2022-02-16 05:28:07 <joe9        > anyone have the backlog of this channel? I was not logged in for sometime 24 hours ago.
2022-02-16 05:28:23 <joe9        > I am not sure if anyone had any suggestions on the security before veltas chimed in.
2022-02-16 05:31:25 <f2a         > check topic joe9
2022-02-16 05:31:27 <f2a         > there are logs there
2022-02-16 06:26:54 <veltas      > RIP person who wrote eqn (the equation macro package for nroff etc that inspired TeX)
2022-02-16 06:27:06 <veltas      > "Lorinda Cherry"
2022-02-16 06:27:29 <veltas      > Oh they wrote dc as well, actually more relevant to channel
2022-02-16 06:28:15 <f-a         > dc is a dumbed down forh :P
2022-02-16 06:28:44 <veltas      > I wonder if it's easier to fit dc in MBR
2022-02-16 06:29:01 <veltas      > I think dc lets you write programs, not sure
2022-02-16 06:29:27 <veltas      > I like eqn, from the couple of times I've used it
2022-02-16 09:50:06 <lagash      > veltas: RIP as in, they passed?
2022-02-16 10:13:44 <veltas      > lagash: Yes, it stands for 'rest in peace'
2022-02-16 10:14:39 <veltas      > Just mentioned it because we've gabbed about troff macros packages in here
2022-02-16 11:27:35 <veltas      > Huh I guess it's not really a 'macro package', it's more like a preprocessor, but it's all the same really. Amazing how any of this stuff worked without horrid syntax conflicts
2022-02-16 17:01:27 <vms14       > guys do you use forth currently for something? even for script-like stuff?
2022-02-16 17:11:31 <remexre     > bare-metal stuff, and occasionally for "exotic" things in normal linux userspace
2022-02-16 17:12:05 <remexre     > writing a properly optimizing forth for userspace linux right now, mostly as a compile target for a different lang
2022-02-16 17:43:34 <vms14       > remexre: any forth tutorial to understand compiling words?
2022-02-16 17:44:27 <remexre     > if you're doing an ITC or DTC design, I'd recommend Moving Forth
2022-02-16 17:44:28 <vms14       > I don't understand what the , does and the inmediate and >does i understand it defines the behavior of the runtime
2022-02-16 17:44:50 <vms14       > no, I just want to understand forth in order to see if I'll like it
2022-02-16 17:44:54 <remexre     > ah
2022-02-16 17:44:56 <remexre     > 1sec
2022-02-16 17:45:17 <remexre     > try http://galileo.phys.virginia.edu/classes/551.jvn.fall01/primer.htm ?
2022-02-16 17:45:50 <vms14       > thanks remexre
2022-02-16 17:45:59 <remexre     > it's written against win32forth, but i don't think anything (maybe some of the floating-point stuff?) is specific to it
2022-02-16 17:46:35 <vms14       > I think it's probably I'll end loving forth, so I want to see
2022-02-16 17:47:29 <vms14       > I love common lisp because it lets me blend the problem into the language and seems forth also does that
2022-02-16 17:48:07 <vms14       > plus forth is so easy to implement that you're likely to create an usable custom forth that does whatever you want
2022-02-16 17:48:36 <vms14       > which means you could implement using any device platform library and language and take stuff from there
2022-02-16 17:49:28 <vms14       > and if you go to bootstrap you can easily make multiple ports of that forth
2022-02-16 17:50:19 <vms14       > which means you'll have the "same" language for doing whatever you need and will always be the right choice for the task
2022-02-16 17:50:40 <vms14       > why do you like forth?
2022-02-16 17:52:46 <remexre     > mostly because C is awful tbh :P like, a forth implementation I wrote myself is infinitely more likely to let me do what I want without exploding on me
2022-02-16 17:53:12 <remexre     > especially "weird stuff" like coroutines, funclets, etc
2022-02-16 17:53:57 <eris[m]     > funclets?
2022-02-16 17:55:40 <remexre     > a bunch of little bits of code that can all serve as "function entrypoints" but you can freely goto between in their implementation
2022-02-16 17:56:33 <vms14       > you implement them by using the return stack?
