2024-11-24 07:00:03 <lilo_booter > been rethinking the string concatenation approach for the specification of my equations - as i said when i first asked about it, i didn't like it, but i still want to abstract things a little - currently thinking that this is perhaps a step in the right direction: y= x ; x-loop y= x 2e f** ; x-loop x= y fnegate y-loop etc
2024-11-24 07:00:40 <lilo_booter > (typo in there - should have ; before y-loop)
2024-11-24 07:06:48 <lilo_booter > the y= and x= words are still a bit of a hack using an evaluate : y= ( -- colon-sys ) s" :noname { F: x }" evaluate ; immediate - no user provided strings, so no messing around with free - and i kinda like the way the input to equation is abstracted away, so there's no reason to use fpick/froll type operations (which are fine for programming, but not necessarily the most natural way to specify an
2024-11-24 07:06:50 <lilo_booter > equation in rpn)
2024-11-24 07:13:14 <lilo_booter > couple of things i dislike about it - the continued use of evaluate and the requirement to coordinate y= with x-loop/x= with y-loop - the latter feels wrong to me and trivial to consolidate in one xy-loop by having y= and x= leave a loop indicator (an int for example) and sys-colon on the stack iyswim
2024-11-24 07:16:33 <lilo_booter > global x/y values are the only things i can think of, but that that feels clunkier than the evaluate
2024-11-24 07:20:54 <lilo_booter > also, been trying to get my head round the introduction of polar coordinates but i'm not really sure of the use cases - would resolve r from theta or theta from r seem sensible?
2024-11-24 08:08:11 <lilo_booter > went for the xy-loop thing - current stuff looks like https://gitlab.com/lilo_booter/plotify/-/blob/main/src/graph.fs - am curious about what people think of my imeplementations of Breshenham's and Cohen Sutherland too if anyone cares to take a look :)
2024-11-24 08:11:59 <lilo_booter > (they're more or less direct translations of psuedo code found found on wikipedia)
2024-11-24 10:19:44 <lilo_booter > hmm - maybe this is prettier? y=(( x x fsin f* )) xy-loop
2024-11-24 10:30:56 <lilo_booter > y=(( x x fsin f* )) :D
2024-11-24 18:49:33 <xentrac     > lilo_booter: that's cool
2024-11-24 18:58:36 <lilo_booter > xentrac: cheers :)
2024-11-24 19:52:48 <nmz         > is the return stack a must?
2024-11-24 20:00:59 <thrig       > no, a must is what you get after crushing grapes
2024-11-24 20:01:31 <nmz         > no that's a mush
2024-11-24 20:01:58 <zelgomer    > nmz: factor doesn't have an explicit return stack
2024-11-24 20:02:21 <zelgomer    > i'm sure it has /something/ because its words are capable of returning, but the programmer doesn't get to manipulate it
2024-11-24 20:03:59 <nmz         > guess its a must
2024-11-24 20:04:20 <zelgomer    > i think the question is incomplete
2024-11-24 20:04:26 <zelgomer    > a must for what?
2024-11-24 20:05:38 <nmz         > was thinking of writing a transpiler
2024-11-24 20:06:09 <zelgomer    > surely your target language has a return stack
2024-11-24 20:07:14 <nmz         > sure
2024-11-24 20:07:20 <nmz         > well, thanks
2024-11-24 21:47:53 <lilo_booter > for polar coordinate stuff, i came to the conclusion that there's no way to derive theta :) - only r (which is already checked in - r=(( 10e )) <- draws a circle with radius of 10, r=(( 1e theta fsin - )) <- heart shaped equation, r=(( theta )) <- basic archemedian spiral etc etc)
2024-11-24 21:50:53 <lilo_booter > general issue with them is that i don't really know how to control theta - currently it just loops from 0 to 2 pi, but the archimedian spiral should allow for controllable bounds (so, like i have x-range/y-range, i guess i could add theta-range?)
2024-11-24 21:53:37 <lilo_booter > amazing how a joke actually made me think about stuff :)