00:00:00 --- log: started forth/05.04.24
00:34:41 --- part: slava left #forth
00:41:13 --- quit: swalters ("User disconnected")
00:45:23 --- join: Topaz (~top@sown-87.ecs.soton.ac.uk) joined #forth
01:10:05 --- quit: Topaz (Read error: 113 (No route to host))
01:14:16 --- join: swalters (~swalters@2416457hfc118.tampabay.res.rr.com) joined #forth
01:41:26 --- quit: Sonarman ("leaving")
07:29:00 --- join: CXDoctor (~rex@pcp0010699353pcs.manass01.va.comcast.net) joined #forth
08:43:02 --- quit: skylan (Read error: 104 (Connection reset by peer))
08:48:10 --- join: skylan (~sjh@dialup-216-211-5-187.tbaytel.net) joined #forth
08:57:25 --- join: snoopy_1711 (snoopy_161@p54A7BEB1.dip.t-dialin.net) joined #forth
09:05:57 --- quit: Snoopy42 (Read error: 60 (Operation timed out))
09:06:12 --- nick: snoopy_1711 -> Snoopy42
09:14:24 --- quit: swalters (Read error: 104 (Connection reset by peer))
09:38:45 --- join: tathi (~josh@pcp01375108pcs.milfrd01.pa.comcast.net) joined #forth
09:54:54 --- quit: Quartus (Remote closed the connection)
10:42:37 --- join: acidx (~leandro@200-158-191-36.dsl.telesp.net.br) joined #forth
10:48:11 --- quit: tathi ("running errands")
11:13:05 --- join: swalters (~swalters@2416457hfc118.tampabay.res.rr.com) joined #forth
11:53:17 <arke> http://www.dotquote.org
12:18:10 --- quit: I440r_ ("Leaving")
12:52:19 --- quit: saon (Read error: 110 (Connection timed out))
13:02:31 --- quit: swalters (Read error: 60 (Operation timed out))
13:06:38 --- join: swalters (~swalters@2416457hfc118.tampabay.res.rr.com) joined #forth
13:11:18 --- join: shoebert (askme@cpe-70-112-100-160.austin.res.rr.com) joined #forth
13:11:19 --- quit: pengy (Read error: 54 (Connection reset by peer))
13:18:27 --- join: Herkamire (~jason@c-24-218-95-147.hsd1.ma.comcast.net) joined #forth
13:18:28 --- mode: ChanServ set +o Herkamire
13:35:04 --- part: shoebert left #forth
13:45:08 --- join: Sonarman (~cleetus@adsl-64-160-165-151.dsl.snfc21.pacbell.net) joined #forth
14:00:22 --- join: pengy (~pengy1@81.6.194.197) joined #forth
14:04:47 --- join: tathi (~josh@pcp01375108pcs.milfrd01.pa.comcast.net) joined #forth
14:12:18 <arke> http://www.dotquote.org
15:40:55 --- quit: Al2O3 (Remote closed the connection)
15:51:39 --- join: slava (~slava@CPE00096ba44261-CM000e5cdfda14.cpe.net.cable.rogers.com) joined #forth
15:56:14 <slava> arke: you might get a few hits from this: http://www.jroller.com/page/slava/20050424#dotquote_a_forth_news_website
16:10:13 <arke> slava: Thank you, I will add you to the "Forth People" Section :)
16:11:14 <slava> i think you could add fts/forth to 'forth systems'
16:11:22 <slava> since you list kc5 in forth people :)
16:11:31 <arke> I will later :)
16:11:38 <slava> have you done any forth coding lately?
16:11:48 <arke> Yes
16:11:59 <arke> if you read the article, you'll see that I wrote the new RetroForth bootsector
16:12:08 <arke> and I'm currently working on an editor for RetroForth
16:12:15 <slava> yes, i already read about the boot sector on crc's weblog
16:12:15 <arke> As well as editor extensions
16:12:19 <arke> :)
16:12:21 <slava> based on crc's editor?
16:12:29 <arke> Inspired, rather
16:12:33 <arke> it wont be much different
16:12:34 <slava> block editor?
