IRC log started Wed Jul 21 00:00:00 1999
[msg(TUNES)] permlog 1999.0721
-:- SignOff AlonzoTG: #TUNES (Have Nice Day :))
-:- eihrul [lee@usr5-ppp83.lvdi.net] has joined #tunes
<Plundis> eirhul!
<Plundis> eihrul even. :)
12:20am
-:- SignOff Mr_Wrong: #TUNES (leaving)
-:- AlonzoTG [Alonzo@209-122-196-115.s115.tnt5.lnh.md.dialup.rcn.com] has joined #tunes
-:- overfien [y0rd4ddy@m242.telcomplus.net] has joined #Tunes
<overfien> y0h
<AlonzoTG> om
05:10am
-:- SignOff AlonzoTG: #TUNES (Ping timeout for AlonzoTG[209-122-196-115.s115.tnt5.lnh.md.dialup.rcn.com])
* Plundis/#tunes is gone. ..
-:- AlonzoTG [Alonzo@209-122-229-117.s625.tnt5.lnh.md.dialup.rcn.com] has joined #tunes
-:- hcf [nef@me-portland-us301.javanet.com] has joined #tunes
-:- ElGato [dave@bespin.dhs.org] has joined #tunes
-:- core [core@mcp.suntech.fr] has joined #tunes
<ElGato> gee i wish tcn was here
<core> people hi
<ElGato> hello core
<core> hi elgato
<ElGato> what's up?
<core> elgato: working on clementine - just synced tons of stuff in cvs, adding a server to openprojects (heinlein.openprojects.net).. not much ;)
<core> you?
* ElGato/#tunes just added some stuff to retro
<ElGato> i guess i better email tcn
<core> elgato: he drops here often tho :) not for long anymore..
<ElGato> :\
<ElGato> i can't wait for clementine
<core> i hope you won't be disappointed :)
<ElGato> :D
<core> nothing will ever be perfect anyway. i think it's not too bad, and some people seem to think it's awesome :)
08:10am
<ElGato> we'll just have to see
<core> elgato: well, at least you'll have actual code to try.
<ElGato> i got retro but im not sure what to do after i boot up@
<core> elgato: it's a forth console.
<ElGato> i know
<ElGato> i don't know much forth :(
<core> elgato: well, that's your problem then ;)
<ElGato> 2 2 + :D
<core> 2 2 + .
<core> :)
<hcf> tut forth get .
<ElGato> it's vocab is very small though
<core> tril could probably run an openprojects server on bespin as well.
<core> ah well, at least I will have a server i can access in 100baseT, and fare will have one much closer now :-)
<core> elgato: well, yeah, but it can be extended at runtime :)
<ElGato> :
<ElGato> tom needs to write but something on his forth
<core> i'm sure he will
<core> he's busy with real life lately
<core> and busy tearing retro apart :)
08:20am
-lilo(lilo@varley.openprojects.net)- [GlobalNotice] Please welcome montpellier.fr.eu.opirc.nu aka fr.eu.opirc.nu aka heinlein.opirc.nu to Open Projects.  Thanks to Emmanuel Marty, the GGI Project and Suntech! :)
<ElGato> heh
<core> :)
<ElGato> "Please welcome montpellier.fr.eu.opirc.nu fr.eu.opirc aka heinlein.opirc.nu to Open Projects.  THanks to Emmanuel Marty, the GGI Project and Suntech! :)"
<ElGato> may i be the first to ask "Can I have an o-line dude?!"
<ElGato> j/k
<core> hehe
<AlonzoTG> om
08:30am
<ElGato> mooo
<core> hi al :)
-:- SignOff core: #TUNES (reconnecting to montpellier*)
-:- core [core@mcp.suntech.fr] has joined #tunes
<core> hi again :)
<ElGato> core: are you any good with pine???
<core> elgato: *shrug* i can use it
<ElGato> do you know how to attach a message?
<core> attach something to a message you mean?
<ElGato> yes
<core> i'd assume ^J works?
<ElGato> nevermind
<core> :)
<ElGato> i got it through
<core> alonzo: around?
-:- SignOff overfien: #TUNES (Write error to overfien[m242.telcomplus.net], closing link)
<AlonzoTG> om?
<abi> somebody said om was in hinduism, a word of affirmation or assent intoned as part of a mantra or as a symbolic mystical utterance during meditation
<core> om.
<ElGato> meow
* hcf/#tunes barks at ElGato
<ElGato> :(
* ElGato/#tunes runs away from hcf
* core/#tunes commits some more.
* hcf/#tunes chases ElGato up a pine tree
08:40am
* ElGato/#tunes ^X's hcf away
* hcf/#tunes gives up, morphs into a cat, rubs around the tree, sprays it, and purrs
<ElGato> :D
* eihrul/#tunes lets a pit bull looes to play with cat.
<eihrul> even worse, a GNU!
<core> oh my.
<eihrul> yes, that was a horribly failed attempt at a joke, excuse my suckiness please
<core> we'll consider it ;)
<core> elgato: are you still working on your language or somesuch ?
<ElGato> yeppers
<core> elgato: what are your goals with that? all i know is that you're working on one :)
<eihrul> what sort of language?
* core/#tunes skims logs in diagonal :)
<AlonzoTG> om
<core> hi al.
<ElGato> oh it's just a pure, fully oo, SEXP language
<ElGato> nothin special
<core> pure is good.
<eihrul> sexp? excuse my ignorance
<core> no method can have side effects?
<ElGato> symbolic expression (lisp like)
<ElGato> non
<core> well, i mean, side effects outside its object
<core> cool
<ElGato> totally pure
<eihrul> core: that's kind of contradictory, don't ya think?
<core> that could be quite nice, elgato :)
<eihrul> considering a methods sole goal is to act on a group of states :)
* ElGato/#tunes likes using buzzwords he learned in this channel
<core> eihrul: no, he means that you can't play with a static heap and such
<ElGato> core: the problem is i've never written a compiler and i don't have any books
<core> eirhul: well, i'd assume you can alter the state of the object the method is part of, or it'd be quite restricted :)
<eihrul> objects have to be stored somewhere
<core> elgato: that's not a light task
<eihrul> and most likely on some sort of heap
<ElGato> i have to find all the info i can on the web
<core> eihrul: yeah. of course. you're not understanding me :)
<core> eihrul: everything is constrained in the object then, there are no "global variables".
<eihrul> elgato: what language you writing in?
08:50am
<core> eihrul: so everything is reentrant, aka pure.
<ElGato> c
<eihrul> core: sure, by enforcing absurd limitations :)
<core> eihrul: how is it absurd?
<eihrul> the easy way to almost solve a problem: ignore it 
<core> eihrul: how do you solve reentrancy, short of having a context particular to the caller?
<eihrul> well, that is the only way
<core> right
<eihrul> but what if you have threads
<eihrul> roaming around in the same object?
<core> what does it have to do with reentrancy?
<hcf> ElGato: still gonna call it lengua?
<abi> hcf: wish i knew
<core> eihrul: you're talking about concurrency
<ElGato> hcf: yep
<core> eihrul: you can make code that is pure and will wreck if it is accessed concurrently with the same context..
<eihrul> core: no, reentrancy, threads context switch too you know
<hcf> abi: lengua?
<abi> rumour has it lengua is a pure, fully oo, SEXP language by hickserv/elgato
<hcf> ElGato: missing any buzzwords? ;)
<core> eihrul: and some code that is not pure, but still properly locked
<core> eihrul: it is unrelated.
* ElGato/#tunes doesn't know how to do dynamic memory allocation
<ElGato> hcf: those aren't really buzzwords
<ElGato> that's what my language will be
<eihrul> core: how is it unrelated?
<ElGato> if i can ever get this compiler written
<core> eihrul: reentrancy and threads context switch aren't the same thing
<core> eihrul: it is unrelated because it's two different things
<core> eihrul: you're talking about locking common data against concurrent changes
<eihrul> core: well, please explain the difference and don't leave me wallowing in my ignorance please :)
<core> eihrul: you can have code that is pure in that it doesn't have global, statically allocated at compile/link time, variables
<core> eihrul: but it can still break if you call it with the same context from two different threads (if you're supposed to in the first place)
<core> eihrul: it's reentrant, but it's not threadsafe.
<eihrul> it sounds like the same problem, on a different scale
<core> eihrul: and, you can have code that is not pure, but still inserts locks in the proper places, and is threadsafe
<core> eihrul: no, it's two different things.
<eihrul> i don't see where...
<core> you don't understand the difference between a context and a lock?
<eihrul> yes i do
<core> okay, so they're two different things, right?
<eihrul> but they're both ways of solving the same problem:
<eihrul> A) allocating many resources, B) guarding access to a single resource
<core> uh, no
<core> you don't allocate contexts just to avoid locks
<core> you do that so that every caller has its state
<core> it's when you want a global state, shared by all callers (like a device driver hardware status or something), that you use a lock.
<eihrul> but you can potentially let every caller have its state by allowing only one caller to access a global state at a time, albeit which is not a good solution :)
<core> read what you just wrote
<core> you can't have every caller have its state with a global state
<core> since it's a global state.
<eihrul> you can create the illusion of it
<eihrul> by making the caller wait until the resource is released by another caller
09:00am
<core> you're talking about global resources
<core> you can have states that are private to a caller, for many purposes; global resources as in related to hardware are much more rare than that.
<core> so again you can't just interchange both ways.
