Slashdot Mirror
.NET Gadgeteer — Microsoft's Arduino Killer?
mikejuk writes ".NET Gadgeteer is a new open source platform, from Microsoft Research, based on the use of the .NET Micro Framework. It brings with it lots of hardware modules that are backed by object oriented software. You simply buy the modules you need — switches, GPS, WiFi etc — that you need and plug them together. The software, based on C#, is also open source, and comes with classes that let you use the modules without having to go 'low level.' Is this a competitor for the Arduino?"
Can anyone think of any example when a [fill-in-the-blank-popular-or-niche-object-of-consumption] killer has ever killed a [fill-in-the-blank]? Calling something a [fill-in-the-blank] killer seems to admit at the outset that the market belongs to [fill-in-the-blank].
The type of person who cares about open anything is the same type who will avoid anything with a Microsoft logo. That alone will kill any potential this platform has.
The Ribbon interface was a productivity killer, if that counts.
Simple reason: The base board looks like it needs connectors best I can tell and costs 4x as much as an arduino board. Plus I'm sure the MS board requires windows. I have an arduino because I can interface with it on different platforms and it didn't cost a ton to get into.
"arduino killer" is not Microsoft's term. They call it a ".NET gadgeteer" or something.
I love that some blogger calls it a "arduino killer" and all of a sudden, "Microsoft's trying to kill the cute little arduino".
Arduino is cool as hell. My daughter and I have been having a blast with a couple of them that we bought just to goof off with.
The .NET Gadgeteer also looks pretty cool, though I don't know much .NET framework. Oh well, I'll let my kid learn that stuff. I'm not that interested, but I don't see any reason why we should find anything negative in this gadgeteer thing from MS.
You know which very rich and successful and famous high-tech company is NOT making an open platform for us to play with?
Seriously, go back ten years, twenty. Now ask yourself which company would come out with something like this Gadgeteer first, Apple or Microsoft. Which company would lock up its handhelds behind a walled garden. Which company would stash profits for a war chest to buy its competitors instead of paying its shareholders a dividend. Sometimes things don't go the way you would suspect.
You are welcome on my lawn.
But what's the power consumption on that? Arduino became popular not just because of the cost, but because of the power consumption and ease of use as well. $120 for something that includes all sorts of stuff that I might not need is hardly a good deal.
The Kinect was intended as a Wii killer and did quite well, but that was only because of Microsoft embracing the maker community that found brilliant alternative uses for it. In the end it was far more effective when separated from the awful excuse for a console platform it was designed for.
So Iguess the answer is no.
I don't think so. The success of the Arduino revolves around how easy it is to get it to reach out and touch something using all the pin interfaces (turn on a LED, control a motor, use PCM to regulate your barbeque temperature. Now there does seem to be a small overlap, it's can't complete with the arduino over it's entire range of applications.
I wont say this will kill anything, but it sounds fun. I'm betting this works with kinect soon. And, as a C# developer (along with many other langs) this sounds like a quick way to start projects to teach my kids robotics. Currently I still lean towards Mindstorm, but options are always good.
Where genius and insanity become confused true wisdom is found
Was that a particularly difficult task to accomplish in the first place?
This is my signature. There are many like it, but this one is mine.
I don't think this will be an Arduino killer. Arduino has too big a lead, and too much traction in the DIY, hacker, and arts communities. But it will appeal to companies that do software and are looking to break into embedded hardware. They're already familiar with .NET, C#, and Visual Studio, and they won't mind paying a premium for the hardware, because it's Microsoft-backed and because they already know the dev tools.
It might also find a home in the industrial space. Lots of manufacturing facilities have bright people who program PLC's and the like, and are quite capable of learning the tools and building simple stuff that can round out a company's automation efforts.
I don't love Microsoft, but kudos to them for branching out creatively in an effort to shore up their sagging fortunes.
'The Economy' is a giant Ponzi scheme whose most pitiable suckers are the youngest among us and the yet-unborn.
