IRC log started Mon Oct 11 00:00:00 1999
[msg(TUNES)] permlog 1999.1011
-:- SignOff air: #TUNES (BRiX [http://www.qzx.com/brix] :: sleep)
-:- shalzac [shalzac@pc-1925326.unisa.ac.za] has joined #tunes
-:- fare [fare@quatramaran.ens.fr] has joined #Tunes
<fare> forum
03:40am
-:- SignOff shalzac: #TUNES (Ping timeout for shalzac[pc-1925326.unisa.ac.za])
-:- shalzac [shalzac@pc-1925326.unisa.ac.za] has joined #tunes
-:- SignOff shalzac: #TUNES (Ping timeout for shalzac[pc-1925326.unisa.ac.za])
-:- shalzac [shalzac@pc-1925326.unisa.ac.za] has joined #tunes
-:- td [x@1Cust48.tnt1.wilmington.nc.da.uu.net] has joined #tunes
-:- fire [no@209-68-229-50.dialup.cust.tfb.com] has joined #tunes
-:- SignOff fare: #TUNES (Ping timeout for fare[quatramaran.ens.fr])
-:- fare [fare@quatramaran.ens.fr] has joined #Tunes
-:- SignOff fire: #TUNES (Leaving)
-:- SignOff shalzac: #TUNES (Ping timeout for shalzac[pc-1925326.unisa.ac.za])
-:- SignOff td: #TUNES (td has no reason)
-:- shalzac [shalzac@pc-1925326.unisa.ac.za] has joined #tunes
-:- SignOff Syntax: #TUNES (Ping timeout for Syntax[01-032.006.popsite.net])
-:- SignOff smkl: #TUNES (Ping timeout for smkl[glubimox.yok.utu.fi])
-:- smkl [sami@glubimox.yok.utu.fi] has joined #tunes
-:- smoke_ [tw026024@zaalf13.twi.tudelft.nl] has joined #tunes
<smoke_> hello
<fare> ole
-:- SignOff smoke_: #TUNES (BitchX-75+Deb1an -- just do it.)
08:40am
-:- Syntax [dma@11-143.006.popsite.net] has joined #Tunes
-:- hcf [nef@me-portland-us935.javanet.com] has joined #tunes
-:- SignOff hcf: #TUNES (Leaving)
-:- Kaufmann [Kaufmann@dial877.infolink.com.br] has joined #tunes
>>> Kaufmann [Kaufmann@dial877.infolink.com.br] requested PING 939669118 from #tunes
<Kaufmann> Shalom!
-:- SignOff Kaufmann: #TUNES ("Jesus would make a good main character for a Seinfeld spinoff" - Amber)
-:- NetSplit: sterling.openprojects.net split from devlin.openprojects.net [12:16pm]
-:- BitchX+Deb1an: Press Ctrl-F to see who left Ctrl-E to change to [sterling.openprojects.net]
-:- Netjoined: sterling.openprojects.net devlin.openprojects.net
-:- Tril [dem@bespin.dhs.org] has joined #Tunes
-:- abi [nef@bespin.dhs.org] has joined #Tunes
-:- Kaufmann [Kaufmann@dial154.infolink.com.br] has joined #Tunes
-:- ServerMode/#Tunes [+o Tril] by sterling.openprojects.net
-:- SignOff Kaufmann: #TUNES ("Jesus would make a good main character for a Seinfeld spinoff" - Amber)
-:- Kaufmann [Kaufmann@dial154.infolink.com.br] has joined #tunes
-:- smkl [sami@glubimox.yok.utu.fi] has joined #tunes
>>> Kaufmann [Kaufmann@dial154.infolink.com.br] requested PING 939669556 from #tunes
<Kaufmann> Anyone care to explain monads to me?
12:20pm
<smkl> they are too hard to explain
<Kaufmann> how friggin' fun.
<Kaufmann> well, can you give it a shot?
-:- Linus_ [andy@pm615-03.dialip.mich.net] has joined #tunes
<smkl> well you just have two operations on the monad, >> and return
<Kaufmann> Hey Linus
<Linus_> Kaufmann: yeah, hey
<Kaufmann> smkl, start from the beginning, please?
<smkl> then you can do liek "read >> \x -> read >> \y -> return (x, y)"
<Kaufmann> Well, not the absolute beginning, as I'm not a complete moron...
<smkl> read has type "IO string" here
<Kaufmann> Wha'?
<smkl> IO is the monad or something
<Kaufmann> ?!?
<Kaufmann> I am the walrus?
<smkl> the monad represents a computation
<Kaufmann> oic
<smkl> then you can bind the computations with >>
<Kaufmann> Now you've lost me
<smkl> return makes the computation to return a value
<smkl> like in my example, you make several computations into one using >> (bind)
<smkl> in haskell it's actually >>=
<Kaufmann> So what has actually happened here?
<smkl> well you make a computation that return x and y that are read from input
<smkl> then you could also execute the computation
<Kaufmann> aaaaaaaaaaaaaah
12:30pm
<Kaufmann> now it all makes sense :)
<Kaufmann> it's still a bit hazy... so the monad represents what, how x and y acquire their value?
>>> Kaufmann [Kaufmann@dial154.infolink.com.br] requested PING 939670363 from #tunes
<smkl> they get their values from computation "read"
<Linus_> Kaufmann: what was that for?
<smkl> if you want to get really confused, i could explain how to implement all this
<Kaufmann> I'm sorry, I'll disconnect for a second
-:- SignOff Kaufmann: #TUNES ("Jesus would make a good main character for a Seinfeld spinoff" - Amber)
-:- td [x@1Cust223.tnt3.wilmington.nc.da.uu.net] has joined #tunes
12:40pm
-:- Kaufmann [Kaufmann@dial212.infolink.com.br] has joined #tunes
* Kaufmann/#tunes hath returned
<Kaufmann> Sorry for leaving... had to disconnect
<Kaufmann> smkl, as you were saying...?
