Re: [Amforth-devel] Why amforth became nonresponsive after reset

[Matthias T. <mt...@we...>]

Pito,

> Hi,
> quite often happens to me that amforth is lost (even reset does not
> help) when running a code under development, so I have to reflash
> the atmega and upload the whole stuff again. What is the typical
> reason for that? My understanding is that unless I (or something)
> reprogram(s) the flash or the eeprom, the amforth shall be rock
> solid and must return to life when reset. 
> 
> Is that because when I loop by accident into flash area it might
> happen it runs a directive which does the reflash of the flash? Is
> there any mechanism available so we may disable such routines in
> order to block accidentaly changing the code?

amforth is not rock solid when it comes to code errors, unfortunatly.
You can make it more solid when you redefine the store operations
using an address checking wrapper, but that will cost runtime 
performance.

And no: I know of no better way but not making errors. "Your" error
with the wrong 1. in colon words may serve as an example for really
hard work to find such bugs...

Matthias