Re: Outcomes of the XG

-------------
SUMMARY

To encourage by every means possible adoption of
compatible systems before any disaster and therefore
to develop tests and protocols with variants for
long-term local administration and short-term relief,
takes explicit legal and social and human rights
analysis and even compromises to the goals of relief
agencies.  I think the mission statement must at least
say that.

I agree generally with the "collaborative, inclusive,
participatory" and "human rights and social justice"
goals and do not believe they can be omitted without
major problems later.  However it is not impossible to
rewrite them in terms that are more operational, more
legally robust, and much easier to argue with, so as
to encourage debate and refinement.  I suggest rather
than omitting the clauses Chamindra objects to the
following language:

 "To encourage collaboration between experts, local
authorities, relief specialists and non-governmental
agencies;  To develop common standards and protocols
for coordinating data gathered in anticipation of a
disaster or for other reasons by authorities or local
groups, early response tasks and gathering of field
data, passing off responsibility for records after an
emergency, and for privacy of records at all phases; 
To simplify the exchange of both the data itself and
best practices and to involve citizens and local
institutions in participatory resilience networks;  To
clarify the role of social software in EM situations; 
To encourage by every means possible adoption of
compatible systems before any disaster and therefore
to develop protocols with variants for long-term local
administration and short-term relief."

"To provide a working example of humanitarian software
designed with explicit design goals of ethical records
handling, privacy and safety of persons in need, and
transparency;  To make available an ontology in open
content form and reference implementations in free
software form;  To insist on disclosures and
share-alike consortium licensing for all participants
who exploit the architecture or ontology in for-profit
or proprietary work;  To define essential tests for EM
software interoperability;  To increase confidence of
donors, victims, aid specialists and governments that
human rights standards are upheld implicitly in the
design, the ontology, the consortium agreements, and
all related outputs."

Detailed rationale follows in which I defer to the W3,
Creative Commons and Java experiences.  Also I have
some suggestions for adopting a draft Charter first
pending literature review, continuing working via
wiki.  I end with some arguments why human rights and
social justice goals must be explicit in charters,
embedded in architectures (citing Lessig), and tend to
increase the accuracy of data and the motivation to
adopt compatible systems PRIOR to any disaster, which
must be among the primary goals.  

I believe the highest level goal is to develop
protocols in common between local administration and
relief systems and render them simply implementations
of the same architecture.  That goal is not explicit
in the mission statement yet.  Perhaps it should be. 
I think this must be discussed early.
-----------

--- Chamindra de Silva <chamindra@opensource.lk>
wrote:
> Renato Iannella wrote:
> > 
> > Thanks all. I will start on the Charter (using the
> W3C template) and 
> > re-use some of the contributions to date at 
> >
>
<http://esw.w3.org/topic/Charter_for_a_Proposed_W3C_Incubator_Group>.

What is the reasoning for not continuing the wiki
process, perhaps dividing a consensus version from an
annotated version ?  
 
> I think our charter is coming out very well in
> general.

I don't believe it's appropriate to finalize a charter
before more visions and questions have been debated
and more references reviewed.  However, a clearly
marked draft charter that will be modified and
finalized could serve the purpose of a "straw man" for
a while.  Say after a literature review of relevant
papers from governments and agencies (with a table of
comparison of goals that each of those has adopted, to
ensure that the final Charter uses the best terms).

> However I think it is important for us to be focused
> on our outcomes on this incubator group and I wonder
if the following
> points on our mission is too generic for this group:

I agree, they are not stated in operational terms and
can be restated much more operationally as follows.

> # To create a collaborative, inclusive,
> participatory emergency 
> management framework that takes into account human
> rights and social 
> equality, aimed to support EM systems designed to
> optimize the use of 
> available resources though the maximization of data,
> information, 
> knowledge and resources exchange

Instead consider:

 "To encourage collaboration between experts, local
authorities, relief specialists and non-governmental
agencies;  To develop common standards and protocols
for coordinating data gathered in anticipation of a
disaster or for other reasons by authorities or local
groups, early response tasks and gathering of field
data, passing off responsibility for records after an
emergency, and for privacy of records at all phases; 
To simplify the exchange of both the data itself and
best practices and to involve citizens and local
institutions in participatory resilience networks;  To
clarify the role of social software in EM situations; 
To encourage by every means possible adoption of
compatible systems before any disaster and therefore
to develop protocols with variants for long-term local
administration and short-term relief."

That will achieve the same goals but the statement is
about HOW it is to be achieved, which is what W3 does.
 It uses the language ("best practices", "resilience
networks") that evokes quality management and
participatory peer networks.  And states intent to
adopt and influence any W3 standards for social
software that evolve in future, and to better unify
long-term and short-term treatment of records/data.

> # To develop and publish an architecture of
> reference for humanitarian 
> software designed to embed and support principles of
> ethics and 
> transparency. with the aim to discourage
> profiteering and personal 
> and/or corporate financial gain from humanitarian
> operations

Instead consider:

"To provide a working example of humanitarian software
designed with explicit design goals of ethical records
handling, privacy and safety of persons in need, and
transparency;  To make available an ontology in open
content form and reference implementations in free
software form;  To insist on disclosures and
share-alike consortium licensing for all participants
who exploit the architecture or ontology in for-profit
or proprietary work;  To define essential tests for EM
software interoperability;  To increase confidence of
donors, victims, aid specialists and governments that
human rights standards are upheld implicitly in the
design, the ontology, the consortium agreements, and
all related outputs."

