Freewrite – The distraction-free draft machine

My Freewrite digital typewriter from Astrohaus arrived this week and I couldn’t wait to open the box and give it a try. It’s been a long time since I ordered it. So long that I had to look up the credit card transaction to remember that I paid for it in February 2015 back when it was still called a Hemingwrite. Listen to the anticipation in my voice when I was a guest on Jon Bruner’s Solid Podcast in September of last year. I’ve been waiting for this thing.

I know it seems ridiculous to spend $400 on a digital typewriter, but I struggle with a deficit of attention, or maybe it’s a deficit of consistent intention, in any case the idea of an utterly distraction-free writing environment was too appealing to pass up. Writing is hard for me because, I don’t know, it just is, and looking at Twitter is easy. Plus I’m a sucker for gadgets.

I’ve experimented with a bunch of software-based attention-preservation mechanisms along the way, but until the Hemingwrite came along I hadn’t thought about using restricted hardware as the path to salvation. Having crossed that line of conceptual demarcation, and facing a long wait for my Hemingwrite’s arrival, I also discovered the Neo by Alphasmart. It’s a really simple word processor designed for kids in a classroom setting that is no longer in production but is readily available on ebay for next to nothing. I bought one from a liquidator that got it from the Houston school district, put three AAA batteries in it, and have been writing with it most mornings since. Even better, I’m still on the original set of batteries.

I mostly use it for my “daily pages,” and sometimes I use it for the first very rough stream-of-consciousness draft of writing projects. It’s light, simple to use, has a decent keyboard, and sync’s via a USB cable.

It’s a word processor, meaning that it’s possible to arrow-key around a document and make corrections and edits, but the little screen only shows a few lines so it’s easier to just focus on what you are currently thinking and worry about structure and editing back on the computer. Also, sync’ing with a computer is one-way. Once text is in the computer it won’t sync back, so editing really is best done there.

Now my Freewrite is here and for $400 it better make my $25 Neo look rudimentary right?

The Freewrite is compact but not tiny; fetchingly retro with its big red power button, two big throw switches and a raised keyboard; and it feels sturdy due to its four pounds of heft and aluminum case. And unlike the Neo, it sync’s with the cloud (Dropbox, Evernote, and Google Drive) wirelessly.
I turned it on and five minutes later setup was complete. All I had to do was connect it to my wifi network, log into my previously created “Postbox” account at, and then link that account to Evernote (it will also link to dropbox or google drive). That was easy.

As soon as I tried it two things stood out. First, the wonderfully clicky and tactile’ly satisfying keys. The Freewrite is built around mechanically switched Cherry MX Brown keys. These aren’t your Mac’s chiclets. They have a long throw, a really nice concave depression that fingertips nestle into, and a satisfying click when they’re depressed. They are the kind of keys that will make you want to type “the quick brown fox” over and over again if you have nothing else to say.

The second impression was more disconcerting. The display is low-power-hungry e-ink and it has a drawback. It renders type slower than my fingers can generate it, so the display tends to always be a little bit behind my thoughts. Every press of the space bar is a tiny little exercise of faith because I’m leaving behind a word I haven’t actually seen yet. Often that faith is misplaced and halfway through the next word I’ll realize I misspelled the previous one. But maybe that’s no big deal. Keep going and correct it later right? This machine is about being in the flow.

And did I mention the keyboard? It’s a revelation. Within an hour of unboxing the Freewrite I was deep into the Reddit forum dedicated to manual keyboards and now I’m looking for a nice Cherry MX-based replacement for my Mac. But back to the Freewrite…

The display shows about a paragraph of text. That’s a lot more than the Neo, and it’s nice because more of the context leading to where you are now is visible. It’s also easy to page up and down to read more if you need to remember the path that you got you there.

But, … and here is where you realize that the Freewrite is an electronic typewriter and not a word processor, it doesn’t even have the Neo’s arrow keys for moving the cursor around. Page up to reorient yourself and notice a mistake? Tough, fix it in the next draft. This cursor stays at the end of the file. So, unless you are willing to backspace the whole way there, the mistake stays. In fact, it might not even be fair to call this a typewriter, even an old mechanical could be rolled back. This thing really is append only.

