In part 1 of this week's conversation, we looked at some of the tradeoffs with different device platforms. Today we're discussing using mobile devices, and "the Community"
Hakim: Of course there are examples of people sticking with the more expensive hardware is the Apple store, who use the iPod touch as their POS terminal. It has wifi, it has a screen, it has touch. But it has a shell, like a shield, which has a credit card reader in it. It doesn't have a printer, so if you want a receipt it sends the information over the network wirelessly, and they go and pick it up for you. It works great for their shop. And it may well be more expensive than another solution, if they had to scale up massively. But they already have a whole load of iPod touches, as they sell them anyway, and it also shows people that they have cool technology, so it's part of the ethos and advertising of it.
Adrian: Yeah, it's more expensive, but they're showing off that they can do POS with it.
Hakim: And if you think about it, the iPod touch is... a hundred quid or so? Probably less for them, so it's perhaps not that much more expensive for them? But if you think that those big, ugly windows phone devices that courier companies like to use, and which can't even really read a signature, because their processor isn't fast enough, cost £1000 or so, even if you're buying a few dozen of them. An iPod costs much less.
Adrian: They are much more ruggedized of course.
Hakim: Absolutely, but if you dropped your iPod touch, you could buy another one for a tiny fraction of the cost...
Adrian: Which, if you're in the Apple store isn't am issue. But if youre a postman who suddenly is without a way of delivering the parcels they've got, than that is an issue...
Now some of my code has lived on some of those devices.
Hakim: Windows Phone?
Adrian: No, the even earlier ones. Psion, when they were still going, and using Symbian. Later they moved to Windows CE. And even before Symbian they had some 16 bit Psion series 3 ones. And we did some of the IP stack that ran on those.
Hakim: I think they're horrible devices though...
Adrian: it depends ont he device... some of them aren't too bad. They haven't got QWERTY keyboards though, which is the most annoying thing in testing, ABCDE keyboards in a weird layout. And if you don't use them that much, which obviously I hadn't, then when you get them in for a test, it's suddenly like "Where are all the keys?" In theory I know where all the keys are but...
Hakim: we were trying to sell a car-rental app, so when you return your rental car the agent can click on an image of the different parts of the car and mark damage and so on. But trying to write code that ran on the windows phone was tough. First of all the version of mobile IE that they had was really limited, and we eventually ended up intalling Opera, which had enough Javascript and CSS that something a bit dynamic would vaguely work. But I kept thinking surely it would've been far cheaper to get an iPhone dev to do it on a modern device. And not just because it's cool and trendy. But... tradeoffs I guess.
Adrian: Of course when I was working on this stuff was the year 2000 and iPods didn't exist. Even the old, clunky scroll-wheel iPod didn't exist then. Different eras!
Hakim: So, after scaling up, we've got "Community". And that' a big thing isn't it?! In some ways, if you look at Arduino say, yes, it's an easy platform, but if you look at the limits of the technology and the costs of it... because even if the chips themselves only cost a couple of quid, when you buy an Arduino board it's £20-30... and you think it's a bit crap in some ways. The memory is very small, and the capabilities...
Adrian: The memory's very small? I don't know, I remember when Arduinos had 1K of RAM!
Hakim: Yes, the limitless possibilities! But, even despite the limitations, the Arduinos have a whole load of mindshare, and available electronics that are compatible with them, and loads of people that know the stuff. So it makes it a very good and vibrant and useful and fun community to be working with.
Adrian: I think that makes it better than quite a few of the more powerful direct competitors, like BeagleBone or chumby hackerboard and Raspberry Pi... And the Raspberry Pi will be quite an interesting one because of the amount of attention it's garnered before anyone's even got hold of one... and if that turns into some sort of community, then you can imagine that it'll be big enough that it should be easy to use and easy to develop for. Whatever project you're thinking of doing, you can generally Google for the project idea and tag "Arduino" at the end of it, and you'll usually find someone that's done it, or something very similar to it. And they've probably shared schematics, or the parts they used, and the code, and a blog post and a video... So there's so much stuff to easily bootrastap things.
And it's interesting that the shields that you plug on, the additional boards, have taken off so well. And there's that "bug" that everyone complains about that's effectively a mistake when they were laying out one of the early Arduino boards, before there was any idea that it would be popular. But because they didn't correct it straight away, presumably because they'd gone and got X number of PCBs made, and didn't have the money to ditch them all and get some new ones, they just put them out there. And perhaps they were going to fix them in the next revision, but then it took off, and became popular enough that people have started building shields that fit the quirky layout of the Arduino.
So now you get boards that are nothing to do with Arduino, like the Netduino (which is obviously riffing on the Arduino name to kinda explain what it does but is all .Net based and 32 bit, with more procesing power) also replicates this weird "bug" so that it can have all the Arduino boards work with it, because they're just electronic circuits really. So you get things replicating the Arduino pin layout and shields.