16:12:36 <arke> Yep
16:12:45 <slava> i tried herkforth for the first time last night, i really liked the editor
16:12:49 <arke> Just with a few different commands, and a few more cmmands, and a history
16:12:55 <arke> Oh yeah, the herkforth editor is nice
16:13:13 <arke> However, crc doesn't want to lose the ability to do uppercase
16:13:14 <arke> :)
16:13:33 <CXDoctor> alot of herkforth doesnt work on macosx
16:13:46 <CXDoctor> I can actually type a capital letter
16:13:53 <slava> arke, i think its due to terminal limitations, mostly. vt100 cannot distinglish as many Control- keys as Shift-
16:14:04 <arke> slava: exactly :)
16:14:28 <slava> i have a passionate hatred for vt100 and especially libncurses
16:14:32 <arke> slava: Anyway, thanks for the publicity
16:14:39 <arke> Oh yes, vt100 is HORRID
16:14:52 * arke has written some simple terminal code using escape sequences .... *shudder*
16:15:05 <arke> Especially the fact that you have to reset the terminal to turn off bold, etc.
16:15:24 <arke> slava: you're added :)
16:15:34 <arke> And, I'm going to add HerkForth in a moment
16:15:59 * slava notices that dotquote.org and the retro news site look identical :)
16:16:16 <arke> I wonder why ..... ;)
16:16:20 <arke> crc hooked me up with dotquote.org
16:16:36 <arke> I'll change the layout eventually
16:17:41 <tathi> arke: btw, dotquote.org has the 25th on that rf8 item -- I thought it was still the 24th everywhere...?
16:17:59 <slava> tathi, this is futuristic shit man
16:18:00 <arke> tathi: blame the server :)
16:18:26 <slava> what services does L4 provide?
16:18:31 <arke> Lots
16:18:35 <arke> IPC, multitasking, paging
16:18:36 * slava reads
16:18:45 <arke> But in a simplistic way
16:18:50 <arke> so you can easily extend
16:19:05 <arke> kc5tja gets a stiffy from it, so it must be good :P :P :P :P :P
16:19:07 <slava> what is the API like?
16:19:25 <arke> No idea
16:19:29 <arke> ^_^
16:20:12 <tathi> arke: ok, just thought you might have some control over that :)
16:21:00 <arke> tathi: I could probqably figure out a way but I'm too lazy :)
16:21:47 * arke hasnt really gotten responses from c.l.f
16:26:22 <slava> arke, why don't you like Joy? its really small and about as simple as a forth/lisp can get
16:27:00 <arke> I don't know
16:27:20 <arke> I just don't really like either of the forth+lisp languages I've tried
16:27:25 <arke> Although I like factor better than Joy
16:27:27 <arke> :P#
16:27:40 <arke> Just personal preference. Don't take it to offense. :)
16:27:48 <slava> oh, i'm not offended at all, just curious
17:00:26 <Herkamire> CXDoctor: You can't use the herkforth command keys? or are you trying to type capitals as part of words?
17:00:46 <CXDoctor> I looked at the page
17:00:52 <CXDoctor> and it looks like you just hold shift
17:01:03 <CXDoctor> right? I do so, and it jsut types the capital letters
17:01:06 <CXDoctor> like H ,etc
17:01:23 --- join: saon (1000@c-66-177-224-130.hsd1.fl.comcast.net) joined #forth
17:09:28 <Herkamire> are you useng ./run?
17:09:58 <CXDoctor> ooh no
17:10:00 <CXDoctor> lol
17:10:07 <CXDoctor> I didnt know that mattered, ignorant me
17:10:22 <CXDoctor> I did cat your tanks file, I just didnt think run was that big a thing
17:10:24 * CXDoctor tries it
17:11:19 <CXDoctor> whoa this is hot!
17:11:19 <CXDoctor> lol
17:11:28 <Herkamire> :)
17:11:59 <Herkamire> the script fixes the terminal settings
17:13:38 <Herkamire> herkforth should be able to do it, but something must be different on osx
17:14:10 <CXDoctor> I think everything works now
17:17:33 <CXDoctor> how different is herkforth?
