 So while Andy is getting mic'd up, I'll tell you the story from how this happened. I'm involved in DevConf because I'm very passionate about open hardware and I know nothing about hardware so I thought I'll organize a conference and learn more about it. While sort of organizing this last year I met Andy and his crew, including Steve McIntyre and B Dale and Mad Dog I think I also attended their talks and that was my first introduction to it. This year when Andy came on Saturday we did a beer trip I guess it went to Brewery. It was awesome. So this is just between Botchever and Hermanus and the brewery is called Hoenenklipp and the guy brewing it is Mark Tamors-Taysen. I think you need to come here. B Dale you need to come here. You can take a seat if you want. You can just pitch in whenever you want to. And he runs his brewery online. So we were talking about how do we get involved in open hardware and the easiest way is to start with a project. So I thought well I'm busy building a house let's make a smart home because then you can have hardware and sensors and monitors and everything and that really got Andy annoyed. Do you want to tell us why? I get really annoyed with silos. Do people know what I mean by silos? You should tell. Like blank faces. Various different companies are jumping on this bandwagon which is being marketed as the internet of things and what they're actually doing is they're trying to get you to buy their product and their product alone. And so they have a device, a hub of some description which will only talk to their own devices or devices in their own ecosystem and their club of friends and they tend to stick everything online. It goes into the buzzword the cloud and then two, three, four years later when the little experiment hasn't succeeded they drop the service services in the cloud and you're left holding a bunch of hardware which may or may not continue to work. That's a bit wrong. We've bought the hardware, we expected it to work, we've been sold something and it doesn't. So that's an issue about how open open hardware is isn't it? Well this is why we need open hardware, we need open software is the culprits are not doing this. But then we had this big discussion about what is open when it comes to hardware because hardware has a physical thing associated to it and there was a thing about that software scales exponentially so once you make it available you could really share it but there's always a cost associated with hardware and that creates attention with how open it is. Is that correct? So next time I mean you've been involved in various projects that you've done as open, genuinely if you push your own design out there the reason you've developed it to the first places you've got a particular rich and you want to scratch it. So you sit and work on something maybe you call some people who have a background in electronics or maybe you're experimenting for the first time you put something together you get it working and you share it with the world. But in order to get any feedback in order to contribute back into that project typically you're going to end up building hardware which has a physical cost particularly for your experimenting on something that you've never done before and there is also a very long lead time with software you can jump straight in you can hack around a bit if it doesn't work you can blow it away you can try again with hardware it's to start off with that's fine you could sit there on breadboard you can couple a few bits and pieces together you can let the magic blue smoke out of the occasional transistor and vet as you build something. But when you start getting to a system on chip or more and more highly integrated devices then you're putting them on quite a complicated PCB and the lead time so that PCB manufacturing is quite hard. So yeah you need to be working together as a group and that got us to the what actually is open is it open if you just put your circuit diagram. Yeah there's a bunch of different ways to think about this right because on one level you know I've been really excited over the last few years with the explosion of you know sort of maker projects people publishing information on the hackaday or wherever else about something that they did that they thought was interesting they thought was cool. The problem of course is that a lot of these are somebody's first attempt ever to do something like this and so on one hand it's really exciting to see them getting excited and learning how to use some sort of building block bits and pieces but a lot of the things that people come up with are very very far away from something that you would want to reproduce on mass either because it's not an efficient way to do it for many copies or because there's something flaky about it that they're willing to deal with in their environment but which people who you know are not as passionate about the making of the thing might not be willing to tolerate. Potentially we've got a series of standards and localised legislation you have to comply with. And the hacker ethic and the maker culture have always been sort of tied up in figuring out how to use what's easily available to you to do something that maybe the people who design those bits didn't originally envision and while that's really really cool it doesn't always scale very well. And so in my experience the thing that people end up sort of stumbling with is how do you go from something that's easy to do in a one-off kind of way to something that you could replicate in a way that would be you know very broadly applicable and lots of people could take advantage of. In fact with the avionics stuff that Keith and I work on the thing that I think has caused us to be successful selling rocket avionics boards is that the stuff we build is actually pretty complicated to make physically. It's you know very fine took surface mount technology and that's not nearly as scary as a lot of folks think it is but to do them and sort of have them be reliable does take a certain amount of experimentation and expertise. And those circuit boards themselves I mean they're not it's got to the point where realistically you're not going to build those yourself you're