel so much more natural — easier on the fingers and easier to remember. I don’t understand why virtually no-one else makes keyboards without the stupid shifted rows (it’s awkward switching between the two to be sure, but worth it IMHO).

Making the keyboard so compact seems unnecessary in my opinion, or rather, some of the layout on the right-hand-side seems a bit strange. I prefer the Thinkpad keyboard for placement of the arrow keys and probably also the shift key. The stated reason for the compact size is so that the mouse can be placed closer to the right edge; I actually find it more ergonomic having my mouse below the keyboard (so it’s just a small bend of the elbow to reach it).

What I’d like to see:

  • A cheaper version, probably based on standard laptop-style keys. Not because the $110 version isn’t worth the price if you’re working day-in, day-out in front of a keyboard, but because a cheaper version would be much more affordable for a second keyboard left at home and to recommend to friends. There are enough people interested in alternative keyboards because of RSI or ease of learning, but unless such models can compete with the dirt-cheap keyboards sold in electronics stores everywhere they don’t stand a chance with most people.
  • A laptop-integration version (specifically, one I can put in my thinkpad).
  • Better separation of the arrow keys (really clear tactile feedback is good).
  • An integrated USB hub. So useful.
Posted in Uncategorized | 3 Comments

Reading is eating

I just stumbled across a bit of food for thought, and think I did (my nice little jog-in-the-rain turned into exercise of a different kind). Have a bite. Go on, eat the whole thing while you’re at it — it’s only 8 minutes. Then it might be an idea to chew-the-cud a bit before we go on — heavy lifting required.

Fed, rested and primed for action now? Good. Now lets get the obvious conclusion out of the way.

I shouldn’t overeat. Did I say that? Come on, not that one. After all, after are stomachs are full eating gets painful anyway. I expect you know the feeling, after having read for a few hours, when your mind blanks out all but the page in front of you, that your vision starts to blur, concentrating on what your reading any more becomes difficult, and it dawns on you, slowely at first, but more and more insistently I need to get up and do something. No, I expect you were familiar with that before you even stumbled across this article.

No, my first conclusion was I may not have a problem with eating and exercising out food, but I sure have put on a bit of cognitive fat, so-to-speak. There’s a few points here:

  1. There’s no point consuming information if you don’t use it. Of course, there’s no point going on a starvation diet either, besides the fact that we need sustenance to grow. But reading something because it tastes nice isn’t a good idea either, nor just reading something because my teacher tells me I need to know this. 
  2. When you’ve got a job to do, concentrate on the job and not the food to fuel it. Sure, you shouldn’t ignore whatever information you’ll need, the same as it’s not nice to come home after an exhausting day’s work and discover you need to cook supper too. But thinking I need to know the theory in this textbook, so I should start by reading the entire book is not the right way to get something done either.

I hope you already knew that.

I did.

In theory at least.

So, what’s the other conclusion I drew?

Well, consider this. You come up with some cool idea, so you write it down. You don’t trust it to your memory, because you know if you do that you’ll forget at least three quarters of the good bits. (I certainly do.) Job done. You think of something else. You right that down too.

A week goes by. A month. Soon you’ve got more pages of notes than you can organise, and you’ve already discovered by now that you’ve written the same idea down twice, more than once. If things go on like this, soon you’ll have a pile of disorganised, highly-redundant notes, and any time you want to find some note in particular you’re going to have a huge pain. So what do you do?

Thankfully you wrote your notes on a computer so you can easily rename them and move them around. So you categorise them. You’ve got quite a few notes about topic X, so you group them together. You notice several are about some
subtopic, Y, so those can go together too. You even tag stuff about topic Z, despite the fact that topic Z pops up all over the place in contrary to your top-down categorisation.

So, some hours of work later, you find you’ve got things a bit better organised. You’ve not categorised everything yet, but it’s just a matter of time. You’ll get to it.

