Analoog aan leading blanks, leading zeroes etc.

--------------------------------------------------
Note added at 4 hrs 37 mins (2005-12-02 19:48:16 GMT)
--------------------------------------------------

@Dave/2hrs25mins:

Hi Dave, I think the header record you're referring to is related to basic file structure. If that is wat Northstar's 'voorlooprecord' is all about, then you're quite right. Normally, however, that type of record would be referred to in Dutch as 'koprecord' [as in Elsevier's Woordenboek ICT] or simply as 'header' (as I knew and used the term during my 20-year IT career). As far as I know, a leading record is related in particular to database structures, as in the following web references:

http://www.ifla.org/IV/ifla64/035-99e.htm
The elements which compose the leading record are as follows:
· five data elements listed above for the Small Database;
· field of application;
· divisions;
· related fields;
· local (narrow) methodological instructions and examples;
· special instructions on common and special auxiliaries;
· "see" notes and "see" numbers;
· "see" notes feedback;
· data of the service classification scheme (additional subject entries and related instructions);
· data on the availability of continued records;
· search image of a dictionary entry in the UDC information retrieval language;
· codes of headings of the State Classification for Scientific and Technical Information corresponding to the main heading of a certain UDC entry;
· codes of DDC headings corresponding to the main heading of a certain UDC entry.


http://www.freepatentsonline.com/5261065.html

DBMS 21 transfers control to the relative byte address source control 21c (step 104 of FIG. 17), in order to convert the page address list code 11a (input in step 123 of FIG. 20), set at the calculation request source control 21a, into the list of relative byte addresses 16 (FIG. 3). The relative byte address source control receives the mis-hit page address list 17. The relative byte address source control 21c checks the data base directory information 15 (steps 124 and 125) controlling mapping information between the buffer control table 12 and the relative byte positions and converts it into the list of relative byte addresses 16. In case they are coincident as determined by step 126, the relative byte position is expressed by the RBA (Relative Byte Address) in the RBA column 15b indicating a relative position from a leading record in the data (step 127) and the next table column is searched (step 128) until the end of the directory is reached (step 105) leading to the end of step 104. DBMS 21 transfers control to the relative byte address source control 21c (step 104 of FIG. 17), in order to convert the page address list code 11a (input in step 123 of FIG. 20), set at the calculation request source control 21a, into the list of relative byte addresses 16 (FIG. 3). The relative byte address source control receives the mis-hit page address list 17. The relative byte address source control 21c checks the data base directory information 15 (steps 124 and 125) controlling mapping information between the buffer control table 12 and the relative byte positions and converts it into the list of relative byte addresses 16. In case they are coincident as determined by step 126, the relative byte position is expressed by the RBA (Relative Byte Address) in the RBA column 15b indicating a relative position from a leading record in the data (step 127) and the next table column is searched (step 128) until the end of the directory is reached (step 105) leading to the end of step 104.

To make things -- seemingly -- complicated, a database file could have two types of "initial" records, if I may attempt to use a generic term that would cover both 'leading record' and 'header record' as a translation of 'voorlooprecord': your more or less computer-oriented "physical" header record, and my more or less application or user-oriented "logical" leading record. The correct translation would depend on Northstar's context, I think. Judging from the application or user-oriented GEA number in the journal (database?) file, I would opt for the latter. But you could be quite right, of course. You would even have a dictionary compiler like Hein Steenis -- who I would in general regard as being quite authoritative -- on your side. In his "Woordenboek Informatica en Telecommunicatie" [Sybex, 4th edition] he renders 'header record' as 'voorloop-record', defining it as: "eerste record van een reeks waarin metagegevens zijn vastgelegd betreffende de erop volgende records". I hope he's adapted that slightly in subsequent editions, though ;-)