The Netduino forums have been replaced by new forums at community.wildernesslabs.co.
This site has been preserved for archival purposes only
and the ability to make new accounts or posts has been turned off.
How long have we been hearing of new MF support coming real soon?
Either people can not estimate things, or they intentionally underestimate. In either case these are the same management teams doing the work, so I have to assume their product is as solid as their estimates, or best case as over sold.
Or is there intentional under funding?
Either way, it does anything but inspire confidence.
I'll try to be as transparent as possible here.
It absolutely is coming. I don't think we ever said "soon" in the past. I'm just about positive Steve didn't, especially given he knows when the development is scheduled during the year. But, if we did, that was our mistake.
We moved the MF team back to the right place as part of the early 2014 reorgs. There was a period of time when we were trying to figure out where NETMF sits in our strategy, and who should own it (it was originally part of the .NET team before we moved it into a different support group). We even called Chris and Gus to provide feedback to us, explain their take on it, etc. With their help, we were able to make NETMF a key part of our IoT/Embedded portfolio and bring it into the right group.
The main delay you've seen is we've been working with our MS Open Tech group to fix the OSS approach to make it far easier for the actual community to contribute to NETMF. We didn't want to start announcing investments until we had that in place.
There will be announcements soon. Small at first, but you'll see that start to build over time. Given the number of ports of NETMF, its reach, etc. these may not come as quickly as with some other products, but that is not representative of any lack of commitment or funding. I can absolutely say to you that we are *more* committed to NETMF now than we were even when you saw the 4 ->4.x releases.
Responding to some other comments I saw:
At Microsoft, we don't think of Windows on Devices and NETMF as competing solutions. They overlap, but each have very distinct sweet spots. Yes, WoD is C++/Win32/Wiring right now. Our intent is to have it be part of our Universal App Platform story (.NET, C++, JS), but we're not there yet. You don't yet see .NET or WinRT on there because we need to create special builds without MMX or SSE optimizations in place. It doesn't make sense for us to do that as a one-off right now when we should focus on the long-term goal.
Knowing this, I challenged Jeremiah Morrill to get Mono working on the Galileo, and for the most part, he has. I hooked him up with another developer in my org so that they can get the GPIO and similar functions working as well. Jeremiah has *zero* device experience. To him, that was just another Windows box. That's the compelling story there. It's Windows.
NETMF, on the other hand, is for people who don't want all the overhead of an operating system, and the chipset support or power requirements to run one. Netduino is to WoD as Arduino is to Raspberry Pi. The Pi didn't make the Arduino inferior or less useful. Similarly, the WoD program doesn't make NETMF or Netduino less useful or somehow inferior. Given that, there's really no need to defend NETMF and disparage WoD, or otherwise form camps -- they're all cool things to build your solutions on.
RTOS:
Windows Compact Embedded is our only certified RTOS, in case you *really* need RTOS in something.
Pete Brown - http://10rem.net (NETMF, C++, Windows, C64, and general geekery) Twitter: @pete_brown I work for Microsoft. Opinions expressed here are my own and do not necessarily reflect those of my employer,our partners or customers.
The post clearly states "in the next few weeks", and was posted in June. I'm sure that was true from the bloggers perspective, but given a lack of anything else, it is also what I made decisions with.
I think many here have expressed the frustration with being left to believe something is coming soon, then hear nothing more about it for a year or more.
Best of luck to you, I truly hope for your success, I have a lot invested in it at the moment. I'm really hoping that was not a bad decision.
Oh well I have not seen anything from it since I registered I think in early April and again after the site was changed.
Then I asked my other co-workers to sign up some where in June and until now none of them has received a kit.
Cheers,
Olaf
We had far more interest than kits. Quite literally, a couple orders of magnitude. We didn't expect that level of interest.
About a week ago or so, we sent out email to everyone who signed up saying that all the kits had been allocated. Anyone who has not yet received a shipping notification or a kit, and who wants to experiment with Windows on the Galileo, will need to pick up one retail and load Windows on it through the WindowsOnDevices site.
The post clearly states "in the next few weeks", and was posted in June. I'm sure that was true from the bloggers perspective, but given a lack of anything else, it is also what I made decisions with.
I think many here have expressed the frustration with being left to believe something is coming soon, then hear nothing more about it for a year or more.
Best of luck to you, I truly hope for your success, I have a lot invested in it at the moment. I'm really hoping that was not a bad decision.
