NMEA 2000 not "Open Standard" in UK

Thanks, Kenyon, but I'm not sure that your conclusion about NMEA 2000 is true. I think the N2K system meets a lot of the U.K.'s standards about open standards, though it does depend some on how you define "low cost" and "royalties":

6. Government defines “open standards” as standards which:
• result from and are maintained through an open, independent process;
• are approved by a recognised specification or standardisation organisation, for example W3C or ISO or equivalent.
• are thoroughly documented and publicly available at zero or low cost;
• have intellectual property made irrevocably available on a royalty free basis;

Ben,
I have to question your assertion that the NMEA meets these criteria. For someone looking to develop an N2K device or software package, they are looking at spending:

NMEA 2000 Standard $2,100 (One time fee)
Manufacturer Registration $2,200 (One time fee)
Product Code $350 (One time fee per product)
Product Certification $750 (One time fee per product)
Certification Tool $2,400 (One time fee)

Are these big numbers for Furuno, Raymarine, or Garmin? No, but they definitely aren't zero or low cost either. Also, the whole certification process per product sure sounds similar to a royalty to me.

I'm not saying that the NMEA or the standard is bad, I just don't think it's "Open" and I don't think they have done a good job fostering development of products. I think N2K would be far more pervasive in the industry if they changed their model and the end users would benefit.

OK, Kenyon, let's go round the mulberry bush one more time, using the UK standards standards:

Government defines “open standards” as standards which:

• result from and are maintained through an open, independent process -- NMEA 2000 definitely yes; written by volunteers from an organization anyone can join.

• are approved by a recognised specification or standardisation organisation, for example W3C or ISO or equivalent -- NMEA 2000 definitely yes: besides NMEA being a recognized standards organization itself, N2K is approved and listed by the International Electrotechnical Commission as IEC 61162-3

• are thoroughly documented and publicly available at zero or low cost -- like I said, this is a vague definition, but given that the costs you listed are a fraction of what one of those companies spends on a single boat show, and a fraction of what's spent on the engineering needed to apply N2K to a product (from what I've heard), I think you'd have hard time making a case that the NMEA fees are "high" costs.

• have intellectual property made irrevocably available on a royalty free basis -- I've looked further into royalties and don't think any of the NMEA fees count. Royalties are invariably defined in terms of usage payments or the sharing of revenues. Once a manufacturer pays its fees, there are no other payments to NMEA no matter how many units it sells.

You also failed to mention how software developers can now work with N2K via the Third Party Gateway for very modest fees. And the fact that several hardware manufacturers are selling N2K "compliant" products that are not certified and yet NMEA does nothing about it, as far as I know, except insist that "NMEA 2000 Certified" not be used unless it's true.

Kenyon, I understand wanting N2K hardware development to be easier for hobbyists and very small developers -- me too! -- but the attempt to turn the UK open standard definition into a condemnation of NMEA seems silly. In fact, I think it backfired on you.

Whoa Ben, easy there. I don't think I am turning anything into a general condemnation of the NMEA, and I don't think I'm being silly. I was just drawing parallels between the announcement by the UK government to many people's opinion of NMEA calling N2K an "Open" standard which I admit is a technicality, but I think it is a big one. Saying something is open or free is like calling someone a virgin. There is no half way, it either is or isn't. It is obviously not productive to continue this discussion as it unfortunately won't change anything about the situation and I regret digging it back up. As I have stated above, my problem with N2K is not the standard itself or really the NMEA, it is with how it is licensed and how that affects innovation within the space and an end user's ability to access and understand their own systems. I will leave you with one final analogy based on your post and we can agree to disagree.

Let's look at the HTML spec defined by the W3C. I have been using HTML since 1998 and have never been a member of the W3C, looked at who is a member, or really cared. All I want is to implement applications within the spec so that my intended audience can use them, that is all. So, I ask you what would the web be today if you, myself, every blogger, app developer, or person with a simple web page had to first pay to join the W3C, pay to become a manufacturer, and pay to have every app or page certified and licensed? I have no answer to that question, but I am fairly sure that it would have stifled innovation and the internet would not be what it is today.

Btw, I find no mention of Third pardy gateway's on the NMEA site.