going to send that away to a third-party company to manufacture. But even more importantly the average person playing around with high-power rocketry has something else that's their primary focus. For Keith and for me it's really rocketry these days is a lot about the electronics and the telemetry and the kind of data we can collect and the ways we can help other people particularly students being able to do cool things that they couldn't do otherwise but we have a lot of friends who are really good at making propellant and their thing is all about changing the chemistry of the propellant and they almost would like for somebody else to build the rocket and they certainly don't want to have to design and build circuit boards. And you've got the hobby flyers too wants to buy an airframe and buy the kit by the propellant from somebody else so you've got the makings of a complete community. Right so the the place that this gets really hard though so you know you talk about smart home things you're absolutely right that every major company out there that would like to play in that consumer space wants to figure out the thing that's going to cause their products to be sticky and desirable to customers. And a lot of times that means that if they had their preference they would build everything with some sort of proprietary local protocol and they would gateway to the internet at one place. There are a couple of working groups that the Linux Foundation now that are trying to fight against that and trying to build industry consensus on standardizing some of these protocols. And if you really care about that stuff at the sort of trying to shape the way the market works level those would be good places to go get involved and try and drive thinking and behavior. The problem of course is that that's it's sort of a macro industrial level right. And so what do you do yourself well I just finished building a greenhouse for my wife to replace the one we lost in a fire in 2013. And it's got a thermal control system and shockingly that's all done using open hardware and software bits. I did cheat a little bit it's not a hundred percent open because there is a raspberry pie in the middle of it and that still has one little binary blob I believe. But you know the control algorithm is about a page of Python that I cobbled together from bits and pieces. The sensors are simple things that are tacked on to GPIO lines the controls are a couple of cheap Chinese solid state relays that I bought through Amazon. And you know there's a fan and electrically operated louvers to open vents and a four kilowatt electric heater sitting in the middle of it. So and the really amusing part is I found some lovely temperature and you know humidity and environmental control plotting stuff that was written by somebody in a marijuana grow facility somewhere in Colorado. It's legal now and there's this whole sub community of people sort of encouraging each other to build more robust infrastructure for growing better marijuana. So I really was just amused when I found this cute little chunk of code for you know creating web plots of environmental data and that's where it came from. And of course since it is for greenhouse project there's been lots of joking around the family about what my wife's actually growing in there. There are plenty of people who are really good at that in Colorado and frankly I would much rather have components for salsa. So there are poblanos and serranos and all kinds of herbs and tomatoes and yeah it's a it's a it's a salsa greenhouse. So coming back to your the greenhouse. It's totally bespoke. No I bought Dallas DS 18 so look I'm I do electronics for fun and you know I used to do it for a living. And despite Keith's beliefs I am actually capable of writing software myself in kind of time. Well I did this time I'll even show you the code if you want to see it. It's the first time I've written I started with like an empty edit buffer and created like a whole page of Python myself it's kind of exciting. This is how you can tell I'm sort of an executive these days right. But no I mean it's fun because simple temperature sensors one way or temperature sensors somebody else had written a library that talks to those sensors on the pie that was easy. The solid state realized come on people one wiggling a GPIO line how hard can it be. It turns out in Python you know it's another library so whatever but it all works just brilliantly. And it took me in a one afternoon to go from getting hardware all sort of hooked together. And by the way if you're interested yes of course like all of my things that's an open hardware design and this I do actually have a schematic for what I wired up around the pie and I have not bothered pushing it out to my web yet because I was busy off playing with my parents and a good friend doing a rocket start last weekend instead. So this fits in really nicely with what we're talking on Saturday and it came down to that things need to be appropriate for what you're doing like sometimes really simple things with just a page of Python code is really all you need and I think that my wife is so happy with the fact that the greenhouse control system is just working. Look it's not complicated code and I say that in part because I used to write you know factory floor automation code and things like this back in the days of my misspent youth but it really it's not complicated. And if anyone here who's ever looked at Python before would look at and probably laugh at my code but you would immediately see just how simple the problem is we're trying to solve. So yes I think on one level what I would love to see is lots more people encouraged to not get hung up about who they have to go buy stuff from and to realize that some of the things you want to do for yourself you really can do for yourself by yourself and it's actually a lot of fun hacking around with little bits of hardware and software like this. Yeah I mean let the budget police work out. But at the other end of the spectrum I think it's really important for us to realize that there is this huge mechanism involved in turning good ideas into mass produced hardware and the support that's required for someone like