-4- 
DATA-COLLECTIN G IN THE SMITHSONIAN 
C. A. Bull and R. Shank have completed a survey of the Smith 
sonian's activities in data collecting, and have summarized their 
results in an unpublished report entitled "Non-conventional File 
Structure Data-collecting Projects in the Smithsonian Institution 
A Survey, Winter 1960 - Spring 1969." The "non-conventional" 
designation is for techniques other than simple alphabetical cat¬ 
alogues or lists which will facilitate rapid handling of data, 
from key-sort to computer. Interest was in proj ects furthering 
research or educational functions as opposed to housekeeping. 
A total of 49 projects were identified, and each is described. 
Some of the findings are: 
Twenty-three proj ects are using or anticipate using some 
sort of machine assistance in storage and retrieval. 
More pro j ects in biology are using automatic means to store 
and retrieve data than are proj ects in the physical 
sciences, history and technology. 
In every case, the highest level of satisfaction with new 
systems and methodology can be found where the curator 
and researcher was heavily involved in system design 
and output control. 
The most critical factor is the validity of the data entered 
and the positive relationship of the degree of expertise 
of the person making the entry with the reliability of 
the entry. 
In assessing the requirements for making their systems viable 
most people who were conducting proj ects felt the need for 
manpower with subj ect knowledge more than the need for 
electronic muscle. 
If anyone would like to know more about the survey, a few 
copies of the report are available for distribution. Write 
to MUDPIE for a copy.--JAP. 
INFORMATION SYSTEMS SURVEY 
"RECON" 
(NASA Information System) 
The NASA Information System, commonly known as RECON, is 
a package of programs for creating, maintaining and querying 
data files. It consists of two major subsystems. The first 
is a batch system which provides for file creation and main¬ 
tenance, batch query and prepares output for publication. The 
second, RECON, is a communication control program, language 
analyzer and search program which permits multi-programmed 
query from a remote console. 
The system operates on variable length data records 
composed of fixed header information followed by a variable 
number of tagged, variable length fields which may be repeated 