<eihrul> core: okay, you're right, i'm sorry :)
<core> eihrul: no problem, i was just starting to wonder if i was going to manage to explain myself :)
<eihrul> well, i'm stubborn... it sucks sometimes
<core> ah well, it's okay
09:10am
* ElGato/#tunes just missed that entire argument
<core> elgato: it's okay :)
<ElGato> im not even sure yet how i'll manage objects
<core> elgato: hmm, that's a question you should start worrying about then :)
<ElGato> indeed
<core> will you split your compiler in a language front-end, and code generation back-end ?
<ElGato> yeah
<core> okay, good
<ElGato> im _still_ working on the lexical anylizer
<ElGato> i've mentioned before that im slow
<eihrul> el: i have some code for a C++ one if you want to look at it
<core> maybe it wouldn't be too stupid to use RTL as the intermediate form, so you can just get away with writing a front-end, and using gcc as back end for now
<eihrul> C is the portable assembly language :)
<core> depends how expressive RTL is, ie. how much latitude you have to introduce new concepts like that
<ElGato> abi: RTL?
<abi> elgato: wish i knew
<ElGato> :(?
<core> abi: rtl is the register transfer language, the intermediate representation of a program, inspired by lisp, used by the back-end of the gcc, g++, g77 etc. compilers.
<core> rtl?
<abi> rtl is the register transfer language, the intermediate representation of a program, inspired by lisp, used by the back-end of the gcc, g++, g77 etc. compilers.
<ElGato> i suppose i'll just write in gcc for now
<ElGato> it's ok
<core> elgato: gcc does that: C -> rtl (frontend)  rtl -> object code (backend)
<core> elgato: it performs a LOT of optimisation on the rtl tree before the code is generated, which is common to all platforms
<core> elgato: then it performs platform-specific optimisation in the generation back-end.
<ElGato> core: do you know how to use realloc()?
<core> elgato: if you can produce RTL with your compiler, you can use it as a front-end and use the rest of gcc.
* AlonzoTG/#tunes needs to learn neural netwerkz.
<core> elgato: yes, and why do you want to use such a crappy function?
* ElGato/#tunes so words can be as big as i want them to be
<core> elgato: sure, but why *realloc* for that?
<core> elgato: realloc() is the nighmare of memory managers
<ElGato> what sould i use?
<core> elgato: malloc()
<ElGato> mkay
<core> elgato: you understand the difference between the two, right? :)
<ElGato> yes
<core> okay.. realloc() will turn your process sbrk into a lot of tiny unusable holes in no time.
<ElGato> i don't know how im going to implement this though
09:20am
<core> maybe for something as complex as a compiler, memory management is something you should really think about first; either using a garbage collector of sorts (the ones for C suck, but..) or some automatic chained memory freeing of sorts like gdb uses
-:- smkl [sami@MCVI.rdyn.saunalahti.fi] has joined #tunes
<ElGato> im getting rid of whitespace then taking real words and putting them into a dynamic array until i run into more whitespace
<ElGato> how can i use mallow to do that?
<ElGato> malloc rather
<core> elgato: malloc() every word?
<ElGato> hmmm
<core> elgato: or think of a better algorithm :)
<ElGato> my older one was:
<ElGato> (char)pWord[i] = getc(fptr);
<ElGato> i++
<ElGato> while(bla bla bla)
<ElGato> when it wasn't dynamic
<ElGato> do...while btw
<smkl> what are you trying to make?
<ElGato> lexical anylizer
<ElGato> i was just thinking of making the word size max 32 byte but people don't seem to  agree with such limitaions
<core> hold on
<core> why do you have to STORE the words you parse
<ElGato> (define 12345678901234567890123456789012)
<core> so, you don't store "define"
<core> you just tokenize that word and store the value it defines i guess
<ElGato> because im sending it from the scanner to the parser
<core> well, the scanner can tokenize on the fly
<smkl> make that buffer extensible
<core> ie, compute a hash as you scan the text, stop on white space, look if you know the hash in the words list, then compare strings to make sure (but that'll probably be one, two at most), and there, you have your index or token or whatever you call it.
<core> store the index, and for what can't be tokenized, a pointer directly into the original text, with the length
<ElGato> oh great i have to write my first hasher
<core> this way no need to realloc() over the place
09:30am
<core> elgato: hashing is easy. if you want you could away rescan your entire word list all the time with strcmp's, but that'd be, er, slow :P
<ElGato> so what i add all the ascii values of each character together then  % it by a prime number and store that value?
<core> you can devise any scheme as long as it has few mathematical probabilities of not being unique
<core> you'll still be doing actual compares for safety on hashes that correspond, but the least probability, the least compares.
<ElGato> well this is my first hashing function i'll have ever written
<ElGato> so i'll need a little help
<core> GNU use that:
<core>    while ((c = (unsigned int) *str++)) {
<core>       hash += c << j;
<core>       j = (j+1) & (8*sizeof(hash)-1);
<core> }
<core> with j initialised to 1 at first.
<core> (and hash to 0 obviously)
<ElGato> ok now i gotta figure out how this works :|
<core> it simply adds the value of every char shifted by j, and j rotates from 1 to the size of the hash value in bits (so probably 32).
<core> i don't know how good this is math-wise.
<core> (well, that'd be 31, not 32.)
<core> it's more prone to false matches than an actual CRC, but it's also way faster.
* ElGato/#tunes is dumb math-wise
<ElGato> the thing is, is at one point i'll have to have a full string
<ElGato> and then i'll have to use dynamic memory
<ElGato> for each word
<core> don't copy the string
<core> just store a pointer into the original text with its length or somesuch.
<core> you should never have to use actual strings for words of your language
<ElGato> a pointer to a     location in the file?
09:40am
<ElGato> microsoft is now encouraging alternative lifestyles
<ElGato> monster truck fetish
09:50am
-:- binEng [Anders@130.244.43.250] has joined #Tunes
<hcf> hoy binEng
<binEng> hi
<abi> niihau, binEng
<ElGato> !!!
<ElGato> binEng
<binEng> hi ElGato :)
-:- mandicg [mandicg@tesla.rcub.bg.ac.yu] has joined #tunes
-:- SignOff mandicg: #TUNES (ircII2.8.2-EPIC3.004 --- Bloatware at its finest.)
10:00am
* ElGato/#tunes is away: (Auto-Away after 10 mins) [BX-MsgLog On]
* ElGato/#tunes is back from the dead. Gone 0 hrs 0 min 5 secs
-:- NetSplit: heinlein.openprojects.net split from verne.openprojects.net [10:16am]
-:- BitchX+Deb1an: Press Ctrl-F to see who left Ctrl-E to change to [heinlein.openprojects.net]
-:- SignOff eihrul: #TUNES (Ping timeout for eihrul[usr5-ppp83.lvdi.net])
<ElGato> core: i can't connect to your sever
<ElGato> server
10:20am
-:- SignOff hcf: #TUNES (Ping timeout for hcf[me-portland-us301.javanet.com])
-:- binEng_ [Anders@dialup43-4-58.swipnet.se] has joined #tunes
-:- SignOff Beam: #TUNES (Ping timeout for Beam[d164.nnb2.interaccess.com])
-:- Beam [nobody3@d164.nnb2.interaccess.com] has joined #tunes
-:- hcf [nef@me-portland-us927.javanet.com] has joined #tunes
-:- SignOff binEng: #TUNES (Ping timeout for binEng[130.244.43.250])
-:- binEng_ is now known as binEng
<AlonzoTG> om
<ElGato> ok
-:- core [core@core.suntech.fr] has joined #tunes
<core> re
<ElGato> !!!
<hcf> gex
<ElGato> wbb
10:40am
<core> re elgato
<core> elgato: you can't connect to my server? how come?
<ElGato> lmme try again
<core> allright
-:- SignOff ElGato: #TUNES (changing servers)
-:- ElGato [dave@bespin.dhs.org] has joined #tunes
<core> seems to work now :)
<ElGato> YES!
<core> not that it'll be any better for you :)
-:- SignOff binEng: #TUNES ( <k!14>)
10:50am
<ElGato> heh
<core> wot?
* ElGato/#tunes is gonna hax0r core
<ElGato> im inside now
<core> i didn't write the motd :)
<ElGato> :D
<core> lilo did
<ElGato> :(
<core> hmm ?
* ElGato/#tunes just popped a big zit
<core> what's that :P
<ElGato> it squirted on the wall
<core> oh, i knew i didn't want to know
<ElGato> you know big red things that grow on your face
<ElGato> acne
<core> well, i've been thru that a long time ago, i forgot :P
<ElGato> pussy
<ElGato> :D
<ElGato> wrong word
<core> pussy is good. not that you'd know
<ElGato> i think i meant pusy
* ElGato/#tunes giggles at the way coresaid that
<ElGato> :)
<core> why? :P
<ElGato> you know why you sick bastard :D
<ElGato> J/K
11:00am
* AlonzoTG/#tunes needs to learn neural wetnurks. =P
<hcf> AlonzoTG: have u read the ann pointer from abi?
-:- eihrul [lee@216.24.141.157] has joined #Tunes
-:- NetSplit: devlin.openprojects.net split from asimov.openprojects.net [11:07am]
-:- BitchX+Deb1an: Press Ctrl-F to see who left Ctrl-E to change to [devlin.openprojects.net]
* ElGato/#tunes is away: (Auto-Away after 10 mins) [BX-MsgLog On]
11:10am
-:- eihrul [lee@usr5-ppp157.lvdi.net] has joined #Tunes
<AlonzoTG> hcf: ann sucked
<AlonzoTG> neuron?
