00:00:00 --- log: started forth/19.01.11
00:19:16 --- quit: dave0 (Quit: dave's not here)
00:35:47 <the_cuckoo> revamped the shebang thing for rpany - has a very strict form now - if the script used defines a __main__ word, this is used after all the following args are handled - switches are words which are introduced by way of an optional __switches__ word
00:35:51 <the_cuckoo> https://gitlab.com/lilo_booter/rpany/blob/master/bin/rpany-shell
00:36:37 <the_cuckoo> $ ./bin/rpany-shell --calculator --prompt="calc> "
00:36:39 <the_cuckoo> calc> words
00:36:41 <the_cuckoo> % * ** + - . / abs acos acosh asin asinh atan atan2 atanh cos cosh e exp ln negate pi sin sinh tan tanh words
00:37:25 <the_cuckoo> as in - lock down - you have a calculator - and that is all :)
00:47:55 <the_cuckoo> fwiw, this is all about the 'domain specific language' thing - the stack parser presented to a user/process can be locked down entirely to the requirements of the domain - internally, it can use anything to satisfy the requirements
00:48:51 <the_cuckoo> this should be a safe thing to expose at the end of a socket - no security holes
02:03:16 --- quit: ashirase (Ping timeout: 272 seconds)
02:07:36 --- join: ashirase (~ashirase@modemcable098.166-22-96.mc.videotron.ca) joined #forth
03:19:02 --- join: dddddd_ (~dddddd@unaffiliated/dddddd) joined #forth
03:19:37 --- nick: dddddd_ -> dddddd
03:53:48 --- quit: dys (Ping timeout: 244 seconds)
04:14:28 --- quit: mark4 (Ping timeout: 252 seconds)
04:15:57 <proteusguy> reepca, re:Advanced Squad Leader - you an old school wargamer? :-)
04:17:12 <proteusguy> the_cuckoo, cudos on the progress!
04:17:39 <the_cuckoo> proteusguy: thanks :)
04:33:09 * Zarutian thought he read "Advanced Squid Leader" first time around. Now, because he is watching a compilation of Japanese commercials on mobile games he has thought up a game premis.
04:34:44 <Zarutian> basically, one is leading an army of cybernetically enhanced squids against minecraft players that has nearly exterminated their kind due to inksack collection
04:37:09 <john_cephalopoda> Whee~
04:37:41 <proteusguy> hahaha
05:12:38 <siraben> Just got a simple menu system working, should be the basis of navigation within my OS
05:12:40 <siraben> https://github.com/siraben/zkeme80/blob/044d80e617819abf0020b491546ededa9c1a03c8/bootstrap-flash3.fs#L150
05:12:43 <siraben> https://imgur.com/a/GSI9lBj
05:14:18 <siraben> Menu entries contain three things: a selector xt, a deselector xt and an "on click" xt
05:14:45 <siraben> So it's easily extensible
05:15:41 <siraben> Thoughts?
05:31:23 <john_cephalopoda> Stylish :D
05:36:45 <siraben> I plan to boot up to a menu that gives the user options to power off the calculator, go into a REPL, see system info or run tests.
05:36:57 <siraben> Maybe a 2x3 grid would work better, hm.
05:45:34 <the_cuckoo> sorry to spam the link again, but i think the form this is taking is really lovely - https://gitlab.com/lilo_booter/rpany/blob/master/bin/rpany-shell - the vocabs used are described in the readme of the project
05:51:40 --- join: mark4 (~mark4@12.41.103.244) joined #forth
05:52:59 --- join: rdrop-exit (~markwilli@112.201.166.158) joined #forth
05:53:29 <rdrop-exit> Proteusguy: Yes, I'm an old schoold board wargamer
05:54:59 <proteusguy> rdrop-exit, same here. Lots of Avalon Hill and other box wargames. And AD&D & Traveller of course.
06:00:27 <rdrop-exit> Me too SPI, AH, GDW, Yaquinto, VG, WEG, etc... On the RPG front mostly FGU titles.
06:01:42 <rdrop-exit> Cool
06:04:39 <proteusguy> Ever play 7-player Empires in Arms campaigns?
