Re: Requirement for 3W interop standard (new proposed schema attached)

Paul

>
> People who want to design from the top down based on how they believe the
> humanitarian community should be operating.

Ah! or: people who want to design by imposing their limited views of
how the humanitatarian community operates and how they think it should
operate by looking only at one flaky example of their choice, which is
not even representative of how the humanitarian community works

> People who want to design from the bottom up based on how the humanitarian
> community actually operates.

Designing from the bottom up means: take into account all
stakeholders, and not just the institutional ones, your are mixing
things up deliberatelly?


> Can you explain exactly what gives you this impression? As far as I can
> tell, nobody has stated or implied that humanitarian responses are static.

I have not yet see a schema of a process, nor even a discussion about
it on this list. Response is represented as something that people
cannot interact with, and frozen into a 'service directory' type.


> A proposed schema should be complete, or it would result in incorrect
> representation of the response cycle
> A big -1 to trying to model the entire humanitarian response cycle,
> especially when the domain representation on this group is so limited.


Paul, by 'complete' I mean complete in the Tarskian sense, whereby
correctness and adherence to reality depend on completeness and
acuracy  etc etc
(I think we just come from different backgrounds and use different 'language')

http://philosophy.berkeley.edu/events/detail/444

> What's wrong with taking manageable slices of the response, building models
> based on actual experience and ensuring maximal flexibility and
> interoperability?

Sure, but that is not what is being done by taking a limited partial
approach and saying 'thats how the humanitarian sector works' Bull.
What is a managemenable slice of the response for the purpose of our
exercise is yet to be decided. I dont remember ever having a
discusssion


 >In complex environments, a single integrated model is  going to be
far more >brittle than a modular one, and people are going to be  less
likely to adopt it.

Paul, it a model that fails to represent the intended reality 'in  a
complete sense' is not truthful then it is flawed, biased, wrong.
Modularity is fine with me, provided the architecture of the module is
sound,

>
> cheers
>
> Pdm
>
>
>
> paola.dimaio@gmail.com wrote:
> Ok Chamindra

I can understand that very practically oriented developers
> find it
hard to work with abstractions, its a part of the problem

I can see
> two possible ways forward

1. we develop two versions of the schema, and we
> make sure that they
are compatible, then you go ahead and work with your
> preferred subset
and leave others to implement differently

2. I work with
> you to help you implement your preferred view of the
schema from a more
> complete and more abstract superset

Would any of the above solutions work
> for you? Do you see other
possible ways that we can reconcile our different
> views of how should
the 3W schema be modelled?


cheers

PDM

On Tue, Aug
> 12, 2008 at 3:04 AM, Chamindra de Silva
<chamindra@opensource.lk> wrote:

> I understand that Paola and I am not trying to constrain the goals of
> the
group, but help ground every decision in reality by having many
> incremental
cycle of the model validate itself with the actual systems it is
> going to be
applied in (currently only OCHA 3W and Sahana). If we come up
> with a very
abstract model and vocabulary that makes it hard to implement in
> Sahana or
OCHA 3W I think we have defeated the purpose of helping to improve
> the
ground realities of efficient information exchange in the times of
> a
disaster. There can be many iterations and multiple versions of this
> model,
but at least lets have a quick win by making sure the first iteration
> is of
value in enabling two systems to work together. This does not mean you
> stop
thinking in a generic way when coming up with the model.

On Tue, Aug
> 12, 2008 at 12:21 PM, <paola.dimaio@gmail.com> wrote:

> Chamindra

the point that is that while it is fully legitimate for you as
> Sahana
developer to create a system in whatever way you chose to model,
> it
would be unfair that your personal choices as a developer constrain
the
> (potential) global scope of this workgroup, which in its small way
is rather
> ambitious:make a contribution to how EM information web
based exchanged can
> be made more meaningful, representative, and
efficient. We can only achieve
> that if we are innovative, critical
and proactive in our modelling
> approach.