11:20am
<hcf> AlonzoTG: http://hera.snu.ac.kr/link.html
<ElGato> yay
<ElGato> fare is almost back
<ElGato> the conf is over
11:30am
<ElGato> if tunes has two memory areas, one that is trusted and one that is not, how is it decided how large either should be
<ElGato> ?
11:40am
-:- NetSplit: tolkien.openprojects.net split from sterling.openprojects.net [11:42am]
-:- BitchX+Deb1an: Press Ctrl-F to see who left Ctrl-E to change to [tolkien.openprojects.net]
* AlonzoTG/#tunes left clicks on hcf
-:- Netjoined: tolkien.openprojects.net sterling.openprojects.net
-:- core [core@core.suntech.fr] has joined #Tunes
-:- ElGato [dave@bespin.dhs.org] has joined #Tunes
<ElGato> core: like the "trusted" space isw one huge segment for the x86?
<ElGato> flat mem management
<core> elgato: s/segment/page directory/
<core> elgato: i suppose you could do it with segments if you like hassle.
<eihrul> core: ack! segments! no way :)
<core> eihrul: don't tell ME
<eihrul> they're impossible to work with
<ElGato> there is really no need for paging
<ElGato> we aren't using virtual memory
<eihrul> ummm, you haven't tried segmentation
<ElGato> yes i have
<ElGato> segmentation is easy
<eihrul> well it makes writing protected device drivers hard
<eihrul> especially for memory mapped devices
<ElGato> it's no harder than paging
!forward.openprojects.net!! Remote CONNECT asprin.openprojects.net 8005 from lilo
<eihrul> and plus it makes better use of the physical memory
<eihrul> re: the realloc () problem core talked about :)
11:50am
<ElGato> sorry i was split off
<ElGato> could you tell me what i missed?
<eihrul> what's last line you saw?
<ElGato> i'll just check the logs
<eihrul> gah
<ElGato> i guess i didn't miss anything
<ElGato> nm
<ElGato> i think for tunes fare just wants a flat memory model
<eihrul> well paging is best way to provide that
<ElGato> no
<eihrul> or else people woul dbe using segmentation
<core> elgato: tunes wants a flat memory model for trusted processes inside a page directory.
<ElGato> just make one large segment 
<ElGato> and work off the offset
<core> elgato: so you can still run untrusted code like linux programs in other processes.
<ElGato> oh
<ElGato> ok
<core> elgato: you'll have to use huge segments if you don't want to reallocate them all the time, so you won't run many processes
<AlonzoTG> om
<ElGato> well paging and and segmentation can coexist
<core> the 4K grain is annoying, but it's the easiest and most efficient over time
<core> segmentation performance degrades over time until next reboot
<eihrul> el: but why bother?
<core> (ie, as memory becomes fragmented)
<eihrul> paging provides all the benefits segmentation does
<core> elgato: you can do all you do with segments, with paging
<eihrul> so it obsoletes segmentation
<eihrul> jinx
* ElGato/#tunes sees a point there
* ElGato/#tunes will have to agree
<core> and segmentation is x86 specific for the most part
<core> if you want your system to be portable to other platforms (and, say, even to the merced), you don't want to use it
<ElGato> okey dokey
<eihrul> core: you could do memory compaction with segmentation, the real crappy parts are that you have to explicitly code for segmentation :)
<ElGato> i don't see how we'll manage two data ares though
<ElGato> areas
<core> eihrul: that (carrying 48 bit addresses which are actually 64 in memory) really blows
<core> elgato: what do you mean?
<eihrul> you could scrunch segments together to eliminate holes
<core> eihrul: memory compaction, then you have to introduce a lock() / unlock() scheme on memory regions like on macOS, and on top of malloc()/free(), this blows goats
<core> eihrul: see what i just said.
<eihrul> exactly, that's my point :)
<eihrul> segmentation sucks
<ElGato> hmmm
<ElGato> segmentation is good ofor pointer arithmetic
<core> why ?
<ElGato> it's faster
<core> er, why?
<eihrul> you sure about that?
<core> tell me how adding 32 bit offsets, is faster than adding 32 bit addresses.
* ElGato/#tunes fears ridicule
<ElGato> i'll just pretend i never said that
<core> heh :)
<ElGato> will th etrusted and untrusted mmemory areas belong in the same page directory?
<core> actually segmentation could be good for garbage collection if you were using one segment for every data and code object in your system, but there aren't enough of them for that.. and it introduces massive overhead.
12:00pm
<core> elgato: no, there will only be one trusted area (i assume), which will be one page directory, and then several untrusted, seperated untrusted ones, which therefore need a page directory of their own for each.
<ElGato> oh
<ElGato> ok
<ElGato> and the entire trusted page directory is 0 will the untrusted ones are 3?
<core> but with migration, if a trusted process grows large, it can be instantly remapped into a seperate directory
<core> elgato: pl you mean? well, the trusted processes in the trusted dir could indeed run at PL 0 for no supervisor calls overhead, since they're safe.
<core> elgato: although it's not required
<ElGato> hell you could run it at 2
<core> there's no point using PL 1 and 2
<ElGato> the untrusted
<core> it's extra annoyance to setup PL 1 and 2 stacks
<core> and it introduces the same syscall overhead as 3
<ElGato> yes
<core> so might as well run only at 0 and 3
<core> well, if we run it at 2, we could run it at PL 3.
<core> i got the chaosdev people to realise that :P
<ElGato> no i mean you could just as well run at 2 and 0 or 1 and 0 or 3 and 2
<core> no
<ElGato> as long as you have two seperate ones
<core> with paging, untrusted code must run at PL 3
<core> the pages can only be protected against PL 3; PL 0-2 is considered supervisor
<ElGato> oh
* ElGato/#tunes must refreash himself with his manuals
<core> PL 1/2 are pretty useless anyway IMHO. most other cpus have only two levels anyway
<ElGato> uh huh
12:10pm
* ElGato/#tunes is examining the picture of core trying to ride the tricycle
<ElGato> intruiging
<core> elgato: i'm intriguing like that.
* eihrul/#Tunes examines the picture of core dumping from the tricycle.
<eihrul> i had to :)
<ElGato> heh spam.abuse.net
<ElGato> there's this game i play where the unwashed villagers stone the spammer
* ElGato/#tunes must be leaving
-:- SignOff ElGato: #TUNES (ElGato has no reason)
12:20pm
-:- _ruiner_ [nate@t-9-180-168.dialup.wisc.edu] has joined #tunes
-:- SignOff Beam: #TUNES (Ping timeout for Beam[d164.nnb2.interaccess.com])
-:- SignOff smkl: #TUNES (Ping timeout for smkl[MCVI.rdyn.saunalahti.fi])
<_ruiner_> greetings all
<core> hey ruiner :)
<_ruiner_> whats the topic for today?
<core> ruiner: nothing special :)
<_ruiner_> hmmmmm......
<core> ruiner: what's up with you today? :)
<_ruiner_> nothing....I'm kinda bored
<hcf> _ruiner_: dont u have an os that needs coing?
12:40pm
<core> ruiner: ah..
<core> ruiner: well, i got work done and a server linked to openprojects today, i'm tired, and nothing else is new :)
-:- smkl [sami@MCVI.rdyn.saunalahti.fi] has joined #tunes
<_ruiner_> hmmmm......
-:- mode/#Tunes [-o Tril] by OperServ
-:- mode/#Tunes [+o Tril] by OperServ
<core> okay i'm done playing :)
12:50pm
-:- SignOff javaks: #TUNES ("Man with hand in bush not necessarily trimming shrubs.")
-:- SignOff hcf: #TUNES (Leaving)
<_ruiner_> I'm taking off....later
-:- SignOff _ruiner_: #TUNES (Leaving)
-:- SignOff core: #TUNES (This is *not* the default quit message.)
01:30pm
-:- hcf [nef@209.94.148.159] has joined #tunes
-:- _QZ [brand@p0wer.qzx.com] has joined #tunes
-:- SignOff smkl: #TUNES (smkl has no reason)
-:- SignOff eihrul: #TUNES (Ping timeout for eihrul[usr5-ppp157.lvdi.net])
>>> hcf [nef@209.94.148.159] requested PING 932594471 from TUNES
-:- smkl [sami@MCVI.rdyn.saunalahti.fi] has joined #tunes
-:- binEng [Anders@dialup47-2-48.swipnet.se] has joined #tunes
-:- tcn [tcn@cci-209150250125.clarityconnect.net] has joined #tunes
<_QZ> tcn: elgato was bitchn about retro
<tcn> hehe
<tcn> what about it, it doesn't have a porn viewer?
<_QZ> it says 'very interactive' but he doesnt know any forth so he thinks its not
<_QZ> hahah
<tcn> heh
<_QZ> do u have any documentation on what words u have for retro?
<tcn> yeah, words.inc
<tcn> I guess I should clarify that, it's very interactive if you know what you're doing, like me.
<_QZ> Plundis: u here?
<_QZ> do u have any links to forth tutorials?
<tcn> as opposed to non-interactive, where you can only control it by editing the source
<_QZ> i think elgato's definition of interactive is that it has a web browser :)
<_QZ> a graphical web browser 
<tcn> yeah, I have a few.. I should post/link them.. I was writing one myself, I guess that'll focus on retro-forth
03:10pm
<binEng> I need some help... how do I pack a whole directory structure in Linux?
<_QZ> 5.      Before anyone else is in the lab, connect each computer to a
<_QZ> different screen than the one it's set up with.