17:17:54 <CXDoctor> If I write a program that works on gforth, and doesnt use any gforht specific stuff
17:18:09 <CXDoctor> generally almost ans compliant
17:18:25 <CXDoctor> will it work on herk? or is herk not trying to be that normal of a forth
17:18:46 <tathi> nah, herkforth is an avant garde colorforth thingy :P
17:19:05 <Herkamire> not trying to be normal
17:19:30 <Herkamire> if there's no parsing stuff, it's pretty easy to port, but you can't just paste stuff in
17:19:50 <Herkamire> tathi: to get it to work on osx I put a noop after each sc instruction
17:19:57 <tathi> huh
17:20:01 <Herkamire> tathi: can you think of any reason that would help?
17:20:42 <Herkamire> and the stupid file size has to be a multiple of 4KB
17:20:44 <tathi> not off the top of my head...but I'm not all that familiar with the exception mechanism.
17:22:08 <tathi> also OS X is BSD based -- maybe their syscall convention is different somehow. (wild speculation here)
17:25:56 <Herkamire> maybe
17:26:01 <Herkamire> it works with the noop though
17:26:14 <Herkamire> (ioctls don't but presumably they just have different constants)
17:26:27 <Herkamire> read, write, open, close, exit work
17:27:48 <tathi> hmm...found the Mach-O ABI document, I think.
17:27:52 <tathi> I'll take a quick look through it.
17:48:33 --- join: aum (~aum@60-234-138-239.bitstream.orcon.net.nz) joined #forth
17:49:03 --- part: aum left #forth
17:49:08 <arke> Any ideas on what I should write about for the next DotQuote story?
17:49:25 <arke> Only suggestion I have so far is RetroForth vs. Reva
17:52:14 <slava> erotic love story involving Herkamire and tathi
17:52:55 <tathi> aha!
17:53:19 <arke> Eww incest
17:53:20 <arke> :P
17:53:22 <tathi> sounds like maybe BSD syscalls return as normal if the syscall *failed*
17:53:34 <tathi> but if it *succeeded*, then it skips that instruction.
17:53:55 <arke> Interesting
17:56:18 <arke> Come on, nobody has any ideas? :)
17:57:21 <tathi> no, sorry
17:57:31 <arke> Hrm :(
18:13:58 <crc> arke: an article looking at thinfu and docl's forthmud work?
18:14:58 <arke> crc: added to list of suggestions. :)
18:18:09 <Herkamire> tathi: oh :) that's interesting
18:18:22 <Herkamire> so you can compile a syscall, then a branch to an error handler
18:20:21 <arke> Herkamire: http://www.dotquote.org
18:20:54 <tathi> Herkamire: I haven't found any hard docs on this, just a vague reference.
18:21:29 <tathi> I'm starting to think the BSD kernel is a little bit TOO modular; it's hard to find the place where the actual work happens
18:24:06 <arke> Heh
18:25:24 <tathi> hmm...or actually, maybe that part isn't in the source that I have.
18:25:31 <tathi> that would explain it :)
18:25:43 --- part: slava left #forth
18:25:52 <arke> :)
18:26:02 <CXDoctor> oh wow
18:26:13 <CXDoctor> normal people like me can post on dotquote?
18:28:26 <arke> You can submit a draft
18:28:28 <arke> but not post
18:28:32 <arke> Unless I screwed up
18:30:21 <arke> yeah
18:30:25 <arke> you can only submit drafts
18:30:36 <arke> I look at the drafts and modify and publish them
18:30:42 <arke> well, me or crc or kc5tja
18:30:52 <arke> But feel free to write a draft
18:33:27 --- quit: saon (Read error: 145 (Connection timed out))
18:48:15 --- quit: tathi ("leaving")
18:49:05 <CXDoctor> ok arke  I might have something one day
18:49:06 <CXDoctor> lol
18:49:33 <CXDoctor> and I must admit , if I do send a draft, it's purely for publicity. but I dont want that right now
18:49:40 <arke> Heh
19:04:17 --- join: Al2O3 (~Al2O3@c-24-1-126-202.hsd1.tx.comcast.net) joined #forth
19:34:14 --- join: sk1p (alex@p548D77FA.dip.t-dialin.net) joined #forth
19:37:55 --- quit: sk1p_ (Read error: 60 (Operation timed out))
20:39:34 --- quit: CXDoctor ("and he jumps into the sky like Neo")
21:35:39 --- quit: Al2O3 ("Leaving")
22:33:56 --- join: mark4_ (~mark4@rrcs-24-242-160-169.sw.biz.rr.com) joined #forth
22:34:09 --- nick: mark4_ -> I440r_
22:36:54 <arke> http://www.dotquote.org
22:40:26 --- quit: Sonarman ("leaving")
22:44:21 <I440r_> lol ."