you know my mother or your grandmother or whoever to be able to call on Sunday night and get help figuring out why the thing doesn't work and those are the sorts of things that cause you to have to stop and ponder where the threshold is between the stuff that you really want and need to have open versus the stuff that you're willing to let somebody have some mechanism for generating some kind of a revenue stream. So in a team discussion over the year we came up with four things about what it means to be open. This was also after many years. So you know I would like some feedback. It got a bit heated. In more ways than one. So what does it mean to be open? It means to be accessible to be hackable to give a how to manual the hardware to say what they did and how the user could do the same thing. So basically explaining how to. No we were talking about in an absolute perfect ideal world. So in your ideal world if you if you've got something you want to start there's enough information somebody else can pick it up and learn. My answer to that is actually yes. To be open you have to have some kind of documentation even if all it is is an assertion of which open standards you're compliant with or something like that. And see from my perspective schematics and source code are documentation. So yeah. So can we get to schematics. Can I do my point. Just to finish my question about the idea that you can at least try to reproduce something doesn't matter how hard it is. You at least have the ability to get in there and try and to share your design and let other people comment on that design. So reproducibility is certainly one potential attribute of open but it's not actually the attribute that most people care about most of the time. Because most people don't actually want to have to build the thing themselves when it comes to physical hardware. What most people want is the knowledge the sure knowledge the assertion the guarantee that if they spend money buying a piece of hardware that somebody else doesn't have the ability to control when it becomes obsolete. And you know in a perfect world you'd buy a piece of hardware that had a completely open design whose documentation came with schematics and builds material and things like this. You know like Hewlett Packard oscilloscopes used to back in the 1970s 80s. Complete theory of operation in the back so that you could actually repair your own instrumentation. That's a lovely point in history. I'm fortunate to have been around then. The what they really care about is that you know somebody else going out of business doesn't all of a sudden make their home control system stop working. I'll be honest with you the things I worry more about are the the the the recent proliferation of devices that depend upon some piece of cloud infrastructure. That just might just go away. It's like oh well that's really cute and it does this really cool thing and I don't have to put any infrastructure together. I just buy six of these and put them down and hey look and I can point my phone at some random website. Now I can see what's going on in my house. Yeah. I mean never mind the security implications of all of that. OK. And we can talk about that in this group for the whole week I'm sure. But just think about what happens if that company goes belly up and that website goes away. It's not even when the company goes belly up. I mean we've already seen examples of I correct me if I'm wrong. I think it was one of the Google control systems where they no longer support the MOT1 API. So people who have gone and put their heating system their lighting controllers and bits of pieces in the hardware works absolutely fine but the cloud services have gone. And so that is where a standard and industry standard could come in place to secure it. Yeah. Open standards really help in this area because you could then change your cloud infrastructure something else. So again this is where to me the definition of open is complicated because it sort of depends on what you're trying to accomplish. A lot of times what I really care about is making sure that if I think something is cool enough that I'm willing to put some of my hard earned money down that it's not going to somehow become unmaintainable or unsupportable later. That to me is one of the biggest freedom. And in fact this is the thing for those of you who weren't there haven't heard about it. I gave a talk at was it this year. Yes. It was this year at Linux conference Australia about the whole house audio system I built for a new house. And you know I my contractor we're building the house. We told him we wanted him distributed high quality audio through that sort of background audio stuff. And you know music in whichever room you wanted it kind of thing. And they expect a really cool looking system fairly high end ridiculously expensive but irrelevant for this part of the discussion. The feature set was pretty cool. You know it had apps for you know your iOS or Android device order you can use to control it and all this sort of stuff. The problem was being the kind of guy I am. I went and did a little quick due diligence on the manufacturer's website and I couldn't find any documentation about what the con software contents of the system were. I found sort of an owner's manual operators manual. And then I went to the support pages on the website and buried down in there were a couple of screenshots of what the administrative interface to the box looked like. And there were terms in use on the screen that were unique to the way you can figure network filtering in the Linux kernel. So does that come back to your schematics conversation. And so when I looked at this I went oh this this is like a 99.99 something percent probability there's a Linux kernel inside this thing. There's not a single mention the GPL anywhere. Now in theory you know until I actually buy a box from them they're not obligated to provide me anything under the GPL right. So in theory I could maybe have spent a lot of money and bought the product and when it arrived maybe it would have come with source code maybe it would have come with documentation of a written offer or something like this. But my expectation based on the