IRC log started Sun Apr 11 00:00:00 1999 No matter if it is on the source or drain? <_QZ> the technical term is emitter, not drain <_QZ> sinks have drains :) <_QZ> and bathtubs [msg(TUNES)] permlog 1999.0411 MOSFETs aren't labled GSD? Gate source drain? <_QZ> yes And transistors are Collector Base Emmiter And tubes are filimnet, plate, grid, cathode? <_QZ> hmm, i tried to email this news article to myself but it only sent the first half?? Weird Try it again? <_QZ> ok now its getting weird huh? <_QZ> i copied and pasted it into a new email to myself <_QZ> same shit <_QZ> weird <_QZ> it had 3 stacked periods that it stopped at <_QZ> . <_QZ> . <_QZ> . <_QZ> it stopped right before them 12:10am <_QZ> KEWL <_QZ> send yerself an email with some text, 3 periods and some more text <_QZ> dunno is ns3.0 is just screwed up or if its an email thing Ok, lemme check Must be ns... the 3 peroids didn't do anything to my e-mail <_QZ> what do u have? Eudora light for win 95 stupid winmodem or i'd be in linux <_QZ> sorry :) :) u know tesla was probly the coolest scientest there was he did a lot <_QZ> no he didnt 12:20am ya AC, Radio, transformer <_QZ> he stole the idea for the 'tesla' coil That is important stuff <_QZ> he didnt discover ac, radio or the transformer <_QZ> he was a drunk Who did then? <_QZ> some guy u never heard of <_QZ> i have it all in a book if u dont believe me Dont say Marconi invented the radio <_QZ> let me get the isbn <_QZ> :) ok abi: lar1 is lar1@earthlink.net ...but lar1 is mailto:lar1@bigfoot.com... ok abi, your right, I won't argue with you... Ack, I need to go.. just send that ISBN to me or I'll get it from you later on IRC... whatever later -:- SignOff lar1: #TUNES (Leaving) 12:30am -:- SignOff _QZ: #TUNES (BRiX [http://www.qzx.com/brix] :: ) -:- SignOff Tril: #TUNES (Ping timeout for Tril[sloth.wcug.wwu.edu]) -:- Tril [dem@sloth.wcug.wwu.edu] has joined #TUNES -:- smkl [sami@CMXCII.rdyn.saunalahti.fi] has joined #tunes !NickServ:*! lilo used GETPASS on j0shua -:- SignOff smkl: #TUNES (dssasasasasdffdgffggh) -:- hcf [nef@me-portland-us321.javanet.com] has joined #tunes -:- binEng [bineng@dialup248-3-18.swipnet.se] has joined #tunes hoy binEng hoy yerself " don't you think abi can be frustrating sometimes?", frustrating how? hi que tal, binEng see? its a bug in u u should type, "hi " instead oh... any patch availible? also, abi responds to hello things only ~30% of the time so you defends her? i could easily disable it entirely if ppl wish it so 11:40am I think it can be quite disturbing sometimes 11:50am <-- leaving... -:- SignOff binEng: #TUNES (Leaving) 12:00pm -:- _QZ [brand@p0wer.qzx.com] has joined #tunes hoy _QZ, sup? 12:10pm -:- SignOff hcf: #TUNES (Leaving) Gakuk! lo 12:40pm -:- hcf [nef@me-portland-us245.javanet.com] has joined #tunes -:- SignOff hcf: #TUNES (Leaving) -:- hcf [nef@me-portland-us245.javanet.com] has joined #tunes -:- Connection closed from clarke.openprojects.net: Success -:- Connecting to port 6667 of server clarke.openprojects.net [refnum 1] -:- BitchX+Deb1an: For more information about BitchX type /about -:- Welcome to the Internet Relay Network TUNES -:- Your host is clarke.openprojects.net[blueberry.inwa.net], running version u2.10.04.resolv9.nmt.egcs4.tok.pten.tlim4.admin.upper4.whisper3.gipl.modeless7 -:- This server was cobbled together Mon Jan 25 1999 at 07 12:10 EST -:- clarke.openprojects.net u2.10.04.resolv9.nmt.egcs4.tok.pten.tlim4.admin.upper4.whisper3.gipl.modeless7 dioswkfcg biklmnopstv -:- [local users on irc(13)] 5% -:- [global users on irc(95)] 37% -:- [invisible users on irc(163)] 63% -:- [ircops on irc(13)] 5% -:- [total users on irc(258)] -:- [unknown connections(0)] -:- [total servers on irc(34)] (avg. 7 users per server) -:- [total channels created(66)] (avg. 3 users per channel) !clarke.openprojects.net Highest connection count: 27 (26 clients) -:- Mode change [+f] for user TUNES -:- Mode change [+iws] for user TUNES -:- JOIN activated by "TUNES #tunes tunes@bespin.cx " -:- TUNES [tunes@bespin.cx] has joined #tunes -:- Topic for #TUNES: Millenia Year Application Software System -- www.qzx.com/myass.txt -:- topic set by _QZ [Sat Apr 10 22:31:42 1999] -:- [Users(#tunes:6)] [ TUNES ] [ hcf ] [ _QZ ] [ Tril ] [ Fare ] [ timestamp ] -:- Channel #tunes was created at Sun Feb 28 08:48:06 