This addresses some very difficult legal constraints
and uses the appropriate language. 

W3 spends a lot of time on legal issues and in this
particular field it should be possible to achieve a
consensus on things like share-alike licensing (NOTE: 
"open source" is NOT all share-alike, many open source
licenses allow extensions to be made and kept private,
and open source does NOT permit non-sharing with those
who will abuse the software or create derivative works
that will abuse persons or fail to remain
interoperable.  Thus it is EXTREMELY UNWISE to use the
phrase "open source" in a mission statement for this
or any other humanitarian project;  All "open content"
and "free software" licenses are "share alike" in the
exact sense of the share-alike clause of Creative
Commons, which means, they CAN include restrictions on
who uses the software and for what, they CAN require
re-integration of improvements to be shared among all
legitimate users, NEITHER of which "open source" can).

If you aren't legally exact in the Charter then there
is no chance that share-alike principles will ever be
completely respected in the follow-on works.  The Java
consortium discovered that being less than completely
clear on what it meant to "be Java" led it into some
conflict with Microsoft.  The solutions they found
were legally robust and included things like trademark
protection and compatibility testing based on a single
test suite to determine what software could be called
"Java".  This is part of what a consortium has to do
and W3 is no exception.  Figuring out the essential
tests for EM software interoperability is thus a goal
as it affects who will be able to claim W3's approval.

Lawrence Lessig explains (in his book "Code") why any
human rights or values that are not explicitly part of
the design goals of a software architecture will end
up being compromised by any applications built on it. 
I believe he is correct.

In other words, the law itself IS a loose software
architecture (and a free and open one at that, mostly)
and software that performs sensitive tasks must always
be designed with ethical considerations clearly in
mind right down to the division of responsibilities of
modules, the algorithms and protocols employed, etc.. 


Where the law is unclear or contradictory, software is
not able to simply ignore it, but must probably
pre-empt it and default to a higher standard than any
law.

> IMO though the above are good objectives, it is too
> generic and goes 
> beyond ontology and interop. We can however include

I agree, but as rewritten, the implications for both
ontology and interoperability are clear to any
software architect.  We need not worry about what is
clear to a non-architect (i.e. programmers who disavow
"ontology"), but if there is anyone who doesn't see
the clear implications of the goals as I've rewritten
them, and agree that without those goals there will be
grave risks of proprietary conflicts and a tyranny of
small differences (between jurisdictions, between the
pre- and during and post-disaster phases of system
use, between various specialists and professions with
different standards for say how they gather personal
data, etc.), then we should probably keep discussing. 

> this in other forums 
> including Humanitarian-ICT
> (http://www.reliefsource.org/foss). In fact 
> we had some initial discussion on interop standards
> here which we can 
> now inherit this work to this group continue instead
> (notes: 
>
http://www.reliefsource.org/foss/index.php/Dev:Interop_Standards

While I'd encourage using wikis to continue to refine
these goals, I do not believe you can produce a good
ontology nor achieve interop goals without having an
explicit mission statement that emphasizes that there
will be as few differences as possible between the way
data is gathered and used by all the various parties.

Again the word "disaster" is problematic as it only
describes the event and relief effort.  Resilience and
recovery goals are seemingly excluded which will not
facilitate (and according to Lessig, not even ALLOW)
even a slight compromise to relief objectives to get
radically better integration, transparency, and trust.

Further excluding human rights and social justice
goals is likewise unwise as extremely personal and
sensitive information is being handled, and errors or
omissions will increase exponentially with mistrust of
the goals of the data gatherers, field agents, etc.. 
This may be especially true in a pandemic or conflict.

Accordingly, facilitating the maximum relevant data to
be gathered by local trusted parties long in advance
(such as lists of vulnerable persons) and immediately
made available to those engaged in relief efforts, and
ensuring that this data will remain updated and useful
during- and post-disaster, will increase its accuracy.

Where those local trusted parties are themselves part
of the relief and response effort, the need for relief
itself may be radically curtailed, the information may
be radically more accurate, making it easier to divert
scarce aid resources to where they're actually needed.

Given that such local trusted parties, such as NGOs or
religious institutions or local self-help groups or
citizens organizations, themselves usually have very
explicit human rights and social justice goals, there
may be significantly better take-up if W3 can say very
clearly that these goals are reflected in
architecture.  It is very hard to imagine
administrations ever adopting compatible systems, say
for the 
registration of vulnerable persons in an earthquake
zone, unless some such universal goals are explicit
and some long-term aid can be directed to this
purpose.  Not just to response but to prevention and
to
resilience.

I believe the highest level goal is to develop
protocols in common between local administration and
relief systems and render them simply implementations
of the same architecture.  That goal is not explicit
in the mission statement yet.  Unless and until it is,
will we not face incompatibilities or
legitimate resistances of all kinds, and is it
therefore not absolutely imperative to make that a
goal?

So if you take just one line, take this one:

"To encourage by every means possible adoption of
compatible systems before any disaster and therefore
to develop tests and protocols with variants for
long-term local administration and short-term relief."


 
____________________________________________________________________________________
Be a PS3 game guru.
Get your game face on with the latest PS3 news and previews at Yahoo! Games.
http://videogames.yahoo.com/platform?platform=120121

Received on Sunday, 17 June 2007 04:26:58 UTC