IRC log started Wed Aug 18 00:00:01 1999
[msg(TUNES)] permlog 1999.0818
-:- SignOff liar: #TUNES (BRiX [http://www.qzx.com/brix] :: sleep)
-:- SignOff eihrul: #TUNES (Ping timeout for eihrul[usr5-ppp44.lvdi.net])
-:- eihrul [lee@usr5-ppp44.lvdi.net] has joined #tunes
-:- fare [fare@quatramaran.ens.fr] has joined #Tunes
-:- SignOff fare: #TUNES (Ping timeout for fare[quatramaran.ens.fr])
-:- fare [fare@quatramaran.ens.fr] has joined #Tunes
-:- AlonzoTG [Alonzo@209-122-204-80.s588.tnt3.lnh.md.dialup.rcn.com] has joined #tunes
-:- twyl [user3544@98AB8553.ipt.aol.com] has joined #Tunes
-:- twyl [user3544@98AB8553.ipt.aol.com] has left #Tunes []
-:- water [water@ppp-tnt-11.tscnet.net] has joined #tunes
<eihrul> lo
<water> hey
05:20am
-:- SignOff eihrul: #TUNES (Leaving)
* AlonzoTG/#tunes left clicks on water
<water> good morning
* AlonzoTG/#tunes is looking at certificate courses to get a better job.... =\
<water> bbiaf
<AlonzoTG> party
05:40am
<water>  bye
-:- water [water@ppp-tnt-11.tscnet.net] has left #tunes []
06:10am
-:- hcf [nef@me-portland-us733.javanet.com] has joined #tunes
-:- smkl [sami@MXCV.rdyn.saunalahti.fi] has joined #tunes
-:- AlonzoTG [Alonzo@209-122-204-80.s588.tnt3.lnh.md.dialup.rcn.com] has joined #tunes
-:- sr [depag@phila-dialup080.nni.com] has joined #tunes
<hcf> hoy sr
<sr> hoy
09:00am
-:- NetSplit: forward.openprojects.net split from irc.linux.com [09:11am]
-:- BitchX+Deb1an: Press Ctrl-F to see who left Ctrl-E to change to [forward.openprojects.net]
-:- Netjoined: forward.openprojects.net irc.linux.com
-:- Crimson [crimson@chaosdev.org] has joined #Tunes
-:- sr [depag@phila-dialup080.nni.com] has joined #Tunes
-:- SignOff sr: #TUNES (Ping timeout for sr[phila-dialup080.nni.com])
-:- SignOff hcf: #TUNES (Leaving)
-:- ElGato [dave@bespin.dhs.org] has joined #tunes
<ElGato> gakuk!
10:20am
* ElGato/#tunes is waiting for the dragon book to come in the mail
* ElGato/#tunes is away: (Auto-Away after 10 mins) [BX-MsgLog On]
-:- ruiner [nate@ppp303.wi.centuryinter.net] has joined #tunes
<ElGato> hi ruiner
<ruiner> hi
<ElGato> what's p?
<ruiner> not much
<ruiner> what are you up to?
11:40am
<ElGato> not much either
<ElGato> im getting the dragon book today though
<ruiner> cool
<ElGato> :D
11:50am
-:- SignOff ElGato: #TUNES (brb)
-:- ElGato [dave@bespin.dhs.org] has joined #tunes
<ruiner> wb
12:10pm
-:- SignOff ruiner: #TUNES (Leaving)
* ElGato/#tunes is away: (Auto-Away after 10 mins) [BX-MsgLog On]
<ElGato> oh
<ElGato> uh
<ElGato> hi
<ElGato> hehheeh
12:20pm
-:- SignOff ElGato: #TUNES (ElGato has no reason)
-:- ElGato [dave@bespin.dhs.org] has joined #tunes
* ElGato/#tunes is away: (Auto-Away after 10 mins) [BX-MsgLog On]
-:- eihrul [lee@usr5-ppp25.lvdi.net] has joined #tunes
-:- SignOff ElGato: #TUNES (ElGato has no reason)
-:- pyro [tcn@cci-209150250120.clarityconnect.net] has joined #tunes
<pyro> abi: seen dufrp?
<abi> pyro: no idea
<pyro> abi: seen fare?
<abi> pyro: i haven't a clue
<pyro> hey.. that feature used to work..
03:00pm
-:- ElGato [dave@bespin.dhs.org] has joined #tunes
<ElGato> hi
<eihrul> lo
<ElGato> :)
<ElGato> my compielr book should get here soon
03:30pm
<pyro> hey
<ElGato> y0
<ElGato> forth is cool cause you don't have to parse anything :D
<ElGato> brb
-:- SignOff ElGato: #TUNES (ElGato has no reason)
-:- ElGato [dave@bespin.dhs.org] has joined #tunes
<pyro> what book did you get?