And I guess the Netduino has a lot more IO capability than the Arduino but I don't know if it exposes it or not. Maybe it has an extra set of headers you can use if you haven't plugged an Arduino shield into it, for example. But it's interesting that the Arduino has taken off so well that it's started to infect other systems with its quirks and even its disadvantages. And manufacturers are replciating those quirks because then they can tap into its community.
It does make it easier to develop stuff with this community. So for example, I've been playing with the Chumby hackerboard recently. It's more expensive, it hasn't been around as long. And there is some sort of community, but it hasn't really taken off... and so there isn't that way of finding answers to problems. There are forums, and people do answer things. But then the wiki talks about 3 different ways to build software... and the first 2 don't work, because they're the old way. So when you come to it for the first time it's quite hard to get into it. There's a lot to be said for the community making it easier to develop things, because there's soemthing to tap into.
That's one of the things of just being on the Internet, that there's lots of explanations about how to do things.
Hakim: I guess one of the things that we'll do in this chapter is... I guess, talk up the advantages of "community". People who are online, in my experience, in open source communities know what the benefit is. But people who don't know may be quite scared of... the idea of looking like an idiot in public... so going to a public forum and saying "I don't know how to do X"
Adrian: Or of going to a public forum and saying "I'm trying to do Y", because they're worried that somebody else will go "Oh! I never thought of trying to do Y! I'll steal your idea and make my fortune, and you won't be able to". Which is often not what happens, because everyone's already got enoguh of their own ideas, too many, to be worrying about stealing yours.
Hakim: so that is a different mindset isn't it? And you can't force people to care about community. But perhaps we'll at least show about communities of makers, and, by showing our case studies about people doing cool stuff, show that it's possible to speak to them, and suggest "Why not speak to your local community of makers doing fun stuff near you?".
Adrian: It makes stuff a lot easier doesn't it? There being a community near you. And I guess we would say that because we've set up a makers' community in Liverpool, so we're biased. But it feels like it makes things a lot cheaper, and easier and faster. And it's what Andy Huntingdon, who works for BERG, and is probably mostly behind the hardware of the Little Printer, says we're in the era of what he calls he "Geocities of Things".
It's the idea that... you had the internet, and then the web came along, and people started to write web pages. And then Geocities came up and everyone wrote web pages that looked really ugly and had animated gifs with men digging "Under construction" and weird rotating "Email me!" signs, or letters that folded themselves up into an envelope. And all this stuff that was quite ugly and vernacular. But a huge explosion of creativity of people realising that they could build shit on the web. And then it all settled down... the era of blogs arrived, where people started to use tools that made it easier to build websites, without worrying about having to build every archive page yourself. And they realised that many people wanted something "a bit like a diary" and without having to learn loads of HTML and FTP stuff.
Hakim: I guess there are a whole load of patterns that emerged, and now you talk about "blogging" and it's a verb and a noun. But with IoT there are still patterns that are starting to get formed and we're at a bit of a frontier that's still to be defined?
Adrian: yeah, there's lots of experimentation, and people who are doing things because they enjoy it, and not necessarily because they want to make lots of money. Though there are people who have ended up building businesses on it and making a living from their hobby, because they were in it early enough. And now people are starting to find it useful.
Hakim: I was going to suggest that in a way, though we can't make people be interested in community, and if you have a corporate or closed source mindset, you might not be interested, but perhaps we can suggest that it might be a "competitive advantage" to have that mindset of sharing a certain amount of things, and trading on the goodwill of people who have the same mindset.
So, for example, on the project that I'm planning to develop as an example of the book (an IoT task timer), I was lent an Arduino to play with by you guys at MakerNight. Then Aaron from Oomlout came to visit us and I was eyeing up the pre-production IOTM kits that you're designing with him, and got given one to play with for some prototyping the device with. And in the Perl community, I remembered that I knew an artist Tanja Orme who might have drawn an image I could use for a logo and got that on a largely goodwill basis (on the understanding that I'll send her one of the first prototypes when I've built it). So I've had personally a lot of benefit for a project that could, theoretically, even be a commercial one. So that's a lot of help with knowledge and advice and initial outlay on kits that I might or might not use.
So there's 2 sides with community, and also with open source. You might not have the IP later that you can sell to earn your millions, but at the beginning it can give you a real boost.
Adrian: It's almost the Internet way of doing things, where you've got more access to get word out about what you're doing. One of the problems is that nobody knows about things. And sometimes the community that you're tapped into can help spread the word about what you're doing. And otherwise you'd have to buy that kind of promotion with advertising.
There are definite examples about stuff "going viral". Bubblino's a bit like that. I built him to demonstrate what Arduino was, and then, everyone seemed to like him. He's been really good for marketing, if nothing else, because people encounter him and like that.
