Home    Bloggers    Messages    Webinars    Resources   
Tw  |  Fb  |  In  |  Rss
Mike Field

The Openness of Open-Source Hardware

Mike Field

I'm on a few emailing lists. I'm not sure about your lists, but each of mine always has at least one person I like to call an open-source fundamentalist (OSF).

You know the sort of person I mean. Whenever anybody says, "I have found X, and it is really neat," the OSF's reply is "Bah! It doesn't work on Linux!" If it does work on Linux, the OSF's reply is "Bah! But they don't publish their API!" If the API is published, then we hear, "Bah! They don't publish their source code!" If they do publish the code, it's "Bah! It's not under an open license!" And if it is under an open license, it's "Bah! That license is no good." And on and on.

Take the Raspberry Pi, a cheap-as-chips ARM-based Linux system based on the Broadcom BCM2835 system-on-chip (SoC). It uses an unconventional boot process: A hidden processor within the SoC loads its firmware from the SD card. Once booted, the hidden processor loads the main ARM processor's boot image into memory and lets the ARM processor loose.

The true OSF will reject the Raspberry Pi because it has closed-source components. Here's a comment from Andrew in the Element14 forums: "Can we just set the record straight: it is not open source hardware -- the schematics and/or Gerber files etc. are not provided and made available under a suitably liberal licence."

Being a pragmatic person, I would have thought FPGAs would be the OSF's nirvana. You can have the full design for your entire system, including the CPU and peripherals -- all laid bare for you to use and improve. What more could one hope for? This isn't pie in the sky. It's all real right now. For example, the OpenRISC core exists and is included in the Linux 3.1 kernel.

Strangely enough, this still isn't good enough for the hard core (pun intended) OSF. The familiar cries are heard once again: The EDA tools don't run on Linux, the EDA tools are not open-source, no open-source synthesis tools exist, the FPGA configuration files are closed, some of the IP modules are closed-source, and so on.

I struggle to reconcile how the same people who laud the Arduino for its "openness" would deride FPGA projects such as the AVR8 processor core by saying the required EDA tools are not open.

My blood boils! Do OSFs have the Gerber files for their laptop PCBs? Do they have the source code to their laptop's BIOS? Do they have the RTL code for their CPU or graphics chips?

I think open-source fundamentalists should embrace projects like OpenCores.org, OpenRISC, Papilio, ZPU, and others. These projects are bringing true openness to the hardware arena, even though they rely on EDA toolchains that may be shrouded in mystery.

I see OSFs as being the sort of people who can discuss over a steak dinner how they could never kill animals for food. They just don't want to get their hands dirty. They should drink a big cup of cement, buy an FPGA board, and start building their open-source world one module at a time!

Page 1 / 2   >   >>
JezmoSSL
JezmoSSL
2/2/2013 4:21:15 PM
User Rank Blogger
Re: gEDA
Look at the project my company is working on, one of the deliverables is all of the design files which would allow our customer to have our design built by some sufficiently competent company should we the originator of the design go bust. Now you tell me is that open or closed source?

50%
50%
JezmoSSL
JezmoSSL
2/2/2013 4:10:50 PM
User Rank Blogger
Re: gEDA
To a certain extent companies make their money from closed source designs but take a look at opencores,org and you see some playing a rather clever game, post a core as Open source only it doesn't do quite what most people are looking for so it's upto you, do you modify the open source to do what you want, or do you phone the publisher and get them to do it for you? Obviously they hope the later As for Linux gimps, well I've dealt with them before and there is no talking to them. The raspberry pi is a good example of how they all delude themselves, it runs Linux so softies get excited, although Linux isn't what you might call a real time O.S ,it's hardware isn't open but they ignore that, because wow!!! It runs Linux!!!!! Sadly opencores forums have been taken over by the sort of people who wonder why they cannot find 50 micro-amp fuses

50%
50%
stanley82
stanley82
8/24/2012 10:09:38 PM
User Rank Beginner
gEDA
gEDA and other free tools I have found to be good for trailing edge technology.  Not all of us make a living from the latest and greatest G bit data or 128 bit differential signal technology.  There must be a huge $ investment in all these tools and chips.

50%
50%
devel@latke.net
devel@latke.net
8/24/2012 1:45:19 PM
User Rank Guru
Re: I'm an Omnivore
Since I make my living designing electronics hardware, the idea of giving it away seems to be a bit problematic.

Also, regarding the open-hardware-fundamentalists/purists. Face it: there are no free (as in speech) FPGA synthesis and implementation tools. And there's no useful free (as in speech) VHDL simulators.

On the PCB side, Kicad and gEDA are not ready for prime time. Sure, people have implemented stuff with them. But compared to commercial tools, it's like using PALASM instead of VHDL. 

