www-legal-discuss mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From "Toby Considine" <Toby.Consid...@gmail.com>
Subject RE: Open Standards and Open Source
Date Sat, 21 Feb 2015 16:13:54 GMT
Since Wi-Fi was introduced to the thread, let me introduce yet another aspect to this thread.

The 802.11 standard is relatively immutable. The years of voting and politics to get to an
IEEE standard are past. It is also next to useless an as interoperation standard. You hand
three teams that spec, and you will get three non-interoperating implementations. 

This is a good thing rather than a bad one. The standard is so loose that it allows plenty
of innovation. Folks are always pushing the limits of the standard further than anyone thought
it could go. The alphabet soup of Wi-Fi standards are interoperable profiles within the standard
that codify emerging best practices at greater speeds without submitting to the process heavy
path to upgrading the standard. 

802.11 is also an incomplete standard. It does not take security into consideration at all.
Security standards evolve at their own pace, driven by forces quite different than generic
communications. Numerous variants of security have been bundled and re-bundled into the underlying
spec over time.

This approach results in innovation by "variation in kind". Because such variations share
base assumptions, it is far easier to design for diversity. Systems negotiate down from the
top-performing variant they know to the one they know in common. Much of the code modules
(or micro-code modules) are identical.

So, invoking the Wi-Fi Model

Innovate freely, but don't expect interoperation.
Occasionally, but rarely, go through the heavy duty process of defining a standard--but don't
make it too tight.
Many of the most commonly used standards are designed for composition. This is the lesson
of the 7 layer ISO stack, but the layers evolve, and the stack is multi-dimensional.
Innovate most of the time within the Standard, and profile when you want interoperation.

This is the approach we used with the smart energy standards, now being developed and profiled
by the OpenADR Alliance and others. (Now if we only had some open source there...)

The notion of rebranding is a useful one. HTTP was called out in this thread. WebSocket is
based firmly in HTTP, but is branded as its own variant. The Internet of Things seems to be
congealing around a couple communication standards, including MQTT. MQTT is even lighter weight
than and quite different from HTTP, but the developers (of the specification) have constrained
themselves so it can be transformed freely into and out of WebSocket, and indeed have requested
that it be considered a variant of WebSocket for IETF purposes.

Standards tend toward two types. Some strive to make a commodity--think standards for paper.
I don't want a special type of paper, I just want paper that won't jam in my printer, and
will look good enough. Once I achieve those commodity function, I am interested in price,
not novelty. Others tend toward increasing innovation--think Linux, Apache, or even 802.11.
Both sorts of standards can support open source, but in different ways. Profiles of an innovation-oriented
standard may tend toward acting as a commodity-tending standard. As a practitioner, one should
understand how this standard is acting in this instance...


tc

"There are seldom good technological solutions to behavioral problems." 
-- Ed Crowley

Toby Considine
TC9, Inc.
Toby.Considine@gmail.com
Phone: (919)619-2104
http://www.tc9.com 
  
Chair, OASIS OBIX Technical Committee
Chair, OASIS WS-Calendar Technical Committee
Editor, OASIS Energy Market Information Exchange (EMIX) 
Editor, OASIS Energy Interoperation
blog: http://www.NewDaedalus.com  


-----Original Message-----
From: Lawrence Rosen [mailto:lrosen@rosenlaw.com] 
Sent: Friday, February 20, 2015 10:30 AM
To: legal-discuss@apache.org
Cc: Lawrence Rosen
Subject: RE: Open Standards and Open Source

Rob Weir wrote:
> But if I fork, say, the HTTP standard, then I potentially harm a network of producers
and consumers of that protocol, by introducing incompatibilities.  

... Or help them by introducing improvements and beneficial features. The point is that the
standards organization and copyright law have no say in THAT matter. Software freedom is the
rule and freedom has its potential costs. 

> I also hurt the brand behind that standard by detracting from the public perception of
interoperability.

Nobody objects to using brands and encouraging interoperability for all software. If you want
to protect the HTTP standard and its interoperability promises by branding it (appropriately)
and insisting that no forks use that brand in their marketing -- do so using trademark law!
That's how Apache projects protect their reputations and software quality without prohibiting
forks. That was the motivation behind the Java brand. That's how Linux allows many forks even
with a (registered trademark) brand. 