22:44:23 <I440r_> thats too cute :P
22:44:44 <I440r_> ." instead of /. haha
22:44:54 <arke> :)
22:44:58 <arke> Go read
22:44:59 <arke> :)
22:45:09 <arke> And, I forgot to add you as a Forth person
22:45:19 <arke> if you give me your personal (not isforth) website, I will.
22:45:38 <I440r_> i dont have a personal site
22:45:41 <I440r_> yet
22:46:26 <I440r_> ill contact some of my friends from the contract i was on before, im sure some of them would love to contribute if they have the time
22:47:02 <I440r_> ima bookmark it
22:47:38 <I440r_> im almost ready to release next isforth
22:48:10 <I440r_> i did some major surgery on the memory manager (due to a change with list code) but now i gotta found out what in there broke the windowing code :P
22:48:34 <arke> Hrm
22:48:39 <arke> You know what
22:48:54 <arke> isForth is quite large
22:49:00 <I440r_> not really
22:49:04 <arke> :)
22:49:05 <I440r_> 99% of it is optional
22:49:14 <arke> True.
22:49:15 <I440r_> plus when im assembling ill be able to make the kernel smaller
22:49:23 <arke> True.
22:49:38 <I440r_> so im not really all that worried about it 
22:49:42 <arke> You should, for fun, make a second forth which is subroutine threaded.
22:49:52 <I440r_> lol the last version of fpc was larger wasnt it ?
22:50:00 <arke> Possible :)
22:50:14 <arke> Well, ignore me
22:50:14 <I440r_> and if you port isforth 100% to dos (except you cant) it would be half th esize
22:50:21 <arke> I just forgot that you can completely strip it :)
22:50:30 <I440r_> so that means fpc was WAY larger lol
22:50:33 <arke> lol
22:50:44 <I440r_> actually stripping headers doesnt cut it down enough for my liking
22:50:49 <arke> Id like to see you make a subroutine threaded foth
22:50:52 <arke> forth*
22:50:56 <arke> that is source compatible with isforth
22:51:02 <I440r_> but the plan that will make turnkeys even smaller will have to wait for metacompiling and the assembler
22:51:06 <I440r_> no
22:51:09 <I440r_> i hate sub threading
22:51:11 <I440r_> its not forth
22:51:15 <I440r_> :)
22:51:18 <arke> It is forth
22:51:19 <arke> :)
22:51:27 <arke> But seriously, you should try it.
22:51:27 <I440r_> lol
22:51:30 <arke> The reason I say this
22:51:42 <arke> is because you have this unfounded fear/dislike of it
22:51:48 <I440r_> it doesnt construct as easilly in my mind. it doesnt decompile at all
22:52:00 <arke> so if you write a small subroutine threaded forth you might grow to like it.
22:52:01 <I440r_> i will never like anything except indirect and direct threading
22:52:16 <arke> You can't decompile?
22:52:23 <I440r_> i would in a pinch on a very small embedded system with severe limitations consider using a token threaded forth however
22:52:23 <arke> I can prove you wrong
22:52:26 <arke> :)
22:52:30 <I440r_> you cannot SEE accuratly
22:52:46 <I440r_> all you can do is disassemble
22:52:55 <arke> Not true once again
22:52:55 <I440r_> but that doesnt show you the forth code that generated that assembler
22:52:56 <arke> :)
22:53:19 <arke> Theres a little more complexity, but not much.
22:53:36 <I440r_> well that complexity is also part of why i dont like sub threading
22:53:41 <arke> And also, merely SEE cant be the only reason why subroutine threading should be dismissed :)
22:53:47 <I440r_> it adds nothing to the equasion except more complexity
22:53:52 <arke> thats the only time there is complexity
22:53:56 <arke> at any other time its simpler
22:54:00 <I440r_> no
22:54:03 <I440r_> i dont think so
22:54:21 <arke> Give me an example then
22:54:21 <arke> :)
22:54:58 <I440r_> Dennis W. Bulgrien - Apr 21, 6:15 am   <-- from CLF
22:55:02 <I440r_> i know him hehe
22:55:09 <I440r_> his brother has been in here :P
22:56:05 <arke> Come on, give me an example
22:56:06 <arke> :)
22:56:18 <arke> Becuase I can't think of anything except SEE
22:56:33 <arke> and the call, which adds $E8 c,
22:56:59 <I440r_> isforth has ,call
22:57:04 <I440r_> which commas a call
22:57:16 <I440r_> just the same
22:57:45 <arke> there you go
22:57:51 <arke> thats all that is added
22:58:07 <arke> And its faster
22:58:09 <I440r_> all words even non coded definitions START with assembler code
22:58:19 <I440r_> a colon definitions CFA is always "call nest"
22:58:22 <arke> say what you want, its quite faster :)
22:58:28 <arke> Oh, I know that.
