Brainstorming

[Howard Berkowitz]

>From: "Gustavo E. Flores" <gflores911 at gmail.com>
>Reply-To: "Trauma &amp; Critical Care mailing list" 
><trauma-list at trauma.org>
>To: "'Trauma &amp; Critical Care mailing list'" <trauma-list at trauma.org>
>Subject: RE: Brainstorming
>Date: Sat, 14 Jul 2007 13:17:00 -0400
>
>Perhaps Karim or anybody else who has experience with Wikipedia can share
>more knowledge about it.
>
>Wikipedia is updated by its own users... namely anybody.
>
>For security purposes, perhaps access to the editing can be
>password-protected to those from the list in order to guarantee accuracy
>and/or privacy regarding what is commented... but that is just a technical
>detail that can be decided along the way.
>
>An article can be created describing the nature of the incident. As users
>log in and add more content, sections within the article are created...
>namely the different sectors within the incident structure. The
>Wikipedia-type layout allows for users to simply click on the menu and jump
>into any specific section within the article. So, theoretically somebody
>could click on the incident article, and quickly jump into the logistics
>section where somebody who was organizing and/or requesting any specific
>resources created a section describing the particular needs.
>
>Anybody with more experience using Wiki can add more information. Anybody?

I do contribute to Wikipedia. For this application, I would require 
individual user accounts, although that might be tied to this list 
membership or something at the trauma.org site. Right now, I'm going in and 
massively reorganizing some computer networking articles that have grown and 
have much misinformation in them. The ability for virtually anyone to edit 
is both a strength and weakness of wikipedia.org.  Wiki software, in the 
general sense, can have more restricted editing capability.

If I may, I'd like to describe the network and servers that we used for the 
US Y2K Information Center, which might be an example of the kind of 
high-availability infrastructure that might underlie a disaster Wiki. This 
example should not be meant to be an absolute model, but I do believe there 
should be some review for rumor management and Wiki changes by people with 
mixed motives/qualification.

The Y2K Information Center Model was also intended for a situation where 
systems might be blowing up nationally and internationally, which is 
probably not the case for the kind of emergency support networks relevant to 
trauma.

The main operations center was in Washington DC, where the user accounts 
were managed, and, as appropriate, staff editors might edit individual 
contributions. There was a limited backup center in the then-FEMA part of 
the White House/Executive Office basement, and another edit facility in a 
hardened facility outside the Washington area.

Whichever operations center was working then forwarded the public (or 
somewhat restricted) information to two identical AT&T data centers, one in 
New York City and the other in San Diego. These web hosting centers were set 
up for high-volume query. To save cost, the data bases were not in strict 
lockstep synchronization, but one would receive new information from an 
operations center and then update the other. This resulted in about a 
1-minute discrepancy between the data bases, which was reasonable.

The operations center also could update other data centers that handled 
classified data, but this was a very small part of the data.

Again, I mention this in a context of an operations center principally being 
responsible for user account management and technical assistance, and the 
distributed web centers set up for high-volume access, and enough 
geographical dispersion that no plausible disaster (I'm not ruling out 
deliberate attack) could take out both faciities.

_________________________________________________________________
http://liveearth.msn.com