way the rest of the material is crafted is I would spend a lot of money and be really unhappy at the end of the day. Both because it appeared to me to be a blatant GPL violation and because they were wrapping it all up as a proprietary thing and not documenting any of this or telling me about the guts which I might like to play with. So of course I rebelled. I ended up designing a cute little board that's sort of USB to 30 watt stereo audio and I built up nine of them and mounted them on a strip and put them up on the wall and bought a cheap Chinese power supply which by the way required I buy an equivalently expensive ball bearing fan to make power supply possible to live with at night. I wish you know another nickel spent on the bearings in the fan would have saved me a lot of hassle. And there's actually a little Intel what's the thing called it's one and I don't know what it's called it's the it's one of their little single board computers that uses anatomy that's not the name of it but whatever it's a little Intel architecture based embedded board it's just sitting there running Debian and I'll hack that version of Mopody and my wife's really really happy but here's an example of where even before I purchased it it occurred to me to care about these things and I went and looked and it was not hard for me to you know I didn't get absolute proof that there was a problem but I certainly wasn't left with a warm fuzzy to spend $12,000 on their hardware and in fact I think the hardware cost of my solution even if I'd had to buy the little Intel board would have been on the order of like 200 bucks so yeah but then you had to assemble it and stuff your time yes exactly and for me of course this is this is the hobby part of it right this is why I was saying earlier this is the thing that sort of differentiates those of us here I mean all of us enjoy hacking on stuff or we wouldn't be involved in Debian on some level and I suspect most of you in this room that the little audio boards yes there were some serious electronic design part of that and I will take credit for the fact that that's not something anybody could just go do trivially but there are lots of projects like this where you really could go do a little bit of it but there's a huge distance between I can do this once and make it work for me and what does it take to actually deliver something in industrial volumes Keith and I struggle figuring out how to get an economical production run done of a tiny little thing at thousand unit quantities that just the logistics I mean you get done with the design you have a working prototype and then you have to do the other 90% of the work oh yeah I mean that's my day job so I'm completely aware of that and and so on some level I have sympathy for people who want to find some way to hook you to ensure that once you they spend a lot of money investing and developing all that that they have some way to recoup that investment but that's something that they does not extend as far as you can lock me out of the ability to have control over the life cycle of that in my own use and to me that's the really important thing is do I or does someone else control what the life cycle is this is going to be so you one of the things we said was documentation we said you know a schematic is a good start point and knowledge of the api's is a really good start point but working on an open on an open hardware project do you just give the schematic and say that that's it as a team as your project we just give the schematic away and that's enough or do we need to give more than just the schematic I mean the number of projects that I've seen which are allegedly open and the schematic is a pdf it's yeah I've got a circuit diagram um they might give away they might give away some gerbers if you're if your purpose is to ensure that someone else can write code that works with that hardware then a schematic might suffice yeah on the other hand I suspect that even if you had the schematics of a tivo you still would need the keys to to cryptographically sign the kernel before the service would actually function for you and so it really again it kind of depends on what you need uh Keith and I make uh schematics and gerbers and drill files and build material stuff available for all of our designs and we know that that's sufficient because we've had people come up to us at rocket launches with clones of our products asking us for some little sheepishly asking if we would be willing to help them with the one little thing they didn't quite understand let's just get somebody's coming up to you they've bought a clone of a board that you've designed no they've they've made it themselves oh they've made it themselves that's okay that's they downloaded the gerbers and the drill files they sent them off to a board shop they had raw boards made they downloaded our bill of materials which has all of the distributor part numbers and everything and digikey and mauser part numbers in the US and they placed orders and bought the parts and they sat down and placed them all themselves and refloated in an electric skillet or something and they turned it on and it kind of seemed to work and they don't understand why they can't hear the rf and it's just because they didn't know they need to calibrate the frequency but that's just the grin on your face when somebody comes up with yeah so as Keith was saying earlier I don't think the first person who did that to us was expecting our response which was immense glee we were so ecstatic that somebody had actually like built a clone of one of the things because it was proof positive that our assertions that we're open hardware guys it was true so it's not just chuck it on the internet on your website here's my here's my um look my files and nobody looks at it people are actually going out there they're searching this one when the the website uh when the search engine wants to crawl up crawl your site I don't know I don't know in this case if it's as much that or as much people at long distance you see somebody else playing go that's really cool and they go look and they go wow this is open I've always wanted to learn how to like place parts and reflow surface mount myself