06:05:21 <rdrop-exit> Only once, the original ADG edition, we never finished it.
06:05:48 <rdrop-exit> Sold my AH edition about 10 years ago.
06:06:28 <proteusguy> Well 120 turns is over 200 hours of game play. I think in college I got to 1812 once. I have 3 copies of the game and laminated maps. That was a big one with our group. Had the map setup for weeks in my bedroom.
06:07:59 <rdrop-exit> Cool, we had World in Flames magnetized on the wall for a year or two.
06:08:52 <rdrop-exit> Got tired of the holes in the rules.
06:09:20 <rdrop-exit> I hosted the Asia-PAC ASL tournament a couple years ago in Manila.
06:09:34 <proteusguy> Wow - how many players showed up?
06:10:00 <rdrop-exit> 25 or so.
06:10:34 <proteusguy> Good turn out. I dabbled in ASL but most of the players in that scene were pretty hardcore about it and I'd get taken out quickly.
06:11:51 <rdrop-exit> It is, overall I prefer operational level games.
06:12:21 <proteusguy> I generally trend towards strategic level that also have significant economic and diplomatic aspects. 
06:13:27 <rdrop-exit> I've taken a break from wargaming for the last year or so, need to get back with my buds and start playing again.
06:13:51 <rdrop-exit> Sorry Forthwrights for the off topic posts :)
06:13:55 <proteusguy> My break has been over a decade but I have managed to get in some AD&D occasionally.
06:14:10 <proteusguy> It's Friday.... we can slack.
06:14:34 <rdrop-exit> :)
06:18:08 <rdrop-exit> I sold off most of my games when I moved from San Francisco to France, and again when I moved from France to Manila.
06:18:32 <rdrop-exit> Still have enough to last me the rest of my days though.
06:20:12 <rdrop-exit> I miss the old Bay Area wargaming cons, and Origins, when it was still wargame oriented.
06:54:58 <Zarutian> looking at the plotQuadBezierSeg function at http://members.chello.at/~easyfilter/bresenham.html I am trying to see if those doubles could just be fixed decimal point
06:59:18 <john_cephalopoda> Zarutian: You got a working Bresenham?
06:59:38 --- quit: mark4 (Quit: Leaving)
07:01:08 <Zarutian> john_cephalopoda: the straight line? yeah, I just translated the C code version from that page into assemblerish Forth
07:06:44 --- quit: dave9 (Ping timeout: 246 seconds)
07:09:25 <corecode> pretty sure you can do bresenham with just integers
07:10:24 <WilhelmVonWeiner> you definitely can
07:10:33 <corecode> can't you just add your xdist and subtract your ydist per step
07:10:43 <corecode> until you get an overflow on y
07:11:01 <corecode> then you move one more point, subtract ydist, continue
07:11:06 <Zarutian> corecode: it is the Bézier curve segment one that I am looking at.
07:11:11 <corecode> oh
07:11:19 <corecode> fancy stuff
07:12:20 <Zarutian> the pdf linked from that page has many variations on how to implement drawing these lines and shapes
07:14:22 --- join: Kumool (~Khwerz@adsl-64-237-235-188.prtc.net) joined #forth
07:14:47 <WilhelmVonWeiner> Zarutian: that algorithm is different to the example I was looking at and makes the thought process much simpler
07:15:15 <WilhelmVonWeiner> all the example code i'd seen had so much duplicate bloat
07:19:20 <tabemann> back
07:20:23 <tabemann> rdrop-exit, I moved wordlists, aside from the linked lists of words themselves, out of the core
07:21:02 <tabemann> and next thing up is implementing marker
07:26:53 --- quit: tabemann (Ping timeout: 260 seconds)
07:47:29 <rdrop-exit> Cool
