00:00:00 --- log: started forth/03.04.12 00:41:38 --- quit: Soap` (Read error: 104 (Connection reset by peer)) 03:19:21 --- quit: Fractal (sterling.freenode.net irc.freenode.net) 03:19:21 --- quit: TreyB (sterling.freenode.net irc.freenode.net) 03:20:36 --- join: TreyB (~trey@cpe-66-87-192-27.tx.sprintbbd.net) joined #forth 03:20:36 --- join: Fractal (vyfgdkf@dont.try.configuring.openbsd.on.stronglsd.com) joined #forth 03:43:03 --- join: mur (ammu@baana-62-165-184-187.phnet.fi) joined #forth 06:16:48 --- join: PoppaVic (PoppaVic@tnt02-67-192.sfld.provide.net) joined #forth 06:53:51 --- quit: PoppaVic () 08:22:52 --- join: deluxe_ (~deluxe@pD950FE42.dip.t-dialin.net) joined #forth 08:27:46 --- nick: deluxe_ -> deluxe 09:19:39 --- join: Serg_Penguin (~z@ppp157-134.dialup.mtu-net.ru) joined #forth 09:22:16 --- quit: Serg_Penguin (Client Quit) 09:26:26 --- join: futhin (futhin@dial-187.ocis.net) joined #forth 09:55:01 --- join: divgrad7 (~wer@81.25.34.174) joined #forth 09:55:12 hey divgrad7 09:55:18 how does it go? 09:55:27 hi 09:55:36 goes as usual 09:55:44 how goes your forth coding? 09:56:11 im still accumulating courage for that 09:56:54 it takes years! :P 09:56:58 i tried to start coding a pair of time 09:57:15 but it made me sick 09:58:03 even coding elementary things brings an awful headache 09:58:03 coding a what? 09:58:10 show me your code 09:58:21 or tell me what you were trying to accomplish 09:58:33 i tried to code a GUI library for swuftforth 09:58:39 ew 09:58:39 swift 10:01:20 so as u see, you guys, masters of forth, real forthers, and me, hopeless coder - better show ur masterpieces 10:01:23 what kind of GUI ? 2d ? 10:01:45 yea, simple Windows/Control 10:01:54 i liked postscript, i've been thinking of coding a similar vocabulary in forth 10:02:10 all the words like MOVETO LINETO RLINETO GSAVE GRESTORE etc 10:02:29 Window 200 300 winPos Z" Hello" winTxt value W1 \ etc etc 10:02:40 er 10:02:49 oops, GUI != graphics 10:02:50 heh 10:02:56 nevermind about the postscript :P 10:05:11 * divgrad7 sighs 10:08:03 divgrad7: maybe code something more simple.. or code something immediately useful? 10:08:24 don't go around creating libraries :/ 10:08:46 the prob is not in GUI, but in the forth itself 10:08:55 no local vars 10:09:16 albeit i think i know how to implement em 10:09:49 aaaaaaaaaaaaa 10:09:56 forget it 10:12:06 --- part: divgrad7 left #forth 10:19:40 --- join: kc5tja (~kc5tja@ip68-8-206-137.sd.sd.cox.net) joined #forth 10:22:42 hey hey hey! :D 10:22:48 how's it going kc5tja :) 10:22:58 I'm in pain, but otherwise OK. 10:23:42 Lots of breakfalls yesterday in aikido. LOTS. 10:24:04 heh, don't you guys just practice punching instead of falling? :P 10:24:10 how long is a session? 10:24:17 An hour to an hour and a half. 10:24:29 Sure we practice punching. 10:24:35 He who punches ends up on the floor one way or another. 10:24:45 That's the point of aikido. 10:25:09 how often do you go to aikido? 10:25:14 Three times a week. 10:25:24 mwf ? 10:25:31 Yup 10:25:47 I could go as much as five times a week (MTWFS) 10:26:13 But three is enough, and that's what I built my schedule around. The Tuesday class is for beginners, and the Saturday class is just plain too early. 10:26:25 * kc5tja isn't even awake until .. oh ... about now actually. :) 10:27:41 heh 10:27:41 yeah 10:28:05 sounds better though. the gung fu classes are only on tuesdays and thursdays 10:28:08 not frequent enough :/ 10:28:13 * kc5tja nods 10:28:14 MWF is better 10:28:43 I don't know if I told you or not, but I got FS/Forth parsing words now. :) 10:28:55 I need to write a string compare routine and the interpretter next. 10:29:10 Then it'll at least be able to interpret text input on the command line. 10:29:16 are you coding in forth or is it still in asm? 10:29:30 FS/Forth is written in FS/Forth using a target compiler written in Pygmy Forth. :D 10:29:44 So it's not the perfect environment, but it's got a lot of potential. 10:29:59 The target compiler is very complete for what it is; I can use it to write any arbitrary software I want for DOS right now. 10:30:17 FS/Forth itself, however, is still very incomplete. 