Warp <war### [at] tagpovrayorg> wrote:
> > The first computer *I* used, the Amstrad CPC, *did* - hehe :P
>
>   Of course even in editors which did have a renum command, it would mess
> up any numbering scheme you might have been using. For example, if you
> started every subroutine using line numbers starting from multiples of 1000
> (so that you could easily remember that "this subroutine started at line
> 3000, and that one at line 5000" etc) the renumbering would mess up all
> that and you lost the semi-logical numbering.

Actually, the Amstrad CPC's BASIC was remarkably sophisticated about that,
allowing you to renumber only a subset of lines, to start from a specified new
number and at specified intervals.

IIRC, really the *only* problematic statements were GOTOs to non-existent lines,
which would by default go to the next higher existing line number; these could
indeed be broken.

Truly computed GOTOs, as far as I can remember, weren't supported by the Amstrad
CPC's BASIC (except with some obscure RSX tools) - only a "ON ... GOTO
line1,line2,line3,..." to pick one of a predefined set of jump targets.


>   Overall, BASIC programs were quite hard to read and understand. Most
> editors in the old days did not support such a basic technique as code
> indenting (iow. even if you tried to indent the code manually, the editor
> didn't support it and you couldn't). This means that every single line of
> code started from the same column, which made reading long nested blocks of
> code a bit difficult.

Still, with some discipline it was possible to write well-maintainable code.

(Though I agree that there are languages that make it far easier.)

Post a reply to this message