How pervasive is the OSHW movement into the professional EE feild?

[zapta]

In my time at both Apple and Google, hobbyist dev boards have been quite prevalent for one-off designs and quick prototypes. With Google, I now work on server designs that will be given to the Open Compute Project as open designs (the schematics, gerbers, BOMs, firmware) will be publicly available under a fairly liberal license. It's not quite OSHWA level but it's a step in the right direction for big companies.

This Wiki page claims that Google also used OSH cameras for its scanning systems https://en.wikipedia.org/wiki/Elphel

[NiHaoMike]

Because projects don't end after that half day Arduino project hack.  The further you get through a project, the harder it is to change directions.  You end up like exmadscientist, stuck with some Arduino thing and having to work around the hardware choices and limitations instead of getting exactly what you need because you designed it.

They're test fixtures, not final products. For any engineer to be able to make a change within minutes (e.g. to accommodate changes in test procedures) is a big asset. Same reason why Python is used so much.

There are times when the need falls outside the scope of Arduino and that's when something else gets used. But the test fixtures in question tend to be relatively simple like toggling and reading GPIOs at rather slow rates. The really advanced stuff like capturing high speed logic are covered by common networked instruments.

[Karel]

The reason that OSHW will never become as popular as OS software, is mainly because of:

- the investment in hardware and instruments needed to design electronics (and als EMC/EMI compliance)

- the amount of available electronic engineers

- expansive to share projects, apart from theoretical design, every new participant needs to order a new piece
  of hardware to test/play with.

Compare the above with the case of OS software:

 - you don't need any hardware or other investments (everybody has a pc)

- there are more software engineers available, so it's more likely that some of them will start or participate
  in an OS project.

- Practically all professional tools to develop professional grade software can be downloaded for free.
  No need to buy (expensive) stuff.

Edit: Available space: compare how much (office) space you need for developing hardware or software...

[EEVblog]

There is also the time honoured "not invented here" (NIH) syndrome.
Using a circuit from an manufacturer app note, fine. But using an open source hardware circuit (or part of) that contains a licence   
If you are a professional engineer working at a big company and you are trying to get a design done to a deadline and you desperately need that part of an open source hardware circuit, you'd just take it and tell no one!

[george graves]

Just as gray beards with slider rulers looked down on calculator guys, and then old school 80-90's guys looked down on internet guys?

[hans]

At work we've used Arduino's and other OSHW occasionally, but my preference is strongly to keep them internal. I am not fond of shipping arduino-based products to customers, unless they are paying for engineering and service fee's for a custom machine installation. In that case it doesn't make sense economically to design custom PCB's for those projects.

For interns that don't have any programming experience, the Arduino is great. For example; let them build a cable tester box with LCD and programmable step programs. It has a bit of electronics in it to drive 40 or 80-pin cables from an Arduino, but also software to read some buttons, think about the programmable sequence routines and drive a LCD text display. Finally they have to write a users manual for production personnel to make use of it.

Other than that, all of the hardware and software I design for work doesn't have a single line of Arduino-based code in it, neither a single CAD entity from an OSHW product. Licensing is a big part in that; I don't think a company wants to open source an "enterprise" or "proprietary" product because of GPL. And that's why they often stay away from it.

I also think there is also a chronic struggle in open-source HW/SW; in that everyone has different needs and ways of doing things. Therefore projects will branch off and merging changes upstream can be either time consuming when not documented, or downright frustrating when ego's collide and both maintainers have different views.

Example in HW: I may not accept schematics with chip symbols exactly like the pinning on the package (like so) rather than somethign that makes sense from a system design perspective (like so). If I take the design and start polishing the design very drastically to adhere to my(or company) standards, and then am required to open source it; okay, will my changes every make it upstream? And in what form?

This is not to say I don't like OS; I run Linux on all my machines and I will chose an open source solution where there is one. However it does need to be considered that open source isn't always "free" in terms of money for a company; it does require maintenance in order to work for you.

Linux is probably one of the most successful OSS projects I can think of. It's used on so many embedded computers, because no one will write their own kernel. In addition, Linux is also a combination of the GNU environment, the kernel and userland applications. In most cases only the first 2 will be open sourced because GPL, which is OK if the companies wants to hide their secret sauce.