<smkl> <> they get their values from computation "read"
<Kaufmann> they... x and y?
<smkl> yes
<Kaufmann> but how does that happen? read is a primitive?
<smkl> it is a compation in monad IO
<smkl> err computation
<Kaufmann> I found a intro to monads at haskell.org...
<smkl> most likely a primitive computation
<fare> hum
<fare> Gakuk!
<smkl> hello fare
<fare> ole, smkl
-:- fare is now known as Fare
<Kaufmann> good morning|afternoon|night, Far�
12:50pm
-:- change_me [user2369@biznaga.entropia.com.mx] has joined #tunes
<change_me> .
-:- change_me [user2369@biznaga.entropia.com.mx] has left #tunes []
<Kaufmann> I gotta admit, that's a cool nick
<Kaufmann> lots of double-entendres\
<Kaufmann> s/\\//;
-:- m [user1773@biznaga.entropia.com.mx] has joined #tunes
<Fare> K: night
<Fare> K: which nick?
<Fare> m: hello
-:- SignOff m: #TUNES (Read error to m[biznaga.entropia.com.mx]: Connection reset by peer)
<Kaufmann> change_me
01:00pm
-:- hcf [nef@me-portland-us1038.javanet.com] has joined #tunes
<Kaufmann> hi hcf
01:10pm
<hcf> lo
-:- Kaufmann is now known as KaufmannBRB
-:- KaufmannBRB is now known as Kaufmann
01:20pm
-:- Linus_ [andy@pm615-03.dialip.mich.net] has left #tunes []
-:- SignOff Kaufmann: #TUNES (Ping timeout for Kaufmann[dial212.infolink.com.br])
-:- Kaufmann` [Kaufmann@dial623.infolink.com.br] has joined #tunes
<Kaufmann`> BLOODY HELL!!!!!
-:- Kaufmann` is now known as Kaufmann
<Kaufmann> As I was saying... anybody know how to print a fraction as a flonum in MzScheme?
<Fare> (display (+ x 0.0))
<Kaufmann> Good idea... I was thinking of something of the sort
<Fare> factor that code properly; also, you may want early type conversion...
01:30pm
<Kaufmann> naw, it's alright
01:40pm
-:- _ruiner_ [nate@ppp054.wi.centurytel.net] has joined #tunes
<Kaufmann> Well, I'll be leaving now...
-:- SignOff _ruiner_: #TUNES (Leaving)
02:00pm
-:- SignOff Kaufmann: #TUNES ("Jesus would make a good main character for a Seinfeld spinoff" - Amber)
-:- Beholder [me@ppp-045.m2-1.sub.ican.net] has joined #tunes
<hcf> hi beh
<Beholder> Hey hcf
<Beholder> hcf: So how goes your day today?
<hcf> fine
* Beholder/#tunes just finished a huge thanksgiving supper...
<hcf> hows AKOS?
<Beholder> hcf: Going quite well now :)  I'm working on a server right now.  Prototyping it in Python.  
<hcf> hows UniOS?
<Beholder> UniOS is not so good...  
-:- SignOff Syntax: #TUNES (Ping timeout for Syntax[11-143.006.popsite.net])
<Beholder> Actually I found a new UniOS project at http://www.unios.org
<Beholder> It's some commercial project... doesn't sound interesting
<Beholder> I'm thinking of telling them the names been used for an OS project already and is copyrighted.  But I'm not one to stir crap.
02:40pm
<hcf> better to stir it now then to have it hit the fan later
<Beholder> ofc, if they make any money... hehe... then I'll stir crap :)
<Beholder> hcf: How goes your efforts?
<hcf> which efforts?
<Beholder> hcf: I'm not sure :)  But I think remember you were working on something for Tunes
<hcf> i'm currently effort-less
<Beholder> PHP3, MySQL web page?
<hcf> afaik
<Beholder> Ahh :)
<Beholder> It's a good place to be :)
<Beholder> My efforts lately have been centred mostly around school work.  Of which there seems to be no end.  But I had some time lately to work on AKOS link layer.  
<Beholder> I'd really like to get a working prototype up asap, so I can get some feedback on the game.  But as always, there are low level, mostly dull things to finish first.
02:50pm
<Fare> hum
<Beholder> Hey Fare :)
<Fare> :(
<Beholder> :(
03:00pm
-:- SignOff Beholder: #TUNES (Ping timeout for Beholder[ppp-045.m2-1.sub.ican.net])
-:- eihrul [eihrul@usr5-ppp158.lvdi.net] has joined #tunes
-:- SignOff td: #TUNES (Ping timeout for td[1Cust223.tnt3.wilmington.nc.da.uu.net])
-:- SignOff eihrul: #TUNES (Ping timeout for eihrul[usr5-ppp158.lvdi.net])
-:- eihrul [eihrul@usr5-ppp158.lvdi.net] has joined #tunes
-:- Kaufmann [Kaufmann@dial835.infolink.com.br] has joined #tunes
-:- SignOff hcf: #TUNES (Ping timeout for hcf[me-portland-us1038.javanet.com])
-:- SignOff abi: #TUNES (Ping timeout for abi[bespin.dhs.org])
-:- abi [nef@bespin.dhs.org] has joined #tunes
-:- hcf [nef@me-portland-us1038.javanet.com] has joined #tunes
-:- NetSplit: varley.openprojects.net split from sterling.openprojects.net [04:16pm]
-:- BitchX+Deb1an: Press Ctrl-F to see who left Ctrl-E to change to [varley.openprojects.net]
-:- Netjoined: varley.openprojects.net sterling.openprojects.net
-:- abi [nef@bespin.dhs.org] has joined #Tunes
-:- hcf [nef@me-portland-us1038.javanet.com] has joined #Tunes
<Kaufmann> So...........