22:58:33 <arke> call _quit
22:58:35 <arke> or nest
22:58:35 <I440r_> but that call is never returned 
22:58:40 <arke> or whatever people wish to call it
22:58:43 <arke> err docol
22:58:47 <arke> Yep
22:58:53 <arke> because the return address is used
22:58:55 <I440r_> th return address is popped off the stack into IP after pusing IP to the return stack
22:59:00 <arke> as the IP
22:59:03 <arke> exactly
22:59:03 <I440r_> there is a faster way but its more obscure
22:59:19 <I440r_> fpc uses it and i avoided using it to save confusion factor
22:59:25 <I440r_> fpc uses jmp nest
22:59:53 <arke> Sure.
22:59:54 <I440r_> because the previous "next" which is "lodsw jmp ax" has ax pointing at the "jmp nest"
23:00:02 <arke> But aren't both very nice and complex?
23:00:12 <arke> :)
23:00:18 <I440r_> so nest adds 3 (a tobody) to ax and puts that value in IP
23:00:47 <I440r_> i think the call mechanism is more intuative that having to think backwards to the NEXT that jumped to the word your trying to figure out
23:01:01 <arke> I think you dont even have to do that
23:01:02 <arke> :)
23:01:04 <I440r_> i can forget that eax points to the call nest 
23:01:15 <I440r_> do youunderstand what i mean ?
23:01:20 <I440r_> lets say im a geginner
23:01:21 <arke> becuase if you can screw nest and just insert $E8 before each cell
23:01:22 <arke> :)
23:01:24 <I440r_> im single stepping
23:01:26 <I440r_> i step a next
23:01:30 <arke> Oh, yeah
23:01:34 <I440r_> that takes me to a jump nest
23:01:34 <arke> I don't disagree with that.
23:01:39 <arke> call is definetely better than jmp
23:01:44 <I440r_> i have to remember that eax holds the address of that jmp
23:01:55 <I440r_> my way i dont have to remember. its simpler for a beginner to grasp
23:01:59 <I440r_> tho less efficient
23:01:59 <arke> but call dup is better than call nest; dd dup
23:02:04 <arke> :)
23:02:13 <I440r_> u miss the point
23:02:18 <arke> I get the point
23:02:25 <arke> I know what you were talking about
23:02:27 <arke> and I agree
23:02:34 <I440r_> my call/jmp has NOTHING to do with subroutine threading
23:02:41 <arke> I know
23:02:46 <arke> I know
23:02:47 <arke> I know
23:02:47 <arke> I know
23:02:48 <arke> lol
23:02:56 <arke> I'm not missing the point
23:02:57 <arke> lol
23:03:28 <arke> call nest is better than jmp nest
23:03:33 <arke> because it is much more reliable
23:03:44 <I440r_> has nothing to do w8th reliability
23:03:45 <arke> because with jmp nest you have to depend on (e)ax
23:03:47 <I440r_> its readability
23:03:56 <arke> and something might screw with it.
23:03:57 <I440r_> eax wont have been touched
23:03:58 <arke> With both, really
23:04:05 <arke> It could have been
23:04:06 <I440r_> but you have to mentally keep track of it which is harder
23:04:07 <arke> err
23:04:10 <arke> no it couldnt have
23:04:12 <arke> :)
23:04:25 <arke> Ok, you were right
23:04:27 <arke> :)
23:04:52 <arke> Now, do tell me what makes call nest; dd dup better than call dup though
23:04:55 <arke> or even inline
23:05:04 <arke> it can't just be SEE - its not that important 
23:06:14 <arke> hrm
23:06:25 <arke> is email a better medium for you?