here's a known working design I can start from and it'll be something really cool to play with when I'm done I don't know exactly what the motivations are should we take some questions comments from the audience yeah it'd be much more interesting for you I think Sam's got the first one so hang on we've got microphone can we take the microphone hi so it's on okay um so Sam Hartman I guess you know this is all great that you hear this this wonderful discussion of open hardware I'm a software guy I would love to have control of the life cycle stuff like bedel's talking about but like even really hooking something up to a GPIO line is actually a little bit more than I'm going to easily be able to do that's okay I can't call white kernel code yeah no right but how do I like how would I go about let's say I wanted to have a you know home audio system that would be really actually totally awesome I'm happy to handle the software side of it what are my options for getting something that meets bedel's open life cycle requirement wow so Keith and I were talking on a flight down about whether I should build and sell some of my audio boards so yeah it's get it find find that community find the people who are already doing it and look at what they're doing so I think the challenge is when you go looking for components to put together to create a system you ought to be asking these questions about all the components you're picking and by components you know I could be talking about you know a server I could be talking about you know some integrated sensor box or something and and there's different levels of openness and this is what I think I was trying to get to earlier really depends on what you're trying to do for some people knowing that I can put this on the net and it speaks IP and if I ask it a query in this particular format I'll get the temperature and the humidity and the barometric pressure and all of that back is open enough they don't actually care about the schematic what they want to know is that they can talk to that box with their own computing infrastructure and Sam you know that might be the sort of threshold that that meets what you're talking about it in other places in the spectrum there are people you know going back to our rocket avionics stuff Keith and I have been totally thrilled that there are high school teams and college teams that are taking our designs and hacking the firmware and adding additional sensors and you know part of the cool thing about being totally open like this and willing to talk to people about what we've done is we're enabling other people to then sort of explicitly go build on top of that and sort of use our stuff as the big piece at the middle of the erector set or the the LEGO set or whatever and and that's just immensely cool it's immensely gratifying and you know at this age and stage in my life seeing people learn from what we've done is almost as gratifying as anything else we do so for some what we're suggesting then find a community of people are already doing much of this do a little research before you whip your wallet out yeah before spending money you know make conscious decisions about the stuff you're buying if you buy it because it's shiny and you know I guess you get what you get yeah in two weeks time it won't be available anymore so yeah hard to say anybody else I used to love to go to ham radio flea markets when I was younger because it was really exciting to find some sort of one-off random piece of stuff and take it home and sort of reverse engineer it figure it out and do something really cool with it yeah and I have to admit that at this later stage in my life I don't find that nearly as exciting as I do figuring out how to create something that satisfies one of my own personal needs and desires and enables other people to be able to do something kind of find the thrill of taking something I myself have done some ham radio stuff and I do find that the technology has moved so far forwards now that it's getting harder and harder to understand uh radio you buy because everything ends up as an soc and it's everything's got higher and higher level of integration and it's getting harder to um produce or hack on something because it is now just so I would I would actually argue that with you um it's absolutely true that my I have I have friends in Cuba that years ago um when they came to a conference in the US I sent them home with a tube full of mini circuits doubly balanced mixers and some oscillator chips and things like this and I knew that every sort of set of rough components I sent home would get one more person on the 50 ish megahertz 6 meter band and that was a totally cool thing to do but at the other end of the spectrum I have a bunch of friends in the ham radio world who all just went in on the crowd supply lime sdr software defined radio boards I mean that's that's a very complex piece of kit but it is completely documented schematics artwork all that stuff is has been released and everyone I know that's buying one of those and wants to hack on it is interested in writing radio code you know software and they're going to hook a couple antennas up to it and that's a building block that they're using and the fact that most of them don't actually really care about the details of the hardware or what it takes to lay out a circuit board so that it's got fidelity from 100 kilohertz to 3.8 gigahertz that this again it's sort of what do you care about what happens so if so I think we could see the same thing from different viewpoints is if you want to hack on something and change its functionality from a software point of view the the higher integration is fine providing you've got access to it right if you're wanting to hack on something from purely as a hardware to experiment to play and actually understand the hardware that's getting harder on off-the-shelf products because they are more and more integrated you know if you're wanting something for longevity don't buy an SOC but it's interesting because on the flip side the availability of well-documented SOCs is now allowing makers to build things that they couldn't have ever built before yeah yeah yeah so I completely hear what you're saying