OK, Kenyon, I did perhaps overstate the situation; I get exasperated. But what you clearly did do was try to turn the UK government definition of open standard into proof that NMEA 2000 is not open. But I think we've seen that you failed to get your facts right in your zeal to ratify your own preconception. NMEA 2000 does not involve royalties and seems like it would qualify under the UK definition fine. (Interesting side question: If it really did not qualify, what would be the alternate choice for a networked marine data protocol?)

And, no, "open standard" is not as black and white as virginity (which is why that UK is applying its own definition). I believe that the primary definition of an open standard is that anyone can use it, which is certainly true of NMEA 2000. A secondary definition is that it is entirely free, even documentation, product registration, etc.

Some people seem hung up on that secondary definition largely for the reason you describe, standards like HTML. What they tend to forget is that several major standards like that were financed largely by the public, via government and academic institutions. There really is no free lunch.

But NMEA 2000 doesn't have any support like that. There is one Coast Guard technical guy who's contributed a lot to the standard but I still think NMEA picks up his expenses for travelling to meetings and so forth. That's the sort of thing those fees go to. If NMEA 2000 was completely free, how do you suppose its developement would be funded?

Info on the Third Party Gateway here:

http://www.nmea.org/Assets/thirdpartyappsnmea_news.pdf

And at Actisense:

http://www.actisense.com/HTML/Products/Gateways/NMEA_2000_PC_Gateway_1/index.php

Sorry, Ben, but I don't think you are accurately characterizing what factors make for an open standard.

Of course there are various definitions of the term -- see the Wikipedia entry for many examples. But virtually all of the definitions include the requirement that the standard itself be publicly and freely accessible, which emphatically is not the case with N2K. The only proof needed of that is Kees' extensive and thorough reverse-engineering of the N2K sentences.

And while it is nice to have Actisense's NGT libraries available, that's only N2K filtered by some manufacturer's market perspective and development process. The actual documentation costs $2,100, making it unlikely that hobbyist developers will be able to write code to the standard (unless they get access to it via another path, such as reverse engineering it).

/afb

I think it's pretty impressive how skillfully the Wiki editors dealt with such a variety of definitions and the strong feelings behind them -- as black and white as virginity? not! -- but it seems a stretch to find zero cost documentation a characteristic "virtually all" definers agree to.

I don't think it's a mistake that the ITU definition of Open Standard is #1 on the Wiki list, and note how the ITU's core definition does not include no-cost documentation:

"'Open Standards' are standards made available to the general public and are developed (or approved) and maintained via a collaborative and consensus driven process. 'Open Standards' facilitate interoperability and data exchange among different products or services and are intended for widespread adoption."

Further along the ITU mentions "reasonable costs" but of course that is an eye-of-the-beholder term.

I think it's useful in these discussions to consider the opposite. I'd like to hear how the folks who claim that NMEA 2000 is not an open standard define a closed or proprietary standard. Will you take a crack at that, Adam and/or Kenyon?

And, again, put me on record as wishing that NMEA 2000 documentation was free or very inexpensive, which I have expressed directly to the NMEA management many times. Let's also note that they have changed the pricing structure over time and would probably object to your $2,100 figure because a developer can supposedly do much with smaller increments. I also support Kee's work and have supplied him with raw data (on the record again!). But I simply don't think that NMEA fees make NMEA 2000 a closed standard.

No problem. A closed standard is one that you cannot freely download, examine, play with, and evaluate.

Perhaps we'll have to agree to disagree here, but I've been doing this a while and have never heard of anything called an open standard that didn't simply have a Web page you could visit to download it. The "gold standards" developed by the IETF and W3C subcommittees are universally free, even when there are corporate representatives on the committees.

Indeed, I might offer a "counter-challenge": can you name an open standard that costs anything to download? Heck, the only standards I know of that are not freely available are ones like Blu-ray that are developed and controlled by corporations for their economic benefit. (Which are fine, by the way; just not open.)

/afb

NMEA 2000 is an open standard, it's approved by IEC and also known as IEC 61162-3.

Adam, please try downloading some of the open standards similar to NMEA 2000, like AIS or CANbus specs. Or most any electrical standard you might find at the ITU or IEC. I suspect that there are lots more standards outside the web world than within it, and that there's standards cultural gap in between. Hence all those definitions on Wikipedia.

Also, if NMEA 2000 is a closed standard, then what are NavNet3D, SeaTalk, etc.?