Time goes on. Notes get better organised. At the same time, more notes get added. You realise from time-to-time that some old categorisation wasn’t very effective, or that you can also categorise items according to some new tag, and
start shuffling around already-categorised notes. You realise that over time, the work of just managing your collection of notes grows in proportion to the number of notes you have. If work goes on like this, you’ll either spend less
and less time doing anything original as more and more of your time is spent keeping things organised, or you’ll have to give up on the organisation and accept that your notes get more repetitive.

Except that that’s not all that will happen if you stop organising. Any plans you had once will get buried. Jobs that you planned out will never get done, not because the information’s not there, but because you can’t find it. Any attempts to stand back to get a wider perspective on all the little conclusions you’ve drawn over time will become extremely difficult, if not impossible, because you can’t _find_ all those little conclusions you’ve drawn.

So what happens?

I’ve been thinking about what happens to society. Already we’re suffering, not from information overload, but from information disorganisation. Physicists need to know about the particles or waves or structures they’re studying, but also need to be able to do some pretty fancy maths in order to achieve anything. Biologists need to know about cells or organic molecules or organ structure or many other things, but in order to do their work, geneticists need complex computer algorithms to analyse anything, pharmacists need a lot of complex chemistry to engineer their drugs, and pretty-much any biologist needs to be able to handle a lot of statistics to prove anything (at some confidence interval).

So, what’s happened? We’ve specialised. Now, more than ever, young people go to university to study maths or English or chemistry or one of many other subjects. Now, more than ever, young people go on to do PhDs — but even if they don’t get that far they’ve already had to specialise from being a computer scientist to focus on algorithms, or language theory, or machine learning, or databases, or computer vision, or encryption, or data transmission and information theory, or one of many other things. There’s no such thing as a generalist any more. Is that a good or a bad thing?

Well, judging by the incessant discoveries in medicine, in computing, in climate science and in science in general, one can hardly say it’s not worked out. I won’t go on about this, because there’s no point — we’ve been developing new drugs, faster computers, better telescopes, etc., for decades, and there’s no sign that this is about to slacken off.

One thing that has got harder though is using results across fields. As the lowest hanging fruit in mathematics has got harder and harder to reach, new developments become harder and harder. That the proof of Fermat’s last theorem took so long is because it required the co-use of so many areas of mathematics. That the Polymath project had such success proving the density Hales-Jewett theorem was due to the fact that it allowed the collaboration of many different mathematicians working in different backgrounds. So what might be possible with massive collaboration across many fields of science? What might be possible by combining the knowledge of the whole of humanity in one place? Proving this point with an example is obviously beyond my capability, but I hope you get the picture that the information processing resources available to individual humans have an enormous effect on society. So big, in fact, that I can only see four possible outcomes:

  1. We continue as we do now. Here and there communication of ideas and results get optimised somewhat, general education may improve, big organisations continue to out-compete smaller ones due, often enough, to being able to employ specialists in more overlapping areas. Scientific advances continue to be made, but we remain a human society, working within human limits. Fundamentally this outcome is unstable due to the following possibilities, but striving as we do for control and economic return we may be able to keep it up for a long time yet. 
  2. At some point, things slip up. Society gets more fragmented as we continue to specialise, and instead of asking people specific questions about there business (you had any problem with foxes lately?) we are more and more reduced to asking peripheral questions (how is work?). Key knowledge is lost as people die or migrate or fall out with each other, university level education harder as professors have to focus on more specific fields, and in the end society fails to  rovide enough young scientists to replace the old, resulting in a spiralling collapse of much of science. We don’t lose our cars or computers or Airbus A380s, at least not immediately, because people already know how to build cars and  computers and Airbus A380s. But scientific advance collapses and maintaining society as we know it becomes a struggle. 
  3. Genetics or implants or some type of human modification enables us to become smarter. People can take in larger amounts of information and process it farther. Temporarily, the new “super people” pick up knowledge across many different topics and produce many new discoveries. But eventually they become the norm, and society — a faster moving, more energy intensive society — has make even better people to keep moving as people have accustomed to.
  4. Computers can already process information a lot faster than us — and computers are, and probably will be for a while yet, increasing in capacity exponentially at a very fast rate. If they get smart enough, they may end up doing our reasoning for us. At first, of course, we’d remain in control — but with them being a lot smarter than us (or at least able to take a much broader point of view), along with the constant pressure to let those best at doing anything organisational do it, it would be almost inevitable that they would end up running society, perhaps leaving us as pets to them as dogs and cats and monkeys are to us.