10:31:06 Currently for DOS, but it'll eventually appear for Linux and Win32 as well. 10:31:10 And yes, I am aware of isForth. :) 10:32:03 In fact, the native, Linux, and Win32 versions will be supported, but the DOS version won't be. After 1.3 is out, that's it -- the last DOS version ever to be released. 10:32:36 I will also write a version for the 65816 processor as well, so I can have a nice software development environment for my 65816-based computer that I want to build. 10:32:49 how is fs/forth written in fs/forth if fs/forth doesn't exist yet? 10:33:01 I bootstrap it in Pygmy Forth using the target compiler. 10:33:35 That is, the target compiler consists of words that exist in FS/Forth, but they're written in PygmyForth. 10:34:25 okay 10:34:39 did you write an assembler in pygmyforth ? 10:34:52 Yes. 10:34:52 or have you managed to avoid writing words in CODE ? 10:35:01 I don't have CODE. :) 10:35:18 My words are compiled subroutine threaded, which makes it trivial to inline raw assembly. So, a "code" word looks like this in FS/Forth: 10:35:44 : + ( a b -- a+b ) [ bp sp xchg, bx pop, bp sp xchg, bx ax add, ] ; 10:36:51 hm 10:37:00 Note that I go into interpret mode with [ to "immediately" assemble the instructions I need. 10:37:03 why does it have to be subroutine threaded to be able to do that? 10:37:27 Because direct threading and indirect threading store definitions in a form that is not directly CPU executable. 10:37:42 They consist of lists of addresses, not lists of actual instructions. 10:37:59 But remember, I can do things like this too: 10:39:04 crap. 10:39:07 I forgot my example. 10:39:09 I just had it. 10:39:25 But I can freely intermix assembly and Forth code at will. 10:39:34 so subroutine threaded = native threaded ? 10:39:45 subroutine threaded is when the following definition: 10:39:49 : FOO A B C D ; 10:39:54 expands to the following: 10:39:57 FOO: 10:39:59 CALL A 10:40:00 CALL B 10:40:02 CALL C 10:40:04 CALL D 10:40:05 RET 10:40:07 in assembly. 10:40:09 yes 10:40:23 Note that it's a sequence of instructions that can be *directly* executed by the CPU. 10:40:29 but when you do the [ asm code ] it doesn't call ? 10:40:32 No. 10:40:33 FOO: 10:40:37 ASM CODE 10:40:38 RET 10:40:40 : FOO A B [ ax inc, ] C D ; 10:40:43 expands to: 10:40:44 FOO: 10:40:46 CALL A 10:40:47 CALL B 10:40:52 INC AX 10:40:54 CALL C 10:40:56 CALL D 10:40:57 RET 10:42:16 Remember that [ and ] switch into and out of interpret mode while compiling. 10:42:17 yeah 10:42:36 --- join: divgrad7 (~wer@81.25.34.174) joined #forth 10:42:48 [ ] switch into compile mode when : FOO A B [ blah ] C D ; is being interpreted 10:42:51 By using subroutine threading, I can eliminate any special distinction between CODE and non-CODE words, and just interject raw assembly where appropriate. 10:43:02 Nope. 10:43:06 [ switches into interpret mode. 10:43:09 ] switches into compile mode. 10:43:19 yeah yeah yeah 10:43:21 mixed it up 10:43:30 Try this: : foo [ ." Hello world!" ] exit ; next time. :) 10:43:33 you can do the same thing with native forth right ? 10:43:49 Native forth is subroutine threaded forth with compiler optimizations. 10:43:59 I didn't bother with compiler optimizations yet. 10:44:05 I will, just not right now. 10:44:11 And especially NOT in 16-bit x86. 10:44:18 The instruction set just isn't regular enough. 10:44:21 ummm 10:44:23 i meant 10:44:28 : FOO A B C D ; 10:44:31 FOO: 10:44:34 A 10:44:35 B 10:44:38 C 10:44:39 D 10:44:40 ret 10:44:44 heh, no ret 10:44:46 jmp 10:44:54 Well, that's illegal. 10:45:00 no, that's native forth 10:45:01 Either you have a list of addresses, or you don't. 10:45:09 uh.. 10:45:21 ret and jmp are CPU instructions, right? 10:45:28 where A B C and D are addresses? 10:45:31 i think there's a kind of forth that compiles the code and stores the actual code in the CODE FIELD 10:45:34 rather than making any calls 10:45:42 no no no A B C D are not addresses 10:45:51 the actual code is compiled into the word 10:45:55 Yes. 10:45:56 I know. 10:46:00 Quartus Forth is an example. 