-:- lar1 [LarMan@dialup-209.244.106.15.SanJose1.Level3.net] has joined #tunes
<lar1> Hey
<lar1> <Flame sheild>
04:30pm
<Kaufmann> *ducks*
<lar1> I was reading the logs for today
<lar1> Doesn't anyone talk about aything?
<hcf> nope, never
<Kaufmann> it's forbidden... by decree from the Pope-Emperor
04:40pm
<lar1> :)
<lar1> asymetric messiging is when a thread is blocked until the data transaction occurs, right?
<lar1> And symetric is when thread aren't blocked?
<Tril> lar1: that would be the other way around
<lar1> Tril: Huh?  In asymetric one thread has to recive and the other has to send, right?
<Kaufmann> Isn't it synchronous/asynchronous?
<Kaufmann> lar, Tril is right
<lar1> I was just reading a QNX faq and it confuesd me... I am still trying to decide on what to do about IPC
<lar1> So how does messaging without blocking work (asynchronous??)
<hcf> lar1: do any web design?
<lar1> hcf:  Like HTML?
<lar1> hcf:  Perhaps the better question would be, "Of what type?"
<Tril> in unix async works in this way: you make a system call and specify in the arguments that it is async. Then after it is done you are sent a signal.
<Tril> If you are actually doing threads programming, you should have a thread setup to handle signals, which makes simpler design.
<lar1> How is a thread alerted that it has a signal?
-:- IMPO [ihatebob@p05-06.hartford.dialin.ntplx.com] has joined #tunes
<Tril> (ie. dont use signal handlers when doing threads programming... do sigwait() instead)
<hcf> lar1: beginner-compatible free/shareware progs for win98 for website design (wysiwyg html editing)
<Tril> lar1:  have a separate thread in a loop that just calls sigwait() to wait on all signals..and have that thread handle all the various ones
<Kaufmann> Aiiirrrrgh... 'serious' UNIX programming annoys the hell out of me
<Tril> In non-thread programming, you setup a special function to be invoked when a signal is received (interrupting the currently running function, and possibly causing all sorts of race conditions)
<lar1> Tril: Oh hey!  That makes sense!
<lar1> Tril:  So you have a thread that sits and waits for a signal and then responds by its self?
<lar1> hcf:  Hmm, I used to use those a long time agao... why?
<Tril> responds, or does whatever appropriate action
<hcf> lar1: whatcha use now?
04:50pm
<lar1> hcf: <Notepad.exe|Edit.exe|Vi|Pico> :)
<lar1> Tril:  So in a sense it is an event driven system?
<hcf> lar1: k, nm
<Tril> which sense? I'm not familiar with the buzzword Event-Driven
-:- SignOff hcf: #TUNES (Leaving)
<Kaufmann> Event-driven, b. A system that sits around in an infinite loop doing nothing until something (called an 'event') happens, in which case it handles the aforementioned event and goes back to its perpetual state of loopiness.
<lar1> Tril:  Like in a GUI, where buttons are set up and stuff and the the program just sits and waits for the user to do somthing
<lar1> loopiness :)
<Tril> well that one thread would be even driven i guess. 
<lar1> Would the signal catching thread ever need to communicate with the thread that spawned it?
<Tril> that depends entirely on the application
<lar1> Persay it did... how would it go about doing so?
<Tril> since threads run in a common address space, they find it easiest to talk just by reading/writing to a variable in memory
<lar1> But the thread reciving must know to check the varible?
<Kaufmann> lar, you have it do so
<lar1> Hmm, ok
<lar1> So if messages are handled by a spawned thread, what does the main thread do?
05:00pm
<Tril> if you dont have something else to do, then you don't even need threads.
<lar1> Good point :)
<lar1> So if the main thread wants to render HTML while reciving image.gif, it creates a thread to retreive image.gif and pass a signal to the main thread when its done.  And the main thread will go loopy until it gets the other threads signal when it finsishs rendering?
<Kaufmann> LOL
<Kaufmann> I'm glad to see my idiom has been widely accepted
<lar1> You don't like my cheesy example?
<Tril> s/pass a signal/write a value into a variable/ .. main can just wait for  the variable WHEN ITS DONE RENDERING 
<lar1> Thats what I attempted to say... :)
<lar1> Kaufmann:  Your idiom being?
<Kaufmann> lar, 'go loopy' :)
<lar1> Kaufmann:  Oh... I like it! :)  Loopy loopyness... very cool terms :)
<lar1> Ok, I belive I now grok this async messaging enough to start implementation
* lar1/#tunes ducks
<Kaufmann> Maybe I ought to submit it to the Jargon File :)
<lar1> Kaufmann:  You should
<Kaufmann> Speaking of grokking, I'm reading _Stranger in a strange land_
<lar1> How is that?
<lar1> Tril:  Are system calls best done via software interupt?
05:10pm
<Kaufmann> Robert Heinlein's sci-fi novel... it's where the verb 'to grok' comes from
<Tril> you mean, if yuou are designing a kernel?
<lar1> Tril:  Yes
<lar1> Kaufmann:  Oh, cool cool!  I'm gonna have to pick that one up
<Kaufmann> abi, kernel?
<abi> kaufmann: i don't know
<Kaufmann> darn
<Tril> lar1: depends on the kernel design
<lar1> Tril:  Monolithic
<lar1> abi: Kernel?