If we as a workgroup produce an integrated model, you will be
> able to
use it and apply according to your preference, including just
> adopting
a subset (part 1), without narrowing its overall
> capability.

PDM


On Tue, Aug 12, 2008 at 12:46 AM, Chamindra de
> Silva
<chamindra@opensource.lk> wrote:

> In Sahana we have these two as separate modules.

1) "Who is doing What
> Where" is the traditional 3W application called
the
Organization
> Registry.

2) "Who _needs_ What Where" is a bulletin board of people
> requesting aid
on
behalf of a victim group in the field called the (Aid)
> Request
Management
System. It also track pledges of aid.

The prior operates
> at a high level of services provided (e.g. medical,
sanitation, food, water)
> by a responder group across the affected area,
whilst the later works with
> units of aid needed specifically by a victim
group (e.g. 100 Tents)

I would
> prefer we stick to the traditional sense of the 3W (i.e. option
1)
to keep
> things simple for now and to help us can quickly get through the
full cycle
> up to an interop standard recommendation. We can always
improve
that
> standard and build it up incrementally from there, though
> I
completely
understand that everything is very closely related.

On Mon,
> Aug 11, 2008 at 9:08 PM, Nigel Snoad <nigelsno@microsoft.com>
wrote:

> Paola,



In the F2F in Washington DC we scoped the 3W/4W as described by
> Gavin.
I
completely agree that there must be a "needs" layer that is
> centered
around
the affected population (I detest the phrasing "victim" and
> can't too
strongly suggest we never use it except for law
> enforcement/human
rights
contexts) as well as the current "response" layer.
> Thankfully, finally,
the
humanitarian clusters are starting to talk about
> this in their data
models,
and definitely affected populations must included
> in the incubator's
data
model from the start.



So – we have a semantic
> confusion about how we should scope "who". One
is
organizational, and one is
> affected populations. In the 3W context for
historical reasons it's the
> organization/group providing
assistance/services
(of course this usually
> includes the affected population themselves,
something usually ignored in
> the UN context). Usefully - from a data
perspective responding organizations
> "need" assistance as well – goods,
staff and services – to continue their
> work, and they, like affected
populations, provide capabilities. I like the
> thought of a symmetric
integrated model along these lines.



So - I's no
> news to all of us that the scope of a solution/application
affects which
> components of a data model are used. The 3W/4W focuses on
"response".



My
> suggestion is that when discussing the 3W/4W use case we confine the
"who"
> to organization providing services, but in the data models that
come
out we
> ensure that the who are subclassed/flagged into both a "needs"
component
> including affected groups and
> organizations
requiring/recieving
support/supplies/services, and a
> "response" component that includes
capabilities and
> activities/outcomes/assistance/services
> provided.



Nigel



From:
> public-xg-eiif-request@w3.org
[mailto:public-xg-eiif-request@w3.org]
On
> Behalf Of paola.dimaio@gmail.com
Sent: Sunday, August 10, 2008 8:38 AM
To:
> Gavin Treadgold
Cc: public-xg-eiif
Subject: Re: Requirement for 3W interop
> standard (new proposed schema
attached)



Gavin


My understanding is the
> 3W is 'just' a directory application, hence the
schema is designed around
> providing directory services.

May I ask what is that assumption based
> on?
Did we as a group discuss/agree on such a constraint?
 Is there any more
> useful purpose for which we need a 3W metaset?
Is the schema for a service
> directory part of our mission ?


assuming 'directory' is accetaptable
> description for everybody, it
should
be designed
to be flexible to
> accommodate for all stakeholder requirements, so we
definetely gotta
> talk




--
Paola Di Maio
School of
> IT
www.mfu.ac.th
*********************************************
>
--
Paola Di Maio
School of
> IT
www.mfu.ac.th
*********************************************
>
>


>


-- 
Paola Di Maio
School of IT
www.mfu.ac.th
*********************************************

Received on Tuesday, 12 August 2008 12:26:47 UTC