2023-02-26 03:28:56 <Stalevar    > Would forth work as a general purpose command line language instead of bash?
2023-02-26 03:29:08 <Stalevar    > if somebody made a forth shell
2023-02-26 05:52:35 <MrMobius    > Sure. Forth doesn't allocate and handle strings automatically like bash so it probably wouldn't be very practical but it would work
2023-02-26 07:16:48 <KipIngram   > Stalevar: I've often thought about trying to do a Forth "shell' for my Linux; the shell is, after all, just a program.  But I've also not yet seen any good way to replicate the capabilities that bash "pipelines" give you.  I've decided Forth really does do pipelines too, but it passes information along the pipe using a stack of integers rather than the way bash does it.  And I don't see how I'd "do the stufF
2023-02-26 07:16:50 <KipIngram   > I do with pipelines" yet.
2023-02-26 07:17:50 <KipIngram   > bash pipelines a different sort of information, I think.
2023-02-26 07:18:26 <KipIngram   > But that doesn't mean you couldn't turn a Forth into a shell - I'm just not sure if I could do everything with it I'd want to do.
2023-02-26 07:19:11 <Stalevar    > but it will at least be better than basic shell on 8-bit computers, right?
2023-02-26 07:19:36 <KipIngram   > I would think so, yes.
2023-02-26 07:20:18 <KipIngram   > bash is just very inherently "file / stream" oriented.
2023-02-26 07:20:30 <KipIngram   > Because of that, it has certain things it does very well.
2023-02-26 07:22:45 <KipIngram   > Maybe the way to say it is that Forth is good at manipulating a stack of integers, and letting you make "pipelines" that work on such stacks.  Bash is good at manipulating a file system full of files, and lets you pipeline those operations.
2023-02-26 07:22:55 <KipIngram   > just different "information domains."
2023-02-26 07:23:50 <KipIngram   > You certainly could write Forth words that manipulated disk blocks or files, and you might arrange for those words to have something like stdin/stdout, and then you could make pipelines of those.
2023-02-26 07:27:49 <KipIngram   > While thinking about this it occurred to me that in Forth, there's a clear distinction between your DATA, which is what's on the stack, and your COMMANDS, which is the stuff you type.  In bash, your commands and your data all kind of go through the same channel.  Commands come into the system through stdin.  On the output side, there is output to "stdout," but there are also manipulations of that stack.
2023-02-26 07:28:14 <KipIngram   > Anyway, I don't think i have it all figured out yet in a really clear way.
2023-02-26 07:29:25 <KipIngram   > There are definitely "capabilities" bash gives me that I wouldn't want to give up, so until I figure out how to capture those well with a Forth shell, it's  kind of a non-starter for me.
2023-02-26 11:28:10 <MrMobius    > Stalevar: you mean basic shell like bash she'll or basic as in BASIC?
2023-02-26 12:32:24 <JITn        > chmod and chown are absolutely BASIC, IMHO!!
2023-02-26 12:33:14 <JITn        > The worst of 8-bit Era.
2023-02-26 12:34:44 <thrig       > me: "you'll fix those messed up permissions, right?" windows admins: <deer in headlights look>
2023-02-26 12:36:36 <JITn        > BTW Dir Permissions in ext4fs are irrelevant for contained files.  What a waste of memory.
2023-02-26 12:43:28 <JITn        > UNIX, and ofcourse BSD and Linux, are 8-bit relics.  They are like Raiders from Mad Max.  Lest We abstract AMD64 into PDP(11).
2023-02-26 12:55:12 <KipIngram   > JITn: Is the "byte oriented" nature of the systems, vs. word oriented, what you're targeting there?
2023-02-26 13:08:04 <JITn        > KipIngram: I meant to say a very idiosyncratic system.  It predates even the ""Structured Programming"".  I see no system to UNIX.  But then X in UNIX is just a pun on CS (in turn Computer System).  Should it be a hint and a warning that UNIX is far from a Computer System?!
2023-02-26 13:10:58 <thrig       > structured programming: make spaghetti of your code so there is only one return
2023-02-26 13:12:56 <JITn        > Python makes use of Hash Tables straightforward.  Is it time for DHTs?!
2023-02-26 13:14:01 <JITn        > thrig: IMHO File System, like ext4fs, is bogus.
2023-02-26 13:16:41 <JITn        > It's Permissions, time stamps (ctime, atime), UID, GUI, and whatnot are a waste of mental effort!!  ""Sticky Bit""!?  ""Security"" is a dirty word!!
2023-02-26 13:20:26 <JITn        > Was SmithForth ( http://dacvs.neocities.org/SF/ ) discussed?
2023-02-26 13:42:33 <KipIngram   > How would you do a file system differently?
2023-02-26 13:43:09 <KipIngram   > If you're going to have shared information resources in a system, it seems like there would need to be *some* method for managing access privileges.
2023-02-26 13:43:24 <KipIngram   > Permissions isn't the only one, but it does work.
2023-02-26 13:44:34 <gordonjcp   > JITn: actually Unix is a pun on Multics, but you've never heard of that
2023-02-26 13:45:07 <gordonjcp   > the way that Unix-y file permissions work is actually pretty okay, it works well for a broad range of applications and is simple to implement
2023-02-26 13:45:27 <gordonjcp   > JITn: almost anything else like ACLs gets complicated and hard to manage way faster than it gets more convenient
2023-02-26 13:47:49 <gordonjcp   > JITn: "more flexible", "more configurable", is not always better
2023-02-26 13:47:58 <gordonjcp   > indeed, I'd say often it's a lot worse
2023-02-26 13:48:17 <gordonjcp   > I could tell you a story about a software synthesizer I wrote, based on FM synthesis that highlights this
2023-02-26 13:48:34 <gordonjcp   > but it's kind of long, and you're probably not into old 1980s synths
2023-02-26 13:49:14 <KipIngram   > I'm just having trouble imagining anything other that permission bits "winning the simplicity contest."  How do you get more basic than that?
