On 14/02/2022 21:46, Rick C wrote:
> On Monday, February 14, 2022 at 5:23:03 AM UTC-5, dxforth wrote:
>> On 14/02/2022 20:40, Rick C wrote:
>> > On Monday, February 14, 2022 at 3:31:58 AM UTC-5, dxforth wrote:
>> >> On 14/02/2022 15:25, Rick C wrote:
>> >> > On Sunday, February 13, 2022 at 7:55:38 PM UTC-5, dxforth wrote:
>> >> >> On 14/02/2022 04:50, Rick C wrote:
>> >> >> >
>> >> >> > I had not seen the FOR AFT THEN NEXT construct. I like it.
>> >> >> Moore accepted FOR NEXT's quirks. Not so others - hence the various attempts
>> >> >> to 'fix it' which made it as complicated as DO LOOP but not as useful.
>> >> >
>> >> > Exactly which quirks where those?
>> >> Loops n+1 times and n=0 isn't a NOOP. Not the behaviour one usually wants.
>> >
>> > That is exactly what DO +LOOP does with a negative index. It works fine for me.
>> Luckily such instances are rare.
>> > Addressing, such as in arrays, works exactly that way. You just need to subtract 1 from N before starting the loop to address n-1 down to 0.
>> >
>> > Considering that FOR NEXT only decrements by 1, it seems logical to me for it to work this way.
>> >
>> Not according to eForth, FlashForth etc who developed workarounds for the behaviour.
>
> So? The thing about Forth is that everyone does what they personally prefer. Writing software is not the same as judging. There's nothing in either of these Forths that says FOR NEXT needs a work around. That's your judgement.
>

My judgement for a Z80/8086 with adequate rstack is DO LOOP. For the AVR FOR NEXT
may be the better choice - but not a bare bones one that shifted the cost to the
programmer to sweat over.

On 14/02/2022 21:07, minf...@arcor.de wrote:
>
> FWIW apart from FOR .. NEXT, I use the following array walking words:
> FOR> ( start number elementwidth -- ) walk forward, N yields the element address
> <FOR ( start number elementwidth -- ) walk backwards, N yields the element address

All these appear to be missing in MinForth ?

On 16/02/2022 05:11, minf...@arcor.de wrote:
> dxforth schrieb am Dienstag, 15. Februar 2022 um 04:39:52 UTC+1:
>> On 14/02/2022 21:07, minf...@arcor.de wrote:
>> >
>> > FWIW apart from FOR .. NEXT, I use the following array walking words:
>> > FOR> ( start number elementwidth -- ) walk forward, N yields the element address
>> > <FOR ( start number elementwidth -- ) walk backwards, N yields the element address
>> All these appear to be missing in MinForth ?
>
> They are in the transpiler.

That means little to a forth user.