Calling something an official "W3C" or "OASIS" standard carries lots of weight because such
names are already well-known trademarks that are already making valuable promises of interoperability.

/Larry


-----Original Message-----
From: Rob Weir [mailto:rob@robweir.com]
Sent: Thursday, February 19, 2015 7:23 AM
To: legal-discuss@apache.org; Lawrence Rosen
Subject: Re: Open Standards and Open Source

On Sat, Feb 14, 2015 at 1:39 PM, Lawrence Rosen <lrosen@rosenlaw.com> wrote:
> To: Apache friends
>
>
>
> I'm going to be at the Linux Foundation Collaboration Summit 2015 in 
> Santa Rosa next week. If there are any Apache folks there, please say 
> hello. There is an extensive legal track with some really good speakers and topics.
> (Sorry, I believe it is too late to register.)
>
>
>
> I'll be speaking on Thursday about Open Standards, a topic that has 
> occasionally interested the participants on this legal-discuss@ list.
> I'm mostly avoiding patents in my talk because important software 
> standards organizations in the US already have strong royalty-free patent policies.
> Instead, I've been focused recently on the gymnastics performed by 
> some standards organizations and their major large members to avoid 
> giving open source communities the copyright to create derivative 
> works. I will also discuss how these policies contradict the same 
> companies' own arguments in the Oracle v. Google lawsuit, which is 
> also about the effect of copyright law on open standards.
>
>

Hi Larry,

I thought the distinction was:

1) Open source as a basis for innovation, to build upon

2) Open standards as a codification of best practices, for interoperability in a given domain


If I fork an open source project and improve it (or make it worse for
that matter) I influence only users of my fork.   But if I fork, say,
the HTTP standard, then I potentially harm a network of producers and consumers of that protocol,
by introducing incompatibilities.  I also hurt the brand behind that standard by detracting
from the public perception of interoperability.

That isn't to say that the technology behind standards cannot and should not be improved,
but that a published standard is a "slice in time" of that technology stream, intended to
give stability and interoperability.

IMH (and personal) O, the "golden median" here would be a licence that allows derivation,
but only in clearly technically incompatible ways.
 For example, if you are going to design an incompatible Wifi standard, then don't put it
on the same frequencies as existing Wifi standards and certainly don't use the name of an
existing standard.

Do you disagree that this distinction and goal
(interoperability/stability) is important?   If so, how else would you
achieve that goal?

Regards,

-Rob

>
> I'm reading an interesting book about The Innovators: How a Group of 
> Hackers, Geniuses, and Geeks Created the Digital Revolution, by Walter 
> Isaacson. He writes:
>
>
>
> "These patent disputes [about the invention of the computer] were the 
> forerunner of a major issue of the digital era: Should intellectual 
> property be shared freely and placed whenever possible into the public 
> domain and open-source commons? That course, largely followed by 
> developers of the Internet and the Web, can spur innovation through 
> the rapid dissemination and crowdsourced improvement of ideas. Or 
> should intellectual property rights be protected and inventors allowed 
> to profit from their proprietary ideas and innovations? That path, 
> largely followed in the computer hardware, electronics, and 
> semiconductor industries, can provide the financial incentives and 
> capital investment that encourages innovation and rewards risk.... In
> 2011 a milestone was reached: Apple and Google spent more on lawsuits 
> and payments involving patents than they did on research and 
> development of new products." [Citing Charles Duhigg and Steve Lohr, 
> "The Patent, Used as a Sword," New York Times, Oct. 7, 2012.]
>
>
>
> Lawrence Rosen
>
> "If this were legal advice it would have been accompanied by a bill."

---------------------------------------------------------------------
To unsubscribe, e-mail: legal-discuss-unsubscribe@apache.org
For additional commands, e-mail: legal-discuss-help@apache.org


---------------------------------------------------------------------
To unsubscribe, e-mail: legal-discuss-unsubscribe@apache.org
For additional commands, e-mail: legal-discuss-help@apache.org



---------------------------------------------------------------------
To unsubscribe, e-mail: legal-discuss-unsubscribe@apache.org
For additional commands, e-mail: legal-discuss-help@apache.org


Mime
View raw message