<_QZ> tar cvf filename.tar directory/
<_QZ> OR tar zcvf filename.tar.gz directory/
<tcn> haha.. nice prank
<binEng> what's the difference?
<abi> the difference is, like, not mainly technical, but mainly political
<tcn> abi, forget the difference!
<abi> tcn: I forgot difference
<_QZ> the second one will gzip the tar
<binEng> :)
<_QZ> tcn: why cant we just leave that shit in abi? it makes for some laughs
<_QZ> and confuses the hell outta ppl that dont know abi
<hcf> plus
<hcf> its part of #tunes' consciousness
<_QZ> ya
<tcn> haha
<tcn> hcf: program abi with, like, a 100 wpm typing delay
<binEng> "tar: Error exit delayed from previous errors"
<_QZ> binEng: u got any files that are not accessible to u?
<_QZ> i had a file with 000 perms and got that error
<binEng> _QZ: um, no..
<_QZ> the tar file is still generated but wont have those files in it
<binEng> I got it, I typed wrong in the parameter list
<tcn> qz: I got a couple emails from elgato.. I guess he likes retro enough to hack it..
03:20pm
<_QZ> how the hell can he hack it if he dont know forth?
<tcn> asm
<binEng> tcn: What does retro do so far?
<tcn> too bad he don't know about diff
<_QZ> tcn: haha, how much did he send u?
-:- ElGato [dave@bespin.dhs.org] has joined #tunes
<ElGato> tcn
<ElGato> !!!
<tcn> hey
<_QZ> wow looks whos here?
<tcn> qz: just 2 files
<ElGato> :D
<_QZ> s/?/ :)/
<ElGato> hey
<ElGato> i wasn't bitching about retro
<ElGato> dumbass
<_QZ> :)
<_QZ> yes u were
<tcn> hey elgato, I got a couple forth tutorials I can dcc you
<ElGato> i don't give a damn about graphics
<ElGato> send em!
<ElGato> although i must agree a porn viewer would be a nice additon
<_QZ> i porn viwer and a 500meg binary image loaded right? :)
<ElGato> nah
<ElGato> um
<hcf> put porn in retro as easter eggs
<_QZ> hcf: hhaha
<ElGato> what should i do once i find that FPU tcn?
<tcn> oh
<tcn> do you need to detect it in real mode?
<ElGato> did you look at what i did?
<ElGato> it's before the jump to pmode
<_QZ> i have fpu detection code
<tcn> yup
<ElGato> right now im just printing a message but i don't  think retro has anything for the fpu yet
<tcn> and yes, nasm has cpuid
<ElGato> good
<tcn> and mmx, 3dnow, whatever
<tcn> anyways the detection code will just set a flag
<ElGato> yeah
<tcn> if it can be done in protected mode, it should be
<ElGato> the message can just be temporary
<tcn> also we need detection code for 386/386's
<tcn> 486's
<ElGato> i can work on that
<tcn> 387/486
<_QZ> i just dcc'd elgato all the detection code
<tcn> yeah, brix detected my 486 right
<ElGato> good for you qz
<_QZ> 386 detection is in boot.inc
<_QZ> er boot.asm
<binEng> tcn: What does retro do so far?
<ElGato> it has some but it's just to detect a 386
<tcn> oh yeah
03:30pm
<tcn> mainly it's a forth interpreter/compiler
<ElGato> i like that pmode message: "You're already in pmode. WTF?"
<tcn> you could get that in dosemu or soemthing
<binEng> tcn: so what can you do with this version of forth?
<tcn> binEng: well, I wrote some gfx routines in forth
<binEng> tcn: but you have not access to all of the machine, or?
<tcn> i've coded most of the important stuff of Forth. Yeah, you have access to the whole machine.. IOPL=0.
<binEng> cool
<ElGato> tcn: what do you think about my revision of start: in boot.asm?
<ElGato> :D
-:- eihrul [lee@usr5-ppp78.lvdi.net] has joined #tunes
<tcn> ElGato: I haven't run it thru 'diff' to figure out what you changed :)
<tcn> ElGato: type 'man diff' and see if you can figure out how to send only the parts of a file that you changed
<tcn> hey, I see all your bitchX's have that highlighted : after the nick
<ElGato> basically you did a " xor ax, ax mov ds, ax mov es, ax"
<ElGato> i did a "xor es, es xor ds, ds"
<ElGato> mabye saved a clock sycle
<tcn> oh, you can do that? :)
<ElGato> i think so
<_QZ> no u cant do that
<ElGato> mabye i forgot about limitations
<ElGato> ah shit
<ElGato> :(
<_QZ> tcn: u might wanna check all the code he changes for u :)
<_QZ> really good
<ElGato> uh uh
<ElGato> i just forgot abnout that limitation
<tcn> yeah, definitely don't commit any changes to CVS that you haven't actually tested
<_QZ> haha
<tcn> nope, you can't xor a segment register
<ElGato> crap
<ElGato> oh well
<ElGato> forget it then
<tcn> download NASM :)
<ElGato> uh
<ElGato> i do
<ElGato> i've had it downloaded for a long time
<_QZ> try using it
<ElGato> and INSTALLED!!! :)
<tcn> 0.98 is out now
<ElGato> god damn both of them versioned me
<tcn> although 0.97 should work
<ElGato> i have 0.98
<tcn> hehe.. FPU support should be easy. The 80x87 uses a stack architecture, just like a Forth chip
<ElGato> :)
<tcn> look how close Intel came to having a Forth CPU
03:40pm
<ElGato> qz: hey and shutup i wrote my first forth program today!
<ElGato> 2 2 + cr .
<ElGato> so HAH!
<ElGato> :D
<_QZ> uhh i told u to write that
<tcn> heh
<ElGato> uhhh sure you did
<ElGato> :)
<tcn> it's an easy language
<tcn> try Visual Basic if you don't believe me
<_QZ> ElGato: u cant deny it cuz we have logs in this channel :)
* ElGato/#tunes shivers at the thought of learning basic
<ElGato> i gotta go help my dad for a half hour
<ElGato> see you later
-:- SignOff ElGato: #TUNES (ElGato has no reason)
<tcn> heh
<tcn> I wonder why he did XOR ES,ES but not MOV SS,STACKSEG instead of MOV AX,STACKSEG ; MOV ES,AX
<_QZ> i wonder if he actually wrote a word (program) or just typed that in
<tcn> : 4 2 2 + cr . ;
<tcn> : four 2 2 + cr . ;
<_QZ> do u have support to change words
<tcn> ever heard of Color Forth?
<_QZ> no
<tcn> yeah, you can redefine words in retro
<tcn> but not retroactively, yet
<_QZ> can u call the old version of the word?
<_QZ> eh?
<tcn> no version tracking or anything
<_QZ> in real forth u can redefine the + word and have yer new version call the old version
<_QZ> but u dont have that?
<tcn> yeah, I can do that
<tcn> but no subsequent definitions can call the old +
<_QZ> does it save to floppy?
<tcn> no but the ide code works
<_QZ> damn
<tcn> I haven't updated the floppy driver yet
<_QZ> i dont think yer suppose to be able to call above the last definition
<_QZ> as long as u can call the previous def yer ok
<tcn> It would be useful to redefine a word and have it affect all previously compiled calls to that word, retroactively
<_QZ> if retro could save words to floppy then i could give elgato a bunch of code that would redefine all his words and he would think retro was broke :)
<tcn> yeah
<hcf> abi: color forth is at http://www.ultratechnology.com/color4th.html
03:50pm
<tcn> : 2 100 ; : 3 2048 ; : 4 write-floppy ; 
<tcn> 1 2 3 4
-:- jal [user1659@borlea1-1-104.abo.wanadoo.fr] has joined #Tunes
-:- jal [user1659@borlea1-1-104.abo.wanadoo.fr] has left #Tunes []
04:00pm
>>> tcn [tcn@cci-209150250125.clarityconnect.net] requested VERSION  from #tunes
<tcn> binEng: what OS you using?
<binEng> tcn: win98 :P
<tcn> did you figure out how to make it boot up faster?
<binEng> me? no..
-:- SignOff smkl: #TUNES (sleeping...)
<tcn> ain't it stupid how windows doesn't give you any kernel messages to make you think it's doing something?
<tcn> it doesn't even have a progress indicator
<binEng> that's the little problem. I'd be happy enough if it just booted faster :)
<eihrul> silly... if it had messages, the users would have to read them....we couldn't have that....
<tcn> brb
-:- SignOff tcn: #TUNES (tcn has no reason)
-:- tcn [tcn@cci-209150250125.clarityconnect.net] has joined #tunes
04:10pm
<_QZ> eihrul: hey i used chaos
<_QZ> eihrul: why does it say yer using fat?
<_QZ> eihrul: and i think u are using a swedish keymap
<_QZ> and it was kinda slow booting
<eihrul> qz: i just joined the project yesterday... these and other interesting questions are still yet unknown to me :)
<_QZ> hmm
<eihrul> i haven't even booted it yet... i'm still assessing source
<tcn> well of course it's slow, it's a microkernel :)
<_QZ> u should boot it so u can see how slow it is and then do something about that, like maybe drop grub :)
<_QZ> tcn: no its a 40k uk
<tcn> grub is slow?
<_QZ> ya it has to make a pit stop to say its is grub
<tcn> microkernel doesn't mean "small kernel", it means pointlessly multithreaded
<eihrul> and macrokernel just means pointlessly unsafe
<_QZ> tcn: qnx has a 16k uk, brix is at 9k right now and got about 512 bytes smaller today
<eihrul> qz: 16k with process manager?