There was an example recently of some people that built a lamp which was a laser-cut yellow question-mark, which is an icon from the Super Mario games, and you turn it on and off by almost punching it, like Mario does (but not as hard, because it's still a lamp). And these guys built it because they thought it would be cool, and it's gone viral and went on Etsy, and got loads of sales. And one of the ways they're scaling it up is that they're part of a local maker community, like a hackspace in the States. And they've almost drafted in everyone else to build bits for them because they've got so much demand. And they can almost give anybody any of the roles, because people who are in the hackspace can usually laser-cut stuff and solder, or put things in boxes. so they can spread the tasks out and vary the tasks, so people don't get really bored. And that helps the community to help people scale and sell things.
Which isn't really IoT, but it is community.
Hakim: Not all of this will go in the book ... but it's maybe a strand of thinking that we're quite favourable to. Perhaps the business models section...
Adrian: Yeah... it might depend on how many they're going to sell too. There's been a spike of interest, so they'll sell lots of them to begin with, but then it'll tail off quite quickly. And maybe they'll have a trickle of orders, but in a way they can cope with that's just a nice little side earner that they can build on demand. And they won't have had to scale up to a huge factory that they then don't need because the demand was short lives. And when it does die down, they can move onto building other things easily. Just build up enough products that they can build... 5 lamps, and 3 table stands.
So perhaps it is scaling up, but it's on the lower scale of it. And you've got communities to tap into to cope with the big spikes. And then you're building stuff up so that over time you might have more staff... or just make them yourself.
Hakim: That almost goes into the ethics chapter... if you think about the workfare stuff... we'd agree that unpaid work experience was a bad thing, but is using the community for work, perhaps unpaid, and for experience, say, because it's a way to learn lasercutting. And is also a bad thing? Or do you work out how to pay all those people for the pieces of work they do?
Adrian: My assumption is that they were paying people to do that stuff. And it's a good way for them to make some money too. But they're not having to scale up and pay loads of ongoing wages. There's probably a whole spectrum from people who just decide it's too much hassle to get paid, and do bits for free, as a community help.
Hakim: I guess with a community thing, it depends a lot on fun... so if at DoES, I had to help John cut a couple of WhereDials, I'd be quite up for that. But if I had to cut 100 of them... I'd be thinking "Is he going to pay me now?" And I understand your points about making it varied but you can see that it's not just about ethics, but also business models and organizational questions that are quite fascinating. And we probably can't give "an answer" to them in the book, but we can introduce them.
Adrian: some of that is me infecting... did you see my blog post about craft versus manufacture yesterday, about the cufflink manufacture? There are lots of things around that. I had an argument... not quite an argument... well, I almost picked a fight with the head of the craft council at the makers' guild event because there's a lot of stuff in the craft council about the "purity of craft" and the "sole maker of things" who doesn't use tools... well possibly does. I sometimes feel that "craft" stuff isn't handmade enough if you've used a hammer, even, that's too high-tech a tool to have used. The more handmade the better...
And there are definitely craftspeople doing things on their own where they do use hightech kit. But the craft council seems to err more, well, not just them, craft in the UK seems to tend more to the idea of people making things by hand, on a really small scale. Or, you get "shit from China". And there's nothing in between. And there should be.
And there are lots of people... like me... and I do like making things by hand, but I also like the idea of having a business that's big enough almost that it's competing with Tesco's on, maybe, one specific item, obviously not on their whole range, but for example a specific IoT device, because that's what I do.
Hakim: But if you can make 10 objects, or 100, or a 1000, or a small number of 1000s of an object without having to go through a massive industrial process then that suddenly opens up whole new markets.
Adrian: And you don't have to try to drive mass consumption of your product to make that market, because you only need to make 1000 of them... but that's still 1000, and that's more than you could make by hand. And if you've suddenly got lots of people making on that scale, then you've got lots more employment in manufacturing locally. Because you often don't get to offshoring things until you're doing many many devices. Well, you might do even at 1000, even. But not nececessarily. You wouldn't have to offshore that, so there'd still be more employment that wasn't in some big soulless sweatshop where you could pretend that it wasn't having a huge social cost.
Hakim: It would be fantastic if DoES was able to... or if some of the companies wokring out of DoES, I should say, as we're not precisely a group or collective, were able to start to employ people to build products on that sort of scale.
Adrian: Yeah... or be employing people still locally, 20 people in Kirkdale, for example, soldering components might be roughly cost effective. And maybe it might be a bit more expensive, but maybe it wouldn't be that much more expensive. And I'd rather have 20 people in Kirkdale, and it adds £1 to a £50 device versus outsourcing to China or India or Indonesia.
But we're straying off the topic of IoT.
Read Part 3