On REST for devices
One things that I don’t understand, is that most people in our field do not really like our approach, or maybe they simply don’t get it. Indeed, we’re kind of stuck between several worlds: wireless sensor networks (WSN) people that find our approach too esoteric, and HCI people that find it not enough sexy. It is a bit annoying, as people don’t take us very seriously, especially not WSN researcher. I overheard comments like “What? You want to use the Web to connect devices? hahaha! I can’t stop laughing, haha it hurts me! Come on, be serious dude! Using the Web for that!! Man, you’re great, I gotta call my colleagues to tell them your idea, they’re gonna *love* it“.
The Web technologies are still so under-exploited, so few projects out there have really used them even though they are good enough for what non-geeks (i.e. over 99% of population) will do with networked devices. But no. We keep being laughed at when we submit papers, and comments are like “Sorry guys, but connecting devices to the web is nothing new. Next.” Sure, we keep pushing our work, so that people can see it working in live. We are at a point where we can finally show researchers that our stuff is working well and can certainly compare in terms of performance with what’s out there, but more importantly it is much more sexy than anything out there in the academia, especially it’s much easier to use and program, and unarguably more flexible. It’s just very small yet and we’re just down the mountain, but we can see the road clearly. We can even see tons of gold shining up there on the top, and we keep finding huge nuggets on our trip, so it’s totally worth it!
WSN people got certainly a point because of course fundamental research is important, but I also think that too many people in WSN are only focused on “let’s get this protocol 2% faster” or “how about a 3% cheaper computational costs of this algorithm”, or “how to locate eastern eggs in burning bushes with WSN” which of course is interesting, but the people out there on the streets (as opposed to the coke-drinking geeks) don’t really care about that. They want to see stuff that works well and they can buy (and if it’s sexy even better). As a consequence, embedded systems remain some NASA-lab/US-Army kind of research, feared by the public who doesn’t want to deal with it (but they do so daily anyway).
Let’s do something concrete with WSN, that works, and that’s usable, and that people will really need.
There’s a huge lack of the “let’s do something concrete with WSN, that works, and that’s usable, and that people will really need” kind of mentality in this area. I find it’s about time people start building toys and real products with WSN, that are easy to use, and do live up to peoples’ requirements. The technology is young and not perfect, for sure. But come on, it’s sufficient for most needs. We just need to empower people to build their own applications and to innovate, and that’s exactly our target at Web of Things. Just like the Arduino enabled tons of designers and other people to access and use digital circuits within their work, we want to push this idea further. I’ve had a huge debate recently on linkedin, where a guy was like “hey, why the hell do you want to connect things? My customers don’t care about that, and I don’t see why you’d do that anyway.” Maybe his customers don’t need it yet (which I doubt), but they will soon. The point the guy was making is that he doesn’t see the reason behind connecting a dog to a fridge. I don’t see why either, but I’m sure that there’s somebody out there who does. And that’s what matters to me! If somebody has an idea a business model behind, then great let’s give him the tools to do it. I think this point is essential, and that’s why
Networked devices have an amazing implicit value for so many disciplines, and I’ve seen fantastic ideas out there (like everything on Yanko design, my daily fix of inspiration), many of which are realistic and feasible today. Still, I don’t see any products like that on the market, and the ones I see simply suck, or are just made for extravagant billionaires. Why is that so? Maybe, because simply don’t know how to (re)use it. More importantly I see the problem being that if you buy two objects that are not from the same company, they just won’t work together. Companies don’t really care. Of course they want to push their own standards, but if a product is based on a closed standard which makes people not wanting to buy it, is that attitude worth it? Even worse, most researchers in WSN don’t care at all about reusing what’s there and they just come up with yet another soon-to-be-dead solution on their own that fails to reach a broader acceptance. Indeed, I’ve seen so many standards that only a few people use drying in the desert of general indifference. So why not just help us reusing the most of what’s there on the Web, instead of reinventing the wheel (and a square one, moreover)? Let’s just stop inventing new standards, and just exploit and explore the ones we have before doing so! Join us today! I guess, I’ve pissed off most WSN researchers – which is really exciting – and I invite you to join the discussion, please send criticism in the comments!