<_QZ> eihrul: i assume it has process management in the kernel
<_QZ> eihrul: my 9k kernel has process management
<eihrul> gcc can't even spit out a version of "Hello, World!" that small
<tcn> hehe
<_QZ> eihrul: i suggest u dump gcc
<eihrul> qz: what else is there?
<_QZ> nasm
<tcn> "Hello world" would be about 25 bytes in Retro
<tcn> in forth
<eihrul> qz: i mean C compilers
<_QZ> i think i make make hello world about 16 bytes in nasm
<tcn> using 5-bit characters?
<_QZ> 8bit
<_QZ> back in #asm we had contests to see who could make the smallest programs
<_QZ> hello world was one of them
<tcn> "hello world" or "hello, world!" ?
<_QZ> hello, world!
04:20pm
<tcn> that's 13 bytes
<_QZ> there is a secret
-:- ElGato [dave@bespin.dhs.org] has joined #tunes
<tcn> in DOS?
<_QZ> ya
<tcn> 18 bytes, right?
<_QZ> no im pretty sure it was done in 16
<_QZ> or 15
<tcn> 3 for mov ?x,address .... 2 for "int x"..
<_QZ> u dont have to store the entire string
<tcn> huh?
* ElGato/#tunes didn't think about just initializing the damn FPU
<ElGato> DUH!
<_QZ> asm is a wonderful and very powerful language
<_QZ> ElGato: u read my source?
<ElGato> im in the process of it
<ElGato> btw though
<ElGato> what do i have to compil to compile retro
* binEng/#tunes is leaving
<ElGato> "nasm boot.asm" doesn't work
<_QZ> if u make any changes send me the diff and not the entire file :)
<tcn> see ya
<ElGato> bye binEng
<_QZ> ElGato: nasm and my boot.asm?
<binEng> cya
-:- SignOff binEng: #TUNES ( <k!14>)
<ElGato> no
<ElGato> for retro
<_QZ> oh
<tcn> make
<_QZ> cuz nasm my_boot.asm wont work without the *.def files :)
<ElGato> "no rule to make target: :(
<tcn> oh
<tcn> I used some stuff in other directories
<ElGato> uh huh
<tcn> in the CVS distribution
<ElGato> well i can't test anything unless i can compile it
<tcn> heh
<tcn> you could type it in machine code in retro
<ElGato> :)
<ElGato> i don't think so
<tcn> hex <opcode> <parameters> , <opcode> <params> , ....
<ElGato> i don't think that's such a great idea
<tcn> well I'll fix it up soon
<ElGato> k
<ElGato> i'll just work on init.inc some more till then
<tcn> you could figure out how to boot it without the full OS, if you wanted to
04:30pm
<ElGato> what do you mean?
<tcn> you'd hafta learn make, first of all
<_QZ> man i found a crap load of optimizations in my setup code. it should shave a good 50-100ms off boot time :)
<tcn> 'info make'
<ElGato> :P
* ElGato/#tunes wants to write an FPU emulator
<tcn> go right ahead
<ElGato> aye aye
<tcn> better yet, how about a multiprecision arithmetic library?
<ElGato> uhhh
<tcn> you could use it for RSA
<ElGato> that sounds big
<ElGato> whatever it is
<tcn> not really
<tcn> instead of using fixed 32-bit numbers it uses as many bits as it needs to hold a number
<ElGato> what would it involve?
<tcn> I wrote a 6812 program to do multiprecision add/subtract, in a few dozen bytes
<tcn> you have to carry between registers
<tcn> there are CPU instructions designed for it
<tcn> like "add w/ carry" and "subtract w/ borrow"
<tcn> dammit. I gotta do laundry
<ElGato> well for now i think i'll just write the FPU emulation and handleing
<tcn> yeah, have fun
<ElGato> :)
<tcn> see ya all
-:- SignOff tcn: #TUNES (tcn has no reason)
<ElGato> how much of brix is in asm?
04:40pm
<_QZ> currently all of brix is asm
<ElGato> heh
<ElGato> typical
<_QZ> i am designing it to be small and fast and like whats his face said "a hello world in gcc is bigger than 16k"
<eihrul> that was more of an elaboration
<ElGato> uh huh
<_QZ> i dont give a damn how readable my source is to others since they will likely never see it
<eihrul> hello world is 4,779 bytes :)
<eihrul> you could fit half a kernel in there
<_QZ> i want brix to be in its gui with networking support,etc in under 5 seconds 
<_QZ> and it must be very small to do that
<eihrul> well what happens when you have to migrate to another platform?
<ElGato> screw boot speed
<ElGato> it isn't important
<eihrul> you're going to rewrite everything?
<_QZ> eihrul: someone ports the kernel to it and they add the binary conversion to the language and have the language make a copy of itself in that binary
<eihrul> if the asm parts are small, then yeah, isn't that bad
<_QZ> ElGato: i leave my system on 24/7 cuz i dont want to wait 5-10 minutes for it to boot up and load my 40 windows
<ElGato> 5-10 minutes isn't that bad if what you're loading is important
<_QZ> the os should boot in the same time it takes the monitor to power up
<ElGato> is there a better way to mask all but the first bit of eax without a bitshift?
<ElGato> im very familier with masking techniques
<_QZ> and eax,~1
04:50pm
<ElGato> why the ~?
<eihrul> bitwise-not
<_QZ> ~1 is faster than typing 1111111111111110b
<ElGato> i'll just use test and jz
<ElGato> seems a little easier
<_QZ> are u testing bit 0?
<ElGato> yeah
<_QZ> then u dont want that
<_QZ> test eax,1 ; jz blah
<_QZ> or test al,1 ; jz blah
<ElGato> how is that any different from "i'll just use test and jz"
<ElGato> now to set the proper flags
05:00pm
<_QZ> if i were to let u have 5 files from the brix source what would u want?
<_QZ> there are 115 kernel files
<ElGato> hmmm
<ElGato> the part that takes care of object persistance
<ElGato> the mmu
<ElGato> after that i really don't know
<_QZ> specify what parts of mmu
<ElGato> well i don't know what parts there are
<_QZ> like ram_alloc?
<ElGato> sure
<_QZ> ram_free, ram_defrag, mem_alloc, mem_free, mem_defrag?
<_QZ> pt_crt?
<ElGato> mem_alloc, mem_defrag, pt_crt i guess
05:10pm
<_QZ> i am moving the persistence code to a thread inside the kernel. do u want the current code that dont quite work or do u want the source after i move it to a thread tonight
<ElGato> the new code
<_QZ> k
<ElGato> oh i get to pick on e more :)
<ElGato> i guess ram free
<ElGato> ram_free
<_QZ> thats 6
<ElGato> oh
<ElGato> sorry 
<ElGato> i thought that was 4
<_QZ> ram_alloc
<_QZ> name 5 files and they will magically appear
<ElGato> mem_alloc, mem_defrag, mem_free, pt_crt, and the persistance thingy
<ElGato> that's 4
<_QZ> bad choices
<ElGato> uh
<ElGato> im dumb
<ElGato> thats 5
* ElGato/#tunes rethinks this
<_QZ> i will be a nice guy and give u a hint, the defrags are finished yet
<ElGato> tell me all the memory management things again
<AlonzoTG> om
<_QZ> press PGUP
-:- SignOff AlonzoTG: #TUNES (changing servers, today's session: 11:21 hours. =P)
<ElGato> um
<ElGato> taht doesn't work
<ElGato> im ona terminal
<_QZ> uhh
<_QZ> well u need to get something better than windows
<_QZ> cuz i can telnet to any linux box and run bx and use pgup
* ElGato/#tunes 's computer is on it's way :D
<ElGato> probably 4 days till i get it
<ElGato> i'll look at the logs
<ElGato> ram_free, mem_alloc, mem_free, pt_crt, and the persistance thingy
05:20pm
<_QZ> ok
<_QZ> i havent yet verified the mem_* or pt_crt functions so give me a couple minutes
<_QZ> i would have picked ram_alloc over mem_alloc tho :)
<ElGato> why not skip the persistance part and i take ram_alloc in it's place?
<_QZ> if u want
<_QZ> ram_* and mem_* are almost identical :)
<ElGato> i think it's for the best
<ElGato> hmmm
* ElGato/#tunes thinks about this
<ElGato> let's see what do i already have?
<_QZ> ram allocates a free block of ram and mem allocates a specific block of memory that is not used by ram
<ElGato> i have ram free
<_QZ> u get 4 more
<_QZ> persistence was a good choice
<ElGato> yeah
<ElGato> i will take that
<ElGato> that leaves me with three more
<_QZ> ram_alloc is probably the best one to take
<ElGato> oh
<ElGato> ok
<ElGato> i will then
<ElGato> now pt_crt
<_QZ> almost done with it
<ElGato> is that the persistance part?
05:30pm
<_QZ> no
<ElGato> :(
<_QZ> page table create
<ElGato> oh
<ElGato> well i'll just reserve that one
<_QZ> u still want it?