23:07:00 <I440r_> you need 5 bytes per token for a start
23:07:07 <arke> So?
23:07:16 <arke> Memory is cheap these days.
23:07:23 <arke> And its not that big of a difference.
23:07:24 <I440r_> your using the system stack for return stack - this makes the parameter stack a software stack
23:07:28 <I440r_> this makes it slower
23:07:41 <arke> Most words average something like 4 calls, so it evens out with the laqck of call nest
23:07:44 <I440r_> it also makes it more complex
23:08:02 <arke> A bit more complex, but not necessarily slower.
23:08:15 <arke> Depends on the register you use.
23:08:26 <I440r_> any time you add complexity your destroying the PRIME thing that forth keeps beautyful
23:08:28 <I440r_> simplicity
23:08:42 <I440r_> the best solution to ANY problem is always the very simplest solution to it
23:08:58 <I440r_> and subroutine threading just doesnt cut it for me, i cant stand it at all
23:09:11 <arke> And the simplest solution is not making an extra interpreter
23:09:21 <arke> the CPU is already a perfectly good interpreter
23:09:23 <arke> why not use it?
23:09:28 <arke> its faster too.
23:11:04 <I440r_> not by much
23:11:35 <arke> quite a but
23:11:39 <arke> bit*
23:11:50 <arke> call ... ret
23:11:53 <arke> versus
23:12:21 <arke> lodsd; jmp ax; call nest; ...; call semicolon
23:12:22 <I440r_> no
23:12:39 <I440r_> you may have made THAT faster but you made parameter juggling slower
23:12:53 <I440r_> and you do more of that than you do nest and unnest
23:12:56 <arke> By how much, on average?
23:13:03 <arke> like 0.1 cycles or something
23:13:12 <arke> you gotta remember how much Intel CPUs pipeline
23:13:19 <I440r_> i would think by more
23:13:33 <arke> not really
23:13:38 <I440r_> only if you do obfuscated crap like preventing agi
23:13:41 <arke> drop is lodsd, versus pop ebx
23:13:43 <I440r_> thus making yor code even harder to read
23:13:48 <arke> heh
23:13:55 <arke> lodsd, pop ebx
23:13:59 <arke> which is faster?
23:14:03 <arke> thats drop
23:14:07 <I440r_> coding efficiently on an intel processor is impossible
23:14:13 <I440r_> TOTALLY impossible
23:14:15 <arke> true
23:14:24 <I440r_> because the instant you learn how do to it they change the rules
23:14:25 <arke> but its possible to do it somewhat
23:14:31 <I440r_> isforth does
23:14:36 <arke> well, the basics of pipelining never change
23:14:44 <I440r_> and i do so without giving a damn about it :P
23:14:54 <arke> so as long as you dont go on a tangent itll be somehwat optimized on every processor since 486
23:14:58 <arke> :)
23:18:15 <I440r_> only if you restrict it to intel
23:18:26 <I440r_> if you run it on an amd your screwed for optimizing
23:18:29 <I440r_> and vice versa
23:18:56 <arke> nope
23:19:07 <arke> im talking about BASIC pipelining
23:19:12 <arke> so basically, dont use a register twice
23:19:35 <arke> that is the same on every intel or amd or whatever since 486 and any RISC
23:20:05 --- quit: Herkamire ("off to bed")
23:23:14 <arke> I440r_: send me an email about this
23:23:22 <arke> chris dot r dot walton at gmail dot com
23:23:26 <arke> and ill get back to you
23:23:34 <arke> I'll convince you eventually
23:23:35 <arke> :)
23:24:11 <I440r_> err can you remind me about that later ?? heh
23:24:28 <I440r_> im kinda in tidy up mode and i gotta go zzz. plus i got a date tomorrow :P
23:24:30 <I440r_> PANIC!
23:25:56 <arke> hehe
23:26:04 <arke> im about to sleep
23:26:08 <arke> so, uuh
23:26:15 <arke> ill send you a msg about it
23:26:20 <arke> and you dont look at it
23:26:27 <arke> and you look later
23:26:32 <arke> and see it
23:26:36 <arke> works?
23:26:37 <arke> :P
23:29:55 <I440r_> :)
23:59:59 --- log: ended forth/05.04.24