WoT keynote (and clones :)

I never know if I need to be happy or sad when I see other people mentioning WoT. Don’t get me wrong: I’m totally happy to see people pushing forward the same ideas we have here, make them better, build upon them and reuse them for their own projects. One year ago, most people I talked to were like “aha, use web on devices…. but why? Why not use something made for devices?”. It didn’t sound right at that time, but when you think about a world of web enabled devices, then you can see tons of bulbs turning on. As a matter of fact, in the last days I’ve seen a proliferation of people talking about WoT, but actually they talk about a different WoT.

To be clear: we don’t claim to have invented WoT; it’s been here since the Web itself, and tons of projects have already connected objects to the Web. This is cool, but doesn’t reflect our WoT. The core concept of WoT is in the W (remember, Web), and it’s about reusing what’s on the Web not only for documents, but with also to interact with things in the real world. In particular, use as much as possible of the HTTP standard to build applications for devices, instead of relying on middlewares or other proprietary protocols. This is completely different from the pre-WoT projects, where the Web was used only to access Web pages with data from devices. What we propose is use as many functionalities of HTTP as possible to make everything in the world both a Web server and a Web client (some of these functionalities might not even be supported by firefox – although they are part of the HTTP specification).

Spot the difference? It’s very subtle, yet fundamental (okay almost philosophical).

The actual debate of what the WoT really is, stems directly the debate between REST and SOAP based Web Services. The point is not which one is the best, but which one is better to fills your requirements. And in the context of WoT more about end-users, rather than big companies, REST might be a better candidate, and that’s what we’re talking about on this site.

Historically, the first time I saw WoT is in these two presentations of Dave Raggett. For me, these slides are the clearest and first conceptualization of the Web of Things. Around the same time, this tech report by Erik Wilde (who should use his account here to post from time to time 🙂 is what I could say the seminal and most inspirational source and documentation about the WoT vision. That was straight to the point, and he laid the finger on exactly what the WoT is about. Our work here is merely a continuation and implementation of Erik’s initial vision. However, the term “Web of Things” itself is older and from what I know, the earliest occurrence can be found in Traversat’s paper.

On the other hand, when I see people publishing and marketing things about WoT without even mentioning that paper (or Dave’s slides), means that either they didn’t google “web of things”, or they just try to hide the fact that they didn’t actually come up with the idea. The same is true for people publishing papers today 2009 about WoT, without acknowledging our papers, or worse none of Erik’s or Dave’s – how about googling “web of things” before proclaiming inaccuracies?

The coolest thing yet for me is to see people popping out of nowhere (such as a totally different area), and being self-proclaimed WoT experts. Check this keynote for example (did anyone of you ever hear about this conference by the way?). I’d sincerely love to see WoT hackers and coders and [insert occupation here] popping out throughout the world and join us to make the WoT a reality, but I don’t think powerpoint is really the tool that would help us the most. We need coders, and people who really want to see what’s underneath the surface of HTTP.

For the sad part, is when I hear people talking about “web of things” without actually caring about the Web (as in HTTP) at all. These folks actually mean Web as in Web 2.0 (you know, facebook, twitter, wiki, bla bla blaaaah), but not a single word about the technology and architecture of the Web. It seems that what is important is to put toilets and things on twitter no matter how (this twitter trend is seriously driving me mad btw). Hmmm, okay, if you say so. This is only the tip of the iceberg which is “putting things on the Web”, and not the real problem which is “making things part of the Web”.

Again, a fundamental difference.

What you guys think? Are we just too religious (and almost fetishistic) about that, instead of focusing on more relevant higher level issues? Would love to hear what you think!

You may also like...

  • ok, i’m taking the bait… i definitely plan to post more frequently over the summer, and i am definitely looking forward to fleshing things out in more implementation-oriented ways. this is where the really interesting lessons are going to be learned.

    just as a comment about “REST vs. SOAP”: the important thing to stress is that WoT does *not* require HTTP/REST on the sensor/device. the back-end implementation is up to the one actually providing access to the managed resources, and it can be HTTP/REST on the devices, or it can be a highly optimized, tightly coupled, RPC-oriented implementation. the important thing is that this back end exposes (some of) its services in a RESTful and loosely coupled way, which then can be mixed and mashed with other such back-ends.

    i think it is very important that we’re careful not to imply that each Thing in a Web of Things is “directly on the web”. it’s not, there is a representation of it with which you can interact over the web, but how this eventually affects the Thing is not something you need to be concerned with.

    when thinking about this, it occurred to me that a nice analogy is that of a Content Management System (CMS). think of the WoT as coming up with ways of how to build Thing Management Systems (TMS), and to make sure they provide services that seamlessly blend into the data and service fabric of the web.

    • oh, sure thing Erik. REST on devices is certainly not a requirement, and could even be a bad idea in some cases. I’m interested into exploring this direction, in particular with WSN to try to quantify the price to use http (or just designing RESTful WSN apps) directly on the devices. Eventually, it will allow a more plug & play experience to do so. I guess…

  • Johann Renner Rouliez

    I can understand the difference between your approach to the WoT and the others. Even though, the thing is your WoT is not a trademark nor an international working group, so it is almost imposible to avoid that other people come up with a similar idea.

    But don’t worry. It seems that you are doing a really seriuos research work so, one day, you will be either remembered for this great effort or the leaders/evangelists of a worldwide accepted WoT.

    In order to answer your question, you should continue modeling and defining your WoT vision towards the development stage, IMHO.