[2N3055]

I use Arduino Mega and Uno instead of Atmel STK series...  Program in C (no Arduino IDE and code) as a test proof of concept ideas..
And by that I mean software side of the thing..

It's great for that. Also, if I have a need for something quick to automate something in the lab, great for that..

It's not a professional tool, nor it was meant to be, despite all the fanboys wishes..
But if you respect it for what it is, great and useful thing to have...

As for OSHW, no thanks... If I  make something, and I feel I would like to share it with the world, I'll publish all the data as is..

Software Open Source ( as open source code as opposed to compiled , "closed" applications ) is a software paradigm...

It has NO applicability in hardware world..  T3sl4co1l put it very nicely... Software people that dabble in lego hardware ( Shieldy thingies of all sorts and types ) don't fully understand problems of quality hardware development...

Linux doesn't have to comply to EMC requirements, has no mechanical requirements (temperature, humidity, vibrations etc..)..

And a funny thing is that Linux works so well because likes of Microsoft for years forced the manufacturers to standardize hardwer to such an extent that everything just works together... No community effort would have achieved that to such a level...

So Open anything I like and support, kudos to all that do it, but hardware is not open to such a treatment... Not for a lack of good intentions, just a very different animal...  Something open might be possible, but not based on software way of thinking...

All the best to all...

[b_force]

I still don't understand what is defined as an 'Arduino'?

Is that just the whole board (like a UNO, or Mega) and the shape, or is that just when you use the Arduino IDE? 
Like myself, I have seen companies using just the IDE.
It saves a lot of time, because you don't have to go into all kinds of datasheets to figure out which registers etc you need.
There are also so many libraries and examples available which are great to start from.

If we are talking about the whole design and board, I also agree that an Arduino is far from a finished professional product.

I only don't agree with the fact that open source can't be professional at all.
There is no reason you can't make an open source design according to all EMI, EMC and safety standards.
The fact that you don't see these kinds of products (take a SMPS for example), doesn't mean it's not possible.
I think it means that most open source projects are done in spare time and as an hobby and therefore people naturally can't be bothered by all these regulations.
Most open source projects (with all the respect) are not much more than some simple add-on modules.
I haven't seen one open source project that was completely done from head to toe (like power supply, proper interface, connectors etc)

Although I really like the idea of open source hardware and software, this is also exactly the point I have problems with it.
In my opinion many (hardware & software!) projects are very far from finished as a professional product.
Absolutely not ready for the bigger audience because it simply is to much of an hassle with manual tweaks, tuning and even programming.
It just simply doesn't work out of the box (plus most interfaces and GUIs are just really really  )
A lot of people don't seemed to be bothered about it, but for a lot of other people they don't have the know how or time for it.
90% of the products I need to use on a regular basis don't work out of the box on Linux.
This is the mean reason I am not going to move to Linux as a main system (I only use it for smaller applications).

In my opinion the open source community really needs to put more focus on finished products instead!
A few posts back, Karel has a few good points about why it's more practical to make open source software easier than hardware.
I'm not quiet sure if that's all true. I mean, a lot of people have enough gear (at home or at work) to do all these tests anyway.
That being said, those tests are not very easy to do, and even developing stuff that is made according to all EMC/EMI and safety rules is not an easy task.
A lot of EE don't even know about all these regulations.

[EEVblog]

Just as gray beards with slider rulers looked down on calculator guys, and then old school 80-90's guys looked down on internet guys?

NOthing really to do with anyone looking down upon anyone else. It's that big companies and their systems are scared of stuff they don't understand, and the NIH syndrome is one of fear of  legality issues mixed in with pride and a bunch of other stuff.

[EEVblog]

I still don't understand what is defined as an 'Arduino'?

It comes down to two things:
1) The Arduino Shield Pinout
2) Compatibility with the Arduino programming ecosystem

Can be either or both.

[b_force]

Just as gray beards with slider rulers looked down on calculator guys, and then old school 80-90's guys looked down on internet guys?

NOthing really to do with anyone looking down upon anyone else. It's that big companies and their systems are scared of stuff they don't understand, and the NIH syndrome is one of fear of  legality issues mixed in with pride and a bunch of other stuff.

