Colin is back, officially.
Colin is back, officially.
There are a few interesting points in that post, I am very interested to see how they address the footprint and performance. Thanks for sharing!
Great news indeed!
Great days ahead ;D
OK no excuses anymore, makers gotta make so go forth and build .NetMF coolness everywhere.
Translation: Microsoft’s IoT strategy is Azure, not NETMF. NETMF is useful as a driver for Azure.
Azure is where Microsoft can make money from IoT. If NETMF helps drive Azure revenue, it will survive and thrive. Without Azure as a potential revenue source, NETMF would likely continue to dwindle.
While I’d love to live in an ideal world where Microsoft continues to invest in technologies simply because they’re awesome, I’m content to live in a world where Microsoft continues to invest in NETMF because it may drive Azure revenue.
So, from my perspective, Azure being the reason for NETMF being useful to Microsoft is a feature, not a bug.
We’ll call it a misfeature, then.
I’m curious…if you were CEO of Microsoft, how would you justify investing in NETMF?
I don’t have any specific inside info from my MS days, but my guess is that there’s little if any revenue to be had from licensing NETMF, so apart from keeping a foot in the embedded world (which has some value, clearly), how do you justify the cost of continued development, marketing, maintenance, etc.?
Not trying to be argumentative here, I’m actually interested in why you object to the idea that Azure is a revenue tie-in here, and how you would do it if you were in charge. Perhaps there’s something I’m missing.
@ godefroi - Lets look at the IoT ecosystem at the .NetMF level, who is going to make money at this level, and the answer is no one as pretty much everything here is open source, so the money will have to be made elsewhere like Azure. Now of course the various players are going to feature their solution as the best and if you don’t play at this level you won’t get your money maker in front of the developers and hence money paying users.
The hardware dudes are going to make money of course but the game is far more complex and bigger at the software level. I will say that a number of hardware guys are going to find themselves out of their element however as they have served the hobbyist market for so long they have no idea of real marketing and are going to end up as road kill in the ditch of IoT.
I wouldn’t justify investment. For Microsoft, NETMF is charity.
To say that nobody’s making money on IoT, however, is plain silly (if that’s what you were implying, @ Duke Nukem). A lot of people are making truckloads of money.
@ godefroi - Some people are making money on IoT, far more will later, but the question was who is making money at the .NetMF level (ie at the software platform)? Arduino or who? Pretty much everything at this level is open source, non-profit, etc so there isn’t money to be made at this level of IoT software food chain, so the money will need to be further up the IoT food chain like Azure. Now if you solution doesn’t include this level of the ecosystem you lose getting your money maker like Azure in front of the developers and hence lose users or the chance to catch the scale up, so while .NetMF might look and feel like a charity investment by Microsoft (who do in fact have a number of true charity investments), .NetMF is actually an important component in their IoT strategy and what is great is Microsoft has finally figured that out and is making the commitment it should have made years ago.
Let me break this down a little further just to be clear here:
Cost to develop an Arduino LED blinking solution would be the motherboard at $25 for a Uno, no cost for the compiler
Cost to develop a .NetMF Gadgeteer LED blinking solution would be $25 for a Cerberus, no cost for the compiler (Visual Studio Express)
So where is the money for anyone other then the hardware vendor and really where is IoT as these are both single non-connected devices. To me IoT is software and hardware, its the analysis, data mining, communications, etc and the role hardware plays in this is the cost of hardware, storage and connectivity have finally dropped to the point that we think we can place enough sensors etc out there to collect enough data to squeeze enough value out of to justify the costs of IoT, but its the software that squeezes the value out of this data, which is why cloud services are lining up to be part of IoT as that is where the money is for software or as they are called now ‘services’.
Of course every capitalist on the planet is trying to figure out what you are willing to pay for so they can create and sell it to you. What would you be willing to buy at the sensor level of IoT besides hardware? For prototyping I’m happy to buy pre-assembled modules with drivers and other hardware components, but on the soft side I’m pretty much happy with what I have now, so where is the money for those guys? Once I have my product built things like Azure, ThingSpeak, OpenSen.se, etc can provide me with a service much like many IoT companies are trying to sell to their clients, a pay as I use service which can look pretty attractive as it removes startup barriers (ie pretty cheap to use initially but as I make money they make money), but unless your development tools make it easy to use, you will likely look for another option. So .NetMF is important as it will be easy to use Azure with .NetMF, it will have a full scalability option which again will be pretty easy to use and frankly the benefit of community as a lot of the development features I’ll need are ready to go now or will be ready to go with large community guidance. Now not to say the community knows what is best all the time as HTML and XML for example are bloated pigs which might not play nice with the network at big number sensor levels, but other more suitable solutions are already in the fire.
A big question is where are the Arduino guys going to go for their services Microsoft, Google, Amazon, ThingSpeak etc or are they going to rely on developing their own inhouse (hello barrier to entry)?
But Azure is built on top of web APIs, to get there you need a great network stack, isn’t that one of the things everyone has been asking for? And if you can call one API you can call any API. Sounds like a win to me.
Hey, I’m not saying it’s not a win, and it’s definitely great that MS is at least talking about NETMF again, but let’s be honest, it’s still pretty much just talk.
We need big changes to be competitive (did you see mBed’s new mBed OS plan?), and we need them quickly.
Completely agree. MS doesn’t get anything but the enterprise. Time will tell, but I’m hopeful.
IoT = Internet of Things
I = Azure
T = NETMF, Windows, and third-party devices.
Our strategy is all of the above. Don’t confuse strategic with “revenue generator”. They are not equivalent in a Satya Microsoft.
FWIW, the IoT group’s charter specifically calls out education and makers in addition to enterprise.
That’s part of it. But if it was the whole story, we’d just supply Azure libraries for Arduino and R Pi and be done with it. Instead, we support those as part of the ecosystem.
There’s a real good story when you have devices that can scale the entire range.
There’s also a realization by those in power, that if you do not have low-cost maker-friendly boards/devices/OS/etc. you lose the education market. Without education, you lose future makers and entrepreneurs. Your stuff isn’t the basis of crowd-funded products. You are simply not relevant.
Irrelevancy bleeds between verticals. If kids don’t see you in IoT, they’re less likely to consider you in other areas as well. Same with makers.
My take on it, anyway