It’s as if it’s saying “I promised you a distraction-free writing environment, but while I was at it I took away your workflow.” This thing is opinionated. It’s not a “writing machine,” it’s a “first draft machine” unless you are willing to return to the workflow you used on your last real typewriter. And that’s not the only way it expresses its minimalistic opinions. Currently, the only way to get text off of the device is via the cloud. Astrohaus intends to enable USB transfer, but it’s not there yet. So for now the only option is for text to pass through Postbox on its way to its ultimate Dropbox, Google Drive, or Evernote destination. But, and here’s the weird part, Postbox keeps a copy of everything you write in your account which you can’t delete because “you should always be moving forward.”

This isn’t a typewriter, this is a typewriter whose manufacturer makes you keep a copy of every page you write in a file cabinet at their office. And best I can tell, there is nothing on the website that describes their security practices or anything else about Postbox’s technical implementation. I am not happy about this part at all. In fact, I’m stunned that they would think was ok. Let me crumble up the paper and throw it away if I want to ok?

Ok, I got that out. Now let’s get back to the workflow implications. I’m old enough to have composed text on a real typewriter and I remember the “make a note and keep going” style of writing that was required. Drafts weren’t just some edits to the thing you already wrote re-saved with a different name, we re-typed them in their entirety from red-ink notes and corrections made directly onto previous versions. I did that for a long time, but that linear style of writing disappeared overnight when I switched to writing on the computer where the relationship between writing and editing is much more fluid. For years I’ve barely made a distinction between the two processes.

Over the last year I have used the Neo on some writing projects and, because of the difficulty of editing on that small screen, I settled on a hybrid workflow that tries to take the best advantage of both environments. I used the Neo as a distraction-free place to do the initial brain dump. I often started with a sketched outline on paper or sometimes a Scapple mind map, and then I just started writing. The mind map gave me some mental structure to work from, but the bare word processor encouraged a kind of unencumbered concentration that helped get those first words down. This is especially important during that first phase of a project, when I’m still working out what I want to say, and the thoughts can be most easily derailed by the distractoverse.

This is the kind of workflow the creators of the Freewrite have in mind, and it will do just fine. In fact, that keyboard combined with that laggy e-ink really are amazing for just letting the mind get free during during that first pass. It is hard to explain, but the feel and clack of those keys bring about an almost meditative state of mind and the “ooh, I need to check Twitter” moments seem to flatten into a much lower frequency. It works.

The problem is that I get distracted during editing too, and since I spend at least as much time editing as I do in initial draft, that’s a problem. For any given project I’m exposed to the distracting environment of my computer for at least half of the time it takes to finish it. I spent $400 and got a $25 Neo with a better keyboard for making first drafts when what I really want is a distraction-free writing environment.

In fairness to the folks at Astrohaus, they never told any other story. This is what they set out to build and they built it. Furthermore, they shipped what they said they’d ship, almost when they said they would ship it. That’s more than I can say for most of the Kickstarters I’ve funded (for the record, I purchased my Freewrite in their first post-Kickstarter sale).

I think what I really want is a Freewrite that would be good for both distraction-free drafts, and distraction-free editing. In other words, distraction-free writing. I want Astrohaus to partner with the folks at Literature and Latte and build a version of Scrivener into the Freewrite’s firmware that seamlessly syncs to dropbox and lets me move between Freewrite and personal computer versions of a project with ease. An 80% implementation of Scrivener on Freewrite would let me stay in its distraction-free environment for much more of the time spent writing and editing. By the way, I also want American politics to moderate and for scientists to discover a microbe that can convert CO2 to a compound suitable for making long-lasting electric car batteries. I can dream.

I like the Freewrite but I don’t love it yet. Their “just keep going” mantra is implemented with too much ideological purity for me. At a minimum I want to be able to move the damned cursor to make a change in that last line, and I want to be able to delete the stuff that I write. Better yet I just want to keep my paper in my file cabinet. They need to enable USB file transfer and soon.