A lot of companies are even scared showing the public what they actually really do.
It is all 'magical voodoo'. It always makes me laugh, because if you're in the field, it's not that difficult to see what you competitors are doing.
Or otherwise you simply reverse engineer a product, like you did with the Brymen multimeter. 

I have done multiple freelance jobs and sometimes I even had to sign fancy looking contracts that I am not allowed to share any kind of 'secrets'.
With even ridiculous kinds of fines and other things to frighten you. 
I always ask them where general EE knowledge stops and something very unique begins.
They never know how to answer to that question.

I still don't understand what is defined as an 'Arduino'?

It comes down to two things:
1) The Arduino Shield Pinout
2) Compatibility with the Arduino programming ecosystem

Can be either or both.

Well in that case, I think there are many more Arduino 'projects'
Because it means that you don't need a board that looks like an Arduino

[madires]

Linux doesn't have to comply to EMC requirements, has no mechanical requirements (temperature, humidity, vibrations etc..)..

That's another misconception between the hardware and software worlds. There are standards and requirements for software too. For example, linux is mostly POSIX compliant (but not certified).

[Karel]

And a funny thing is that Linux works so well because likes of Microsoft for years forced the manufacturers to standardize hardwer to such an extent that everything just works together... No community effort would have achieved that to such a level...

Quite the opposite. There's the microsoft way of doing things, and there's the way the rest of the world is doing things.
Microsoft has always reinvented the weel for the sole purpose of incompatibility.
Computers and software have evolved, not thanks to microsoft but despite microsoft.

[zapta]

Well in that case, I think there are many more Arduino 'projects'
Because it means that you don't need a board that looks like an Arduino

Even the Arduino Mini Pro ($2.16 shipped, on ebay) doesn't look as Arduino:

[2N3055]

And a funny thing is that Linux works so well because likes of Microsoft for years forced the manufacturers to standardize hardwer to such an extent that everything just works together... No community effort would have achieved that to such a level...

Quite the opposite. There's the microsoft way of doing things, and there's the way the rest of the world is doing things.
Microsoft has always reinvented the weel for the sole purpose of incompatibility.
Computers and software have evolved, not thanks to microsoft but despite microsoft.

LOL , you seem  have alergic reactions to mentioning of Microsoft........ 
I hope your EPI pen is at hand.... I didn't want to upset you.. 

It's just history how it went, you might be young enough not to be there when it happened...

I was referring to IBM PC compatible platform that was used from a day one to develop original Linux...
That PC platform was designed with the SOLE purpose to run IBM OS/2 and Microsoft DOS/WINDOWS/WINDOWS NT..
Nobody gave a flip-flop about some students playing with kernels...
And it was designed according to Microsoft PC platform specs, to ensure compatibility.

And it was a good thing because it brought standardization to computer world...
That enabled stable platform for Linux to evolve trough cca. 20 years
That part of Microsoft bullying hardware vendors actually was benefit to Linux...

Other of their practices not so much... 

And before you start bashing me, I had a PC running Linux Slackware in 1997.  I had to compile kernel to make it work with my PC..
 I love Linux, and like BSD even better...

And for my day to day work I use Windows.. All the software just works etc.. I don't have time to fiddle with things..

I guess that will clear up some confusion.. I'm not very good with English..

All the best,

[b_force]

Well in that case, I think there are many more Arduino 'projects'
Because it means that you don't need a board that looks like an Arduino

Even the Arduino Mini Pro ($2.16 shipped, on ebay) doesn't look as Arduino:

But it's still an 'only one chip on board' idea.
I was more talking about a full working concept like a bench supply, were a Atmega 328 (or something) is being used as the core.

[Karel]

And a funny thing is that Linux works so well because likes of Microsoft for years forced the manufacturers to standardize hardwer to such an extent that everything just works together... No community effort would have achieved that to such a level...

Quite the opposite. There's the microsoft way of doing things, and there's the way the rest of the world is doing things.
Microsoft has always reinvented the weel for the sole purpose of incompatibility.
Computers and software have evolved, not thanks to microsoft but despite microsoft.

LOL , you seem  have alergic reactions to mentioning of Microsoft........ 
I hope your EPI pen is at hand.... I didn't want to upset you.. 