1999 -:- BitchX+Deb1an: Join to #tunes was synced in 6.240 secs!! -:- Mode change [-s] for user TUNES -:- abi [abi@bespin.cx] has joined #tunes -:- veve [h@bay-039-b6.codetel.net.do] has joined #Tunes -:- veve [h@bay-039-b6.codetel.net.do] has left #Tunes [] -:- SignOff hcf: #TUNES (Ping timeout for hcf[me-portland-us245.javanet.com]) -:- tcn [tcn@cci-209150250103.clarityconnect.net] has joined #tunes <_QZ> hello tcn <_QZ> how goes retro? oh, I'm finishing the forth interpreter now. Took yesterday off to make some dough :) any progress on brix? <_QZ> ofcourse <_QZ> ugh, my fingers are cold and i cant type :) 04:50pm <_QZ> im almost finished redoing the object system and paging <_QZ> the new paging will require more memory space but will be alot faster gakuk! bonjour <_QZ> gakuk hum, did not have much time to work on tcn. I've been thinking on the virtual machine for retro-scheme and the problem is, it doesn't map simply the model for retro-forth that would mostly be an inefficiency problem, if it were not also a problem about memory management in presence of multithreading (and we DO want multithreading, don't we?) yeah so there is a problem with our having to define a memory model. <_QZ> whats the problem? i think the problem is ... (fill in problem here) haha My documented sources of inspiration are 1) Rscheme 2) LiSP LiSP is the book by C. Queinnec that everybody should have bought about LISP implementation. s/bought/read/ Do you think Linear LISP would help with your problem? (I don't know enough about LISP variants..) except that there is no such thing as a standardized linear LISP. It's mostly an internal compiler language in hbaker's compilers. 05:00pm also, there is the question of the invariants that the FORTH ought to respect or not: must it respect the GC invariants? if not, this means that any call to FORTH must be "atomic" with regard to LISP whoa.. too much CS jargon there :) don't be afraid of jargon. Jargon is here to help sentences be short and accurate. ok, what's an invariant? in that, it opposes buzzwords. I'm just a hacker, I didn't go to school for this stuff :) <_QZ> :) plus we've got the french-english language barrier. The french seem to use one big word where americans would use a couple short words. fare, are you typing a big explanation of 'invariant'? invariant: a property that always holds, or a value that doesn't change with time abi: an invariant is a property that always holds, or a value that doesn't change with time, in a given system abi: an invariant is also a well-established word in the CS community okay, Fare. abi: an invariant? an invariant is a property that always holds, or a value that doesn't change with time, in a given system or a well-established word in the CS community 05:10pm hum. Ok, if Forth respects the GC invariants, that means Forth stuff is garbage-collected? well, garbage-collectible, anyway nah, let's not get too fancy with Forth. also, FORTH may have laxer invariants than scheme wrt GC for instance, only be GC-ready at PAUSEs and ALLOTs Forth doesn't seem very GC-friendly no it doesn't. maybe that's good. still, if we're having a GC, this _will_ have an impact on FORTH programs. *Later on* we should work on unifying postfix/prefix/infix syntax, using the same underlying stack machine that's our HLL oh, the syntax is just irrelevant (except that postfix is mucho mucho better for interaction) infix is great for math, though (and that prefix is mucho mucho better to describe structure) infix is not THAT great for maths I want to write an infix parser that's almost as simple as a Forth/LISP parser infix is what most people are accustomed to it does have some readability advantage in some cases of medium complexity, but it is soon lost outside of this limited range we speak that way, as in "two plus two".. hmm, we use prefix in speech too, "add two and two" (although those cases are precisely those used for school exercises) 05:20pm tcn: there are lots of existing infix parsers for FORTH and LISP... the major impacts of GC on FORTH programs with be with 1) dictionary access and 2) multithreading yeah, but infix parsers always seem to be so *complicated* they *are* complicated, but given the proper formalism they are very manageable. What is much less manageable is the horrid _semantics_ and _ambiguities_ of languages such as C. tcn: btw, commit? I don't have anything new didn't work on Retro yesterday I'm doing the math stuff now <_QZ> tcn: u got a binary of retro that does anything? math stuff? qz: It has a forth interpreter, but don't bother looking at it till I flesh it out more. oh, btw, we will need varying-length integer arithmetics at some near time in the future... Fare: + - * / MOD /MOD 2* 1+ 1- etc... did you look at eForth? it has all that in Forth, from a fe (30 or so) primitives... no I'm looking at 16-bit retro-forth <_QZ> 2*? 