Email from Daniel Kjar to Jorge Soberon, 20 July, 2005

Date: Wed, 20 Jul 2005 12:10:09 -0400
From: Daniel Kjar
To: Jorge Soberon
CC: Richard Smith , John Pickering ,,,,,,,,,,,,,,,,,,,,,,,,,,,, Dr S Soetikno ,,,,,,,,,,,
Subject: Re: Technology centers + search tool

Hello Jorge,

As Pick is off in Malaysia and Borneo I wanted to reply to your
email(and I am sure I will get a nasty email from Pick later about how
in fact he is trying to take over the world, and I should just keep my
mouth shut).

I am one of the collaborators working with Pick (on ants for the most
part) and do quite a bit of database work (currently working on a
project with the ant types of the Smithsonian). I do not have anything
to add on the international relations questions, as that is beyond my
scope, but I can clear some things up about what Discoverlife is, and
what it is not.

GBIF, Digir, and the other acronyms mentioned are all good efforts, and
discoverlife is not attempting to dispose of them. On the contrary one
of the main advantages of discoverlife is that current data formats or
applications can continue to be used if your institution or lab decides
to work with discoverlife. All that discoverlife does is go into your
existing database structure and present it seemlessly along with any
other persons data, without concern for how your fields are labeled, or
what format you like to use for a particular field. For example, my
data from Georgetown University is in a much different database
structure than that from MOBOT. Although Discoverlife could and can
replace these other services, that is not the intention. Discoverlife
is meant to be a painless way to add more functionality to existing
projects, and incorporate that information with useful tools like the
IDguides, label making utilities, and the mapping program. The other
value added to participation in discoverlife is a very user friendly
interface, which is a rarity in many of the older endeavors.

As far as technology/servers/storage is concerned, that is also another
*advantage* of the Discoverlife initiative. We use proven open
technology (PERL, Solaris, APACHE, all of which are free of charge) that
consumes very few resources, and is scalable to the nth degree. We have
over 10 servers online at the University of Georgia, and MOBOT, and a
large server at Georgetown University that has not even been brought
online yet. Storage is down to less than a dollar a GB and total
operational costs, when running full tilt in the future, are estimated
to be very low, in the 150,000 dollars range, plus end user costs
(internet connection). The entire platform is centralized with all
photo editing, and other processing tasks occuring on the servers, so
end user platforms only require a web browser (and even a very old one
will work).

Participation with Discoverlife can be centralized or distributed. You
can maintain your own current work and also include your data in
discoverlife, or you can move everything to discoverlife, or a little of
both. Data possession is always with the institution. If pick upset me
today, I could delete my local database and tomorrow all of my
information would be gone from Discoverlife. If I am worried about a
species and do not want it to appear on discoverlife, I could arrange a
way to flag sensitive data with Pick, or remove it entirely.

Although I am sure Pick would like it if everyone wanted to put their
information on discoverlife's servers, that is not necessary. Although
in some cases where large files (such as pictures) are not being served
because of technological and financial constraints in the institution of
origin, that may be a very good idea.

Daniel Kjar

Daniel Kjar
Graduate Student
Biology Department
Georgetown University

Discover Life | Search | Prospectus - International centers | Correspondence | Top