Data architecture
Una Smith
una at DOLIOLUM.BIOLOGY.YALE.EDU
Tue Dec 1 11:08:37 CST 1992
I agree completely with Jim on the issue of data base architectures;
perhaps I would go even further, and say that storage of data sets
in data base package formats may be equivalent to putting your data
in an iron box and burying it on Treasure Island. I've seen too many
valuable, old data sets transfered from cards (remember those?) to
binary files in some XYZ data base format, stored on a diskette and
forgotten again. The data might be better off left on paper cards!
I wish it were otherwise, and perhaps in a few years something new
will come along that solves this problem (but I'm not holding my
breath). In my own work I have gone to the inconvenience of keeping
the authoritative data sets in plain ASCII files, and loading them
into data base packages as needed for analyses. I keep a paper trail
of programs for each step, and I do not keep the formatted data sets
once I am done using them. I realize this is a rather extreme point
of view, but I think it is a sound long-term position. The choice of
format is of course a personal one, but it should also be as informed
and long-sighted as is practical.
Una Smith smith-una at yale.edu
More information about the Taxacom
mailing list