2022-02-16 17:56:34 <remexre     > this is a subset of tail-call if you have that
2022-02-16 17:57:06 <vms14       > how should I represent words in a language other than asm
2022-02-16 17:57:27 <remexre     > my forths has a primitive, (BRANCH) (also called (BR) by other people I think?) that sets the forth IP to a constant; I just use that
2022-02-16 17:57:39 <vms14       > I mean words are primitives which are native code or lists of words, but they're really memory address and lists of memory address
2022-02-16 17:57:55 <vms14       > so they're kind of an asm label
2022-02-16 17:58:27 <vms14       > how should I emulate this behavior in order to implement later the return stack using those adresses
2022-02-16 17:58:34 <vms14       > I have no idea
2022-02-16 17:58:39 <remexre     > possibly something like https://git.sr.ht/~remexre/rtf/tree/trunk/item/bs/word.h#L15
2022-02-16 17:59:15 <vms14       > also about a repl with a transpiler in C, should I compile and load dynamically with dlopen or recompile everything?
2022-02-16 18:01:19 <remexre     > like compiling forth to C?
2022-02-16 18:02:48 <vms14       > remexre: yes, something like words are just blocks of code that when called write c code that ends in a .c file
2022-02-16 18:04:08 <remexre     > like... why do that
2022-02-16 18:04:31 <remexre     > that sounds like you're putting the compile vs execute boundary in a weird place
2022-02-16 18:05:22 <vms14       > cause it would let you provide new words that are "native code" (target code in this case) in runtime, but I have no idea about incremental compilers or alike
2022-02-16 18:07:06 <vms14       > actually a repl would only be for dev purposes and if you want to make executable then it gets transpiled, but providing a repl for a transpiler, forces you to compile stuff in the targe language and execute it, or make a whole new interpreted version
2022-02-16 18:07:40 <remexre     > if you do want to understand JIT compilers (both the JVM/V8 flavor and the CL:COMPILE flavor), IMO https://dl.acm.org/doi/pdf/10.1145/1706299.1706313 is a good thing to read to "get a flavor for it," just ignore the verification parts that don't sound relevant to "how does this actually work"
2022-02-16 18:08:21 <remexre     > but what I'm doing is accumulating chunks of an IR when COMPILE, gets called
2022-02-16 18:08:50 <remexre     > (and other compiler-only magic used for IF THEN ELSE RECUR etc)
2022-02-16 18:09:15 <remexre     > and executing ; triggers actual compilation
2022-02-16 18:09:39 <remexre     > to a big rwx memory region
2022-02-16 18:10:15 <remexre     > but I always generate position-independent code, so I can fairly easily either just jmp to it (well, B since I'm on aarch64, but same difference) or put it in an ELF file
2022-02-16 18:10:33 <remexre     > I would highly highly highly recomment implementing a DTC forth system before you try something like this though
2022-02-16 18:10:39 <vms14       > makes me subscribe to read that page
2022-02-16 18:10:51 <remexre     > ugh, 1sec
2022-02-16 18:11:46 <remexre     > try https://cdn.remexre.xyz/files/52d2ed3238f6d3c4fb6715a97329bb64b2247e6c.pdf
2022-02-16 18:12:10 <vms14       > thanks :D
2022-02-16 18:13:56 <remexre     > I'd say sec3-5 are the useful ones for "how does a jit actually work"
2022-02-16 18:19:58 <joe9        > vms14: regarding does, etc. check out f83.
2022-02-16 18:20:10 <joe9        > it makes things a little more simpler.
2022-02-16 18:20:19 <joe9        > Dr. Ting wrote a book about it too.
2022-02-16 18:23:10 <vms14       > joe9: http://forth.org/OffeteStore/1003_InsideF83.pdf this one?
2022-02-16 18:27:37 <vms14       > remexre: I'll share your pdf
2022-02-16 19:12:29 <vms14       > joe9: I'm trying it f83 with dosbox
2022-02-16 19:12:38 <joe9        > the book is good.
2022-02-16 19:13:25 <vms14       > https://github.com/ForthHub/F83/blob/master/f83.com took that and just typed "dosbox f38.com" and it works :D
2022-02-16 19:13:57 <vms14       > I have pforth, but now I can follow the book and everything should work
2022-02-16 19:14:13 <vms14       > thanks for the recommendation
2022-02-16 19:44:15 <vms14       > joe9: why do you like forth?
2022-02-16 19:47:36 <vms14       > and why in f83 ." text " says ok but does not print anything nor touches the stack?
2022-02-16 20:19:44 <vms14       > ooh, I'm looking at dx forth
2022-02-16 20:20:00 <vms14       > trying the spiral demo in dosbox
2022-02-16 20:20:55 <vms14       > are there games written using dx forth
2022-02-16 20:20:59 <vms14       > ?