Thanks. I see that on the blog post now. I know the OSS stuff took quite a bit longer than they had thought it would, but sorry it has been more than a few weeks.
But we've already scheduled the recording for Channel 9, so it really is close for the first updates. As I mentioned before, the updates are small, but the team will be working on more as we go forward.
Let me reiterate: NETMF is in a MUCH better spot now than it was just a year or two ago. The team is larger. Its position in our strategy is solid, and investments are approved all the way up. Personally, I'm excited.
Pete
Pete Brown - http://10rem.net (NETMF, C++, Windows, C64, and general geekery) Twitter: @pete_brown I work for Microsoft. Opinions expressed here are my own and do not necessarily reflect those of my employer,our partners or customers.
Hi Pete thanks for the updates. I am glad things are happening in NETMF space.
BTW, I enjoyed listened to you on .NET Rocks episode from last week.
I waited for my WoD kit but I just ended up buying my own Galileo board and loading Windows on it. The dev experience is pretty nice (compared to Arduino).
We had far more interest than kits. Quite literally, a couple orders of magnitude. We didn't expect that level of interest.
About a week ago or so, we sent out email to everyone who signed up saying that all the kits had been allocated. Anyone who has not yet received a shipping notification or a kit, and who wants to experiment with Windows on the Galileo, will need to pick up one retail and load Windows on it through the WindowsOnDevices site.
Thanks. I see that on the blog post now. I know the OSS stuff took quite a bit longer than they had thought it would, but sorry it has been more than a few weeks.
But we've already scheduled the recording for Channel 9, so it really is close for the first updates. As I mentioned before, the updates are small, but the team will be working on more as we go forward.
Let me reiterate: NETMF is in a MUCH better spot now than it was just a year or two ago. The team is larger. Its position in our strategy is solid, and investments are approved all the way up. Personally, I'm excited.
Pete
Hello!
And what happens to those of us who have registered early? I did so as soon as the location was announced earlier this year. Then not too long ago I get an e-mail that I need to confirm my physical address for something via an appropriate form which was sent to my business e-mail address. Pete I am actually wondering what will happen further along.
I am currently configuring my backup work machine to pose as a Windows 8.1 system, with the usual tools for building applications. My other concern is one of doing stuff on Windows 8.1 and Dot Net MF for the Netduino devices.....
Hi Pete thanks for the updates. I am glad things are happening in NETMF space.
BTW, I enjoyed listened to you on .NET Rocks episode from last week.
I waited for my WoD kit but I just ended up buying my own Galileo board and loading Windows on it. The dev experience is pretty nice (compared to Arduino).
Thanks
The dev experience isn't *quite* as good as Netduino (hard-coded deployment paths and device names in the config), but it's still really good. I enjoy working with both.
And yes, I love Visual Studio, so happy to use that rather than the Arduino IDE. Personal preference, of course.
Hello!
And what happens to those of us who have registered early? I did so as soon as the location was announced earlier this year. Then not too long ago I get an e-mail that I need to confirm my physical address for something via an appropriate form which was sent to my business e-mail address. Pete I am actually wondering what will happen further along.
The IoT team tried to be as fair as possible, but we simply couldn't cover everyone. We tried to make sure there was decent global coverage, and a number of other factors. I'm sorry you did not get a kit. They really did the best they could.
On the plus side, the level of interest helped us convince Intel to make the UEFI firmware available, which then made it possible to let anyone install Windows on a retail Galileo. Note that we'll also have support for the Galileo 2 in the future -- something you might find appealing.
I am currently configuring my backup work machine to pose as a Windows 8.1 system, with the usual tools for building applications. My other concern is one of doing stuff on Windows 8.1 and Dot Net MF for the Netduino devices.....
Always prudent to test, but I wouldn't anticpate any Windows 8.1-specific problems. The only thing to pay attention to right now is that we don't currently support NETMF development with Visual Studio 2013. That will change soon.
Pete
Pete Brown - http://10rem.net (NETMF, C++, Windows, C64, and general geekery) Twitter: @pete_brown I work for Microsoft. Opinions expressed here are my own and do not necessarily reflect those of my employer,our partners or customers.
The dev experience isn't *quite* as good as Netduino (hard-coded deployment paths and device names in the config), but it's still really good. I enjoy working with both.
And yes, I love Visual Studio, so happy to use that rather than the Arduino IDE. Personal preference, of course.