Astrohaus says they are going to make it possible for third party developers to contribute and build new capabilities for the device. I hope Literature and Latte is listening, because they need to get busy. A machine for getting focused on the first draft is nice, but a writing machine would be amazing.

• • •

Centralized Blockchain as a Service?

I ran across this piece about Microsoft Azure’s Blockchain as a Service and I just thought it was weird. What is the point of a centralized blockchain that requires you to trust a single provider? Isn’t this just a really slow and limited version of SQL Azure? I can’t fathom why an organization would use a blockchain as a service rather than a database as a service. Maybe I’m missing something.

Technologies like Ripple let a group of trading partners put together a distributed network where trust is required, but distributed among the group enough to prevent any single entity on the network from being able to modify the blockchain. And no one is trusted beyond the value of their counter party risk.

When IBM, Accenture et al talk about the Open Ledger project, they are only talking about the source code for an alternative blockchain that (presumably) will offer better data support than OP_RETURN and a variety of ways that consensus can be established (e.g. mining etc). What isn’t stated but (I think) generally understood, is that each use case will require thinking through how many nodes will be required, and who should host them, to adequately distribute trust appropriate to the problem being solved. In practice, this will probably look a lot like Ripple. A bunch of trading partners decide on an Open Ledger solution, and implement a secured network by having multiple trading partners independently run nodes.

Which brings me to this. Blockchain as a Service is probably a cool idea, but not if it is implemented by a single company that you have to implicitly or explicitly trust. Imagine, if Microsoft, instead of putting the whole thing on Azure, had rallied all of the major cloud providers to together host a Blockchain as a Service. If Azure, AWS, IBM Bluemix/Softlayer, DigitalOcean, Google, Rackspace etc. all agreed to host a common service you’d get the benefit of a hosted blockchain, but with at least a reasonable level of distributed trust.

In fact, a scheme like this would be great for “scaling” trust to go with your needs. Apps with low trust needs (but that still want to be a blockchain because of sidechaining or whatever) could run on a single provider. If you needed to scale up the level of trust for a particular application, you could pay more to have the nodes distributed across more of the participating providers but still billed through a single entity.

There’s a lot more to think about here. Like, can organizations that are already agreeing on a platform be trusted not collude, etc. But conceptually this makes a lot more sense to me than a centralized single-provider blockchain.

• • •

Read on paper, love better?

Five years after it was published, I finally just finished reading The Shallows by Nicholas Carr. It was a sobering read. Sobering in the sense that he makes the compelling case that digital media is changing us, in possibly irreversible ways. We are adapting to distraction but may be losing our ability to think deeply in exchange. What he doesn’t address as much is the anxiety that many of us feel as part of the change.

The the book is a downer in that if offers no solution. It painstakingly describes how our brains are changing, and what we are losing, but it doesn’t offer any magic bullets. This is the neurological equivalent of all those conversations we have about dying newspapers. We know that digital media is killing them, but ten years in we have no clear idea how to keep journalism alive. Digital media, it seems, puts us in a perpetual state of epiphany-delayed in whatever area of endeavor it attacks. When it comes to dealing with digital distraction I think we’re all be waiting for our epiphanies for a while.


• • •

A World Without End – I wish it were more real

This piece, World Without End, in the New Yorker got me really excited. Briefly. And then it disappointed me. Mostly. This is the opening that grabbed my interest:

“The universe is being built in an old two-story building, in the town of Guildford, half an hour by train from London. About a dozen people are working on it. They sit at computer terminals in three rows on the building’s first floor and, primarily by manipulating lines of code, they make mathematical rules that will determine the age and arrangement of virtual stars, the clustering of asteroid belts and moons and planets, the physics of gravity, the arc of orbits, the density and composition of atmospheres—rain, clear skies, overcast. Planets in the universe will be the size of real planets, and they will be separated from one another by light-years of digital space. A small fraction of them will support complex life. Because the designers are building their universe by establishing its laws of nature, rather than by hand-crafting its details, much about it remains unknown, even to them.”

