[EpiData-list] Status and plans for development (as of February 2010)

epidata-list at lists.umanitoba.ca epidata-list at lists.umanitoba.ca
Mon Feb 22 14:45:31 CST 2010


This is a brief status of developments - (rather technical in content)

It is reassuring to see, that the interest in testing is rather large
(Approximate download counts for latest version: Linux 50 , Mac 75 and
Windows 300). Few comments have been given on the list, but I take that
as a positive indication - I hope I am right, otherwise please notify on
the list.

The short story is:
We are definitely moving forward - the basic way of working in graphic
mode with data definition is more or less in place. But we must spend
time on implementing related data structures and finding a final data
structure. Therefore the first EpiData Entry Client based on the new 
principles will not come out until some time from now.

The longer story is:
During the most recent decade use of standardised data formats in public
sectors and research has become an increasingly important issue. The
idea of this is that data should be usable in any application and system
based on published verifiable standards (XML based). XML is one of the
structured information formats for machine readable information
published by the W3C organisation - and is very complex - other ones
from W3C are css and html. The published standards for XML based data
documentation and storage of data as such are unfortunately plenty, a
few of these are DDI (see DDI-alliance.org), CDISC ODM  (see cdisc.org),
SDMX (see sdmx.org). It seems as medicine, national statistics and the
social sciences each have developed their own. The only common
denominator seems to be in relation to overall descriptions such as
"title", "contributor", abstract etc. These terms are well known from
citations and are based on a metadata set called Dublin Core Metadata
(see http://dublincore.org/specifications/).

For implementation of a XML based file format in the EpiData Software it
would seem obvious to NOT develop a new one, but instead use one or more
of the mentioned ones. Therefore there have been discussions with data
documentation specialists at the Danish Data Archives (see
http://samfund.dda.dk/default-en.asp).

Following these discussions and after studying several published
standards it has been decided that the EpiData Software XML format will
conform to the overall principles of the mentioned standards, but also
that we must work with a restricted set of these.

The nature of the EpiData project is to be focused on simplicity and we
must stick to this principle. Otherwise we would be lost in months of
generalisation, of no interest to the general user and also completely
unrealistic within the economical and personal resources available.

The time frame for the next phases are uncertain, but I will give
orientation on this list as improvements are made and obstacles removed.
Most likely there will be a meeting in the second week of March in Paris
for detailed discussions on some of these issues. More information on
that later.

regards
Jens Lauritsen
EpiData Association



More information about the EpiData-list mailing list