08:04:15 <the_cuckoo> if true else false then <- always find this a bit clumsy for some reason - going to add ?{ true }:{ false } which gives a nice ?{ } for the if then case, and even the invert isn't that bad as ?{ }:{ false }
08:08:38 <rdrop-exit> Similar to the control flow words from LaForth
08:13:33 <WilhelmVonWeiner> the_cuckoo: I usually rdrop out of another word to avoid that
08:13:48 <WilhelmVonWeiner> or `if true exit then false`
08:15:24 <rdrop-exit> Me too, I use mostly `?;` and `0;`
08:16:06 <rdrop-exit> I can't recall the last time I used else
08:23:14 <the_cuckoo> yeah - it's a funny one - don't really know why if else then feels wrong other than a vague notion that else and then are the wrong way round i guess
08:23:40 <WilhelmVonWeiner> else considered smelly
08:23:41 --- quit: proteusguy (Ping timeout: 258 seconds)
08:23:47 <the_cuckoo> :)
08:24:10 <WilhelmVonWeiner> http://wiki.c2.com/?ElseConsideredSmelly
08:27:08 <zy]x[yz> that's why in my forth i called it OTHERWISE instead of ELSE
08:29:30 --- quit: Kumool (Quit: EXIT)
08:43:35 <the_cuckoo> another non-conventional think i think.. <int> pick - 0 pick is tos, 1 pick is over etc... but i have -1 pick as the very bottom, and 0 - depth as the top
08:45:21 <rdrop-exit> The road to hell is paved with PICKs :)
08:46:24 <the_cuckoo> 1 2 3 .s <3> 1 2 3  ok
08:46:26 <the_cuckoo> -1 pick .s <4> 1 2 3 -1  ok
08:46:32 <the_cuckoo> rdrop-exit: :)
08:46:53 <the_cuckoo> i don't get why gforth does that with the -1 pick
08:48:07 <the_cuckoo> drop  ok
08:48:09 <the_cuckoo> -2 pick .s <4> 1 2 3 6395112  ok
08:48:11 <the_cuckoo> huh?
08:48:21 <WilhelmVonWeiner> PICK means your code could probably be far simpler
08:48:28 <WilhelmVonWeiner> use the return stack more.
08:50:09 <rdrop-exit> A stack is not an array
08:50:42 <the_cuckoo> don't have one :) - not yet anyway - may add it for word definition as it's kinda nice there - but for iterating through n items in push order, i have a rather funky: 0 n do 0 i - roll . loop (and yes, that runs from n to 1)
08:51:36 <the_cuckoo> oh i'm using it as one for sure :) - inverted as the 0 is the top of stack
08:52:39 <the_cuckoo> but then, i have fixed size objects on my stack, so that works for me
08:52:51 <rdrop-exit> =8-O
08:54:18 <the_cuckoo> yeah - it's weird - the data stack is std::deque< boost::any > so no matter what the object is inside the any, the stack grows in a fixed way
08:57:46 <the_cuckoo> but i guess that answers my original question - a normal forth stack wouldn't be that easy to navigate
08:58:27 <the_cuckoo> but surely then a -ve pick should be an error?
09:03:02 --- join: dave9 (~dave@193.060.dsl.syd.iprimus.net.au) joined #forth
09:07:48 <rdrop-exit> A stack machine's stack is not ncessarily in program addressable memory
09:11:01 --- join: dys (~dys@tmo-104-204.customers.d1-online.com) joined #forth
09:14:43 <the_cuckoo> fair point
09:40:17 --- join: Kumool (~Khwerz@adsl-64-237-235-188.prtc.net) joined #forth
10:37:05 --- quit: rdrop-exit (Quit: Lost terminal)
10:40:24 --- join: proteusguy (~proteus-g@cm-58-10-154-246.revip7.asianet.co.th) joined #forth
10:40:24 --- mode: ChanServ set +v proteusguy
11:00:38 <siraben> Overhauled boot sequence: https://github.com/siraben/zkeme80/blob/master/demo.gif
11:35:46 --- quit: gravicappa (Ping timeout: 246 seconds)
11:41:28 <proteusguy> siraben, wow so old school! looks cool.
11:45:35 <proteusguy> you make your own font or was that already existing?