The IoT team tried to be as fair as possible, but we simply couldn't cover everyone. We tried to make sure there was decent global coverage, and a number of other factors. I'm sorry you did not get a kit. They really did the best they could.
On the plus side, the level of interest helped us convince Intel to make the UEFI firmware available, which then made it possible to let anyone install Windows on a retail Galileo. Note that we'll also have support for the Galileo 2 in the future -- something you might find appealing.
Always prudent to test, but I wouldn't anticpate any Windows 8.1-specific problems. The only thing to pay attention to right now is that we don't currently support NETMF development with Visual Studio 2013. That will change soon.
Pete
Hello!
I understand. Really. They show on their site that they will be present, next weekend at Maker Faire here in Queens NY, at the NY Hall of Science site. I know Chris will be there,
Will there be anyone from the Windows on Devices unit to answer my questions?
On a side note I did find for sale two open box candidates for the platform at my Brooklyn Micro center. I am debating if I should attempt to buy one of them sometime after the events.
You also said,
Always prudent to test, but I wouldn't anticpate any Windows 8.1-specific problems. The only thing to pay attention to right now is that we don't currently support NETMF development with Visual Studio 2013. That will change soon.
I'm going to also install the VS2012 full kit on the system as well so I'll probably be moving my Generation one device over to NETMF 4.3 as soon as work out those details.
You also mentioned,
The IoT team tried to be as fair as possible, but we simply couldn't cover everyone. We tried to make sure there was decent global coverage, and a number of other factors. I'm sorry you did not get a kit. They really did the best they could.
On the plus side, the level of interest helped us convince Intel to make the UEFI firmware available, which then made it possible to let anyone install Windows on a retail Galileo. Note that we'll also have support for the Galileo 2 in the future -- something you might find appealing.
Unlike some members of this august gathering I'm not annoyed, surprised and rather pleased however, not annoyed at things, which they sound like. And I never got one of those e-mail messages. Which are described here:
We had far more interest than kits. Quite literally, a couple orders of magnitude. We didn't expect that level of interest.
About a week ago or so, we sent out email to everyone who signed up saying that all the kits had been allocated. Anyone who has not yet received a shipping notification or a kit, and who wants to experiment with Windows on the Galileo, will need to pick up one retail and load Windows on it through the WindowsOnDevices site.
I subscribed to the WoD time ago, but no any incoming e-mail so far. Anyway, I don't mind for that.
What you are pointing is pretty interesting to me, Pete, but also confusing.
I don't ask you to "disclose" too much, but I was surprised to read about the "hierarchy" MF < WoD < WEC. How many "layers" should be, and why?
Netduino, as the MF implementation I know better, is both a pretty powerful hardware board and a weak software device, at the same time.
Talking about "Plus 2", you have a blazing-fast 180MHz-CPU with lot of flash and lot of ram (if you compare to the typical C-language device). However, you must fight against free ram, speed, and some severe language/framework limitations. Finally, it's almost impossible (realistically speaking) to create a low-level native driver, unless recompiling the whole sources.
I really can't blame Chris for such a board, because it's very nice indeed: I'd re-buy again! But I wonder how could be a supposed "Netduino 3" hardware without a huge overlap (even surpassing) a Galileo board (or hopefully, a more serious and reliable Edison board)?
I only could interpret the current (incoming) moves by MS (and other companies around IoT) by creating "kinda" super-compact-embedded. Why wasting human resources creating several different flavors of the same thing, and not going toward *just one*, very-well designed, scalable, upon a decent hardware compromise, maybe with high-end features?
Really, can't blame who chooses a Pi (or a far better BeagleBone) even for small home/hobby projects.
Please, explains us, as far you can, maybe with a "comparison table", which typically comes quick-and-easy to understand.
I'm just a drop in the ocean of users, but I really don't see the point on the described "diversity".
Final question: in a three weeks there'll be the great Maker Faire in Rome, which is the biggest in Europe. Since MS is a huge partner+sponsor of that event, do you think/know whether something new should be presented?
We're working on improving performance on NETMF so Chris and others can take better advantage of the MCU (it's not a CPU, there's a difference in mindset there) power available.
I don't know of anything new coming in time for the Maker Faire in Europe. All of our OS products are on the same shipping train at this point.