<abi> Kernel is the stuff at the bottom of the popcorn bag
-:- fire [no@209-68-229-146.dialup.cust.tfb.com] has joined #tunes
<lar1> :)
<fire> :D
<Tril> you wanna make a traditional kernel? like linux? system calls through software interrupt? Well there is hardly a faster way.
<lar1> I thought Linux was based on a uK?
<Kaufmann> lar, not at all
<Tril> no, not that i know of.  HURD is based on uK
<Kaufmann> Linus is Finnish
<Kaufmann> *ducks*
<fire> heh
<Kaufmann> Bad pun, I know
<fire> my spanish teacher is finnish 
<abi> i already had it that way, fire.
<fire> she's an air head :D
<lar1> Oy vey
<lar1> Whats the difference between Monolithic and uK again... I got confused
<Kaufmann> Well, /my/ extended ASCII supports the letter mu
<Kaufmann> Here, �K
<Kaufmann> What do you see that as?
<lar1> The greek u thing
<Kaufmann> Good!
<fire>  � is for micro :D
<lar1> �K is the letter mu?
<Kaufmann> lar, yep
<fire> �m
<Kaufmann> LOL fire
<Kaufmann> lar, � is the letter mu
<lar1> ok
<Kaufmann> � is delta
<lar1> Tril:  How are system calls best done in a uK?
<fire> messaginh
<lar1> Kaufmann:  that I can't see
<fire> er messaging
<lar1> How is that <better|worse>?
<fire> beats me :)
<Kaufmann> � => beta; � => sigma; � => phi; � => pi; � => weird-pi (also known as the paragraph symbol)
05:20pm
-:- Connection closed from irc.us.openprojects.net: Success
-:- Connecting to port 6667 of server irc.us.openprojects.net [refnum 0]
-:- BitchX+Deb1an: For more information about BitchX type /about
-:- Welcome to the Internet Relay Network TUNES (from adams.openprojects.net)
-:- Your host is adams.openprojects.net, running version u2.10.05.18.(ipcheck4-5) (from adams.openprojects.net)
-:- This server was cobbled together Wed Apr 28 1999 at 13 24:47 EDT(from adams.openprojects.net)
-:- adams.openprojects.net u2.10.05.18.(ipcheck4-5) dioswkfcg biklmnopstv 
-:- [local users on irc(33)] 6%
-:- [global users on irc(239)] 46%
-:- [invisible users on irc(283)] 54%
-:- [ircops on irc(13)] 2%
-:- [total users on irc(522)]
-:- [unknown connections(0)]
-:- [total servers on irc(33)] (avg. 15 users per server)
-:- [total channels created(140)] (avg. 3 users per channel)
!adams.openprojects.net Highest connection count: 49 (48 clients)
!adams.openprojects.net Welcome to Open Projects!  You are on 5 ca 1(2) ft 14(14) tr.
-:- Mode change [+f] for user TUNES
-:- Mode change [+iws] for user TUNES
-:- JOIN activated by "TUNES #tunes tunes@bespin.dhs.org "
-:- TUNES [tunes@bespin.dhs.org] has joined #tunes
-:- Topic for #TUNES: Reflective Computing System - http://www.tunes.org/
-:- topic set by IMPO [Sat Oct  9 15:59:37 1999]
-:- [Users(#Tunes:11)] 
[ TUNES     ] [ fire      ] [ IMPO      ] [ lar1      ] [ abi       ] 
[ Kaufmann  ] [ eihrul    ] [ smkl      ] [@Tril      ] [ shalzac   ] 
[ Fare      ] 
-:- Channel #Tunes was created at Sun Feb 28 08:48:06 1999
-:- BitchX+Deb1an: Join to #tunes was synced in 7.195 secs!!
<fire> like im writing an interpreter and the programmer inputs from a prompt
-:- Mode change [-s] for user TUNES
<fire> i must first create a buffer
<Kaufmann> writing an interpreter... in C.
<fire> then the programmer inputs into the buffer
<lar1> For log:  Please note a 6 message hole exists in this log due to a ping timout for the logger
<Kaufmann> How painful^H^H^H^H^H^H^H interesting.
* lar1/#tunes ducks
<lar1> Heh
<fire> no i just need to bootstrap it
<fire> i am writing it in itself
<fire> For log: i said some naughty things about fare ;)
<Tril> for input I always use fgets and you specify the size to read...
<fire> what im saying is that size is static right?
<fire> Kaufmann: well if i new more scheme i'd do it in scheme
<fire> knew rather
<lar1> Where can I find a document comparing the pros and cons of Monolithic and micro Kernels?
>>> Kaufmann [Kaufmann@dial835.infolink.com.br] requested PING 939688101 from #tunes
<Kaufmann> fire, well, for bootstrapping, C is actually a good idea
<fire> Kaufmann: and no i am not lagged
<Kaufmann> :)
<fire> yes yes
<Kaufmann> lar, I'd give you the Tunes OS Review subproject page... but it's probably biased :)
<fire> fare wants me to write a scheme->FORTH interpreter
05:30pm
<fire> heh it'll be a while before that's done
-:- Kaufmann is now known as KaufmannBRB
<lar1> Kaufmann:  Biased to uKs?
-:- KaufmannBRB is now known as Kaufmann
<Kaufmann> lar, no, biased to kernel-less design... don't ask :)
<lar1> Ohh yeah... :)
<fire> ok then if i used fgets with stdin then how would i check to see if the programmer entered in too much text
<Kaufmann> fire, well, the good thing about FORTH is that it's automatically compatible with every stack machine... I mean, I myself could design one, given a bit of time...
<fire> yep
<fire> im writting a forth like language
<fire> so little syntax make fire a happy boy
<Kaufmann> Scheme-FORTH is better described as compiling, I guess
<Kaufmann> rotfl... http://www.they.com/disclaimer.html
<Kaufmann> Hey, isn't the Symbolics Lisp Machine hardware designed as a stack machine?
