[sc34wg3] WG3 Focus

Graham Moore gra at networkedplanet.com
Wed Nov 2 03:16:59 EDT 2011

Morning Patrick, All

My contribution for Friday:

SDShare really should be completed and for me things like SDShare are
the kind of standardisation work that WG3 should focus on. Below I
breakdown SDShare into several general aspects that I think could
inform the work that WG3 does.

Use of Existing Standards
There are lots of standard out there and creating new ones just adds
to the pile. However, reusing, incorporating and referencing existing
ones strengthens both the new standard and helps to focus adoption of
the others. The incentive to organisations to adopt a given standard
increases if the number of general uses it has it bigger.

SDShare uses the ATOM, RDF and XTM, standards to deliver its value. It
didn't reinvent new semantic formats, or payload formats, it used what
was there.

Small Amount of Innovation

In WG3 we tend to invent rather than standardise and this raises the
question why are these invention activities are happening in a
standards body? I'm not necessarily against this invention aspect, as
the WG brings together lots of smart people and that should result in
something good. Part of the problem with this in the past, HyTime,
Topic Maps, is that these are BIG things and the lots of smart people
have lots of different ways and ideas on how to invent these things
big things. As they are inventions rather than standardisation there
is no real deployments to reference, and no communities of practice to
unite. This sometimes :) leads to quite differing positions, which are
hard to resolve and slow down / prevent us finishing work.

SDShare came together originally in a CEN standards workshop and even
with its limited goals, sharing semantic descriptions / metadata
between government agencies, we had 'tough' technical discussions over
things such as SOAP versus REST. But these were resolvable issues not
fundamental disagreements or intractable positions. SDShare is limited
in newness, but I think it has just enough newness to make it exciting
and shiny. It glues together well known and 'mature' standards to
deliver value, to provide guidance and address real issues.

So perhaps in general WG3's innovation should be about how to we reach
some of the goals that Steve N talked about by adding the bit of WG3
smarts to things that are out there. Inventing just enough to empower
a new community, inspire a new generation of developers, provide clear
guidance on how stuff can work together.

Clear Business and User Value Proposition

Topic Maps, RDF, HyTime, SGML, XML. Yep they all have business value,
but in general its not immediatley clear to many people (managers,
organisations, developers, bob on the street). I'm not saying WG3
should ignore the larger problems that need more meta solutions, but
that we can do a lot of good by addressing things lower down the
complexity order.

SDShare started out focused on Government Agencies and how they share
data across boudaries. It dealt with the political issues that people
don't want to share access to their underlying systems, nor do they
want to cede control of their data into another silo. It was aimed at
offering an alternative to copying and siloisation of EU data,
reducing costs. Now, its starting to get traction in large
mutli-national organisations who internally have many of the same
political and technical issues.

SDShare is a meta-technology but started out with some fairly well
focused, middle ground, goals. What I mean by this is that we didn't
focus on one government agency with one problem and fix it, we tried
to understand the space and come up with a generalised solution. But
we didn't focus on how to allow any system of metadata to talk to any
other system of metadata.

Clear Standardisation Value

There needs to be communities that will benefit from the
standardisation of something. There will always be some people who
benefit from invention, but to benefit from standardisation is
different. Adoption of a standard increases the fluidity, speed and
accurrancy with which organisations through the systems they have in
place can communicate. I think this goes a long way to meeting some of
SteveN's goals.

Easy / Concrete Enough To Implement

SDShare has traction as there are lots of tools out there to deal with
ATOM, RDF and XTM and the extra bit can be implemented quickly by most
developers. WG3 should consider the burden in places on developers and
implementors with the work it produces. Some things are simply
complicated and cannot be reduced, but maybe through better packaging,
by considering who will implement standards we can help increase

Mindshare with the Developers

Still today you say triple or generic object or SPARQL / TMQL to the
average developer on the street and they look back blankly. The world
of information and information management is controlled by the Java,
Python, .NET developers of this world. They do not think in terms of
information, concepts, subjects, they think in terms of UI, web
services, functions and classes. RDF has marginally better mindshare
than Topic Maps, but not really much. OData (Microsofts Open Data
Standard), which is a nicely written standard, uses ATOM, REST URI
etc, but not a very good Open Data one.) is already getting more / has
the same traction as RDF / Topic Maps. Why? I don't buy that its just
becuase Microsoft are pushing it, yes it is in SharePoint, yes there
is an Excel plug-in for it, yes there are Visual Studio tools for it.
All of these developement things could have been done by a
non-microsoft company. No, the key thing is that the deveopers have
been embraced. The tools are not alien, the concepts are an evolution
of what they know and understand.

Although we are dealing with information management in WG3 the people
who build the systems are developers and architects and to get
adoption of these standards we should be embracing and extending how
they work, much as we should reuse and add the magic dust to existing
standards and best practice.

Principled Approach or Why this way?

There are many ways to solve a problem and part of what WG3 should be
focused on is using experience and insight to put together standards
that embody the wider way we see things. For example, SDShare problems
could be solved by installing a monster IBM or Microsoft Message Queue
system. Why don't we do that? Well SDShare embodies openess, any
client can ask what has changed, all resource representations are URL
addressable, new clients can be added and join the game with minimal
overhead. The message formats are not proprietary, the messages can be
processed by many different tools of many platforms.

We embodied our values of openess and transparency in the way we put
together the standard. WG3 standards will get adoption firstly if they
help communities and orgisations but also because of the inherent
values they exude. Adopters of WG3 standards believe in what we

On 2 November 2011 00:55, Patrick Durusau <patrick at durusau.net> wrote:
> Greetings!
> In case you hadn't noticed, today is 1 November. Which by my calendar means
> that this coming Friday is the 4th of November.
> We are scheduled to have a WG3 meeting on Friday and I am going to send out
> the teleconference details tomorrow.
> Just wanted to give anyone who is typing furiously a little nudge about
> typing even faster. ;-) Marks are not taken off for typos.
> BTW, I deeply appreciated the submissions made thus far (Steve N. and
> Michel) but also look forward to additional submissions.
> Hope everyone is having a great week!
> Patrick
> --
> Patrick Durusau
> patrick at durusau.net
> Chair, V1 - US TAG to JTC 1/SC 34
> Convener, JTC 1/SC 34/WG 3 (Topic Maps)
> Editor, OpenDocument Format TC (OASIS), Project Editor ISO/IEC 26300
> Co-Editor, ISO/IEC 13250-1, 13250-5 (Topic Maps)
> Another Word For It (blog): http://tm.durusau.net
> Homepage: http://www.durusau.net
> Twitter: patrickDurusau
> _______________________________________________
> topicmapmail mailing list
> topicmapmail at infoloom.com
> http://www.infoloom.com/mailman/listinfo/topicmapmail

Graham Moore, Director, Networked Planet Limited
Editor XTM 1.0, ISO13250 (TopicMaps) -2,-3, TMCL
e: graham.moore at networkedplanet.com
w: www.networkedplanet.com
t: +44 1865 811131
m: +47 90056479   (Norway)

Networked Planet Limited is registered in England and Wales, no. 5273377

More information about the sc34wg3 mailing list