2024-05-30 16:15:40 <yakubin     > Hello. Is this a bug? SwiftForth: > -3 2/ . -3 2 / . => -2 -1  ok
2024-05-30 16:18:36 <olle        > 2/ ?
2024-05-30 16:18:50 <olle        > Or 2 / ?
2024-05-30 16:19:17 <yakubin     > I mean the discrepancy between the 2/ word and the / word with the top element 2.
2024-05-30 16:19:39 <yakubin     > gforth and pforth give the same answer in both variants.
2024-05-30 16:19:51 <yakubin     > -2
2024-05-30 16:20:12 <yakubin     > IMO a bit unintuitive, but at least consistent.
2024-05-30 16:29:13 <GeDaMo      > 2/ is usually a right shift by 1
2024-05-30 16:30:43 <zelgomer    > 2/ rounds toward negative infinity, yes
2024-05-30 16:31:18 <yakubin     > That’s alright. But why does 2 / then round in the opposite direction? Is this a bug or allowed behaviour?
2024-05-30 16:31:34 <zelgomer    > because / rounds toward zero
2024-05-30 16:31:49 <yakubin     > Not with gforth and pforth.
2024-05-30 16:32:28 <zelgomer    > that's weird
2024-05-30 16:32:44 <zelgomer    > not all forths are the same
2024-05-30 16:32:46 <yakubin     > Oh. Wait. pforth does. I looked wrong at it.
2024-05-30 16:32:56 <yakubin     > gforth doesn’t though.
2024-05-30 16:33:24 <yakubin     > On gforth: -3 2/ . -3 2 / . => -2 -2  ok
2024-05-30 16:33:47 <yakubin     > Alright. I’ll assume gforth is the odd one out then.
2024-05-30 16:33:54 <zelgomer    > what does gforth produce when you do this?  -15 10 / .
2024-05-30 16:34:16 <yakubin     > -2
2024-05-30 16:34:55 <zelgomer    > weird
2024-05-30 17:02:47 <veltas      > yakubin: Shift doesn't round
2024-05-30 17:02:50 <veltas      > 2/ isn't 2 /
2024-05-30 17:02:56 <veltas      > Confusingly
2024-05-30 17:03:12 <veltas      > 2/ is really a signed shift of 1 bit, that's all it is
2024-05-30 20:03:20 <Al2O3       > shift right?
2024-05-30 20:03:23 <Al2O3       > :)
2024-05-30 20:03:51 <Al2O3       > 3 2/ 1 . 1  ok
2024-05-30 20:04:11 <Al2O3       > not confusing to me at all, 2\ should be obvious :)
2024-05-30 20:04:33 <Al2O3       > sorry, meant 3 2/ . 1  ok
2024-05-30 20:04:36 <Al2O3       > :)
2024-05-30 20:04:52 <Al2O3       > love the left shift doesn't work.
2024-05-30 20:04:58 <Al2O3       > on gforth
2024-05-30 20:31:21 <yakubin     > It works. Division doesn’t work.
2024-05-30 20:33:18 <yakubin     > (Also, when the dividend is positive, integer division by 2 is indistinguishable from right shift.)