It's just history how it went, you might be young enough not to be there when it happened...

I do remember very well how it went:

Microsoft's Campaign To Destroy DR-DOS
Microsoft's Anticompetitive Per Processor License Fees
Microsoft's Retaliation And Price Discrimination Against IBM
Microsoft's Organized Collective Boycott Against Intel
Microsoft's Elimination Of Word Perfect
Microsoft's Deceptive WISE Software Program
Microsoft's Elimination Of Netscape
Microsoft's Attempts To Extinguish Java
Microsoft's Elimination Of Rival Media Players
Microsoft's Campaign Against Rival Server Operating Systems
Microsoft's Failure To Comply With The Final Judgment
Microsoft's Campaign of Patent FUD against Linux and Open Source Software
Microsoft's False Promises of Interoperability

The EC is also investigating Microsoft's actions to manipulate the vote
of the International Organization for Standardization / International Electrotechnical
Commission on the recent standardization of Office "Open" XML ("OOXML"). As reported
widely in the press and on the Internet, Microsoft's manipulation of the standards setting process
in favor of OOXML included financial inducements, threats, misleading information, and
committee-stuffing.committee-stuffing. These investigations are compelling examples of Microsoft's continued
misconduct related to its monopolies in operating systems and other products.

[T3sl4co1l]

I love it when people hate something so much, they become blind to anything that may be beneficial...

2N3055 was trying to bring to light some of the positive aspects.

When there's a very powerful player in some space, they tend to dictate the rules.  That might not be a good thing for the little players trying to compete directly.  But on the other hand, objects in that space (in this case, the IBM-PC and clones) tend to be very consistent, which can be advantageous to the little players as well.

And 2N3055 is not blind to the above reasons, as allowed for here:

Other of their practices not so much... 

That he chose not to dwell on them is simply a matter of preference.

Tim

[2N3055]

I do remember very well how it went:

Microsoft's Campaign To Destroy DR-DOS
Microsoft's Anticompetitive Per Processor License Fees
Microsoft's Retaliation And Price Discrimination Against IBM
Microsoft's Organized Collective Boycott Against Intel
Microsoft's Elimination Of Word Perfect
Microsoft's Deceptive WISE Software Program
Microsoft's Elimination Of Netscape
Microsoft's Attempts To Extinguish Java
Microsoft's Elimination Of Rival Media Players
Microsoft's Campaign Against Rival Server Operating Systems
Microsoft's Failure To Comply With The Final Judgment
Microsoft's Campaign of Patent FUD against Linux and Open Source Software
Microsoft's False Promises of Interoperability

The EC is also investigating Microsoft's actions to manipulate the vote
of the International Organization for Standardization / International Electrotechnical
Commission on the recent standardization of Office "Open" XML ("OOXML"). As reported
widely in the press and on the Internet, Microsoft's manipulation of the standards setting process
in favor of OOXML included financial inducements, threats, misleading information, and
committee-stuffing.committee-stuffing. These investigations are compelling examples of Microsoft's continued
misconduct related to its monopolies in operating systems and other products.

You are, of course, 100% right, but....

None of that has any bearing to what I sad... And I did said "their other practices not so much.."

I didn't come her to play advocate to Microsoft corp.. They have they legal team for that...

What I said is still 100% true... Linux was lucky to take advantage of a mature professional hardware platform many very rich companies spent billions developing...  Without that platform, no Linux as we know it today would happen..
Maybe something else, but not what we have today...

Cheers!!

[2N3055]

I love it when people hate something so much, they become blind to anything that may be beneficial...

2N3055 was trying to bring to light some of the positive aspects.

When there's a very powerful player in some space, they tend to dictate the rules.  That might not be a good thing for the little players trying to compete directly.  But on the other hand, objects in that space (in this case, the IBM-PC and clones) tend to be very consistent, which can be advantageous to the little players as well.

And 2N3055 is not blind to the above reasons, as allowed for here:

Other of their practices not so much... 

That he chose not to dwell on them is simply a matter of preference.

Tim

Thanks a lot Tim, that's exactly what I wanted to say... 