2* is a left-shift 05:30pm <_QZ> why make a 2* when u can just make a left shift 2* is quicker to type and it's used alot is it logical or arithmetical left shift? ;-> oh, definitely logical jk <_QZ> so if u have a 2* do u also have a 2/? <_QZ> and what about 4* and 4/? <_QZ> and... :) if you want them, sure <_QZ> i would like upto 65536* hmm.. then you just type 65536 * or 16 LSHIFT <_QZ> no that too much to type haha <_QZ> and 65536 * is slower <_QZ> to process anyways I'm done w/ the math <_QZ> so the binary is ready? 05:40pm no I'm gonna do logic, conditionals, and loops <_QZ> ugh oh yeah, I should add variables & constants ASAP <_QZ> no, just make a binary allright you win <_QZ> :) adding them to the dictionary.. tcn: another tricky thing is garbage collection of code... and/or dynamic replacement of running code... <_QZ> what Fare: let's not do that in the early LLL :) if we want to be able to dynamically replace part of the code, we have to be able to atomically reify all reusable state, unlink the code, link the new one, and initialize it with the given state. tcn: right I need to write a "mission statement" for Retro so people will stop thinking so far ahead :) abi, retro? retro is at http://bespin.cx/~tcn/retro.html <_QZ> is that binary ready yet? lemme test it 05:50pm looks good <_QZ> kewl commit commit commit ok, binary's up.. it's under ~tcn/retro/kernel crap yes? qz, grab the 'writeboot' util from ~tcn/retro/util I'm gonna change it so 'dd' will work <_QZ> why wont dd work? http://tunes.org/~fare/penguindex.html the 4th byte of the bootsector says how many sectors to load modules are appended to the image, so that number has to be changed. <_QZ> and writeboot must set thatvalue? yeah, currently that's the only thing writeboot does :) why not set the value in the image directly? <_QZ> i solved that problem long ago as part of make? Fare: that's what I will do think of writeboot as a relic from the past <_QZ> mk_ofs makes a default ofs filesystem image and adds the boot sector to it and adds the kernel as an object and points the bootsector to the kernel object <_QZ> add_o adds objects to the ofs image QZ: why invent yafs? I don't have a fs now.. what's ofs, then? object fs <_QZ> goto the brix page and read up on ofs <_QZ> basically it has a directory for the system and each user and each directory is flat <_QZ> no sub-dirs 06:00pm <_QZ> the first couple objects in each directory (user space) are database objects that hold info about the objects in the user space Our fs in Retro/Tunes is gonna be almost like the persistent store in RAM (which is subject to change a lot) so I'm not bothering with the fs just yet When I start writing Forth in itself, I'll create the text file in unix, "incbin" it NASM, and 'cat file.rdf >>image' what's incbin? tcn: for now, don't bother about persistency Fare: look at kernel/help.asm Ok, I'm gonna add CONSTANT, VARIABLE, and the compiler words ':' and ';' now, so we can write forth in itself. Hey, we can use eForth.. tcn: how do you determine the size of the blob? Fare: I don't :) could you? hang on 06:10pm maybe with XCOM can't NASM do a foo: ... bar: ... mov ax, foo-bar ? RDF only specifies the size of the .code, .data and .bss segments, not each individual record (I did that with TASM) Fare: yeah, you can do that, but I thought you were thinking about GC (and as86, too) btw, it still reboots hard on my PMMX laptop do a 'dichotomy boot' are you using the laptop now? <_QZ> wonderful <_QZ> seems i incorreclty hooked up my floppy drive don't you have like 10 computers? <_QZ> yes hey if I had 10 computers they wouldn't all be PC's I'd have 2 or 3 PC's, an acorn, an apple 2, a C64, a LSI-11, an X-MP, uhh.. :) <_QZ> hmm, i have the hdd of