Less Is More

or: how the funny source names came to be


In the beginning - oh, long before that - when there were real computer, real data was kept in IMS databases and real programs were kept on punch cards. That left few needs for sophisticated file systems with long file names or directories. When the first version of SAP DB was written (or 1) as it was called), there were about 20 files with the name COMMON and many additional name clashes. So someone came up with the following file name scheme:

v[a-z][a-z][0-9][0-9]

Because linker were limited in the number of significant letters in an external identifier, all external procedures had a prefix consisting of the first letter of the layer and the two digits.

This proved remarkably efficient:

But the Iron Age was followed by the Dark Ages 3). People raised on Windows could remember two bytes at most (therefor the change from 'Pascal' to 'C' 4).

So this naming scheme is fading slowly into the mist of history. Possibly to be revived by the next generation of software engineers who think that software development is less like building construction and more like chemistry 5) - and that the notation should demonstrate this.


1) Translation courtesy of Torstar
2) some proposed hex digits to get a bigger range and to weed out lesser programmers - the weed won
3) so called as most effort went into the customization of screen savers
4) it wasn't until 32bit Windows appeared that 'C++' became possible
5) actually it's more like Alchemy: people only have to claim to transform lead into gold to rake in ridiculously amounts of cash


last changed 2001-05-02
Daniel Dittmar