And I'm the last one to bash any useful platform... Started with computors, at 15 something, now more than 30 years... (darn I'm old  )

Been administering and repairing (hardware) all kinds of stuff, UNIVACs, DECs, HP Risc, IBM Risc (now Power), Intel platforms, Apple 68k, Power and Intel, ... XENIX (Microsoft Unix LOL), SCO, Ultrix, AIX, Linux..... Used to be MCSE, but it expired long time ago....
I still manage AIX and Linux servers for a client...

Those are tools, and all of them had their quirks and good sides... And was going where job took me...

like DeNiro in "Ronin" says: "... what do you mean my favorite weapon?  I don't care, it's a tool.. You put it in a tool box and you go to work... ""

 

All the best,

Sinisa

[Karel]

What I said is still 100% true... Linux was lucky to take advantage of a mature professional hardware platform many very rich companies spent billions developing...  Without that platform, no Linux as we know it today would happen..
Maybe something else, but not what we have today...

You can say that about all operating systems. In addition, Linux has proven to be flexible enough to adopt all kinds of platforms.
Not just one or two like most os's.
That hardware platform you were talking about, was already available before DOS was ready.
And Linux was able to (stable!) multitask while microsoft was still offering win 3.11...

[b_force]

We can debate about a lot of details.
In the end it comes down to the fact who has the best PR people and enough luck at the right time.

[2N3055]

What I said is still 100% true... Linux was lucky to take advantage of a mature professional hardware platform many very rich companies spent billions developing...  Without that platform, no Linux as we know it today would happen..
Maybe something else, but not what we have today...

You can say that about all operating systems. In addition, Linux has proven to be flexible enough to adopt all kinds of platforms.
Not just one or two like most os's.
That hardware platform you were talking about, was already available before DOS was ready.
And Linux was able to (stable!) multitask while microsoft was still offering win 3.11...

Your timelines are wrong...

First, Linux needs 386+ processor and at about 1995, first Linux wasn't running on 8 bit 8088 from 1985, but on a 10 years old PC platform, that was quite mature by then.. So, yeah, it benefited from Intel/IBM/COMPAQ/HP/MICROSOFT and bunch of clones industry, that was allready a many billions industry...

And it wasn't flexible enough at first, it had to rewritten to became so..

And it didn't have real SMP (symmetric multiprocessing) capabilities until kernel 2.6 (cca 2003.). It had them, but it didn't work well.
And those SMP fixes that were first steps to making Linux serious OS we have and admire today were contributed  by IBM..
And,no, they are not so good, they did it because they needed alternative to Windows, and for their own development..
They were decent enough to publish it though, so it was merged to Linux...

And at the time of first Linux, and Windows 3.11,  we were running Windows NT 3.51.. Real SMP system, designed by Richard Rashid (the inventor of MACH microkernel, the version of which is NT kernel), and Dave Cuttler designer of VAX VMS, a legend in OS world...

So no Linux wasn't some magical thing that dropped from the sky.. For years was no more than a toy... It grew to be real OS, and that's fantastic thing... But it benefited that PC were done by somebody else at somebody else's expense and effort...

That's all what I said..

[Karel]

What I said is still 100% true... Linux was lucky to take advantage of a mature professional hardware platform many very rich companies spent billions developing...  Without that platform, no Linux as we know it today would happen..
Maybe something else, but not what we have today...

You can say that about all operating systems. In addition, Linux has proven to be flexible enough to adopt all kinds of platforms.
Not just one or two like most os's.
That hardware platform you were talking about, was already available before DOS was ready.
And Linux was able to (stable!) multitask while microsoft was still offering win 3.11...

Your timelines are wrong...

Nothing wrong with my timelines.

First, Linux needs 386+ processor and at about 1995, first Linux wasn't running on 8 bit 8088 from 1985,

I never wrote that. What I meant is that the pc was ready before DOS was ready. So, it was not microsoft who invented the pc platform or who made it possible to exists.

but on a 10 years old PC platform, that was quite mature by then.. So, yeah, it benefited from Intel/IBM/COMPAQ/HP/MICROSOFT and bunch of clones industry, that was allready a many billions industry...

It benefitted from the IBM clones, not from microsoft. Without microsoft, we still should have had pc's, but probably running
DR-DOS and later OS2-warp.

And it wasn't flexible enough at first, it had to rewritten to became so..