<ElGato> i want it when it's finished
<ElGato> that leaves me with one more
<ElGato> i think i'll take mem_alloc
<_QZ> still stuck on that mem_alloc :)
<ElGato> yeah well it'll give  me some idea about mem_*
<ElGato> just send it on over
05:40pm
* ElGato/#tunes is away: (Auto-Away after 10 mins) [BX-MsgLog On]
-:- Netjoined: mccaffrey.openprojects.net koontz.openprojects.net
-:- _QZ [brand@p0wer.qzx.com] has joined #tunes
* ElGato/#tunes is back from the dead. Gone 0 hrs 9 min 59 secs
<ElGato> sorry i had to sand a stair rail
<ElGato> do send that file qz
-:- SignOff hcf: #TUNES (Leaving)
-:- hcf [nef@me-portland-us318.javanet.com] has joined #tunes
06:00pm
-:- _BC [bmcbrine@pine.Alberni.Net] has joined #tunes
<ElGato> hola
<abi> what's up, ElGato
<_BC> , has anyone seen/tried some.net irc server?
<_BC> Hi ElGato, how you doing?
<ElGato> fine thanks
* _BC/#tunes some.net is another opensource place, but i am unable to connect to the servers. just wondering if it was only me.
-:- SignOff _BC: #TUNES (I toss myself out of an airlock and try to sing....)
<hcf> i checked it a while ago
<hcf> for "some" reason i ignored it
<hcf> well, its opn now, and seemingly works
06:10pm
<ElGato> :
<ElGato> :\
<ElGato>   -
<ElGato> :/
<ElGato> ::
<ElGato> ah shit
<ElGato> i screwed that up
<hcf> im not surprised ;)
<ElGato> :D
<ElGato> oh hcf you tickle my FPU
06:20pm
* ElGato/#tunes is away: (Auto-Away after 10 mins) [BX-MsgLog On]
* ElGato/#tunes is back from the dead. Gone 0 hrs 1 min 5 secs
* ElGato/#tunes is away: (Auto-Away after 10 mins) [BX-MsgLog On]
* ElGato/#tunes is back from the dead. Gone 0 hrs 0 min 7 secs
-:- s_r [sr@phila-dialup552.nni.com] has joined #tunes
<ElGato> hello s_r
<s_r> hi
<ElGato> what's up?
-:- s_rr [s_r@phila-dialup044.nni.com] has joined #tunes
<ElGato> re: what's up?
-:- SignOff s_r: #TUNES (Ping timeout for s_r[phila-dialup552.nni.com])
<s_rr> re?
-:- s_rr is now known as s_r
<ElGato> how are you, what have you been doing?
<s_r> reading logs
<ElGato> ahhh
<s_r> :)
<hcf> s_r: hows ur hyperprogramming compiler?
<ElGato> wtf is hyperprogramming?
<hcf> abi: hyperprogramming?
<abi> hyperprogramming is where you make programs by connecting nodes in a persistent database to represent data or code, instead of using a text-based language
<ElGato> uh
<ElGato> i don't recall ever doing that
<s_r> * ElGato/#tunes just added some stuff to retro
<s_r> heeh hcf it's vaporware
<s_r> what did you expect?
<s_r> ElGato: what did you add to retro?
<ElGato> im addING FPU support
<hcf> s_r: got specs?
<s_r> hcf: i'm going to keep everything to myself until it gets finished, i don't want to raise anyone's expectations... forget that i told you, i might not finish it :P
07:00pm
<hcf> s_r: u brought it up 1st :P
<ElGato> im afraid i must leave in five minutes
<s_r> ElGato: your brother bothering you?
<ElGato> s_r: no i must go to myy tae kwon do lessons :)
<s_r> hcf: are you working on a compiler>
<s_r> ?
<hcf> s_r: no
<s_r> bah jeet kune do dude
<ElGato> uh
<hcf> huh?
<ElGato> ok
<ElGato> eres bavaso
<s_r> what does hashing do in a compiler?
<ElGato> it's not specific to a compiler
<ElGato> i may as well use it though
<ElGato> bye 
-:- SignOff ElGato: #TUNES (Oh my hairy ass!)
07:10pm
-:- kermit [kermit@rrosd28.impsat.net.ar] has joined #tunes
<kermit> hello all
<hcf> hoy kermit
<s_r> hmmm i need to learn more about compilers
<kermit> may i recommend good reading about life in other planets ?
<hcf> s_r: uv got the urls i spewed at hs right?
<s_r> hcf: would you like a good hyperprogramming compiler?
<hcf> s_r: probly not
<s_r> hcf: what language do you use most often?
<hcf> s_r: right now, perl
07:20pm
-:- SignOff kermit: #TUNES (Ping timeout for kermit[rrosd28.impsat.net.ar])
<hcf> s_r: whatcha using currently?
<s_r> compiler?
<abi> compiler is more interesting, perhaps
<hcf> pl
<s_r> tasm32
<hcf> dont like nasm?
<s_r> it's harder to make win32 binaries with nasm
<s_r> never quite figured it out
07:30pm
-:- _ruiner_ [nate@ppp327.wi.centuryinter.net] has joined #tunes
<s_r> _ruiner_: did you obtain your name from the Nine Inch Nails song of that title?
<_ruiner_> yeah....pretty much
<_ruiner_> why?
<_QZ> s_r: i gave brix source code to hs
<s_r> that's a fine CD
<s_r> cool
<s_r> QZ: well there's probably not a lot, it's far from finished, right?
<_QZ> uhh
<s_r> the OS as a whole
<_QZ> if 115 files for a total of 8800 lines of code is not alot then yes
<s_r> not the kernel
<_QZ> well i have a gui
<s_r> do you have any servers?
<_QZ> u can drag components around in it
<_QZ> servers as in drivers? ofcourse i do
07:40pm
<_ruiner_> do you like skinny puppy s_r?
<_QZ> ruin: ofcourse he does, they are easier to hold onto while doing them
<s_r> ruiner: i haven't heard much of their music but i probably would
<s_r> hahah 
<_ruiner_> cool
<_ruiner_> don't mind qz..he's just bitter...
<_ruiner_> about his os, about life
<s_r> QZ must hate me for some reason
<_QZ> abi: be s_r
<abi> qz is brix out yet? is it? come on! i need it NOW
<s_r> case in point :)
<_QZ> hate is such a strong word
<hcf> hate isnt a strong enough word
<s_r> hcf: for what QZ feels about me?
<_ruiner_> qz hates everyone....
<hcf> for what he would feel if i were him ;)
<_ruiner_> he's a lot like me in that way
<_ruiner_> hcf....play nice or I'll make you sit in the corner
<s_r> hcf: am i really that annoying?
<hcf> s_r: yes
<_ruiner_> no
<_ruiner_> don't mind them
<_ruiner_> so...._QZ...when is brix coming out? I can't wait to see it
<_ruiner_> heh heh heh
<_ruiner_> ;-p
<_QZ> hcf: hahahah
<_QZ> ruiner: next week
<_ruiner_> one would think if you were going to write a full operating system, you'd be wise enough to back up all your source code
<_QZ> my source code is backed up plenty
<_ruiner_> uhhuh
<_ruiner_> I hear you lost the whole thing when you were drunk one night
<_QZ> i put a copy on my isp, on cdr, and on the other machines here
<_QZ> ruin: that was a joke to get hs and sr off my back
<_ruiner_> ohhhh
<_ruiner_> my bad....sorry
<s_r> QZ: well i'll say that i'm looking forward to brix's release... but i won't say more
<_ruiner_> maybe you're not *QUITE* the idiot I thought you were
<_ruiner_> ;-p
<s_r> QZ: i apologize for being annoying
<_QZ> :)
<hcf> _ruiner_: most of what _QZ is a lie, some more ammusing than others
<_QZ> ya hcf is right
<_QZ> i am a habitual liar
<eihrul> qz: how do we know that's not a lie?
<hcf> s/is/says is/
<s_r> pathological liar
<_QZ> eihrul: u dont :)
<_ruiner_> the thing about liars is they tend to be ignored completely after a while
<_QZ> i call family members and relatives all the time to inform them of deaths in the family
<_ruiner_> I'd laugh if you put out brix and nobody believed you
<_QZ> ofcourse no one has ever died
<hcf> theyre more like inside jokes than lies
07:50pm
<_QZ> ruin: i have been lying for years and no one has ignored me yet :)
<s_r> QZ: do you want source code to my OS?
<_QZ> s_r: no
<_QZ> s_r: i dont look at OS source code
<eihrul> qz: liar
<s_r> i am going to pick up where the Merlin project left off
<_QZ> hahah
<s_r> i'm going to call my OS "Gandalf"
<_QZ> u cant fill jecel's shoes
<s_r> jecel < me
<hcf> cant fill his condom either
<_QZ> hahah
<s_r> hcf: you've got a warped mind
-:- SignOff eihrul: #TUNES (Ping timeout for eihrul[usr5-ppp78.lvdi.net])
<_ruiner_> hmm.....
<s_r> hey ruiner have you heard of Sister Machine Gun?
<_ruiner_> yeah...chicago band.....
<_ruiner_> named after a lyric in a skinny puppy song
<_QZ> i just uploaded a pre-release binary to the web site for anyone that cares
<_ruiner_> brb...gotta free up the line for a minute
-:- SignOff _ruiner_: #TUNES (Leaving)
<s_r> QZ: are you going to start releasing patches for it too as Microsoft releases service packs
<s_r> like BRiX Service Pack #1
<_QZ> s_r: no
<_QZ> s_r: brix has an auto-download feature so it can patch itself without u knowing
<s_r> or you could try another Microsoft practice, needlessly partitioning an OS's functionality
<s_r> BRiX Server edition
<s_r> BRiX Professional edition
<hcf> BRiX Bathroom edition
<_QZ> BRiX Professional Server edition
08:00pm
<_QZ> oh an i will give them year versions
-:- _ruiner_ [nate@ppp380.wi.centuryinter.net] has joined #tunes
<_QZ> nothing will ever be added or changed between versions, just file formats so it will force u to upgrade
<s_r> QZ: don't worry, if you don't follow microsoft tradition, i will with my OS
<_QZ> hah
<s_r> VapourOS '99
<_QZ> i do not feel that u have what it takes to write an OS, no insult intended
<s_r> QZ: you don't know me entirely and you're saying that?