Creating a devices that come with software bits that enable the control and interfacing between your programs and the hardware devices which more easily enables you to make the devices do useful things!! I can't believe no one has thought of this before!! Next thing you know, someone will tie all of these software-hardware interface modules (let's call them drivers for short) into a bundle that lets them share a common pool of resources such as processor time and memory more efficiently... "in an object oriented way" of course because unless it was done with "objects" it couldn't be new or novel could it?
I know -- laugh me right off of slashdot right here and now. The second programming language I learned was assembly language for 8 and 16 bit Motorola processors. (The first was BASIC.) Assembly language taught me how computers really work. All this "object oriented" crap is just an abstraction that helps people see programming in ways that aren't even natural to the machines and how they run. I prefer to see things as they are, not how I want them to be. It's all still bits, bytes, signals, registers, inputs and outputs. Sure, you can imagine it's all some sort of flowers and sunshine microcosm where causes and effects happen magically just like they do in the real world (hint: the real world isn't magical either and there ain't no god that makes it all work either). It's all a bunch of tiny, tiny operations that accomplish bigger things. (Hell, for that matter, even the complex instructions in today's processors are really just simple instructions running on smaller processors... anyone ever wonder what "microcode" is and why/how we always see that "microcode update" line in the booting of Linux? Yeah... I knew and have known for a long time... long enough to chuckle to myself when the "RISC vs CISC" debates were going on. (Hint: It's ALL "RISC" now even if it wasn't in the early 8/16 bit days... it would have been horribly more difficult to scale processors to the size and performance we see today without building it all out on RISC elements.)
It has all been done before and they will continue re-doing it again and again because new ideas are truly rare. I wonder how long it will be before we see an " -- over the internet" or an " -- on a handheld device" versions of these same things.
Productivity was a Ribbon interface killer?
First: MS. Hobbyists, especially the microcontroller crowd, are usually aiming for independence, interconnectivity and freedom of choices. Most microtinkerers I know were even shy to touch the Arduino because it came along with its own development tools that smelled like "you need them to do anything with it". Only after reading the specs, seeing the PCB around the chip and noticing that it is pretty much simply a (rather well designed) pimped out devboard, essentially a "standardized breadboard plus programmer", they started to use it. Many I know still refuse to use the compiler that came with it and stick with AVR Studio or GCC. Some even consider that "too far from the metal" and stick with ASM, personally I think one can overdo his zeal for independence and "feeling your controller", but I'm not judging them. Case in point, microdevs hate being locked into something. Despite the perpetual ATMEL vs PIC battle (and the self-chosen lock-in with either platform, since few people I know really want to work with both).
Second: Microcontrollers are still very, very tiny in their specs. The average affordable model measures their clock in the Megahertz and their flash rom (program memory) in the kilobytes. And for that a .net platform? Are you kidding? Now, I might be prejudiced in this matter, but unless they somhow then turn that .net program into very tight assembler, the 72MHz Arm will feel like a 8MHz Atmel. Now, that Arm implementation MS is offering has 4500kB of flash. Pretty much, considering most AVRs still measure their flash ram in the single and double digit kilobytes. But will that .net compiler spit out native code? Or will a good deal of those 4.5MB be taken up by some virtual machine that then tries to run the object code? Essentially the question is, how much "work" can you push into the flash, how many instructions can you possibly put into it before you're running out of space?
And finally: As a extension from the first point, MC developers love to tinker and toy with their gadgets. And they love expanding on them. Having a wide selection of addons is nice, but how easy is it to roll your own? In case I do not want that Ethernet expansion, can I make my own? Are the specs known? What about the legal shit, can I publish what I create without paying MS for it?
I'd be wary to take the information provided at face value. 72MHz look far more than the measly 20-48MHz Arduino offers (depending on the board you choose). And 4.5MB certainly is far more than 128KB of flash rom. The key question is, though, how much of that rom is usable, how do the processors perform in comparison, and how easy is it to roll your own expansions.
We used to have a Bill of Rights. Now, with the rights gone, all we have left is the bill.