03:40pm
<ElGato> dragon book
<ElGato> brb
<ElGato> back
<ElGato> i'll get it when the ups man gets here
<ElGato> im waiting
<pyro> heh
<pyro> sorry can't chat much, i'm trying to get stuff done on retro
<pyro> i've decided a LISP or Scheme interpreter would be a nice addition; i'm getting used to it at work
03:50pm
<pyro> it's a good language for graphics
* ElGato/#tunes is away: (Auto-Away after 10 mins) [BX-MsgLog On]
04:00pm
-:- SignOff pyro: #TUNES (pyro has no reason)
* ElGato/#tunes is back from the dead. Gone 0 hrs 7 min 19 secs
<ElGato> sorry
<ElGato> yeah 
<ElGato> im gonna write lengua for retro :)
<ElGato> when is that bloody ups man gonna get here!!!???
04:10pm
<eihrul> he died
<ElGato> nooo!!!
04:20pm
-:- liar [brand@p0wer.qzx.com] has joined #tunes
-:- ruiner [nate@ppp103.wi.centuryinter.net] has joined #tunes
<ElGato> hello liar and ruiner
<ruiner> greetings
<liar> hello
<abi> salut, liar
<liar> eir: u here?
<liar> eihrul: u here?
<eihrul> yeah
<liar> do u still have that brix binary i sent u?
<eihrul> i believe so
<liar> could i have it back? :)
<liar> thanks
<eihrul> you deleted something you shouldn't have?
<ruiner> heh heh heh
<liar> well i have begun adding all the objects to the brix image but its not working yet and someone needs a binary
04:30pm
<liar> oh what cpu did u run it on again?
<eihrul> 486sx
<liar> and yer sure its a 20mhz?
<liar> and not 40?
<eihrul> lemme check bios again
<fare> gakuk
<eihrul> hrmm... QNX says 25 mhz :)
<liar> image 636 was reporting all 486's and up with correct mhz but it reported 386's at half there speed so i made it double the mhz for 386 but it might have a bug and be doubling the 486 value too. but there is no way it could be quadrupling it
* ElGato/#tunes is away: (Auto-Away after 10 mins) [BX-MsgLog On]
<liar> i think u have an amd 486
<fare> 80x86 CPUs are HELL
<liar> cuz amd/cyrix optimizes loop instructions so the speed test reports outrageous speeds on them
<liar> ah crap
<liar> it reported my 33 as a 108
04:40pm
<liar> ok the speed test works now
<liar> bochs is a 7mhz 386 on a celeron 450
04:50pm
-:- SignOff ElGato: #TUNES (Hey!  Where'd my controlling terminal go?)
-:- SignOff smkl: #TUNES (sleeping ...)
-:- SignOff ruiner: #TUNES (Leaving)
-:- water [water@ppp-tnt-85.tscnet.net] has joined #tunes
<water> hey all
<liar> hi
05:40pm
<eihrul> liar: hey, is it better to test for 0 with test reg, reg | or reg, reg?
06:00pm
<liar> or
<eihrul> hrmm...
<eihrul> it seems that either one is generally as fast as the other
<eihrul> atleast when doing reg, reg
* AlonzoTG/#tunes left clicks on eihrul
<liar> i use or
* water/#tunes drags AlonzoTG to the Trash icon.
<liar> water: :)
<eihrul> hmm... opt manual says test reg, reg is pairable and both take 1 cycle in that form
<liar> hmm
<eihrul> http://www.agner.org/assem/pentopt.htm
<water> interesting...
<water> actual evidence that the x86 architecture is harder to work with than risc or equivalent
<eihrul> yep :)
<eihrul> but it's cheaper
<water> yeah, whatever
06:40pm
<water> man-hours cost money, too
<eihrul> but unfortunately
<eihrul> the barrier to entry is the cost of the processor
<water> just the time saved by programmers doing compilation-edit-test cycles alone is probably worth the difference
<eihrul> if i don't earn money in working with the processor
<eihrul> i surely cannot earn money to put forth towards a good risc system
<water> iMac?
<eihrul> ha
<eihrul> i can buy a $500 bargain box
<eihrul> that performs thrice as fast
<water> bs
<water> but the point is that you enjoy investing your life in a platform destined for extinction
<eihrul> name one platform that isn't destined for extinction?