I won’t say that one of these scenarios is better, or more likely, or more preferable than another, because I didn’t write this article to talk about armageddon. But I will say that information is more important to us than ever.

Posted in Uncategorized | Leave a comment

mobile “keyboards”

Since I’m doing a project on chording based input devices next term, I thought I’d ask for a few opinions.

We already have a working 5-button keyboard which can be learned within one hour. Sound interesting?

There are quite a few ideas for small chorded input devices to fit in the hand out there. The basic motivation is: you have a very small keyboard, operated with a single hand, which can be used standing, sitting, running, diving, more-or-less wherever you want. I’ll put some links to existing ideas on the web below; unfortunately only one is available to purchase and that’s not exactly great value for money (I’ll post a review if anyone’s interested).

So, do you think something like this would be marketable? Which designs do you like best, and would you want an integrated mouse (thinkpad-style stick or ball) too?


Posted in Uncategorized | 4 Comments

Split brains and unified cognition

Very interesting talk. So is this kind of division of labour necessary in a creature which can both reason and decide what’s important in life? And how come I can coordinate the actions of my hands at all if they’re controlled by two largely separate entities?

Reminder to self: think a bit more about this.

Posted in Uncategorized | Leave a comment

My keyboard layout

For a long time I’ve been using a modified keyboard layout — I think it’s time I finally shared it with the world. You can read the details here:


Posted in Uncategorized | Leave a comment

N900 + mappero + OSM + wanderreitkarte.de = map

To any other maemo users wanting good walking/cycling maps: I’ve been trying to add any map with contour lines to mappero for a while now, and finally figured it out. Probably it shouldn’t be that difficult but for the lack of information on how to do this, so here’s a guide.

Mappero uses tile-based maps, so what does a map actually consist of? A base plus optional layers. You can probably work out how to edit the map resources mappero uses (menu->maps, tiles and repositories sub-menus). If you look at the existing tile resources mappero comes with, each has a name, a unique identifier, and a URL.

Tile sources

You can see a list of a few tile servers on the OSM wiki. What format do these use? The first URL helpfully comes with an example scheme to access tiles, though it’s not the same scheme mappero uses. I wanted to add the Reit- und Wanderkarte maps, but couldn’t work out how to access individual tiles. In the end I just had to guess. The URL scheme for accessing OSM tiles (taken from existing sources) has the form:


where 12 is the zoom level and the other two numbers are longitude and lattitude coordinates (details on how to calculate these coordinates can be found here). It didn’t seem unreasonable, therefore, to assume a similar scheme would be used to access the wanderreitkarte tiles; with a bit of trial and error I found that the following worked:


There are three layers here: the base map, an overlay with roads and place names (topo) and finally an overlay with contour lines (hills). To add these sources to mappero, create new tile sources, enter a name and the URL (replacing each number with %d), set the type to XYZ_INV, and set “layer” to true for the two overlays.

Mix and match

Now you’ve added some tile new sources, you need to set up a “repository” (I’d call it a map configuration) to view them. From mappero’s maps menu, click the “repositories” button, add a new repository, change the zoom levels if you wish (the tileservers page lists the maximum zoom for most sources as 17-18), then set the base tiles and any overlay layers. After clicking “save” you need to close the maps menu completely before reopening it in order to select the new map tiles. Voilà!

As you may have realised at this point, there’s no requirement to only use wanderreitkarte overlays with the wanderreitkarte base (or whichever new tiles you just added). Since the default OSM maps look nicer than wanderreitkarte’s base+topo layers, I combined the OSM base with the wanderreitkarte hills layer. Hey presto, exactly the maps I wanted, and I only need to download one more layer!

Posted in Uncategorized | 3 Comments