borg nfs mounted on p0wer and my /dev/fd0 is owned by me but i couldnt access it nah the X-MP would use too much electricity 06:20pm <_QZ> wow fatal errors up the butt <_QZ> says "Loading... FATAL ERROR" over and over hmm <_QZ> j/k <_QZ> but 'bye' does not work <_QZ> and why wont writeboot write to my nfs mounted floppy? i dunno <_QZ> i can 'cp image /B/dev/fd0 but I'm gonna get rid of it anyway I just did 'LITERAL'.. I think I'll do 'bye' now tcn: I'm almost always on the laptop :) it's my main computer... what's a LSI-11 ? isn't an X-MP a cray? yeah :) an older one I believe 06:30pm an LSI-11 is a PDP-11 made out of LSI chips, fits in a single rack case I could just see a corporation putting their old supercomputer out on the curb :) Is there a PDP-11 emulator somewhere? A friend of mine would love to run ITS... probably oops, ITS runs on the 10, not the 11 I think I saw a post about how to do that.. search for "Incompatible Timesharing System" on altavista. It'll only find 40-50 pages, don't worry :) tcn: is that you having a page on ITS? yeah, I started one at bespin.cx/~tcn/its.html 06:40pm add these pages to it: http://metalab.unc.edu/pub/academic/computer-science/history/pdp-10/docs/ http://www.siconic.com/vcf/vcflinks.htm http://www.inwap.com/pdp10/index.shtml thanks http://www.36bit.org/ http://bony.umtec.com/pdp10.html 06:50pm haha: http://www.ee.ryerson.ca:8080/~elf/hack/academic.html yeah 07:00pm hehe, "Rabid Prototyper" thought that was more for a sec s/more/me I find a bit of all of them in me :) yup "if a language can compile its own compiler it isn't a toy" :) 07:10pm macro programs: "they usually take as long to run as to write" as an exception, I'm definitely not number crusher (well, after all, there's the remembered O-level math) -:- AlonzoTG [Alonzo@client-151-200-125-84.bellatlantic.net] has joined #tunes I'm not a WACF, either howdy alonzo are you a MPS, fare? neither a ICFR I'm a MPS, definitely! not an OFAP yet fare: I gots me a book on lambda calculus and that stuff is like WAY over me head, u know what I mean? 07:20pm I got an idea, alonzo.. learn a little LISP.. that's what made lambda-calc "click" for me yeah That's what I'm up to now... :) 1337 $#!+ :) (lambda x (+ x 1)) lx.[x+1] what book was that, and how is it over you? what is the definition of lambda? it's a greek letter. Does it mean "execute" umm.. lemma try another tack alonzo, you know f(x)=5x kinda notation from math, right? \x -> E means "function that given an object, binds x to that object, and evaluates E in the environment extended with this binding" "lambda calculi: a guide for 1337 hax0r doodz" ATG: attracting title actually it is for computer scientists... if f(x) = x+1, then \x.[f(x)] = f a method for abstracting functions is... 07:30pm when you say f instead of f(x) ? I guesso really not sure... when you take the "lambda" of something, you're abstracting it okayz There's a There's a "LISP tutor" infocom game.. you should try it there's a link to it from the Review/Languages page phunkie! =P sleep time! later * Fare/#Tunes is away (rrrr zzzzz) 07:40pm hey alonzo, you gotta try Retro I'm too lame to try retro ... I got last november's version of it... haha there's a binary you only use DOS right? yep and 'doze 9.11 In a life or death emergancy I could fire up my linux partition and limp onto the net... but that thing is MONDO fucked up... -:- hcf [nef@me-portland-us807.javanet.com] has joined #tunes 07:50pm how much RAM/mhz, alonzo? 32mb P100 oh that's enough for linux you should get a decent linux setup Great plan!!!! Debian 2.1 is good.. I'm using it, bespin.cx is running it.. it's great Okay. I'll do that... Now who will I get to admin the sonbitch? just install the base system and let the install program run overnight how much will it cost me? admin?? you! $0 I've tried to admin it, Its a nightmare! It has me screaming within minutes, each time I fire it up! what distribution? redhat 5.0 no wonder that sucks! How was I supposed to know? I only have 33.6... and I SUCK REALLY HARD at getting linux on the net... so downloading it isn't an option... redhat looks nice and it was on the shelf... so I picked me up a coppy... then I realized what a piece of shit it was... I have a 56k but I only connect at 31.6.. too far from the CO Go to www.debian.org and read the install instructions <_QZ> 31.2 I didn't know you were anal retentive haha <_QZ> :) anyways, as long as you don't pay by the minute for phone calls (and if you did why are you on IRC?) you might as well get it over the modem nah, I'm too lame... you just pick the programs you want, and it does the rest automatically It'll be just another distribution that I wouldn't have a clue as to how to administer... 