10:46:01 FOO: 10:46:05 (code of A) 10:46:08 (code of B) 10:46:09 etc 10:46:21 But when you have: : FOO a b c d ; in such a Forth, the "code" will just be a set of CALL instructions. 10:46:43 Unless the compiler is smart enough to realize that a, b, c, and d are used just once or twice, and then will inline them automatically (I think SwiftForth does this). 10:47:02 Native-code Forths start their lives as subroutine threaded Forths. Always. 10:47:10 Because subroutine threading *IS* a form of native code generation. 10:47:24 Anything above that is considered an optimization. 10:47:45 For example, a lot of compilers can realize that 1 + is the same as "INC". 10:48:05 or that 4 * is the same as SHL r,2 or whatever (where r is some kind of register or stack location). 10:48:18 My compiler isn't that smart ... YET. 10:48:30 I am aiming specifically for correctness right now, not speed. 10:49:24 What gets compiled "inline" usually are primitives like +, -, *, /, etc. Conditionals and looping constructs of course (even in mine they do) get compiled inline as well. I also inline literals. 10:49:27 But that's it. 10:51:10 Still there? :) 10:52:34 yeah sorta 10:52:40 i'm multitasking a little bit sorry 10:52:47 heh 10:52:49 i started a bit of a flamewar on #osdev about forth 10:52:51 it died out 10:52:53 thankfully 10:53:01 and now it's some stupid argument about community and perfection 10:53:13 * futhin is a troll ;) 10:53:26 Well, when I write my native-hardware Forth implementation, I can assure you that many elements from Dolphin will make it in there, including the GUI concepts that I have in mind. 10:53:46 there CAN be a forth that inlines all code instead of making calls, that's what i call "native forth" :P 10:53:47 And preemptive multitasking, and ... 10:53:57 There is no such thing. 10:54:02 There can be no such thing. 10:54:07 heh 10:54:09 okay 10:54:14 what do you think about 1:1 compiling? 10:54:18 You'll run out of memory instantly as soon as you compile the definition for QUIT. :) 10:54:21 doesn't an optimizing compiler affect that? 10:54:27 heh 10:54:38 Well, define 1:1 compiling? 10:55:05 I want to make sure that your meaning and my interpretation are the same. 10:55:24 um.. i sorta forget .. it's where you can decompile and produces the same source i think 10:55:29 OK 10:55:34 or something else maybe? 10:55:37 i440r rants about it :P 10:55:38 : back ." re" ; back 10:55:56 OK, I know what you're talking about. 10:56:05 And I agree with I440r that 1:1 is a GOOD THING. 10:56:18 But I don't follow that precept to the exclusion of all other philosophies. 10:56:38 As I indicated earlier, I'm specifically targeting correctness for my FS/Forth compiler environment. That means, subroutine threading *ONLY* for the time being. 10:56:43 The code it produces is 1:1 with the source. 10:56:47 HOWEVER, ... 10:56:55 this is known to produce abhorently slow code. 10:57:09 At least in comparison with code that is hand-written in assembly language. 10:57:34 And especially in comparison with other compiled languages, like C, where they have a higher level view of the source code available to them. 10:57:58 if the coder is good, he can avoid doing things like "1 +" and instead use "1+" 10:58:03 One method Lisp uses to get around this is to take hints from the programmer. These hints are thoroughly ignored while interpreting the code directly, but they affect compilation. 10:58:13 Hold that thought. 10:59:12 Actually don't bother: the problem with using 1+ and such is that you have a HHUUGGEE proliferation of words that compile to definitions with only a few CPU instructions. The x86's subroutine call overhead (including that for RET instructions) is IMMENSE; quite often, it'll dwarf the execution time of the word's definition itself. 10:59:34 So, in FS/Forth, there are several ways around this today, but none are programmer friendly. 10:59:48 Instead of 1 + or 1+, you can write [ ax inc, ] and that'll increment the top of stack. 