<_QZ> damn these holiday weeks tire me out
<_QZ> drag my ass outta bed at 10am goto work at 11am get off at 3pm and im tired at 9pm
<_ruiner_> its safer to say it s_r
<_ruiner_> kinda a prove me wrong type thing
<s_r> eh screw QZ :)
<_QZ> ya prove me wrong
<_QZ> i am yer god biotch :)
<s_r> QZ: do you think HickServ will complete his compiler work and his OS work?
<_ruiner_> no
<_QZ> no
<_ruiner_> I don't think it'll get much further than it is now
<_QZ> i say he can write a the language for brix but that is just so he dont cry
<_QZ> i am secretly writing it behind his back :)
<_ruiner_> lol
<_QZ> i shouldnt be on irc when im tired, i let out too much
-:- eihrul [lee@usr5-ppp108.lvdi.net] has joined #Tunes
<_QZ> crap, someone modify the logs so hs dont read that
<s_r> well i guess QZ is right in a way
<s_r> i don't know enough about the intel architecture yet to write an OS from scratch
<s_r> i still have a lot to learn
<_QZ> u want the brix source to help u out?
<s_r> QZ: when it's officially released
<eihrul> qz: and not so long ago you were speaking glowingly of the merits of it's grokkability :)
<_QZ> eihrul: eh?
<hcf> eihrul: only 1 'k' on that ;)
<s_r> well i'm spending enough time already looking at Solaris source, and linux source, freebsd, etc
<hcf> s/on/in/
<eihrul> grokability doesn't seem right for some reason
<eihrul> it looks like you'd pronounce that groakability
08:10pm
<_QZ> eihrul: eh?
<eihrul> qz: something to the effect of: "i don't care if others will have trouble reading it"
<_QZ> oh i dont care
<_QZ> that doesnt mean it is unreadable
<eihrul> i'm just picking nits, don't mind me
<_QZ> it is laid out quite nicely
<_QZ> s_r: keep up man
<_QZ> im gonna start killing the sends :)
<s_r> why the f*** do you have everything in .incs?
<_QZ> why ask why
<s_r> @$@!
<s_r> stop sending files
<_QZ> what dont u want brix source code?
<_ruiner_> write your own code s_r
<_ruiner_> that underscore is pissing me off....I'm gonna call you sir from now on
<_QZ> ruiner: hahah that underscore pissing everyone off
<hcf> _ruiner_: nick completion broke?
<hcf> _QZ: not me
<s_r> ruiner: i have a lot to learn first
<s_r> hahah
<s_r> heh look who's talking _QZ
<_QZ> sr: my underscore is BEFORE my nick
<eihrul> sr: doing is the process of learning
<_QZ> eihrul: stealing is the process of getting ahead
<_ruiner_> call up intel....they'll send you stuff on the x86 architecture
<eihrul> qz: and redundancy is the process of stating the obvious
<_ruiner_> free even
<s_r> i need to learn to do, and thus i need to learn more
<_QZ> look at all those idle dcc's that are being closed
<eihrul> ruiner: you can download stuff free from the intel site... nice manuals they are too
<hcf> abi: nickometer for s_r
<abi> 's_r' is 14% lame, hcf
<_QZ> u learn coding by example, and u get examples by stealing them
<_ruiner_> blah blah blah
<s_r> abi: nickometer for hcf
<abi> 'hcf' is 0% lame, s_r
<_ruiner_> only if you're a pud
<s_r> abi: nickometer for _QZ
<abi> '_QZ' is 23% lame, s_r
<_QZ> ha
<hcf> abi: nickometer for _ruiner_
<abi> '_ruiner_' is 26% lame, hcf
<_ruiner_> I think somebody is playing with the results....
<hcf> _ruiner_: nope
<_QZ> ruiner: :)
-:- _ruiner_ is now known as billgates
<billgates> suck on that
<_QZ> abi: nickometer for fare
<abi> 'fare' is 0% lame, _QZ
<_QZ> yep someone is playing with the results
<_QZ> abi: nickometer for tril
<abi> 'tril' is 0% lame, _QZ
<billgates> abi: nickometer for billgates
<abi> 'billgates' is 0% lame, billgates
<_QZ> the top 3 seem to have no lameness
-:- billgates is now known as ruiner
<hcf> http://www.cs.cmu.edu/~infobot/misc/nickometer.pl
08:20pm
<_QZ> s_r is lame
<ruiner> yeah well....look at what it came up with for billyboy....
<_QZ> ruiner: it goes by nicks it has seen
<_QZ> i think u and i had it out
<hcf> _QZ: no
<ruiner> bah!
<_QZ> that is why our nicks are lame
<hcf> it goes by percentage of annoying chars in nicks
<ruiner> ah....
<ruiner> no....
<ruiner> qz would be 33 then
<ruiner> as would sr
<hcf> abi: nickometer for ruiner
<abi> 'ruiner' is 0% lame, hcf
<_QZ> and ruiner should be 66
<ruiner> abi: nickometer for ruiner
<abi> 'ruiner' is 0% lame, ruiner
<ruiner> hmmmmmm.....
<hcf> see no annoying chars, 0% lame
<_QZ> so one underscore is 14 but 2 are 13 each?
<_QZ> or is it more lame to have it in the middle
<hcf> placement matters, also case
-:- _QZ is now known as queuezee
<hcf> placement doesnt matters
<hcf> abi: nickometer for _sr
<abi> '_sr' is 14% lame, hcf
<hcf> abi: nickometer for s_r
<abi> 's_r' is 14% lame, hcf
<hcf> abi: nickometer for sr_
<abi> 'sr_' is 14% lame, hcf
08:30pm
-:- ElGato [dave@bespin.dhs.org] has joined #tunes
<queuezee> elgato
<ElGato> hi
<abi> hoy, ElGato
<queuezee> i just barely pressed the send button
<s_r> hi ElGato
<s_r> how goes work on the tokenizer?
<queuezee> ElGato: read yer email
<ElGato> k
<ElGato> it hasn't yet recieved anything
<ElGato> s_r: i noticed what i have right now sucks so i easrased it
<ElGato> which email address did you send it to?
<ElGato> oh
<s_r> ElGato: what's wrong with it? what are you going to do now?
<ElGato> send all other mail the dave@bespin.cx or dave@tunes.org
<queuezee> k
<ElGato> s_r: i don't know at what point to started recording words in the file
<s_r> hey what about dave@persiankitty.com
<ElGato> i could tokenize directly from the file without recoreding each word
<s_r> and dave@sexsexsex.com?
<ElGato> uh no
<ElGato> i really want people to see progress 
<ElGato> but i don't know completely how to go about this
<s_r> just read in a portion of the data into a buffer...
<s_r> tokenize it and put it in anotehr buffer
<s_r> do the same and put it into a diff. buffer
<queuezee> ElGato: what are u trying to tokenize?
<s_r> and send it to the asm generator
<ElGato> im tokenizing a text file
<queuezee> what language
<ElGato> lengua
<queuezee> lengua?
<abi> lengua is, like, a pure, fully oo, SEXP language by hickserv/elgato
<s_r> HickServ|elgato
<s_r> sort of like route|daemon9?
08:40pm
<ElGato> well i guess i could start storing in the scanner but im not all that knowledgeable about doing dynamic memory allocation
<s_r> ElGato: you want to store the text words?
<hcf> ElGato: have u looked at other scanners?
<ElGato> yes (whatever isn't whitespace)
* ElGato/#tunes ignores the comment made by hcf
<queuezee> show me the syntax
<ElGato> (<function> <argument> <argument> ...)
<queuezee> lisp?
<abi> i guess lisp is ((simply) (perfect))
<queuezee> if its OO then how do u handle objects
<ElGato> im taking non whitespace characters out of the file until i run into more whitespace but i can't make a dynamic sized array to put it in when i don't yet know the size of the array
<ElGato> each word should be able to be as large as it wishes
<hcf> ElGato: words have wishes?
* ElGato/#tunes is for word rights
<ElGato> no
<ElGato> the programmer
<ElGato> she can make words as  large as she wishes
<queuezee> everyone knows that chicks cant code so u can just use 'he' :)
<hcf> size should matter to her ;)
<queuezee> hcf: haha
<ElGato> uh
<ElGato> it's just my style
<ElGato> and i know a chick that's a good programmer
<ElGato> so shut your sexist hole
<queuezee> this nick is too annoying
-:- queuezee is now known as liar
<liar> ElGato: has she always been a chick?
* hcf/#tunes is only here to make _QZ laugh and poke fun at hickserv
<ElGato> i believe so :
<ElGato> D
<ElGato> :D
<liar> did u have sex with her before or after her surgery?
<ElGato> i've never had sex with her
<liar> so how do u know?
<ElGato> she's way older than me
08:50pm
<ElGato> and i don't think any girl would want to have sex with me anyway
<hcf> ok, lets get off her and back to business
<ElGato> really my only problem is with how i will do dynamic memory allocation
<liar> ElGato: in C?
<ElGato> yes
<s_r> <ElGato> and i don't think any girl would want to have sex with me anyway
<s_r> hhaha
<ElGato> it's true :(
<ElGato> well i reckon i must leave again
<liar> do u look like el gato after it got run over?