11:52:00 <the_cuckoo> that's pretty nice indeed :)
12:00:48 <presiden> siraben, nice one
12:20:50 --- quit: X-Scale (Ping timeout: 245 seconds)
12:23:03 --- join: X-Scale` (~ARM@46.50.4.5) joined #forth
12:55:51 --- join: mtsd (~mtsd@94-137-100-130.customers.ownit.se) joined #forth
13:13:35 * Zarutian muses about bit-skip-vector trie struct for keeping lots of sha256 hashes
13:15:06 <Zarutian> first cell is a bitmap of what nth hexdigits are actually present. (N being the current 'level/depth' into the hash being looked up)
13:15:57 --- quit: mtsd (Quit: WeeChat 1.6)
13:16:00 <Zarutian> followed by offset pointers, to next level/depth, that are present or not per that bitmap
13:18:08 <Zarutian> now, use this trie construct a 'compression' of a datablob by incrementally hashing the blob 4-8 bytes at a time
13:19:20 <Zarutian> to decompress you bruteforce the first 4-8 bytes by looking if a matching hash exists in the trie
13:20:15 <Zarutian> then you bruteforce the next 4-8 bytes of the blob and check if there is an hash for the blob so far
13:20:25 <Zarutian> and so on
13:21:26 <Zarutian> until no more hashes are found present in the trie, or the expect length of the datablob is reached
14:02:31 <zy]x[yz> isn't the end result worse than a full sha256 brute force?
14:03:16 <zy]x[yz> by the end you've still brute-forced the entire thing, except with some added intermediate steps
14:03:31 <zy]x[yz> maybe i don't understand it.  i don't have a strong cryptography background
14:05:18 <john_cephalopoda> I've been trying all morning to make a pun with the words "difficult" and "Diffie cult" but haven't found a good one yet.
14:06:20 <john_cephalopoda> "Crypto is too difficult? Join our Diffie cult!" or something.
14:09:24 <Zarutian> zy]x[yz: the analogy is like those combination locks on briefcases
14:10:42 <Zarutian> zy]x[yz: if you can find the first digit by ten tries and next by ten tries and so on to the fifth then you have done 50 tries instead of 100000
14:22:54 <zy]x[yz> ah, yeah, i see that now.  my storing intermediate steps you've defeated the avalanche effect where information is normally lost
14:23:00 <zy]x[yz> s/my/by/
14:26:02 --- quit: dne (Remote host closed the connection)
14:28:29 --- join: dne (~dne@jaune.mayonnaise.net) joined #forth
14:54:16 --- join: PSnacks (~PSnacks@p200300CB2704ABFC803B515519F4E611.dip0.t-ipconnect.de) joined #forth
15:03:08 --- quit: Kumool (Quit: EXIT)
15:08:18 --- join: Kumool (~Khwerz@adsl-64-237-235-188.prtc.net) joined #forth
15:54:02 --- quit: dys (Ping timeout: 244 seconds)
16:11:14 <siraben> proteusguy: the OS logo is split up into 4 (8 by 8) chunks https://github.com/siraben/zkeme80/blob/87edca4e52a516c465b5b36017e86b708b4cdb12/bootstrap-flash1.fs#L52 otherwise I just reused https://github.com/KnightOS/kernel/blob/master/src/01/font.asm
16:11:33 <siraben> Wasn't planning on making my own, hehe
16:12:58 <siraben> the icons you see in the menu buttons are drawn by me
16:15:30 --- nick: bhodgins -> FatalNIX
16:26:46 <siraben> Looks like Forth made the front page of HN again https://news.ycombinator.com/item?id=18885493
16:35:49 <WilhelmVonWeiner> First comment is always someone complaining about how they can't understand bad code and applying that to the entirety of forth
16:38:52 <PoppaVic> Or they didn't have a book, blah-blah
16:41:15 <ttmrichter> WilhelmVonWeiner: I always get the urge with those twerps to put up something from Obfuscated C and complain that C is unreadable.
16:42:55 <siraben> Any language is unreadable in the hands of a bad programmer
16:43:42 <Zarutian> something I have noticed with many C inspired programming languages programmers is that what they consider a step is quite huge granularity wise.
16:45:28 <PoppaVic> About the only reason I really liked forth in the Long-Ago - which always peeved others - was the blockfile size.. It enforces decomposition.
