... <j2thomas@cavtel.net> wrote Re: R> DROP in GForth [..] So the new method adds essentially nothing to the complexity of...addr n 0 do \ m addr dup r @ execute \ m flag if \ m \ a...r> cell+ @ execute unloop exit then r> cell+ cell+ \ m a loop drop ; \ USAGE EXAMPLE : ?9 dup 9 = ;...

On Tue, 1 May 2007 23:01:04 +0200 mhx@iae.nl (Marcel Hendrix) wrote: Jonah Thomas <j2thomas@cavtel.net> wrote Re: R> DROP in GForth [..] So the new method adds essentially nothing to the complexity of the system. It provides powerful flow control for slightly less overhead than you'd use for errors. I ...

Jonah Thomas <j2thomas@cavtel.net> wrote Re: R> DROP in GForth [..] So the new method adds essentially nothing to the complexity of the system. It provides powerful flow control for slightly less overhead than you'd use for errors. I remember from the old CLF BacForth threads that this leads to quite impenetrable, very difficult to ...

Jonah Thomas <j2thomas@cavtel.net> writes: Does R> DROP work correctly on other Forth systems, with some reasonable restrictions like no locals and careful loops? Some optimizations that break using R> DROP for non-local exits are inlining and tail-call elimination, so I guess that there are several Forth systems around where you cannot...

... <j2tho...@cavtel.net> wrote: Does R> DROP work correctly on other Forth ... like no locals and careful loops? R> DROP works correctly on all ISO ...systems I know of, as long R> as one is reasonable, and don...it must be preceded by some R. Groetjes Albert I'm pretty sure he is referring to R> DROP as a factor of EXIT-CALLER. Yes, but as I ...

....6784b1f1.j2tho...@cavtel.net>, Jonah Thomas <j2tho...@cavtel.net> wrote: <SNIP> Does R> DROP work correctly on other Forth systems,...>restrictions like no locals and careful loops? R> DROP works correctly on all ISO Forth ...e. it must be preceded by some R. Groetjes Albert I'm pretty sure he is referring to R> DROP as a factor of EXIT-CALLER...

...<20070430095927.6784b1f1.j2thomas@cavtel.net>, Jonah Thomas <j2thomas@cavtel.net> wrote: <SNIP> Does R> DROP work correctly on other Forth systems, with some reasonable restrictions like no locals and careful loops? R> DROP works correctly on all ISO Forth systems ...I.e. it must be preceded by some R. Groetjes Albert -- -- Albert van der Horst, UTRECHT,...

... notice that R> DROP to leave a calling word works in GForth, but if the calling word uses ...haven't looked at that.... Does R> DROP work correctly on other Forth...t nl \ conditionally exit : ?EXIT IF R> DROP THEN ; : ?EXIT ]] IF EXIT ... when this word completes, like R> DROP : ?SKIP-CALLER \ even clumsier name IF NIP R THEN ; If it isn't ...

I notice that R> DROP to leave a calling word works in GForth, but if the calling word uses ...I haven't looked at that.... Does R> DROP work correctly on other Forth ... for it? \ conditionally exit : ?EXIT IF R> DROP THEN ; : ?EXIT ]] IF EXIT THEN ... word when this word completes, like R> DROP : ?SKIP-CALLER \ even clumsier name IF NIP R THEN ; If it isn't worth ...

...the start of the first definition in the chain. The purpose of this is case constructs, piecewise functions, etc: sign : drop 0 sign : dup 0< then drop -1 sign : dup then drop 1 This set of definitions yields 1 for positive, -1 for negative, otherwise 0. In effect...