And it didn't have real SMP (symmetric multiprocessing) capabilities until kernel 2.6 (cca 2003.). It had them, but it didn't work well.
And those SMP fixes that were first steps to making Linux serious OS we have and admire today were contributed  by IBM..
And,no, they are not so good, they did it because they needed alternative to Windows, and for their own development..
They were decent enough to publish it though, so it was merged to Linux...

And at the time of first Linux, and Windows 3.11,  we were running Windows NT 3.51.. Real SMP system, designed by Richard Rashid (the inventor of MACH microkernel, the version of which is NT kernel), and Dave Cuttler designer of VAX VMS, a legend in OS world...

So no Linux wasn't some magical thing that dropped from the sky.. For years was no more than a toy... It grew to be real OS, and that's fantastic thing... But it benefited that PC were done by somebody else at somebody else's expense and effort...

Let's say that, at that time, Linux was already in a better state than win 3.11 and win95.

[2N3055]

What I said is still 100% true... Linux was lucky to take advantage of a mature professional hardware platform many very rich companies spent billions developing...  Without that platform, no Linux as we know it today would happen..
Maybe something else, but not what we have today...

You can say that about all operating systems. In addition, Linux has proven to be flexible enough to adopt all kinds of platforms.
Not just one or two like most os's.
That hardware platform you were talking about, was already available before DOS was ready.
And Linux was able to (stable!) multitask while microsoft was still offering win 3.11...

Your timelines are wrong...

Nothing wrong with my timelines.

First, Linux needs 386+ processor and at about 1995, first Linux wasn't running on 8 bit 8088 from 1985,

I never wrote that. What I meant is that the pc was ready before DOS was ready. So, it was not microsoft who invented the pc platform or who made it possible to exists.

but on a 10 years old PC platform, that was quite mature by then.. So, yeah, it benefited from Intel/IBM/COMPAQ/HP/MICROSOFT and bunch of clones industry, that was allready a many billions industry...

It benefitted from the IBM clones, not from microsoft. Without microsoft, we still should have had pc's, but probably running
DR-DOS and later OS2-warp.

And it wasn't flexible enough at first, it had to rewritten to became so..

And it didn't have real SMP (symmetric multiprocessing) capabilities until kernel 2.6 (cca 2003.). It had them, but it didn't work well.
And those SMP fixes that were first steps to making Linux serious OS we have and admire today were contributed  by IBM..
And,no, they are not so good, they did it because they needed alternative to Windows, and for their own development..
They were decent enough to publish it though, so it was merged to Linux...

And at the time of first Linux, and Windows 3.11,  we were running Windows NT 3.51.. Real SMP system, designed by Richard Rashid (the inventor of MACH microkernel, the version of which is NT kernel), and Dave Cuttler designer of VAX VMS, a legend in OS world...

So no Linux wasn't some magical thing that dropped from the sky.. For years was no more than a toy... It grew to be real OS, and that's fantastic thing... But it benefited that PC were done by somebody else at somebody else's expense and effort...

Let's say that, at that time, Linux was already in a better state than win 3.11 and win95.

Well I see...

After all this now you invent what I said... LOL

I didn't say Microsoft invented platform... 

And although, technically it was better architecture than WFW3.11, WFW 3.11 was useful as a platform with thousands applications..
You could use it to do work... Networking too.. Primitive, but file sharing and printing worked just fine..
And Windows 95 was actually quite good... And good GUI for the era... It was expensive Macs or Windows 95...


Linux then was a joke if you wanted to use it for something useful... I know, I had separate PC running Linux so I can play with it.. It was fascinating learning tool.. nothing more..  Much different from today where today you can have a workstation running Linux and do actual work...

And Microsoft Reference Hardware platform was de facto industry standard, especially for those clones you mentioned.. Intel would make reference platform for motherboard and clone makers were at first simply implementing reference designs without changing a thing...

It is not I say it was Microsoft that we have to thank for being so great and worship it... I didn't say that.
All I said that hardware manufacturers wanted to be able to have "Compatible With Windows" sticker on their boxes  ( and they still do ) because they wanted to be compatible with 98% of PC computer owners...

That created standardization effort in industry, that drove volume, price drops etc., that, thankfully helped other platforms...