<lar1> That URL was /. ed
<Kaufmann> lar, not anymore
<fire> In the unlikely event that this merchandise should contact antimatter in any form, a catastrophic explosion will result. 
<fire> lol
<lar1> Brb
-:- air [brand@p0wer.qzx.com] has joined #tunes
05:40pm
<air> anyone know what the lowest prices are for web page design for businesses?
<IMPO> yes
<IMPO> zero dollars - read an html tutorial
<air> uhh no
<Tril> zero- steal someone else's design
<lar1> Kaufmann:  What do you mean not anymore?
<air> this business wants me to create a web page for them and maintain it 4 times a month
<Kaufmann> LOL!
<air> about 2 hours a week every week
<Tril> air: why are you asking about low prices then?
<air> i have to scan in about 50 images each week
<air> Tril: cuz i dont want to go below them
<air> but i want to give them a good rate
<fire> ok then if i used fgets with stdin then how would i check to see if the programmer entered in too much text
<Kaufmann> lar, I accessed it the first time I tried, without a problem
<fire> this one's for tril :)
<lar1> Kaufmann:  Oh... I ment the site was on slashdot, not that is suffered the /. effect.  Sorry, unclear pronoun refrence :)
<eihrul> fire: define 'too much text'
<Kaufmann> LOL lar
<fire> like 514 characters when the limit is 513
<eihrul> what is this limit for?
<air> eihrul: he wants error checking
05:50pm
<eihrul> if you're going to answer someone else's questions, make sure you give the right answers :)
<fire> ok you tell me how not to have limit at the prompt
<eihrul> fire: what is this limit on?
<fire> the buffer size
<air> fgets(buffer, size, file);
<eihrul> no, i mean what is this buffer for dude
<eihrul> what are you reading in?
<air> he said text
<eihrul> well, is this any arbitrary text?
<fire> keyboard input
-:- SignOff Kaufmann: #TUNES (Ping timeout for Kaufmann[dial835.infolink.com.br])
-:- Kaufmann` [Kaufmann@dial851.infolink.com.br] has joined #tunes
<Kaufmann`> FUCK FUCK FUCK FUCK FUCK
<Kaufmann`> Brazil is Internet hell
<Tril> water's ISP is worse
<eihrul> fire: don't bother with fgets (), just fread () greedily if you don't want limits, expand the buffer as needed
<fire> *sigh*
<fire> i've spent so much damn time on c io
<eihrul> what are you trying to make?
<fire> an interpreter to bootstrap a language
<eihrul> mmm, why bother?
<fire> so that i can tes tthe functionality of the language
<eihrul> you can write code without an interpreter
<fire> yeah well there aren't any interpreters/compilers for this language
<eihrul> doesn't matter
<eihrul> you can still execute code in your mind
<fire> bah
<fire> i would really like to implement this thankyou :P
<air> anyone here know perl?
<Kaufmann`> air, yep
<Kaufmann`> although you might find #perl more useful for your purposes
<fire> now the thing is with fread is that you are having the _typist_ enter a specific amount of text non the less
<eihrul> no...
<eihrul> you read as much as you can
<air> Kaufmann`: if i call a perl cgi program with program.cgi?username how do i get that into a variable in the program?
<eihrul> if you fill the buffer, realloc () to a greater size
<eihrul> then read more
<Kaufmann`> air, go to www.perl.com, search for a Perl module called CGI.pm
<fire> so like how does it know when it has reached the limit
<fire> to realloc
06:00pm
<eihrul> umm, fread () returns the number of elements read
<eihrul> er actually, hmm, perhaps fgets () would be better in this situation
<fire> but that is after the programmer types enter, no?
<eihrul> eh? no
<air> Kaufmann`: thanks
-:- Kaufmann` is now known as Kaufmann
<eihrul> if you just want to hack something up quickly
<eihrul> just use a fixed size buffer
<fire> seems like such a sin though ;)
<eihrul> shrug, if you're prototyping, prototype
<eihrul> otherwise, you'll be bogged down over little implemetnation problems
<eihrul> moreso than what you're implementing
* fire/#tunes is tempted
<eihrul> and you won't get shit done, i know :)
<lar1> I gotta eat
<lar1> Later
-:- SignOff lar1: #TUNES (Leaving)
06:10pm
<Kaufmann> Can anyone check this URL? http://laurent.penguinpowered.com/~aspect/interp-tcl.html
<Tril> the server (laurent.penguinpowered.com) seems to be down
<Tril> btw that's on Fare's ISP  cybercable.fr  ..  They have a limit on upload bandwidth.. running servers is impractical.  I wonder if they also have dynamic IPs.
<Tril> yes they must, penguinpowered.com has a 10 minute refresh.
<Kaufmann> hrrrrm.
<Tril> I suggest you wait 10 minutes and try again- his IP may have changed.
<Kaufmann> I've already been trying for a while
06:20pm
<fire> it's times like this i am tempted to use goto
<Kaufmann> heh
<air> anyone here use freebsd?
<Kaufmann> fight the urge, fire!
<Tril> i have used freebsd.. we have some machines with it here
<fire> why couldn't somebody do programmer a favor and make a function where you can input text and it returns a string the size in characters you typed
<Tril> echo enter text now, ctrl-d Enter to end ; cat > file; ls -l file
<Tril> that's the easy way :)
<fire> heh
<fire> in c silly :)
<Tril> wrap it in a system()
<Tril> use perl
<fire> no!
<fire> damnit this is in c
<fire> and it will stay in c
<fire> im holding my ground
-:- mode/#Tunes [+b *!no@*.tfb.com] by Tril
-:- fire was kicked off #TUNES by Tril (Bitch-X BaBy!)