and it's very interesting that after the fire when I went to rebuild my RF test bench I found myself buying test equipment mostly that was designed in the 1980s and the reason is that was about as recent as you can go when things are still really general purpose I mean I bought a vector network analyzer and I bought a spectrum analyzer that just allow you to look at the RF yeah since then the thing that's become really cool is to put more and more levels of abstraction and sophistication on top of that so now if you go buy an RF analyzer you sort of need to go in knowing am I going to use this for for looking at gprs am I going to use this for you know what what what radio thing am I trying to do and what protocols do I care about because what the vendors want to sell you is something that's optimized for helping you understand a deeper slice through a taller stack and also and also the thing that you're buying is you can't actually just look at the spectrum from DC to 26 I think you're also buying is a is a is a broad range sensor yes and it gets straight into the digital main and it's processed in software on the flip side today you can go to somewhere like seed studio and buy a little digital signal oscilloscope or a digital sampling oscilloscope or a little logic analyzer board for 50 bucks or something that will allow you to diagnose and troubleshoot things that I could not have afforded the test equipment for in 1986 so it goes both ways yeah great so closing statement where do you think hardware is going open hardware is going in the next couple of years or anything else you would like to close the session well I'm really excited that they're now a number of companies in the world that are making a point of being open hardware companies that are north of 10 million a year in revenue I either like becoming real companies I'm personally really pleased to be on the board of directors of aleph objects which is the parent company of lullspot who provided the 3d printer stuff that you see out in the hall in fact my son is working there this summer so I have a fair amount of connection with them and by the way I had nothing to do with the fact that they sent hardware that's immensely cool and done on their own part but I think we are now seeing enough examples of real companies that are being really successful following a truly open model to believe that they're not just random unicorns that this is a structural model that can work but I think we're beginning to see it in the commercial space but I think so far we've only seen it work in specialty kinds of product spaces it would be really nice to think it's going to ripple out more you know but when I can go into a retail outlet and have my choice of open design wi-fi access points or open design home automation pieces then I'll be a believer right now I think there's a whole bunch of hurdles left before the industry at large understands that they can be more open and still make money and from the hobbyist point of view where are we now well I personally you know everything I hack on at home is open I just life's too short for secret source great Andy last statement from you it's it's I pretty much agree entirely we'd be there on this perhaps I'm still seeing from the the hacking side of from individuality that we're you know open hardware is I still think 20 years behind the open software movement we're still in the shareware stage we're just beginning to come out of shareware start to see collaborative projects but finding them is still few and far between I will work on it because the other thing I would think about is that a lot of things that used to be perceived as really difficult complex manufacturing processes are now available in job shop form not only are there places to make spaces and that type yeah but I mean there are companies now where you can upload a design file and get back laser cut or 3d printed or cnc machined parts and one-offs it costs some money but I mean you can design that is true of pcps as well I mean there's plenty of people doing circuit boards if you want to get raw circuit boards if you want to give somebody build materials and have them do surface men assembly so by the way there's job opportunities on both sides of that you know you can find another niche that somebody's not filling and go provide a service like that become an expert on something and on the front side it's actually more possible to hack than it used to be I think we've got one last question yeah one two questions do we need io2 look look I mean you get to decide what level on a universal platform I don't know that I care so much about our landing on a singular one of anything as much as I care about the things that we end up adopting on mass having some degree of openness I mean my biggest concern with iot isn't necessarily even openness it's the the massive commercial pressures to have your sensors and actuators so cheap that any semblance of security goes out the window and people who are peddling those sort of sensors and it's becoming to be included in clinical you know health care at home do not care about any security of any type the platform is so cheap you can't actually put it on there it's being thought of too late I guess you get these platforms that are developed within big corporations or big vendors but then are we going to is it going to be a big vendor that comes up with the platform that everybody ends up using or what what is it I mean what is the space for an open source project to or a free software project to guide so there are at least two working groups at the Linux Foundation that are tackling open protocols for iot space and they have a long list of corporate participants I would strongly suggest you go take a look at those projects and see what's happening because those are places where people who are passionate about it and have some knowledge and interest can actually help define the standards that industries will end up adopting great and on that point yeah I hope that this creates a lot more conversation in the corridors and throughout the week yeah if I'm not too busy we'll talk over and I want to thank you very much for joining me on this panel great thank you thank you