2023-02-26 13:49:22 <gordonjcp   > KipIngram: CP/M
2023-02-26 13:49:37 <gordonjcp   > KipIngram: no directories, no permissions other than "read only" and "hidden"
2023-02-26 13:49:46 <KipIngram   > I mean, I guess you could... ok - that.
2023-02-26 13:49:53 <KipIngram   > you could "simplify the scope of the permissions."
2023-02-26 13:49:56 <gordonjcp   > CP/M 3 introduced "user areas"
2023-02-26 13:50:02 <KipIngram   > But I meant an equally capable system.
2023-02-26 13:50:23 <gordonjcp   > but there wasn't anything that distinguished between them other than a number, there was no concept of users and passwords
2023-02-26 13:51:00 <KipIngram   > That sounds similar to Forth's "user variables."
2023-02-26 13:51:13 <KipIngram   > Which is not really "security"; it's more "convenience."
2023-02-26 13:51:41 <gordonjcp   > yeah "user variables" are like vocabularies but for variables, I guess
2023-02-26 13:52:27 <KipIngram   > The other real access control approach I know much at all about is capabilities.
2023-02-26 13:52:40 <KipIngram   > I've always found them interesting, but I've never actually worked with a system that uses them.
2023-02-26 14:20:47 <gordonjcp   > I think OpenSolaris has them
2023-02-26 14:23:41 <JITn        > gordonjcp: Isn't Cloud a utility that MULTICS should have been?
2023-02-26 14:26:11 <gordonjcp   > JITn: that sounds a little incoherent
2023-02-26 14:26:46 <JITn        > KipIngram: Tangentstorm coined ""an ancient curse"", refering to ""localStorage"", when you are limited to byte string.
2023-02-26 14:27:04 <gordonjcp   > what's wrong with byte strings?
2023-02-26 14:27:23 <gordonjcp   > everything any (practical) computer deals with is a byte string
2023-02-26 14:27:44 <JITn        > Can we have simple (for a start) Records rather than just byte strings?
2023-02-26 14:29:52 <JITn        > Nothing wrong.  Memory devices provide them, so why would I need a ""File System"" at all?
2023-02-26 14:31:31 <JITn        > I meant to say they finally made MULTICS and they call it ""Cloud"".
2023-02-26 14:52:20 <KipIngram   > Ok, so you want a way to embed more organized structure down at the lower levels?  Because of course you CAN have records - that's what databases are.  But you'd rather have them at a more fundamental level?
2023-02-26 14:53:57 <KipIngram   > I certainly wouldn't want to "give up" simple character files - sometimes it's exactly what you want.
2023-02-26 14:57:08 <JITn        > KipIngram: Why don't we have File System""s and ""s in RAM?
2023-02-26 14:57:15 <JITn        > 
ACTION sigh

2023-02-26 14:57:29 <JITn        > KipIngram: Why don't we have ""File System""s and ""Database""s in RAM?
2023-02-26 14:59:00 <thrig       > sqlite3 db :memory: -create true
2023-02-26 16:24:34 <KipIngram   > Yeah, I have no idea about exactly why we have certain things and don't have other things, but I imagine most "big" ideas have been tried out at some point along the way and either "caught on" or "didn't catch on."
2023-02-26 16:25:34 <KipIngram   > Regarding databases, I don't think they offer a lot of value until you need to manipulate large amounts of data, and then you're not really "in RAM" anymore.  At least I never had any appreciation at all for database tools until I was faced with needing to deal with large amounts of data, and then I realized just how great they are.
2023-02-26 16:26:34 <KipIngram   > As far as smaller jobs go, we do have tools for manipulating structured data in RAM; that's exactly what structs are, and a lot of languages have such facilities.
2023-02-26 16:27:13 <KipIngram   > We just get it at the level of compiled languages, rather than at the processor / assembly language level.
2023-02-26 16:58:56 <Zarutian_iPad> CHINA-NUMER-ONE: is this nick of yours a teaset reference? or a name of a O’Neal cyl?
2023-02-26 16:59:30 <thrig       > maybe someone is jaded
2023-02-26 18:24:57 <nmz-        > CamelCase or under_score_per_word or justkeepeverythingasasingleword?
2023-02-26 18:25:11 <nmz-        > does anybody use hungarian notation?
2023-02-26 18:26:06 <thrig       > the hungarians, presumably
2023-02-26 19:41:29 <crc         > nmz-: I generally use CamelCase for variable names, UPPERCASE for constants, and lowercase (with - if a compound name is needed) for regular word names.
2023-02-26 23:15:30 <KipIngram   > I am still delving into Tolkien lore.  Now, this isn't full time by any means - it's been weeks since I first mentioned it and that would be a little extreme even for me.  But it's like a damn bottomless pit.
2023-02-26 23:15:41 <KipIngram   > The dude was just amazingly prolific.
2023-02-26 23:16:09 <KipIngram   > LotR is just the teeny tip of the iceberg.
2023-02-26 23:20:25 <thrig       > Moria did go pretty deep