Look at the specs. Arduino's "beefy" MCU is 16 MHz, 8 bits. This is 72 MHz, 32 bits. Arduino draws a sub-10 uA sleep current. This thing draws a 40 mA (yes, milliamp) sleep current. They're completely different devices targeting completely different markets. Talk of "killing" Arduino is just meant to draw eyeballs and clicks.
I'm, I guess, what you'd call a professional arduino programmer. been working exclusively for the past 2 years or so on a combo hardware/firmware project. I did the hardware design, proto testing and driver+apps coding.
and so, I'm pretty familiar with hardware and operational costs of getting arduinos up and running from scratch (we did our own board called the LCDuino-1). when I look at the stuff mentioned in the article, I see more zeroes behind non-zeros than should be there. just too expensive for an 'arduino killer' platform or system.
I'm not against this *because* its MS; and I'm willing to consider other alternatives, but this does not at all look cost effective, just from the pure cost of boards and parts and cables POV.
otoh, the ARM systems (so-called 'plug computers' from marvell) are the next open-source and viable step up from arduino-land. those seagate dockstars and pogoplugs are common examples of those (and nice and hackable, too). with those, you get a full debian linux stack in there; not some mickeymouse 'other thing' inside that you now have to learn and deal with (and debug).
for me, its arduino for the extreme low end; and small/tiny/fanless linux boxes for the "$50 and up" kind of range. they even mix well; if you need a tcp/ip presence, use one of those dockstars and have full ip-tables and all that neat protective stuff there; then have a serial link to the arduino for when it needs to report 'back up'. great way to add remote web control safely to the arduino realtime systems.
--
"It is now safe to switch off your computer."
Does it work from linux?
All of the above was encrypted with a Quad ROT-13 method. Unauthorized decryption is in violation of the DMCA.
16MB - sure, but .NET isn't the most compact code in the world. Nor is the framework - even the "compact" framework sucks up several megs.
My Other Computer Is A Data General Nova III.
14 .NET Gadgeteer compatible sockets
And how about compatibility to something I dream up? Can I attach whatever I wish to it, and still continue to develop in that comfy .net environment?
Arduino's main appeal to the microcontroller hobbyist crowd is that it offers simple access to AVRs without limiting you. Meaning, you basically get an environment that lets you use the microcontroller as if you didn't have it embedded in the Arduino platform if you so desire, but allows you to use it if you so please. How does Gadgeteer fare in comparison?
What microcontroller is it, anyway? I can't find that information. It's an ARM7 CPU, ok, but is it a microcontroller at all? Or just the CPU and some MS-invented design around it?
I might be extra wary when something has an MS label attached, but let me reiterate that: Arduino's appeal stems for no small reason from its openness. It's, in its bare bone, only a PCB that exposes the AVRs pins in a standardized layout. Nothing more, nothing less. You can, when you're fed up with the training wheels that their development environment is, simply hack them off and use it as a simple AVR with a PCB around that exposes the pins in a standardized layout. The crucial question is: Can you do the same with Gadgeteer?
We used to have a Bill of Rights. Now, with the rights gone, all we have left is the bill.
I think Sega killed the Dreamcast by releasing so many consoles and console mods in such a short time. SegaCD, Sega 32x, Sega Saturn, and the Sega Dreamcast in less than a decade.
This is my signature. There are many like it, but this one is mine.
I went to a BSD conference and didn't see either one. Those stupid geeks were even running open source BSD on laptops with wireless and multimedia; anyone with proper upbringing knows you don't do that! I suppose the 30 million GNU/Linux users haven't bought their macs yet, but soon will. And experts predict IOS will dominate the mobile market, right? Oh no wait, it's to be two-thirds Linux by 2013
So it's real FOSS. Cool. No matter how much corporate MS sucks, MS Research is usually great and their use of AL2 instead of some "Shared Source" license makes Gadgeteer fully Free. One could even port it away from .NET and MS could do nothing about it. AL2 even includes a royalty-free patent license.