I will say that SDCC, at least for the 8051, works well and is supported by Silicon Labs for their (non-free-as-in-speech) debug tools.

So until the open-source developement tools for embedded systems mature, the open-hardware-Taliban will have to shut up.

100%
0%
Duane Benson
Duane Benson
8/20/2012 1:35:16 AM
User Rank Blogger
Other rants
Here's what amazes me about some of the OSHW fanatics:

In case you haven't heard about it, Makerbot is an inexpensive (under $1,000) 3D printer. It's become pretty popular in the maker community. As of this writing, there's a person on Kickstarter trying to raise money for a lower cost clone. According to the Kickstarter description, it's exactly the same and all of the parts are interchangeable with parts from the Makerbot.

The purist circles are up in arms about this. The license says it's perfectly legal to copy a design bit for it and sell it, which is what this person is doing. However, these folks consider it bad form and a violation of the unwritten rules to just copy it exactly without adding some level of additional value.

I'm not big on exact copies of someone else's design, but what he's doing is fully compliant with the license. If the folks who came up with the open source hardware licenses didn't want exact copies; they should have written that into the license. It's too late to complain now. If the rule is so important, it should be written.

100%
0%
William Murray
William Murray
8/19/2012 4:11:44 PM
User Rank Blogger
I'm an Omnivore
I get by on many free tools -- some open -- some closed -- mostly free --  When your working in a 50 year old building with Bad A/C in August - A fan may have a higher priority than a fancy S/W package/license fee --

50%
50%
Crusty
Crusty
8/18/2012 4:17:21 PM
User Rank Blogger
Re: Excellent rant!
Hi Hamster,

I think I will put OSF in my lexicon alongside Luddite.

I have a suspicion that OSF very rarely produce any open source of their own.

I have no problem with black box technology in hardware or software, as long as it works and the bugs, if apparent  are dealt with fast.

 

50%
50%
jacklsw86
jacklsw86
8/18/2012 4:21:55 AM
User Rank Blogger
Re: Thought provoking, Hamster..
using or working with the open source stuffs is not so bad. the main @55holes are those legal staffs defining it.

50%
50%
Scorpiion
Scorpiion
8/17/2012 6:54:14 PM
User Rank Beginner
Good post
Hi, I just wanted to say it was a good post on an interesting subject.

I have posted similar posts on google plus before but people do not always want to discuess the question, or I simply have to few reader...

Not sure if there are any rules agaist links (?) but people who thinks that's Mike's post was interesting might like reading these posts of your truly:
https://plus.google.com/101372835757809463996/posts/9R6nvS3pAhk
https://plus.google.com/101372835757809463996/posts/6Jzp181BVxF


As I said, I don't mean to breaks any rules by linking to other sites (if there are such rules here), I just thought these links could be interesting to people that find this post interesting.

50%
50%
KenwickVS
KenwickVS
8/17/2012 6:33:59 PM
User Rank Blogger
Thought provoking, Hamster..
Your "rant" is really about a group of people that we can call 'extremists' - rather than the pros and cons of the Open Source "movement" itself. Yet, most comments so far are discussing the validity of Open Source for all but the hobbyist venture. I think it's a different subject - but worthy of discussion.

Back in the day, as a young designer, I was enamored by Digital Equipment Corp - primarily because they openly published detailed specs, including schematics, for their PDP11 computer systems and operating systems for all to see - unlike most other computer companies of the time. As a result, many university programs used the PDP as a teaching and lab tool and when students went to work, they applied what they learned and PDP's ended up in many applications that DEC couldn't address alone. While DEC wasn't producing Open Source hardware and software by today's definition, it was this openness that helped them prosper in those days.

Back to your Blog point: A DECUS (DEC UserS group) convention was full of 'extremists' believing anything but DEC was crap. And, so it goes.

50%
50%
Page 1 / 2   >   >>
More Blogs from Mike Field
Over the past few days I have managed to create quite a few designs using the Xilinx Memory Interface Generator (MIG) and they all seem to work.
If I were an evil genius working on a plan for world domination (with regard to enterprise-level data storage solutions) I would be seriously considering building my design around a Zynq All Programmable SoC.
I would like to present to fellow readers of All Programmable Planet a new technique that I have invented to serialize data within the FPGA's main fabric at 1.5Gb/s.
It's time to look at the receiver portion of the link. After a few false starts, Mike Field has found what looks to be a workable design.
As with most things, my feeling is that there is no better way to understand high-speed serial links than to implement one from the ground up, so that is what I've set out to do.
flash poll
information resources
sponsored content by Xilinx
All resources
follow us on twitter
follow Xilinx on twitter
like us on facebook
like Xilinx on facebook
All Programmable Planet     About Us     Contact Us     Help     Register     Twitter     Facebook     RSS