16:46:00 <siraben> \
16:46:59 <Zarutian> lets take an example: in C arr[idx] is thought as single 'step' but in Forth it is ( idx arr -- ) >R getArrItemSize * R> + @
16:47:48 <siraben> I've seen something like *x[foo++] before
16:47:48 <PoppaVic> you must mean idx[arr]
16:48:04 <PoppaVic> siraben: so?
16:48:36 <Zarutian> PoppaVic: same difference, but you see what I am getting at
16:48:39 <siraben> PoppaVic:  it made me loathe trying to work with low level memory stuff like that, until I found Forth
16:48:54 <siraben> @ and ! are really nice notations
16:49:00 <PoppaVic> Zarutian: I suspect you mean 'operation' perhaps when you say 'step'.
16:49:32 <PoppaVic> siraben: I get along fine with C - just keep the standard-whores away from me. I do like forth tinkering
16:49:37 <Zarutian> PoppaVic: nope, I mean 'step' in the sense of unit of the programmers thinking.
16:50:06 <PoppaVic> Zarutian: oh, hell - yeah.. ##C has some funky term for that nowadays - always made me want to slap someone.
16:50:15 <PoppaVic> ?? points
16:50:27 <PoppaVic> something-points.. damn.. I can't recall it
16:50:56 <PoppaVic> Zarutian: I just think of what I am trying to reach/grab/stuff.
16:51:18 <WilhelmVonWeiner> Maybe I'm partially to blame for the bad overview of Forth `: G ( R: tgt --) here r> r> swap >r dup >r - 1 cells / r> ! ;`
16:51:21 <Zarutian> so what many C programmers do is do that full Forth sequence every time
16:51:39 <Zarutian> instead of say making arr@ and arr!
16:52:02 <PoppaVic> <shrug> Whatever floats their boat.
16:52:45 <PoppaVic> WilhelmVonWeiner: yeah, I figure yer the posterboy for "those needing to be beaten with a frozen salmon"
16:52:57 <Zarutian> if you factor the array access into that you can actually spend two frontmost cells of the array as length and item size
16:53:23 <Zarutian> WilhelmVonWeiner: let me guess, you have done a lot of mathematics
16:53:39 <PoppaVic> Zarutian: certainly there are multiple methodologies there, but folks have to learn somehow.
16:55:21 <Zarutian> PoppaVic: yeah by reading Starting Forth and Thinking Forth and really think about what those two books say
16:56:13 <PoppaVic> I've even told C/php/ruby/perl/whatever folks about "thinking".. THey see "Forth", sneer, and go back to their trash
16:58:32 <Zarutian> another thing I have noticed is that people with hands on electronics background are more likely to like Forth than various Algol68 syntaxed languages
16:59:37 <Zarutian> probably they have written MCU code in assembly for various archs
17:00:13 <Zarutian> (why? Because C/C++ compilers STILL produce crap code when targetting MCUs)
17:00:25 --- join: TheCephalopod (~john@unaffiliated/john-cephalopoda/x-6407167) joined #forth
17:00:45 --- quit: john_cephalopoda (Disconnected by services)
17:00:47 --- nick: TheCephalopod -> john_cephalopoda
17:01:02 <Zarutian> s/probably they/probably because they/
17:02:08 <WilhelmVonWeiner> Zarutian: lol, I'm just bad at coming up with names.
17:02:27 <PoppaVic> Well, if you are writing a small program for a "fat" mcu, go for it - get 'er doone... But, yeah: looking at the .S file makes you want to scream
17:02:34 <Zarutian> WilhelmVonWeiner: oh, one of the hard problems of compsci, naming things.
17:02:57 <WilhelmVonWeiner> PoppaVic: It's code to store the counted pattern matches in the started of a CREATEd data structure 
17:03:25 <PoppaVic> WilhelmVonWeiner: seriously - too ugly and obtuse for me to care.
17:03:42 <WilhelmVonWeiner> Don't know a better way to do it. Never bothered to make sata structures before,
17:04:15 <WilhelmVonWeiner> x/sata/ c/data/
17:05:08 <Zarutian> PoppaVic: yeah, specially when one is targetting an MCU with 32KiBiInstr of Program memory but the compiler complains that it has run out of target-memory.