For those prices, why wouldn't I just get a gumstix and run Mono on it? The gumstix boards also host ARM CPUs that are clocked 10x faster.
Higher Logics: where programming meets science.
TL;DR: "You kids get off my lawn"
May apologies, but you are on the wrong side of history. In the 50's, there were "old guard" programmers who wanted to program in octal instead of assembly so they could really understand what the computer was doing. In the 60's, the "old guard" fought COBOL and FORTRAN in favor of assembly so "they could understand what the computer was doing". In then 70's, they fought virtual memory because "only with real memory could you understand what the computer was doing". In the 80's, they fought SQL and wanted to keep COBOL so "they could understand what the computer was really doing". In the 90's they fought GUIs because "only with a command line could you really understand what the computer was doing". And in the last decade, they fought bytecode and interpreted languages because "only with a compiled language can you really understand what the computer is doing".
This is not to say that every proposed new language and concept is good -- they aren't. There was an research computer where the compiler was in hardware (yes, individual gates and thing to parse your source code), along with the entire OS. There have been visual languages by the dozen; almost all were losers.
But, overall, history isn't on your side. The higher level languages and abstractions actual make people more productive programmers. Both Java and .NET have been accepted as "good" by an enormous number of working programmers and their hard-nosed managers; they are here to stay.
Want a sig like mine? Join ACM's SigSig today!
"LED module ($15)"
fuck off
They use a common port connection, so that you can use any cable to talk to any board. Sometimes it's overkill - sometimes it's not. But it simplifies connectivity with a common connector.
Browsing at +1 - no ACs, I ignore their posts. So refreshing!
If there is a part store conveniently located near you, then by all means go there. Most people no longer have that luxury, and are at the wits of the free market.
Meh, for $120 this thing is probably in trouble anyways. Though that's a shame, more stuff is always better.
.net koolaid". The Netduino has been around for a long time. And it's $35.
And lets remember, there's already an arduino for "people who've drunk the
http://netduino.com/ http://www.sparkfun.com/products/10107
Going by above statements, you probably don't know anything about the Arduino except the the name. The Arduino is a simple cheap and very limited 8-bit micro-controller, where important part is the limited costs (below $30 to get started) an easy way to hook it up to the computer and a moderately useful development environment where even hobbyists can switch things easily. The big limitations of the Arduino is very little RAM and very little processing power.
This platform by Microsoft doesn't even play in the same league as the Arduino. It costs 5 times as much and can do many things the Arduino even can't dream about, things like handling audio or video data. This product should be seen more as a contender against the small ARM platforms, Propeller and similar. And then, it's similar priced than those.
Will it succeed? I hope so.
Will it push out the Propeller or some ARM products? Perhaps
Will it push out the Arduino? Definitively not.
Seeed Studios makes Grove System, a connectorized bunch of input and output parts that plug in to an Arduino shield, and also sells Arduinos, Netduinos, Zigbee things, and *duino clones. Grove is $39 for the basic set of ~10 things (and has various extra frobs available.) (And realistically, you'll still want to get a breadboard and bag of assorted LEDs and some resistors.) They previously made a similar system called Electronic Brick, which is a 3-wire interface; Grove is 4-wire so it can support either I2C or analog.
Bill Stewart
New Fast-Compression-only CPR http://preview.tinyurl.com/dy575ks
Yep. For less then the price of the main board of those I can get something like the Sparkfun Inventors kit which is the sort of thing everybody should have.
If building little arcade machines like in the article is your thing then you can get (eg.) Arduino+Gameguino (again for less than the price of *just* their main board).
comes with classes that let you use the modules without having to go 'low level.'
Um, so does Arduino. Using a servo (or whatever) is two lines of code.
Arduino killer? Maybe for .Net hipsters with over-rich parents...
No sig today...
The big limitations of the Arduino is very little RAM and very little processing power.
There's bigger Arduinos available. There's even ARM boards which are compatible with Arduino shields, etc.
If the low-powered Arduinos are so popular it's probably because people figured out you don't need much RAM or processing power to do what people are using them for.
No sig today...