10:59:56 Ditto for 2* and 2/ and whatnot. 11:00:03 But this doesn't go far enough. 11:00:43 (Case in point: to implement the words < <= = => and >, I had to define a set of primitives: IF and >IF in the target compiler to produce code that was even half-way efficient) 11:01:11 maybe you could make 1+ immediate 11:01:19 Well, that's what you'd have to do. 11:01:21 so it acts like [ ax inc, ] 11:01:34 But that doesn't eliminate the *massive* proliferation of such words. 11:01:59 And it would be rather massive considering how rich the x86 instruction set is (especially in 32-bit mode). 11:02:14 yeah 11:02:19 less words is more 11:02:20 What optimizing compilers in Forth do is recognize patterns. 11:02:21 :D 11:02:46 i think it's important to separate words (not necessarily in forth, but in the mind of the coder) 11:02:59 And if it can detect a certain pattern of instructions that it knows it can replace with another sequence of instructions that fits a certain criteria (e.g., it's known to execute faster without taking much more space), then by all means, it should do this. 11:03:07 hey hey kc5tja 11:03:10 and to keep the core set of words small 11:03:34 The principle of concatenativity and the principle of referential transparency not only permit this, but even encourages it. These principles together combine to form another science which we're already familiar with: algebra. 11:03:39 re mur 11:04:07 So, if I were to add an optimized compiler to FS/Forth, I'd probably do it like this: 11:04:13 : FOO [Optimized] A B C D ; 11:04:41 The [Optimized] "flag" is really a word that subverts the "dumb" compiler, and proceeds to compile the definition as IT sees fit. 11:04:51 Without the [Optimized] flag, it'd be 1:1. 11:05:20 Again, the *user* is in full control. In order to optimize the definition, the user must consciously make the decision to write in the [Optimized] flag. 11:06:36 I think this is the best system available. 11:06:52 --- join: crc (12345678@AC937834.ipt.aol.com) joined #forth 11:06:54 coding the optimizing compiler must be quite tricky.. 11:06:54 The idea is to write the program without ever using [Optimized]. 11:06:58 Test it, profile it, etc. 11:06:58 hey crc 11:07:04 hello 11:07:13 Then for those words that look like they need speeding up, recompile them with the [Optimized] flag. 11:07:17 Test it, profile it, etc. 11:07:27 If any more speed up is necessary, then it's time to re-write in assembly. 11:07:28 crc: do you have a forth? 11:07:51 crc: i think i saw your nick on some site, but i forget 11:08:01 futhin: Tricky in that you need to remember a number of details, but it's actually relatively simple. At least in comparison to C's optimizers. :D 11:08:09 Yes, I develop RetroForth 11:08:14 --- quit: divgrad7 (Read error: 113 (No route to host)) 11:08:15 oh right yeah 11:08:32 crc: have you seen the cliki entry about "forth is NOT intrinsically slow" 11:08:33 --- join: kitsune (foxchip@adsl-209-182-168-45.value.net) joined #forth 11:08:46 Yes 11:08:52 recently? 11:08:55 i added an entry myself 11:09:07 hmm.... The iki implies a Wiki; what's the CL stand for? 11:09:14 Not recently 11:09:19 * kc5tja is thinking Common Lisp? 11:09:53 kc5tja: cliki.tunes.org . possibly common lisp i guess.. i'm curious what you think of cliki.tunes.org/Microkernel and cliki.tunes.org/Exokernel 11:10:06 I *LOVE* exokernels. 11:10:13 those entries seemed a little bit poor 11:10:16 I like microkernels, but exokernels have them beat soundly. 11:10:22 * kc5tja checks 11:10:39 maybe you could enter some entries if you don't like theirs ;) 11:10:50 the microkernel entry seemed to be wrong on some counts 11:12:20 --- quit: TreyB (Read error: 113 (No route to host)) 11:12:59 check out http://cliki.tunes.org/Forth%20is%20NOT%20intrinsically%20slow too. i'm particularly proud of my entry ;) 11:13:38 my 2nd entry anyways 11:20:37 you make some excellent points 11:21:19 thanks :) 11:27:04 kc5tja: hello there? what's up? :) 11:44:13 * kc5tja is writing a response. 