My mistake was to assume that they were using real physical laws to create this universe. I jumped to the wrong conclusion though. Like every video game, they are relying on made up algorithms to create realistic-looking-but-ultimately-artificial places. The distinction here is probably minor, because no application of math could recreate the universe as it is, but I was hoping, naively I realize now, that they were attempting to re-create a universe at least based on some basic physical laws.

Why do I care? Why not just have a game that, like any fiction, conforms to an internal logic that doesn’t need to be derived from the real world? Maybe it doesn’t much matter, but I increasingly find fiction and games that play fast and loose with basic physics empty and dissatisfying. The tropes of infinitely available energy, hyper drives, and the like, in my mind at least, feed a kind of false optimism about our prospects off of this planet that are at some level counter-productive. We are not going to save human kind by finding another planet. If we can’t make a go of it on the one we evolved on we’re screwed. Let’s take better care of this one.

Also, and maybe more pragmatically, I grew up in a world of physical things. I made model airplanes, worked on cars, and took apart every toy I ever got. Working on (and breaking) things in the real world gave me useful intuition about how that world worked. Intuition that was useful to me later as an engineer. Today, virtual worlds like Minecraft are a substitute for a lot of the real world play of my day. They are amazing playgrounds that let the imagination go wild, which is really cool, but I don’t think they help intuit how our actual world works. In fact, I think they create false intuitions in the way they tweak physicality.

Anyway, this still promises to be a really interesting game when it becomes available. The sheer scope of the universe they are building will be vast, and probably the biggest problem they’ll have to face with gameplay (just like we experience in real life with our still silent SETI endeavor) will be coaxing interaction between players who start off billions of light years apart.

• • •

Run a vanilla Minecraft server on Raspberry Pi

My partner has a teenage son, Michael, and like a lot of adults with a kid in their life I’ve come to know Minecraft. I even kind of like it. I was scared the first time night fell in survival mode, but I’ve made it from uber-noob to mere noob now and can usually make it through the night huddling in a basic wood block structure. I won’t get points for the aesthetics of my creations, but I survive and even have fun with it. Michael still makes fun of my noob’ness and will regularly switch to creative mode, grab a diamond sword, and off me just for fun. But whatever, I’m learning.

Michael and I have been playing with a Raspberry Pi lately. Actually, we bought two of the 1Gb Pi 2 Model B’s. These things are actually quite powerful little quad core computers so after a few simple projects we decided to use one of them as a dedicated Minecraft server.

The Raspberry Pi already comes with a cut down version of Minecraft that you can play directly on it, but that’s not what interested us. We wanted to create a “real” Minecraft server that we could leave running indefinitely and connect to from our local LAN anytime we wanted to.

There are quite a few tutorials out there on how to get this working, but most of them are of out of date. For example, we wasted a lot of time on our first try installing Java onto a Pi that already had it as part of the OS install. So, having gone through the process a few times, we figured we’d write this up for anyone else looking for a how to.

We should start with a note of caution. This is an installation guide for a vanilla server, and it frankly doesn’t perform very well on the Pi. On our first installation we got lucky and ended up with a seed for a really simple island world. The server runs flawlessly with that simple world. However, other worlds with more complexity have consistently crashed the server with two users. We’ll include the seed for “Michael and Jim’s island world” later in this post so you can use it if you’d like. Or, you may decide to modify this tutorial and install a Bukkit server. It should perform better but we haven’t tried it.


• • •

Buy an alarm clock

I’ve blogged recently (and often) about my struggles with technology addiction. Since writing this piece in March I’ve made some progress. I make good use of a simple word processor when I’m trying to write, I regularly use a dumb feature phone sans data plan to stay in contact without staying connected, and I’ve been meditating almost every day.

So when I read Turning off technology is about mental wellbeing – not becoming a digital hermit about managing ones digital lifestyle I thought I had it covered. Different person, same journey. But the advice at the end of the piece stood out. “Buy an old-fashioned alarm clock so you can leave your phone in the other room at night.”

I already put my phone on do not disturb at night, but sometimes I still reflexively grab my phone in the morning and check my email despite my stated intention not to look until noon. So I like the idea of just not having the phone in my room at all. I think I’ll try this.