17:06:16 <Zarutian> PoppaVic: turns out that the static text strings for the LCD display are copyed into wram before being sent out.
17:07:39 <Zarutian> PoppaVic: meanwhile most assemblers just keep them in program memory and put in an accessor routine. (Gimme nth char of str x basically)
17:08:21 --- join: rdrop-exit (~markwilli@112.201.166.158) joined #forth
17:08:25 <PoppaVic> Yeah, some of the logic gcc used for the atmel stuff makes me want to hurt them.
17:08:44 <PoppaVic> happily, I found the 'naked' attribute and a few other stupidities
17:10:10 <PoppaVic> Zarutian: it's also irksome because a keystone facet of C has always been "never copy data if you can help it". So, accessing it from wherever it lives is almost always "better".
17:11:51 <Zarutian> or the time when the compiler 'optimized' away the call to fast_memcopy() because it thought it had no effect when in fact it was setting up various special regs and initializing global variables that followed
17:12:22 <Zarutian> it was a keystone facet of C, now it is "cache locality trumps all! YOLO!"
17:17:06 <Zarutian> I have started to consider the first memory device that a cpu instruction fetch hits to be the program memory and same with load/stores to be the work memory
17:17:36 <Zarutian> what most people call main memory nowdays I just look at as fast ramdisk
17:34:22 --- join: dave0 (~dave0@193.060.dsl.syd.iprimus.net.au) joined #forth
17:34:28 --- join: tabemann (~tabemann@h193.235.138.40.static.ip.windstream.net) joined #forth
17:34:48 <dave0> hi
17:35:06 <tabemann> hey
17:35:54 <tabemann> I should implement MARKER for hashforth, now that I've moved much of the functionality for wordlist out of the core
17:36:34 <tabemann> *wordlists
17:44:02 <siraben> Forth would be a great teaching language
17:44:33 <siraben> What better way to teach how computers work than the entire system, top down?
17:45:28 <siraben> At first I didn't like how Forth was devoid of safety checks, and that forced me to write code that was testable, modular and simple.
17:45:30 <tabemann> your assignments over the course of a semester would be to build a fully functional, multitasking forth implementation
17:45:54 <siraben> How would I do multitasking on a Z80?
17:46:06 <siraben> I have coroutines, I suppose
17:46:10 <tabemann> the simplest way is cooperative multitasking
17:46:15 <tabemann> each task gets its own stacks
17:46:15 <siraben> Yeah
17:46:22 <tabemann> and they use PAUSE to switch tasks
17:46:35 <PoppaVic> you could switch entire register banks, too
17:46:43 <siraben> The Z80 has shadow registers
17:47:00 <siraben> I'm not currently using them for anything, but they're used to handle interrupts
17:47:26 <tabemann> all you really need are SP@, SP!, RP@, and RP!
17:48:23 <tabemann> saving and loading TOS-in-a-register would be handled by those already
18:12:28 --- quit: tabemann (Ping timeout: 250 seconds)
18:19:01 --- quit: PSnacks (Ping timeout: 250 seconds)
18:30:35 <rdrop-exit> While on a real or FPGA-based stack machine, you could have built-in support for a small fixed number of cooperative tasks by having dedicating a pair of hardware stacks to each, and using a task selection register to switch between them.
18:33:41 <rdrop-exit> e.g. on a stack machine with support for two tasks, the task selection register would be a single bit.
18:35:44 --- join: tabemann (~tabemann@2602:30a:c0d3:1890:7d04:1e92:6395:7db) joined #forth
18:35:54 <rdrop-exit> Each stack would have its on bank of ip, sp, rp registers (optionally an extra rp for exceptions)
18:36:14 <PoppaVic> Or.... you could buy a 2nd processor!
18:36:31 <PoppaVic> or... Add more pixie-dust!