08:00pm I does most of that for you s/I/it BS <_QZ> haha, some plane dumped its toilets over a neighborhood here It may abstract the administering process but there is no way that I can andminister something *THAT* complicated... <_QZ> that would suck to have been outside at that time haha AlonzoTG: try it! when you have time my os will divide the administration task into byte sized chunks! :) Unix is a pain in the ass but it's better than DOS unix is worse than DOS because it is a pain in the ass. it is kinda slow starting, but faster than windows Anyone seen Win98? That shit is sloooooooooowwwwww yep It needs a disk scheduler... I did a defrag on the gremlincomputer and it took and hour! It takes a brand new machine w/ 64M RAM longer to boot than an old P90 with 16M (constantly swapping) to boot Win95 if it had a good scheduler it would have taken 10 minutes... my machie swaps in bursts... Wait till win2000 It decides that it needs to swap for 30 seconds and chills out... it'll need 2000 megs of disk space for itself (I have 3.11) haha 08:10pm My OS will be about the size of the active part of my DOS partition (12 megs) om that is just for the core, the aditional modules sold seperately... special high intensity training... :) Retro is 29 sectors holey moley... the 12 megs counts all drivers, libraries, and utilitys... a full distribution basically.. no networking no multiuser... no application packages... 12 megs is alot for all that compared to what? well, I guess if you have drivers for all 50,000 video cards... DOS is about 6mb on my system!!! not to mention all the other components of it! With Retro or Tunes we're gonna have a web page where you pick the drivers and stuff you want, and it doesn't send you anything you don't need. uh webpage? Most of DOS is crap, though, like MSAV and QBASIC.. that's lame, here is proof: Why make the user go to one webpage to make his computer run? What if it is just a word processor? no,that's just to install the OS will you make that user get a modem, an ISP, and all that stuff, just to get DRIVERS?!?!?!! c'mon! With Linux & cousins, you end up downloading a lot of crap you don't need. The kernel source is now over 12 megs, and it's mostly drivers. you can put suplemental stuff on the page but the rest should be included... I was thinking of just putting VESA driver on mine and let it take care of ALL the cards... If you get Tunes on a CD, it would come with everything. But if you're getting it over the net, why bother. Pretty soon everything's gonna be wired to the net, anyway 08:20pm tcn: when exactly is 'pretty soon'? hcf: it seems like everything is already on the net! nah That is just a fad... tcn: define everything People will soon put the net in perspective and it will be just another utility... everyone It'll be like the telephone I agree. But not everybody does all their banking over the phone or the net!!!! My OS will be distributed on CD... There is no other way to distribute an OS... Well, Tunes is free, so payment is not an issue without consuming outragous ammounts of bandwidth... I tried downloading BSD but I couldn't connect it to the net! Even installing a bloated piece of shit like Linux over the net, I use less bandwidth than all the lamers who browse commercial websites with images, java, etc turned on. BSD isn't very admin-friendly :) name a linux that is! It took me a few hours to figure out its PPP system BSD ain't Linux any unix... that is user friendly to the administer! Debian is great It's easier than Windows <_QZ> uhh windoze 9x but not 3.11 you ever try installing win95 or win98? Debian is easier! anyway... 9x sucks 3.11 is okay... yup <_QZ> my little brother can install win98 that's why I use it. It still pisses me off though It keeps thrashing and wearing out my drive! :*((((( yeah, 3.11 is even less reliable than 95 but better than 98 :) yep! :)