2024-12-03 01:38:52 <Zarutian_iPad> damnit, I saw a yt vid recently on an magnetic cores based rom recently, where you wire up as many 'speed dial' numbers as you could fit sense wires for. And now I am thinking of how one would have done such for say Apple ][ or such
2024-12-03 01:40:35 <Zarutian_iPad> and I have come up with a scheme.
2024-12-03 01:41:16 <Zarutian_iPad> lets say I wanted to wire up the wozmon rom
2024-12-03 01:42:33 <Zarutian_iPad> I need 256 ferrite core rings in an 16x16 grid.
2024-12-03 01:43:16 <Zarutian_iPad> this allows for 16 wires horizationally and 16 wires vertically
2024-12-03 01:43:35 <Zarutian_iPad> these are the addressing and driving wires
2024-12-03 01:44:34 <Zarutian_iPad> then per nybble I need 16 sense wires
2024-12-03 01:45:15 <Zarutian_iPad> so for a 8 bit system 32 wires or 64 wires for 16 bit system
2024-12-03 01:46:30 <Zarutian_iPad> only one nybbles sense wire of 16 goes through a core for that address
2024-12-03 01:50:53 <thrig       > didn't they get grandmothers to wrap the apollo memory wires
2024-12-03 01:54:14 <Zarutian_iPad> indeed, but I thought of this as a product for computer hobbyist, that is, what half a century too late to market
2024-12-03 01:56:10 <Zarutian_iPad> and that damn This Museum is (not) obsolete had me thinking on how to make a electronic code door lock using a punched card, uniselector and a few relays
2024-12-03 01:56:53 <Zarutian_iPad> the punched card stores the current unlock code
2024-12-03 01:57:44 <Zarutian_iPad> I am in half of mind sending them a schematic of the damn thing
2024-12-03 05:16:01 <KipIngram   > One advantage of that kind of low-tech stuff can't really be hacked.
2024-12-03 05:16:26 <KipIngram   > And you could "program it" with a piece of card stock and a hole punch, so that info never passed through anything hackable.
2024-12-03 14:19:57 <veltas      > Okay well the first advent of code this year is not something I can write portably in a lunch break, but it seems fun enough to write a solution for anyway
2024-12-03 14:20:33 <veltas      > I will share when I've got it working
2024-12-03 14:26:39 <xentrac     > :)
2024-12-03 15:07:12 <user51      > veltas: Are you scanning the input, or inlining it?  Asking about day 1.
2024-12-03 15:07:52 <user51      > More specifically the given example in the day 1 page.
2024-12-03 15:48:33 <veltas      > I'm thinking of doing a retro solution so will provide a utility for entering the input
2024-12-03 15:49:18 <veltas      > In gforth I will probably sed the input to how it would be entered, rather than entering 1000 lines manually
2024-12-03 15:49:45 <veltas      > I'm also interested in running it on the ZX Spectrum, so might have to enter it manually for that.... we'll see
2024-12-03 15:50:53 <user51      > veltas: I usually just cat|gforth
2024-12-03 15:51:03 <user51      > It's a bit messy, but here's how it looks
2024-12-03 15:51:04 <veltas      > If I do write this I'm hoping it will be interesting
2024-12-03 15:51:11 <veltas      > Will look tonight!
2024-12-03 15:52:25 <user51      > I'm not finished yet, just my progress so far
2024-12-03 15:52:44 <user51      > code http://0x0.st/X7A5.forth
2024-12-03 15:53:10 <user51      > output http://0x0.st/X7AR.txt
2024-12-03 15:54:14 <user51      > I used a linked list, but honestly now that I think about it 'A B , ,' might suffice, just count each one
2024-12-03 15:56:05 <veltas      > Many ways to skin a cat
2024-12-03 15:59:57 <user51      > yep, it is http://0x0.st/X7mr.txt
2024-12-03 17:24:19 <user51      > I think I'll bench forth this time, the python solution is much cleaner.
2024-12-03 17:25:27 <user51      > It's not too difficult, but manually writing low-level data structures and subroutines gets old pretty fast.
2024-12-03 17:34:45 <xentrac     > I can bench 50kg but Forth may be too much for me
2024-12-03 17:56:48 <MrMobius    > there is that whole "right tool for the job" thing but it contradicts the "one true language" thing a lot of other people think
2024-12-03 19:06:59 <veltas      > user51: I recommend circular lists if you do linked lists again
2024-12-03 19:07:55 <veltas      > It is basically doubly linked lists but easier
2024-12-03 19:10:54 <veltas      > user51: You were saying it's messy but you've taught me stuff I didn't know about Forth
2024-12-03 19:11:04 <veltas      > I've never seen the TH word before, looks useful
2024-12-03 19:11:15 <veltas      > I like it
2024-12-03 19:13:36 <user51      > veltas: I can't find it, but last year right when I was getting deeper into forth I wrote some fancier list-handling code, words like LIST-APPLY ( xt list -- ) and such.  That could probably make some loops redundant.
2024-12-03 19:14:21 <veltas      > A lot of Forth code ends up being very 'functional'
2024-12-03 19:14:33 <veltas      > And I think passing xt's around like that fits the bill
2024-12-03 19:15:55 <user51      > Found it http://0x0.st/X7Bu.forth
2024-12-03 19:37:43 <GeDaMo      > http://www.murphywong.net/hello/comus.htm#th
2024-12-03 20:50:53 <veltas      > I've not read through that whole thing, maybe I should