18:40:47 <rdrop-exit> or go the Parallax Propeller route
18:44:34 <rdrop-exit> PoppaVic: FPGAs are pixie-dust! :)
18:44:46 <PoppaVic> yeah,
18:45:53 <siraben> My OS takes around 13.5 seconds (real time) to compile for the first time and then 0.424 seconds (real time) for subsequent recompiles
18:45:53 <siraben> Well assuming only changes in the Forth layer
18:47:25 --- join: PSnacks (~PSnacks@p508BC31E.dip0.t-ipconnect.de) joined #forth
18:49:25 <dave0> i was reading this on wikipedia   https://en.wikipedia.org/wiki/Lua_(programming_language)#C_API    it has an interesting way to interface with C... it uses a stack to call and return values to a C function.. very forthy
18:50:25 <PoppaVic> it's how C used to do it - slowly - before they usurped registers.
18:50:46 <dave0> oh!
18:50:58 <dave0> brb gotta do laundry
18:51:14 <rdrop-exit> I feel for you
19:02:13 <rdrop-exit> BTW, they now make a version of the 8-core Parallax Propeller with Forth in ROM
19:02:49 <rdrop-exit> Parallax P2X8C4M64P 8-core 32-bit Realtime controller with Forth in ROM
19:07:15 <rdrop-exit> There's some videos on the Parallax Youtube channel of them receiving the first test batch of chips
19:20:58 <rdrop-exit> https://www.youtube.com/watch?v=x7-gGOBnk68&t=2094s
19:26:23 <tabemann> back
19:42:47 <tabemann> question
19:42:58 <tabemann> does MARKER typically reset the states of the stacks?
19:50:36 --- quit: dddddd (Remote host closed the connection)
19:51:00 <crc> tabemann: I don't believe ao
19:52:04 <tabemann> does it only reset the dictionaries and wordlists?
19:52:12 <tabemann> and the user space
19:53:24 <rdrop-exit> IIRC, LAST and DP
19:53:55 <rdrop-exit> More depending on the internal structure of your Forth
19:55:10 <rdrop-exit> On a VM-based Forth you could just save and restore the VM image
19:55:18 <tabemann> yeah, my MARKER resets the dictionaries, wordlists, wordlist order, LATESTXT, and user space pointer
20:00:46 <rdrop-exit> On a VM you can save a snapshot of the dictionary to disk and reload it later, taking care to shutdown any IO.
20:03:34 --- quit: PSnacks (Read error: Connection reset by peer)
20:04:00 --- join: PSnacks (~PSnacks@p200300CB272FF3FC18381BC7823A490C.dip0.t-ipconnect.de) joined #forth
20:06:06 <tabemann> except that hashforth is IO limited, to avoid encoding IO in the set of opcodes
20:06:13 <tabemann> it only has terminal IO
20:06:48 <tabemann> but yes
20:07:14 <tabemann> it would be easy to take a snapshot of hashforth's state and save it to disk
20:21:26 <tabemann> yes - marker seems to work
20:25:29 <tabemann> I need a better way of doing this
20:26:10 <tabemann> loading programs which aren't in assembled images into hashforth literally requires piping/redirecting in the code from elsewhere, and then no other user input is possible
20:40:59 <tabemann> hmm
20:42:00 --- join: gravicappa (~gravicapp@h109-187-2-216.dyn.bashtel.ru) joined #forth
20:43:49 <rdrop-exit> On my umbilical host I store the source code in blocks.
20:44:38 <rdrop-exit> The blocks are part of the VM image.
20:45:31 <rdrop-exit> They all reside above a certain address: STORAGE
20:45:38 <rdrop-exit> : block ( blk# -- a ) k STORAGE + ;inline
20:47:15 <tabemann> the thing is for me I'm regularly rebuilding my image, so I don't want to save code in the image itself that isn't getting rebuilt
20:47:26 <tabemann> maybe I can build forth code into the image
20:47:39 <tabemann> alongside the VM assembly code built into the image
20:52:15 <rdrop-exit> When I lauch the VM executable, the VM loads a bin file from disk that contains everything.
20:52:39 <rdrop-exit> * launch
20:55:25 <rdrop-exit> Gotta walk the dogs, catch you all later
20:55:46 --- quit: rdrop-exit (Quit: Lost terminal)
21:23:15 --- quit: proteusguy (Remote host closed the connection)
23:59:59 --- log: ended forth/19.01.11