11:45:04 yay :) 11:45:08 did you disagree with any of it? 11:45:15 Tons 11:45:32 yeah.. i thought so 11:45:54 and i'm sure you'll write an excellent write up :D 11:46:05 s/write up/response 11:46:10 Well, it'll take time. It's not really well written yet. 11:46:45 yeah took me an hour to write my response to the "forth is NOT intrinsically slow" one 11:46:52 i kept redoing it 11:47:05 even after i entered it 11:47:41 if you don't enter anything into the "summary of changes" it won't show up that you've been modifying it on the "recent changes" page 11:49:09 --- quit: deluxe (Read error: 54 (Connection reset by peer)) 11:49:28 --- join: deluxe (~deluxe@pD950FE42.dip.t-dialin.net) joined #forth 12:01:41 --- join: fridge (meldrum@zipperii.zip.com.au) joined #forth 12:05:20 That's fine. 12:05:23 I prefer that. 12:05:34 I'll edit the summary of changes when I have something that I'm happy with. 12:06:59 yep 12:08:08 what did you thhink about the microkernel entry.. it says a smaller kernel adds more overhead which seems a bit ridiculous.. 12:08:40 I just finished my response. Read it, let me know what you think. 12:09:04 Well, it sort of does, in a way, but if you stop thinking in the Unix-way of doing things, you can often circumvent such overheads. 12:09:12 yep reading now 12:09:20 In other words, if you take a Unix kernel and shrink it, it WILL be dominated by overheads. 12:09:27 And that's the basic problem with Mach. It's UNIX, but smaller. 12:09:34 NO! That's not how you write a microkernel! 12:10:43 Thats noot how you make porridge! 12:10:51 i recommend throwing in an

between your entry and the original 12:11:04 fridge: noot? is that an accent? 12:12:11 kc5tjas statement reminded me of an ad I saw when I was younger, some scottish lad saying "Thats noot how you make porridge" =) 12:12:42 heh :P 12:13:58 Where'd I wrote noot? 12:15:01 no, that's fridge 12:15:04 not you 12:15:07 ahh 12:15:10 --- join: deluxe_ (~deluxe@pD9E4E718.dip.t-dialin.net) joined #forth 12:15:17 Sorry. :) 12:15:26 I misread the name next to it. Same number of characters. 12:15:29 i like your entry :D 12:16:03 nicely concise, succint, gets the point across & demonstrates understanding of the subject :D 12:16:33 That's what I've been going to school for.. .:D 12:16:40 Now if only I can learn how to type... :D 12:20:04 seriously though, if you can write like that in the 30 minutes it took you, you should be writting articles for magazines and getting easy moola ;) 12:20:51 i might be wrong, but writing technical articles for magazines might be easy & good pay 12:22:11 You're wrong. 12:22:13 did you add anything to Exokernel ?? 12:22:19 It's both hard (DAMN hard) and you don't get paid very much. 12:22:21 Yes 12:22:22 i don't see anything different with Exokernel 12:22:27 Refresh. 12:22:35 I wrote that an exokernel is not a microkernel. 12:23:33 --- quit: deluxe (Read error: 60 (Operation timed out)) 12:24:10 i wonder who the author of the microkernel entry was 12:24:11 heh 12:24:29 did you see my entry with "forth is not intrinsically slower" ? 12:24:34 --- nick: deluxe_ -> deluxe 12:26:18 I don't know, but it's mostly an invalid argument. 12:26:25 The idea is right, but there's no supporting evidence. 12:26:30 Anyway, I gotta go -- getting some food. 12:28:22 my argument ? 12:28:24 or his argument? 12:31:01 --- quit: fridge ("http://lice.codehack.com") 12:34:13 i suppose i could've put more examples/supporting evidence for my arugment ?? 12:41:23 --- quit: kitsune () 12:48:06 --- join: flyfly (~marekb@ip164.ktvprerov.cz) joined #forth 12:57:30 --- quit: crc ("Leaving...") 13:06:25 --- join: bwb (~bwb@ip68-4-121-108.oc.oc.cox.net) joined #forth 13:06:46 kc5tja: hey... just got back.. seems my powersupply on my main computer died :( 13:08:02 --- quit: sifbot (Ping timeout: 14400 seconds) 13:08:42 kc5tja: and I need to go get something to eat and go drop off my sister some where.. So... I may be able to talk later... :/ 13:08:44 bbl 13:19:10 hi bwb 13:25:36 --- join: deluxe_ (~deluxe@pD9E4E718.dip.t-dialin.net) joined #forth 13:26:39 --- quit: deluxe (Killed (NickServ (Ghost: deluxe_!