-:- SignOff ElGato: #TUNES (You can parse my tokens any time baby!)
09:00pm
-:- Mr_Wrong [seanl@adsl-63-192-213-146.dsl.snfc21.pacbell.net] has joined #tunes
<ruiner> I'm bored
<eihrul> it's a conspiracy then, because so am i
<ruiner> don't we need 3 for a conspiracy?
<ruiner> isn't 2 coincidence, 3 conspiracy?
<s_r> eihrul you're bored? but you have so much coding to do
<s_r> !
<eihrul> sr: i can barely stack quite awake enough to do it... on hour 28 right now
09:10pm
<eihrul> it's the "i'm going to pass out soon anyway so i don't feel like doing anything right now" kind of boredom
<eihrul> s/stack/stay
<ruiner> hey sr...got that new nin single?
<s_r> yea
<ruiner> send it to me?
<s_r> ok
<liar> send it to me too
09:20pm
<s_r> get it from ruiner
<s_r> i'm on a 2600 baud modem
<ruiner> yeah.....I can tell
<liar> so, im on a 300 baud modem
<ruiner> I could send you nothing compares 2 u by sinead o connor if you want qz
<liar> uhh, no thanks
<ruiner> you're sure?
<ruiner> its a killer song.....written by prince no less
<s_r> that's a good song
<liar> what else u got?
<ruiner> bowie/reznor I'm afraid of americans
<ruiner> ummmm.....
<ruiner> a couple wierd manson remixes
<s_r> depeche mode = the best
<ruiner> depeche mode sucks
<s_r> why is this send so slow?
<ruiner> oh.....that reminds me...I've got a few kmfdm songs
<s_r> ruiner: hmm i'm shocked to hear that
<liar> do u guys know about audiofind?
<s_r> ruiner: what don't you like about them?
<ruiner> listen to the lyrics
<liar> ya depeche mode sucks
<s_r> ruiner: let me send you a better song than this NIN song
<ruiner> no
<s_r> i have one that you might like (it's not depeche mode)
<liar> s_r: send it to me damnit
<ruiner> don't even dare think of disconnecting this
<liar> what song?"
<s_r> do you like Nation of Fear
<ruiner> no
<s_r> i like this one Nation of Fear song
<s_r> it sounds nin-ish
<ruiner> yeah, except its crap
<liar> s_r: what nin song u sending him?
<s_r> day the world went away
<s_r> bah you people have odd standards
<ruiner> I don't tolerate rip off bands
<ruiner> nor do I tolerate career opprotunists
<s_r> wtf is this so slow
09:30pm
<s_r> ruiner: do you like Kid rock?
<ruiner> no clue.....I figured it was on your side
<ruiner> oh hell no
<ruiner> I hate him
<s_r> do you like 16 volt?
<ruiner> all kid rock lyrics are is a bunch of sentence fragments
<eihrul> eh, and that's just limited to kid rock? goes way beyond him...
<ruiner> true
<s_r> ruiner: what music do you like besides NIN?
<liar> rage
<ruiner> I like new age.....especially if there are wolves howling in the background
<ruiner> or crickets.....crickets rule
<liar> uhh, u got problems
<s_r> liar: do you still use LiNuKs?
09:40pm
<liar> [BitchX-74a14 by panasync][Linux 2.2.7]
<s_r> [BitchX-75p1 by panasync][BeOS R4.5]
<ruiner> bitchx
<liar> hmm i think i know why netscape and bx have the wrong time
<ruiner> heh heh heh
<liar> they are both libc5 apps and im running them on glibc2
<liar> s_r: that is amazing that u modified yer sv to have output identical to mine :)
<s_r> liar: i love be
<s_r> i'm a bedope
<s_r> hehehe
<liar> whats more amazing is that u have never before seen my sv
<liar> atleast u had enuf brains to disable version :)
<liar> oops i spoke too soon
<liar> === CTCP VERSION reply from s_r : mIRC32 v5.51 K.Mardam-Bey
<s_r> liar so?
<s_r> liar i was kidding about the sv
<liar> NO really? :)
<s_r> heh
<s_r> liar do you think i should switch to BeOS?
<liar> no
<liar> its not ready
<liar> and its stupid IDE kept crashing on me
<s_r> want source code to commercial OSes?
<liar> uhh
09:50pm
<liar> 91meg? i think i'll pass
<s_r> hehe i was kidding
<s_r> i don't have the source
<liar> damnit, i just found a bad hard to fix bug in brix
<s_r> describe it
<liar> in the kernel init before paging has been setup i make calls to pt_crt and map_method
<liar> but both of those functions disable and re-enable paging
<liar> the disable part is ok but when it tries to re-enable it will crash
<s_r> why does it crash when it re-enables it?
<liar> oh wait
<liar> nevermind
<s_r> are you running it through Bochs?
<liar> i forgot i read cr0, save its contents and then clear the bit to disable. when i re-enable i just copy the saved contents back into it
<liar> so it will never change cr0
<liar> yes i use bochs
<s_r> what kind of servers have you written for BRiX?
<liar> bochs is telling me that im issuing an HLT but the only hlt is never executed
<liar> ftp, telnet, http
<s_r> you need a NIC server first
-:- SignOff hcf: #TUNES (Leaving)
<liar> i have ne200
<liar> er ne2000
<liar> did u write my tcp/ip stack?
<s_r> hell no
<liar> why?
-:- hcf [nef@me-portland-us318.javanet.com] has joined #tunes
<s_r> isn't it already finished?
<s_r> considering that you have ftp,telnet,http?
<liar> no
<s_r> well i *might* do it once brix is released
<s_r> don't count on me
<liar> it will be really easy
<hcf> darn, im sure he was about to count on u
<liar> tcp/ip is just a storage driver to brix
<s_r> hehe hcf
10:00pm
<liar> hcf: :)
<liar> u just need a load and store function and a function that the ne2000 driver can call
<liar> 3 very simple functions
<s_r> when you run brix through bochs, is it fast?
<eihrul> li: hrmm, bochs is free?
<liar> it boots at the same speed that it does on my real machines
<liar> eihrul: yes
<liar> bochs?
<abi> hmmm... bochs is a PC emulator at http://www.bochs.com that runs on win32, un*x, beos, mac, os/2 and amiga
<s_r> is it fast on your real machines?
<eihrul> li: url?
<liar> s_r: ya
<eihrul> ah, definitely useful then
<liar> the speed that bochs looses with emu is made up by the fact that is gets the image from teh hdd
<liar> eihrul: it also has a built in debugger
<liar> if yer not using bochs then yer doing it the hard way
<eihrul> yeah, definitely :)
<eihrul> rebooting over and over gets kind of annoying
<liar> u can step thru the code line by line and change registers and memory content
<s_r> it's 1:12 AM
<liar> s_r: so?
<s_r> damn it's late
<s_r> liar what time is it where you live?
<liar> s_r: do u not know howto to get my time
10:10pm
<liar> guess where i live?
<s_r> california?
<abi> hmmm... california is nice Beholder
<liar> what time did it give u?
<s_r> 5 something
<liar> so its 1am where u live on the east coast but its 5am here in california?
-:- BlackPhoenix [black-phoe@adm.univd.kharkov.ua] has joined #tunes
<liar> how old did u say u were? and how many grades have u been held back in?
<s_r> eheh 
<hcf> hoy BlackPhoenix
<BlackPhoenix> hi hcf
<liar> s_r: i live in corvo flores
<s_r> i have it mixed up
<s_r> you must be in europe
<liar> no
<liar> corvo flores
<liar> its an island west of portugal
10:20pm
* s_r/#tunes is moving to corvo flores
<liar> uhh
<liar> why would u wanna move here?
<s_r> sounds nice
10:30pm
<liar> im part of an army ranger platoon stationed here
10:40pm
<eihrul> finally got bochs working...
<eihrul> tis cool
10:50pm
<liar> did u get the version with dynamic translation?
<liar> something 13 i think
<eihrul> i just got whatever was dated july 18 :)
<liar> did u compile in dynamic translation and debugger?
<eihrul> probably not if it had to be done explicitly
<eihrul> i'll figure it out after this
<liar> u have to tell configure to use them
<liar> type: ./configure --help
<eihrul> well let me get this microkernel to actually compile first :)
<liar> i couldnt get 13 to compile with debugger
<liar> i had to use 12 and it doesnt have dynamic translation
11:00pm
-:- SignOff Mr_Wrong: #TUNES (Read error to Mr_Wrong[adsl-63-192-213-146.dsl.snfc21.pacbell.net]: EOF from client)
-:- Mr_Wrong [seanl@adsl-63-192-213-146.dsl.snfc21.pacbell.net] has joined #tunes
<liar> setup=2702,  kernel=8162
<liar> its not crashing :)
11:10pm
<ruiner> this song is taking forever
<ruiner> I started a manson one half an hour ago and its done already
<liar> heh
<BlackPhoenix> heh (backward)
<liar> hmm i dont feel like i did any work on brix today but i feel like i got so much working :)
<ruiner> I feel like shit
11:40pm
<liar> u smeel the same way too :)
<ruiner> talking to an ex girlfriend will do that though....I guess
<liar> smell
<liar> im going to sleep
<liar> cya
-:- SignOff liar: #TUNES (BRiX [http://www.qzx.com/brix] :: sleep)
-:- SignOff BlackPhoenix: #TUNES (Leaving)
11:50pm
[msg(TUNES)] newlog 1999.0722
IRC log ended Thu Jul 22 00:00:00 1999