Pick the board that works best for you. Some like Arduino. Some like Netduino. Some prefer to work with a full OS like Linux or Windows. Boards that can run full Windows Embedded (like Minnowboard Max) are getting smaller and more powerful as well. At the same time, Pared down versions of Linux can work on MCUs like the one on the Pi, and Intel is doing things to shrink previous CPUs down to MCU SOC form factor and function. No table is going to cover all that, and I'm not going to try to convince you to pick one over the other in any broad way. Go by the requirements of your project and the skills you have or want to build.
Pete Brown - http://10rem.net (NETMF, C++, Windows, C64, and general geekery) Twitter: @pete_brown I work for Microsoft. Opinions expressed here are my own and do not necessarily reflect those of my employer,our partners or customers.
Although I prefer to work with .NET MF, cheapest chip is about $7 - $10 and above that can fit it.
And I know you are probably all about putting windows on devices but what about the lowest tier, I would like to have a higher programming language even on that tier.
For instance I'm now using an ATTiny85 to do some LED lighting, the chip is about $0.8 - $1.2, and can run what I would call a subset of Arduino.
Neither I nor Microsoft are "all about" putting Windows on devices. Sorry I didn't make that clear. NETMF is a big part of the strategy.
NETMF is fundamentally different from something like the Wiring language and Arduino. When you write for Arduino, you are basically using an extremely thin layer of abstraction over what the MCU natively provides. If you code using C on an AVR chip, you'll quickly notice that.
NETMF, on the other hand, provides a number of services above that, including CLR garbage collection, memory management, threading, and more. It's also interpreted at runtime, not something that's natively compiled.
For those reasons, it's impossible in its current state, to target 8 bit devices, or low memory devices with NETMF. Maybe we'll find a way to change that in the future, but I don't see a way to do that with .NET and C# as people currently recognize them.
Similarly, you can't get Windows on an 8 bit device like that, because it's a full modern operating system.
It's possible to get NETMF running on smaller 32 bit ARM devices. Chris has done that himself over the years as he's evolved Netduino and related. I'm not sure the cost savings are worth what you give up, though. The vast majority of NETMF is the core runtime and support, not the individual functions.
On the IoT team, our plan is to support, whenever possible, these third-party devices when it comes to connecting with the rest of our ecosystem and services. We're also looking at ways to make NETMF faster and more efficient. Some of that may help here. TBD.
Pete
Pete Brown - http://10rem.net (NETMF, C++, Windows, C64, and general geekery) Twitter: @pete_brown I work for Microsoft. Opinions expressed here are my own and do not necessarily reflect those of my employer,our partners or customers.
It absolutely is coming. I don't think we ever said "soon" in the past. I'm just about positive Steve didn't, especially given he knows when the development is scheduled during the year. But, if we did, that was our mistake.
(....)
There will be announcements soon.
(....)
Pete
Pete, thank you very much for providing some information! We truly appreciate it!
I think that one of the biggest problems of NETMF is the lack of information. The www.netmf.com contains so little news that anyone going there is led to believe that the project is dead. And the fact that Sal has had to write a blogpost just to say "we're still here" is worrying as well. Reminds me of all those personal blogs where the last blogpost is "it's been a while since I've blogged, but I'll get better". And that is the last thing you ever see on those kind of blogs.
The community is information-hungry, and we gather as much as we can from any source available. And that usually means the hints given to us from either Chris or the guys over at GHI. The slightest hint they give that something is about to happen gives us a sugarrush, but when nothing happens after that we crash and get grumpy...
I believe that one of the most important abilities any person can have, regardless of the domain, is expectation management. That you know how to control other peoples expectations is sooo important. When my kids ask "are we there yet?", my reply is always "No, we've still got a long way to go". They accept the reponse, and when we arrive at the destination ten minutes later they say "Oh, are we here already?! Yey!". Had I said "Yes, we're almost there" I would have been asked again every two minutes "Are we there yet? Are we there yet?", and they would have felt the last ten minutes took forever. The drive was always the same, but the expectations was managed differently.
That is kinda where we are with the information flow about NETMF. Two months ago there was lots of information about something happening "soon".
Soon to us means days, maximum a week. Here we are two months later. The drive feels forever and we are getting grumpy.
There was supposed to be a Channel9 video with Steve in June so we got our hopes up, but that was all of a sudden cancelled, leaving us in the dark again.
You yourself indicated a few days ago that a Channel9 video is upcoming with the comment "But we've already scheduled the recording for Channel 9, so it really is close for the first updates". That leads us to check the Channel9 site regularly from now - and you can imagine the disappointment if another month passes without anything happening on that end. I truly hope it doesn't.
