Re: 0 SET-ORDER why?

On 6/28/24 03:04, albert@spenarnc.xs4all.nl wrote:

In article <v5kde0$2sasd$1@dont-email.me>,
Krishna Myneni  <krishna.myneni@ccreweb.org> wrote:

...

The solutions are:
>
1) leave everything as is, and live with the contradiction and the
hazard of performing 0 SET-ORDER in interpretation state.
>
2) make SET-ORDER state-smart, and live with the contradiction. This
will potentially break code.
>
3) disallow zero as an argument to SET-ORDER e.g. throw an error for zero.
>
Am I missing any other options?

 4) Use wordlists as a pure stack.  Also ASSEMBLER  .. PREVIOUS.
Define the minumum search order as only handling numbers and other constants.
Declare SET-ORDER and GET-ORDER obsolete.
 

The whole Search Order word set is clunky and has a cobbled together feel about it. It is also difficult to integrate a named modules system into which provides Public/Private definitions into the standard words e.g. making ORDER list the names of the modules isn't easy.
A stack model for the search order may be the way to go. It would be more intuitive to Forth users than having to remember ONLY ALSO PREVIOUS etc.
Currently the standard defines what a minimum search order should contain but then promptly disregards it by standardizing 0 SET-ORDER. This is dangerous in interpretation mode, and, while it may have uses in compilation mode, within a definition where the prior search order can be restored (or not), the notion of a well-defined minimum search order should be a strong guarantee to the Forth programmer and not allowed to be violated easily.
--
Krishna