1. more for the wish list
While we are suggesting things,, i have another wish which seems at first
glance to be useless. Access to the procedure/function call stack. This way
a program can check it's own flow, to see how it got to where it is, and if
possible to change the stack, to have more control over where it backs up
to. I don't remember why this is on my pascal gripe list, it seems there is
a way to get this data covertly, and it's possible to record it in a user
record, but there wasn't a clean way to alter it, to skip over a lot of code
when returning to the original calling line. I think i was getting really
really deep into nested calls, with a huge call stack ( the debugger shows
the call stack ), and wanted to pop straight back to the place that started
the recursion, rather than returning one procedure at a time, and testing at
each return to see if i should return to that calling procedure, and testing
at that line to see if i should return again, and .. ad infinitum. This
could also allow a procedure to see who called it, to allow the calling proc
permission to use it or not. It would also allow a user-defined form of
function overloading in a way, dealing with passed parms in a way dependant
on who passed them. Or seeing if it called itself and how many times. Put
this one at the bottom of the wish list, it's a luxury, i suppose.
Kat,
rambling.
2. Re: more for the wish list
A spooky feature that would be. But most certainly interesting.
Check out Neil´s -(old) ememcopy´s command list techniques.
They provide sort of that, but to gain speed and to declare a procces of
graphical actions in a more readable way. Before-hand, while in the actual
game-loop (or ..) you only change parameteres from the command list, or
insert-remove command (that were created before hand as well) .. the effect
is you can gain lots of lots of speed, since for each frame a lot of things
don´t change .. and now they don´t have to be calculated for each frame as
well ...
Maybe you can do something simerlar using routine-pointers and a sequence ?
Ralf N.
> While we are suggesting things,, i have another wish which seems at first
> glance to be useless. Access to the procedure/function call stack. This
way
> a program can check it's own flow, to see how it got to where it is, and
if
> possible to change the stack, to have more control over where it backs up
> to. I don't remember why this is on my pascal gripe list, it seems there
is
> a way to get this data covertly, and it's possible to record it in a user
> record, but there wasn't a clean way to alter it, to skip over a lot of
code
> when returning to the original calling line. I think i was getting really
> really deep into nested calls, with a huge call stack ( the debugger shows
> the call stack ), and wanted to pop straight back to the place that
started
> the recursion, rather than returning one procedure at a time, and testing
at
> each return to see if i should return to that calling procedure, and
testing
> at that line to see if i should return again, and .. ad infinitum. This
> could also allow a procedure to see who called it, to allow the calling
proc
> permission to use it or not. It would also allow a user-defined form of
> function overloading in a way, dealing with passed parms in a way
dependant
> on who passed them. Or seeing if it called itself and how many times. Put
> this one at the bottom of the wish list, it's a luxury, i suppose.
>
> Kat,
> rambling.
>