~deluxe@pD9E4E718.dip.t-dialin.net))) 13:26:43 bwb: do you do forth or turbines or what? :D 13:26:50 --- nick: deluxe_ -> deluxe 13:27:01 --- join: deluxe_ (~deluxe@pD9E4E718.dip.t-dialin.net) joined #forth 13:27:25 --- quit: deluxe_ (Client Quit) 13:29:13 --- join: ASau (~asau@158.250.48.197) joined #forth 13:29:36 hey asau :) 13:29:53 Good "late evening"! 13:30:00 :) 13:31:26 I've heard of a good proposal about how Gforth should be constructed in c.l.f. 13:31:58 constructed ? 13:32:04 tell me about it 13:32:05 Also there was Gforth EC referenced. WHere is it? Have you seen? 13:32:14 no idea 13:32:21 A proposal was approx this: 13:32:47 starting from debugger/assembler make small forth 13:33:01 based small version make more elaborate 13:33:25 at end make full-featured forth. 13:33:58 what assembler 13:34:02 gforth has an assembler? 13:35:04 Possible way is to define simple virtual machine and write a compiler for it 13:35:05 --- join: TheBlueWizard (TheBlueWiz@207.111.96.119) joined #forth 13:35:13 hiya all 13:35:27 hi TBW 13:35:32 THis VM should be very simple to code it in asm 13:35:44 Hello TBW! 13:35:55 hiya futhin 13:35:59 hiya ASau 13:36:06 asau: yes, this method for writing new forths is pretty simple and easy to understand, it's nothing new 13:36:18 terve TheBlueWizard 13:36:30 asau: i440r could've done that, but he doesn't like coding assembly in RPN 13:36:40 Of course, it is not new. 13:36:59 does gforth have an assembler? 13:37:02 But I have not seen it running. 13:37:49 kc5tja is writing FS/Forth in FS/Forth ontop of a target compiler in pygmy 13:38:13 terve ja hyvää päivää mur 13:38:21 Back 13:38:26 hiya kc5tja 13:38:31 heh, did it beep you ;) 13:38:48 s/;)/:) 13:39:02 * futhin seems to use emoticons randomly 13:39:44 * TheBlueWizard thinks random emoticons are a nice antidote against order ;) 13:39:48 more practise required for Master of Smileys! 13:39:49 gforth has several assemblers available for it. 13:39:53 However, it itself is written in C. 13:40:13 this is serious line ;) 13:40:18 and this happy :( 13:40:18 yeah, but it's possible to write a "pure" forth using an assembler in gforth 13:40:32 mur: heheh :( 13:40:44 hih :c 13:40:44 futhin: Absolutely. I intend on doing just that in writing FS/Forth for Linux. 13:40:51 it was not that funny :D 13:41:18 re TheBlueWizard 13:41:18 TheBlueWizard: Still back-reading.. :) 13:41:18 i know :*D 13:41:29 kc5tja: back-reading what? 13:41:37 quick, type lots so that kc5tja can't catch up! :D 13:41:37 I was away getting food. 13:41:47 back-reading the #forth talk 13:41:48 I'm catching up pretty fast. 13:41:52 oh, the conversations here :) 13:41:56 Yes. 13:42:17 kc5tja: were you talking about my argument being invalid? 13:42:18 still out of job? 13:42:36 futhin: No, the original author's argument. His point is valid, but offers no supporting evidence for it. 13:42:51 which point is valid ?? 13:42:55 I tried very, very hard not to attack the author in my message. I'm not sure how well I've succeeded. 13:42:56 * mur is planning to do some graphics now 13:43:01 That microkernels are inherently flawed. 13:43:01 i disagreed with almost everything 13:43:05 ohhh 13:43:06 From a performance perspective. 13:43:06 no 13:43:10 i was photoing a while ago. i'll maybe do next some art crafts :) 13:43:41 kc5tja: have you checked http://cliki.tunes.org/Forth%20is%20NOT%20intrinsically%20slow ? 13:43:42 I don't disagree with his point, but I must disagree with his assertion that *ALL* microkernels are bad. 13:43:54 futhin: Yes, I posted a small response. 13:43:59 Look for KC5TJA on the page. 13:44:01 what did you think of my entry? 13:44:13 I am not sure how it fit in with the original point. 13:44:28 Sorry, I'm back. 13:44:38 the 2nd entry 13:44:42 my 2nd entry 13:44:55 At the bottom of the page? 13:45:02 yes 13:45:14 you don't think it fit in with the original point ? 13:45:22 --- quit: deluxe (Read error: 104 (Connection reset by peer)) 13:45:55 * futhin will probably remove the 1st entry of his 13:46:02