Theo van Hintum
Centre for Genetic Resources the Netherlands (CGN), Centre for Plant Breeding and Reproduction Research, P.O. Box 16, 6700 AA Wageningen, The Netherlands
 

The easiest way of providing access to a database via the Internet is by creating a downloadable version of the database, or parts of it, and putting it ‘on the net’. It allows any user connected to the Internet to download it and use it the way (s)he likes.

Pros and cons
Providing downloadable databases should be considered a complementary service to providing on-line searchable databases. Both approaches have their pros and cons.
Some advantages of downloadable databases are:

• Downloadable databases are very easy to prepare. Often an off-line version of the database is already available for distribution via floppy.
• Down-loaded databases can be analyzed and processed by the user. Many users need the possibility to load the database on their local computer. For instance, if genebanks would provide downloadable databases per crop with passport data in a format following the ‘Multicrop Passport Descriptor List’ it would become possible to create a central crop database of any crop within a few days.
• Downloadable databases allow sending additional documents along with the data sets. These documents can provide information for a proper interpretation of the information in the data sets, but also information about how to order material, etc.

• The access is indirect; the user must take an extra step before actually accessing the data. This costs time and requires some knowledge.
• Downloadable databases are, generally, snapshots. Updates of the database made from the moment the files were generated will obviously not be included.
• Data sets can start leading their own lives. Especially in the case of coding systems the dangers of downloading, changing and further distribution will be evident.

• a well-known format for the file containing the data (DBF or TXT) and the accompanying text files (DOC or TXT),
• a well-known or easy interpretable structure of the data file, e.g. the IPGRI Multicrop Passport Descriptor list,
• standard of interpretable coding, i.e.
• standard coding systems such as the codes defined in the IPGRI Multicrop Passport Descriptor list, the ISOŠ country codes or the FAO institution codes, or
• ‘local codes’ accompanied by decoding information in a separate file,

• the data file(s),
• the file(s) containing the information needed to decode the codes used in the data file(s),
• the file describing the structure of the data and decode file(s), and
• a file called something like READ_ME.TXT with some nice words and a short description of the content of all other files.

Final steps
Once the ZIP files are available they have to be put on the net, and referenced in HTML documents on the net. If no local server is available and/or knowledge of HTML is lacking, a member of the ‘Internet Advisory Group’ (e.g. CGN,IPGRI, NGB, ZADI) should be approached. They will provide you with a simple solution within minutes.
 
Conclusion
Downloadable databases, accompanied by information for their interpretation and compressed in ZIP files, provide an attractive way of providing access to databases. They should be considered as a option in addition to providing access via on-line searchable databases. Since the technology needed is available in any genebank, the databases can be ‘on the net’ within weeks.