Of course, there are other good reasons not to have your phone in or near your bed.

Anyway, I’m getting that alarm clock. I still have a landline nearby (which is crazy, I know) if anyone really needs to get hold of me at night.

• • •

New home

Hi, welcome to the new home for Limn This. I decided to move it from Typepad to a self-hosted WordPress instance. I’m still trying to fix and clean up some issues so I apologize if it’s a bit of a mess around here.

• • •

The Dash Button – Can you say lock in?

Amazon announced the Dash Button this week and most of the buzz was of the “is this an April fools joke?” variety. That’s unfortunate because it’s a really interesting development and deserves more careful consideration. The Button is simple but it demonstrates the impact of inexpensive inter-networked things.
Simon Wardley’s take was probably the most interesting I’ve seen, if a little bit breathless. The Dash Button is best suited for recurring consumables and while that will hurt local grocers, I don’t see it cratering their businesses despite the importance of paper towels and milk to their bottom lines. I also don’t buy into the idea that Bezos bought the Post (and Buffett bought Media General) to turn it’s printing presses and delivery trucks into a Button manufacturing and distribution engine.
To put my money where my mouth is I bet Simon a cup of tea that Warren Buffet won’t be making a “small fortune” by 2030 on the manufacture and distribution of digital electronics via printing presses. He should have said “Berkshire Hathaway” because I’ve got not only the weight of my argument on my side but the actuarial probability of Warren’s expiration by then as well. If I lose, I’m counting on the mists of time to intervene on my behalf and maybe Simon will just forget our bet by then.


• • •

Getting out of my Skinner Box

When I was a kid my teachers told me I was “hyperactive,” and they gave me lots of demerits to emphasize their point. I’m pretty sure, had I been born later, I would have been considered ADHD. I coped reasonably well for most of my life, but the advent of the smart phone was not good. Since then I’ve struggled to stay focused long enough to write and do other things that require stretches of solitude and flow. These days the one two three punch of laptop, smart phone, and Twitter are enough to keep me nearly always distracted.

I have tried all of the things people try. I used RescueTime for a while but I really didn’t like all the tracking. So, I tried Freedom but I tended to just re-boot my computer when I thought to myself “I just need to look up that one reference, I’ll turn Freedom back on as soon as I do.” An hour of surfing later I’d ruefully contemplate the inadequacy of a reboot to inhibit distraction – it simply doesn’t cost enough.

More recently I set up a second computer that doesn’t even have a working web browser on it and started working in an attic room with the wifi turned off downstairs. I figured at least then I’d have to make a trip down and back up the stairs to turn the wifi back on. Sometimes it works. Sometimes I just get more stair climbs on my fitbit. I’ve also used all the minimalist text editors. Whatever. They are nicely distraction free in their design, but they’re just a screen away from the distractions. When the itch starts to build they do nothing to prevent the quick alt-tab into the rabbit hole.


• • •

Public vs. Private Cloud: Price isn’t enough

Last October Simon Wardley and I stood on a rainy sidewalk at 28th St. in NYC arguing politely (he’s British) about the future of cloud adoption. He argued, rightly, that the cost advantages from scale would be overwhelming compared to home-brew private clouds. He went on to argue, less certainly in my view, that this would lead inevitably to their wholesale and deep adoption across the enterprise market. 
I think Simon bases his argument on something like the Rational Economic Man theory of the enterprise. Or, more specifically, the Rational Economic CFO. If the costs of a service provider are destined to be lower than the costs of internally-operated alternatives, and your CFO is rational (most tend to be), then the conclusion is foregone.
And of course, costs are going down just as they are predicted to. Look at this post by Avi Deitcher, Does Amazon’s Web Services Pricing Follow Moore’s Law? I think the question posed in the title has a fairly obvious answer. No. Services aren’t just silicon, they include all manner of linear terms, like labor, so the price decreases will almost certainly be slower than Moore’s law, but his analysis of the costs of a modestly sized AWS solution and in-house competition is really useful. 


• • •
1 2 3 16