-:- mode/#Tunes [-b *!no@*.tfb.com] by Tril
-:- fire [no@209-68-229-146.dialup.cust.tfb.com] has joined #tunes
<fire> :P
<Tril> losers get banned!
<fire> :(
<fire> i think i'll just use a static sized buffer for now
<Tril> write function yourself.. use realloc a lot
<fire> it's for the better
<fire> that's what im trying to do!!!
<Tril> while((c=getc(stdin))!='\n') { realloc(buf,size++);buf[ptr++]=c;}
<fire> ...
<fire> wow
<fire> that was simple
06:30pm
* fire/#tunes pets tril
<fire> :D
<Tril> yes but whether it works is another question!!
<IMPO> whether or not it works is unimportant
<IMPO> just give it a nice looking gui and you'll be fine
<Tril> s/ptr++/size/
<Kaufmann> wow, Tril, now just shorten the variable names to one character each, and you've got a ready-to-order C Obfuscation Contest entry!
<Kaufmann> LOL IMPO
<fire> buffer[size] = c;
<fire> not size++
<Tril> thats waht i said
<fire> oh :(
<Tril> read carefully
* fire/#tunes SORRY
<Tril> at least you knew it was wrong
<Tril> :)
<Kaufmann> Well, I'm leaving now
<fire> bye bye mr brazil :D
<Kaufmann> :)
-:- SignOff Kaufmann: #TUNES (42)
06:40pm
<fire> YES!
<fire> finally read() is done
-:- SignOff fire: #TUNES (Leaving)
06:50pm
<Fare> readline();
-:- SignOff air: #TUNES (Ping timeout for air[p0wer.qzx.com])
<eihrul> doh
<eihrul> i forgot about that
07:00pm
<Fare> enter_FOO(); /* replace FOO with your favorite embedded language */
<Fare> (e.g. guile scheme, perl, lua, tcl, python, slang, etc)
<eihrul> you didn't mention forth
07:10pm
-:- air [brand@p0wer.qzx.com] has joined #tunes
-:- SignOff IMPO: #TUNES (Leaving)
<eihrul> air: squeak is bootable man.... :)
<air> so
<eihrul> just letting you know people are beating you to brix...
<air> no they arent
07:30pm
<eihrul> abi: seen core
<abi> core was last seen on IRC 13 days, 4 hours, 11 minutes and 45 seconds ago, saying: what's up? did you write more of EMK? [Tue Sep 28 15:20:31 1999]
07:40pm
-:- Syntax [dma@08-182.006.popsite.net] has joined #Tunes
-:- lar1 [lar1@sdn-ar-019casfrMP170.dialsprint.net] has joined #tunes
<lar1> Hey
08:00pm
<air> anyone know howto top align shit in tables?
<Tril> VALIGN=?
<air> ah thanks
<lar1> Tril:  Do you know how big an octet is?  (In refrence to the size of e-mail messages)?
<Tril> octet = a byte
<lar1> air:  I am in Linux!  Rejoice!
<lar1> Tril:  Ah, ok thats what I thought.  Thanks mucho
<Tril> It's called octet because some machines may not use bytes, and on those, a sequence of 8 bits would be called an octet.
<eihrul> why does fetchmail use octets?
<eihrul> i mean what commonly used machines actually have non-8 bit bytes?
<eihrul> or atleast, modern day :)
<Tril> well, that's what I think anyawy.  It's called octets in RFCs somewhere. 
08:10pm
<air> before rasterman came to america he had a page with a program that mirrored an image around so it could be used as a background. anyone know what it was called or if that page still exists?
* eihrul/#tunes ponders ditching his assembly code-base and writing in part C/C++.
<air> eihrul: i wouldnt write it in C++
<eihrul> why not?
* lar1/#tunes tells eihrul to not ditch ASM!
<lar1> air:  Any new features in BFE I should be aware of?  I am going to heavy devlopment now
<eihrul> air: eh?
08:20pm
<air> lar1: i have been adding to the todo list in bfe of things it could use to make it better but im not ever going to do any of them. it does what i need to ok
<air> eihrul: C++ sucks
<eihrul> ok, why?>
<eihrul> for all it's faults, it works...
<lar1> air:  Is the latest Bfe available on the site?
-:- hcf [nef@me-portland-us815.javanet.com] has joined #tunes
<lar1> Hey hcf
<hcf> hi lar1
<lar1> What is a Journaling file system?
<hcf> lar1: see the comp.os.research faq
08:30pm
<lar1> Where can that be found?
<hcf> where faqs r found
<hcf> like faqs.org
<lar1> Ok, thanks
<lar1> Will winstupid programs requiring TCP/IP work in WINE if I am using pppd?
<hcf> ok, i guess i was wrong
<hcf> lar1: do get and read the c.o.re faq
<hcf> but that doesnt have journalling fs poop in it
<lar1> Ok
<hcf> put "Journaling file system" in alta and dig thru the 90hits
<lar1> Uhggg
<hcf> uhg my ass, stop being lazy
<lar1> I'd rather google and dig through the 10-20 hits :)
<lar1> I am mucho lazy!
<lar1> At least I got in Linux today :)
<lar1> Don't you think it at least semi-rediculous that one needs to do so much work just to find out what a Journaling file system is?
<lar1> It should be much simpler
08:40pm
<hcf> how do u get 10-20 hits from google on "Journaling file system"?
<lar1> I don't know.... Gogle (almost) always does a very good job
<hcf> estimated 257 matches
<lar1> Like I type in comp.os.research and the first link is what I want
<lar1> Wow!  This FAQ answers a lot of my questions!