I hope I don't come of as being negative, but we're just soo excited about NETMF that we want to see it succeed!
Pete, thank you very much for providing some information! We truly appreciate it!
I think that one of the biggest problems of NETMF is the lack of information. The www.netmf.com contains so little news that anyone going there is led to believe that the project is dead. And the fact that Sal has had to write a blogpost just to say "we're still here" is worrying as well. Reminds me of all those personal blogs where the last blogpost is "it's been a while since I've blogged, but I'll get better". And that is the last thing you ever see on those kind of blogs.
The community is information-hungry, and we gather as much as we can from any source available. And that usually means the hints given to us from either Chris or the guys over at GHI. The slightest hint they give that something is about to happen gives us a sugarrush, but when nothing happens after that we crash and get grumpy...
I believe that one of the most important abilities any person can have, regardless of the domain, is expectation management. That you know how to control other peoples expectations is sooo important. When my kids ask "are we there yet?", my reply is always "No, we've still got a long way to go". They accept the reponse, and when we arrive at the destination ten minutes later they say "Oh, are we here already?! Yey!". Had I said "Yes, we're almost there" I would have been asked again every two minutes "Are we there yet? Are we there yet?", and they would have felt the last ten minutes took forever. The drive was always the same, but the expectations was managed differently.
That is kinda where we are with the information flow about NETMF. Two months ago there was lots of information about something happening "soon".
Soon to us means days, maximum a week. Here we are two months later. The drive feels forever and we are getting grumpy.
Thanks for the information.
Do understand, though, that you've just completely made the case for the approach of not announcing *anything* until it is built and ready to ship. This is consistent with the feedback on the GHI forums as well. Steve even took a pretty big risk by announcing the investments for NETMF at Build; we usually don't confirm things like that that far in advance.
Product teams miss deadlines for a variety of reasons. They also have a different definition of "soon" compared to what most normal people do . Typically "soon" in the PM world is "next few months". That's because so many of us have our heads in some time in the next quarter or so, so in our minds, that's actually "soon". It comes even faster for the people who are trying to build the stuff in a limited time. For those on the outside waiting, though, it feels like forever.
Nevertheless, the expected dates were not hit by the team here; they did expect to ship something earlier this summer.
We'll look at other ways to get information to the community without announcing upcoming work.
I'll likely be traveling when the Channel 9 video goes up, and so probably won't announce anything on the boards. As I mentioned earlier, this is a small announcement, just the start of the work. It's my hope that it shows the beginning of the investment.
Sites like netmf.com:
At Microsoft, we tend to spin up too many little satellite sites, IMHO. Quite honestly, I don't think netmf.com should be out there. It should be part of the dev center on MSDN, or part of the embedded/IoT site. I feel the same way about several other sites we run. We've run across this before where the site is a passion project, and then the person moves to another group or something, and the site just sits there. We need to move that information into a proper dev center, IMHO.
Pete
Pete Brown - http://10rem.net (NETMF, C++, Windows, C64, and general geekery) Twitter: @pete_brown I work for Microsoft. Opinions expressed here are my own and do not necessarily reflect those of my employer,our partners or customers.
BTW, if folks have NETMF feedback for the product teams, I'm happy to take actionable feedback to them.
Actionable feedback for a product team comes in the form of a scenario rather than a specific feature request. For example "I am building an X. Currently, I'm blocked because I'm trying to accomplish Y and there's no way to do that in NETMF. I was expecting to see <some feature>."
Best way is to get it to me at pete dot brown at microsoft dot com. (put NETMF feedback or similar in the subject line to help me identify it)
Actual bugs should be logged in Connect. I know that system is a bear, but it creates a ticket, and forces the product teams to deal with the issue, even if that means marking it as by design or not to be fixed. It also provides a facility for you to comment on it.
Scratch that. The team has decided that Codeplex bug reporting is best for transparency and to keep everyone involved. So please continue to report there, vote things up, etc.
Thanks.
Pete
Edited by Pete Brown, 17 September 2014 - 08:07 PM.
Pete Brown - http://10rem.net (NETMF, C++, Windows, C64, and general geekery) Twitter: @pete_brown I work for Microsoft. Opinions expressed here are my own and do not necessarily reflect those of my employer,our partners or customers.
And I never got one of those e-mail messages. Which are described here:
