Morning,
I heard that some NASA engineer actually invented the core stuff. The big difference was the fact that the files could be defined external to the source code, and the buffer/structure definition was made available to the prg (.CMD) files throught the use statement.

But you could not have more than two files open, with only 32 fields per file and 64 memory variables active at anyone time. 64K COM file did it all - whole applications on floopy drives, the users would be prompted to swap diskettes at the appropriate moment. Lotus and dBase were both assembler - then they were ported to C - and got bigger and bigger and bigger:-)

Tricks like combining data as strings in a big (128 byte max) character field and the parsing out what you needed were the norm.

We had the INDEX and FIND and locate - but no PEMS (like when or valid) and no procedures (everything was a "DO")

I had a couple of FoxBase Xenix (dBase III) projects and a couple dBase II/CCPM gigs.

Working with those old systems sure gave me an appreciation for conservation of clock and resources - I also think I write faster code - but I was a baby then - I did not get an inkling of where things would go until I saw my first gui on a portable compaq with a hercules card:-). Everything was in BASIC - the cursor masks - I even cracked visicalc (by accident) once - it was also basic.



>Hello,
>all comments are very interestingg...
>
>I remember those good old XBase days... imho they used the Word Fox mainly because of the "X" in Xbase.
>there are a just few words ending with X. Just an idea.
>Quoted from foxprohistory.org: 1984 (December) Fox Software releases FoxBASE., This was a dBASE II work-alike
>
>Warm regards,
>Pablo V