<lar1> Thanks for pointing it out
<air> a journaling file system is a filesystem that keeps a record of what is being changed so it can repair any damage done by a power outage
<air> bfs is a jounaling fs and sun just open sourced their xfs for linux
<air> er not sun, sgi
<air> http://oss.sgi.com/projects/xfs/
<hcf> abi: bfs =~ s/sun/sgi/
<abi> OK, hcf
<air> hcf: can u fix that journaling typo
<lar1> air: Isn't that what RAID is for?
<air> lar1: no
<air> raid: is for drive failure
-:- Zhivago [brian@th.merddin.com.au] has joined #tunes
<hcf> air: can u fix it, xxx =~ s/wrong/right/
<air> if the power goes out on a raid system it will have a corrupt fs
<air> so u need journaling to keep a list of all changes and whether or not they completed
<air> bfs =~ s/jounaling/journaling/
<air> abi: bfs =~ s/jounaling/journaling/
<abi> OK, air
08:50pm
<air> ah kewl
<air> that sure beats retyping it like i have done in the past :)
<lar1> Doesn't a Journaling file system take up a lot more space?
<air> no
<air> u just need a little scratch pad on the drive, maybe a sector
<air> before u do something u add an entry to the scratchpad
<air> then u do the thing and remove the entry
<lar1> Ohh
<air> if the power goes out it will compare the entry info to whatever was suppose to be done to see if u even started
<lar1> What if the power goes off while witing to the scratch pad?
<air> if u didnt start it will remove the entry. if u had started it will undo what u did or attempt to finish
<air> well u burst the entire sector
<lar1> Hrmm
<air> if the entry isnt complete then it will know
<Zhivago> journalling can be fairly expensive, depending on the resolvution grain
<lar1> So it helps stop fs corruption, but it doesn't totally make it bullet proof
<Zhivago> it depends on how often you unjournal
<lar1> Zhivago:  Should it be done in the manner air just described?
<lar1> Zhivago:  Which woudn't involve much space...
<air> yes
<Zhivago> well, air's technique is a little small
<Zhivago> and expensive, since it requires seeking to a particular sector before all writes
<lar1> How would you do it Zhivago?
<Zhivago> which means that read and writes are now likely to be far apart
<Zhivago> I would do it like this
<Zhivago> I would partition my space generationally
<Zhivago> say 100 megs for ephemeral data, and the rest for older data
<Zhivago> when I do now is to write to the end of the ephemeral data section
<Zhivago> we always write to the end here so its fast to do - little housekeeping, although we can get a lot of fragmentation
<Zhivago> when we exhaust the emphemerial region, we walk though it, and copy any files which still exist into the next generation up, defragmenting at the asme time
<Zhivago> we stick modifications into the ephemeral region as well, and when we resolve things, we suck the complete file into the next-past emphemeral generation
<Zhivago> if that makes sense
<Zhivago> this is a garbage collection derived technique, which works nicely because there tend to be few references between generations
<Zhivago> and we have a lot of ephemera.
09:00pm
<Zhivago> we can also have multiple ephemeral regions scattered though the disk, and then go to the nearest one in order to write a new object out
<Zhivago> hmm, if that makes sense
<Zhivago> anyhow, for journalling it depends on the commit/rewind resolution
<Zhivago> with one sector, then you'll still get corruption when writing more than one sector's worth
<Zhivago> the journalling pad should be as large as you can make it in order to avoid corrupting objects
<lar1> Sorry, dad needed me
<lar1> What does ephemeral mean?
<Zhivago> ephemeral means short-lived
<Zhivago> I'm working from a somewhat different fundamental model
<Zhivago> which will make things a bit strained here
<Zhivago> I don't have files
<Zhivago> I have a million billion objects
<Zhivago> the vast majority of which have very small life expectancies
<lar1> Oh yes, just like OFS
<lar1> I plan to use ofs, I like the idea
<Zhivago> so I allocate them in the emphemeral pools, and every-so-often I garbage collect the emphemeral pool
<lar1> Makes sense
<Zhivago> anything I have left, is probably not having a tiny life expectancy
<Zhivago> so I reallocate it into a more permanant store
<Zhivago> so I gc emphema a lot, gen1 sometimes, gen2 occasionally, gen3 almost never
<Zhivago> and each time I promote objects I defragment
<Zhivago> I also prefer immutable objects where possible
<lar1> Cool.  I have to go... parents bugging me.  Thanks mucho mucho mucho.  I might be back in an hour or so
-:- SignOff lar1: #TUNES (Leaving)
09:10pm
-:- SignOff Syntax: #TUNES (Ping timeout for Syntax[08-182.006.popsite.net])
-:- Syntax [dma@08-182.006.popsite.net] has joined #Tunes
-:- SignOff Zhivago: #TUNES (Leaving)
<air> anyone know howto use giftrans?
-:- td [x@1Cust166.tnt3.wilmington.nc.da.uu.net] has joined #tunes
09:40pm
<air> damnit, E keeps crashing when i save a gif as interlaced in gimp
10:10pm
-:- Zhivago [brian@th.merddin.com.au] has joined #tunes
-:- SignOff hcf: #TUNES (Leaving)
<Zhivago> hmm, how irritating, a power outage
<eihrul> anyone know where i can find any good documentation on various scheduling algorithms?
<Zhivago> um, all over the place
<Zhivago> the www is probably a good starting place
<Zhivago> but there are a few classic types
<Zhivago> priority queue based
<Zhivago> round-robin
<Zhivago> are the two most common
10:20pm
<Zhivago> for batch oriented scheduling, it becomes a variation of the backpack problem
<Zhivago> with the idea to minimise average completion time
-:- _ruiner_ [nate@ppp185.wi.centurytel.net] has joined #tunes
<eihrul> well my current scheduler is a mix of round-robin and priority queue
<eihrul> the problem is my scheduling algorithm allows deadlocks and other nasty things to happen
<eihrul> so i need a better algorithm that gives all threads a chance to run
<_ruiner_> poor guy..
<_ruiner_> keep trying and you will choke that chicken
<eihrul> why only chickens?
<_ruiner_> I dunno
<eihrul> why not other avian creatures?
<_ruiner_> I hated that movie anyhow
<_ruiner_> I guess cuz they only practiced on chickens
10:30pm
<eihrul> pigeons are much easier to choke
<_ruiner_> I'm not in ninja training so it doesn't really matter to me either way
<eihrul> you never know
<_ruiner_> I always know
<eihrul> to know is to not know
-:- SignOff _ruiner_: #TUNES (Leaving)
<eihrul> Zhivago: what algorithm you using?
10:40pm
* eihrul/#tunes hmmms.
* eihrul/#tunes ponders not implementing a scheduling algorithm at all.
-:- SignOff eihrul: #TUNES ([x]chat)
-:- _Secret [scrat@20.pitt1.pitt.dialup.nauticom.net] has joined #tunes
<air> hey
<air> long time no see
<_Secret> y0
<_Secret> What is up
<air> not much
<_Secret> How is brix?
<air> its doing quite well
<_Secret> How much is left to do
<air> until what? :)
<_Secret> It's finished
<air> finished is relative
<_Secret> hehe
<_Secret> Who is all working on it now?
<air> its got a long way to go before it will have a replacement for every app in every os
<air> me
<_Secret> How much of the OS itself is done?
<air> kernel, keyboard, floppy, ofs, vga,
<air> text screen
<_Secret> Does it have TCP/IP stack?
<air> speaker
<air> no
<_Secret> I could design that
<air> ok
<air> i need to finish the language first
<_Secret> Ok
<air> everything has to be written in it
<_Secret> Ya, well, it's going to take some time to just desgin the OSI model
<air> osi?
<_Secret> Have to determine an interface from the data link to the network, then the TCP/IP transport layer
<_Secret> ISO rather
<_Secret> then the sockets on top of that
<_Secret> The transport layer is where we have to put the protocols
<_Secret> and tcp/ip stack
<_Secret> NIC -> Data Link -> network/Protcols (transport)/tcpip stack -> Session (sockets) -> application
<air> do u know that in brix the tcp/ip is a storage driver?
<air> it has to appear as an ofs filesystem
<_Secret> Ya, I have to find out the implementation details of brix
<_Secret> But, unix/windows do it the same way I'm talking about
<air> ya i know
11:30pm
<_Secret> So, you want TCP/IP to work like the file system
<air> yup
<_Secret> That can still be an implementation at the protocol level hehe
<_Secret> Do you have super vga support?
<air> no
<_Secret> I can write that
<air> i do have partial vesa
<_Secret> ok
<air> vesa2
<_Secret> We could write it up to use vesa2/vesa1 transparently. i.e.
<_Secret> the user won't know, if they have vesa 2, uses linear, if not, uses bank switching, but the application wouldn't know
<air> the app just needs to know about a framebuffer
<_Secret> Ya
<air> or the presence of gfx functions
<_Secret> We can give them one either way
<_Secret> but we wouldn't let them write directly to the video screen
<air> no
<_Secret> We can have some gfx function primatives
<_Secret> In BRiX, does it assume FPU?
<air> well im thinking about requiring pentium
<_Secret> Ok
<air> but brix does know if it has an fpu
<_Secret> Ya, if we don't have an FPU tho, we'd have to write up some emulation routines
11:40pm
<air> ya
<_Secret> For sound, We'd have to figure out a common interface that could be exported from any sound driver and have applications interface with that and then write seperate drivers for each card
<air> sorta
<_Secret> Does Brix still incoperate the MOM ?
<air> a driver in brix is a collection of methods
<_Secret> int 60h?
<air> so if 2 cards have the same part we dont need to write one for each
<air> no
<air> all cards that are sb compatible can use the sb wave methods
<_Secret> ok
<air> same with the other parts like FM
<_Secret> Since the OS is developed without a legacy, we can exclude support for ancient hardware :-)
<_Secret> like BeOS
<air> just things like emu10k need there own methods
<air> heh
11:50pm
<_Secret> How much of the code is 100% assembly?
<air> everything right now
<_Secret> Hehe
<air> after i get the language done i will convert everything to it, except the kernel
<_Secret> Have to be careful, may have to convert
<_Secret> if the x86 dies soon
<air> its not gonna die
<_Secret> hehe
<Zhivago> just write your os to be portable :)
<air> oh its protable
<Zhivago> x86 probably has another 5 years left in it though
<_Secret> Ya
<_Secret> Amd is even comming out with an uncompatable cpu 
<air> u just port the kernel, add support for the target into thelanguage and have it recompile itself and everything to the target
<_Secret> So, it's not just the Merced that intel can't get working you have to worry about
<Zhivago> x86 is a nasty arch anyhow
<Zhivago> only useful because its ubiqitous.
<_Secret> it's so twisted it's fun. Look at other computers, their hardware interfaces are too easy and standard to have fun 
<air> heh
<Zhivago> i would hav epreferred more registers :)
<Zhivago> ppc looks sweet
<_Secret> that's like cheating tho
<_Secret> have so many registers you don't need variables
<Zhivago> yes, that's what I'd like :)
<Zhivago> enough registers and I can get rid of the stack
<Zhivago> (for most programs)
-:- CodeRot [scrat@94.pitt2.pitt.dialup.nauticom.net] has joined #tunes
<CodeRot> hehe
[msg(TUNES)] newlog 1999.1012
IRC log ended Tue Oct 12 00:00:01 1999