<water> exactly
<water> all of them are
<water> :)
<eihrul> and so are we :)
<eihrul> and if i become extinct before the platform itself
<eihrul> than i am investing my life in something seemingly not destined for extinction
<water> that's a restrictive view of the point of humanity, to say the least
<water> i.e. "our children are our only future" is not what humanity is about
<eihrul> we are about nothing, by definition, i would think
<water> my point is that spending time optimizing code for a processor is for compilers and their designers
<eihrul> i'm well aware of this
<eihrul> but i wanted to know why
<eihrul> call it explorative masochism
<eihrul> i'll walk away from this with a better appreciation of what compilers do, that's all
<water> ok, i will :)
<eihrul> and even so, how do you know i'm not going to write a compiler some day, eh?
06:50pm
<water> you probably will write some c-compiler, being a masochist ;)
<liar> ugh my if flag is set but im not getting any interrupts now
<eihrul> ack... no way
<eihrul> let C die with gcc :)
<water> good
<eihrul> at the very least, maybe a D
<water> bleh
<eihrul> notice i said "at the very least"
<eihrul> i don't aspire to just rip C and add a few goodies
<eihrul> plus infix notation is annoying to parse
<water> no kidding
07:00pm
<water> but then, so is manual typing of code
<eihrul> how else do you propose to write code?
<water> graphically
<water> language primitives could be clicked-and-dragged from a sort of toolbar
<eihrul> that would be slower
<water> not necessarily
<water> besides, writing code is usually a very low-resource process
<eihrul> i would say there is no way it could be faster
<water> well, sure
<eihrul> anymore than manipulating something with a rather abstract textual language
<water> i think that we're contradicting each other out of channel boredom
<eihrul> or maybe we're just very contradictory by nature :)
<water> that, too
<water> i just have ideas i'd like to discuss, and you're definitely the wrong person to talk to
<eihrul> not so...
<water> yeah, whatever
<eihrul> if you mean by the fact that i won't accept many blind assertions
<eihrul> than yes, i won't
<eihrul> er rather, then
<water> no, i mean that you're not educated about the stuff that i think about
<water> it would take too long to educate you via irc
<eihrul> and by not communicating about ideas of which i do not know
<eihrul> i will remain similarly as ignorant as before
<water> why would i communicate to you ideas that you _have_ heard before?
<eihrul> unless education by osmosis and similar theories have truth to them
<water> this is pointless. have someone mail me if they want to talk
-:- water [water@ppp-tnt-85.tscnet.net] has left #tunes []
07:10pm
-:- Mr_Wrong [seanl@adsl-63-192-213-146.dsl.snfc21.pacbell.net] has joined #tunes
-:- water [water@ppp-tnt-85.tscnet.net] has joined #tunes
<liar> eihrul: so what would cause interrupts to not be set?
<liar> s/set/sent/
<eihrul> no idea
07:50pm
<AlonzoTG> interrupt flag...
<liar> nope
<AlonzoTG> the IRQ controller...
<AlonzoTG> =P
<AlonzoTG> sumpfin on the mobo.
<liar> shut yer mouth if u cant help
08:00pm
-:- SignOff AlonzoTG: #TUNES (Have Nice Day :))
-:- ruiner [nate@ppp019.wi.centuryinter.net] has joined #tunes
<ruiner> hey liar.....can I pm you some ?'s?
08:50pm
<liar> pm?
<ruiner> private message
<liar> ya
<ruiner> k
<liar> well my ints work now
<liar> it helps if u unmask them after masking them
09:00pm
<liar> u gonna ask?
<ruiner> for agreeableness I rank in the 2nd percentile....I'm classified as equivalent to emperor palpatine
<ruiner> yeah....was doing that and had to move something for my mom
09:10pm
<liar> uhh now
09:20pm
<liar> get a notebook with cdrom drive
10:00pm
-:- lar1 [LARMAN@dialup-209.245.142.131.SanJose1.Level3.net] has joined #tunes
<lar1> Hey
<abi> niihau, lar1
<water> hi lar
<liar> brix is multitasking
10:40pm
<liar> and as soon as i find the event bug it will be interactive
<water> cool
<liar> oops
<water> lol... what?
<liar> maybe it wasnt a bug
<ruiner> ?
<liar> its not considered a bug if the programmer forgets to put in a section of code is it?
<eihrul> it's a feature
<ruiner> lol
<ruiner> depends on who you work for
<liar> i kinda forgot to create event lists for the 15 irq's
<water> doh!
<liar> so the irq handler calls event_signal for event 0 cuz thats what it was set to
<liar> and event 0 doesnt exist
<liar> well just label that under stupid programmer mistake instead of bug
10:50pm
<lar1> liar:  What os are are talking about?
<water> brix
<lar1> Isn't brix _QZ's department?
<water> liar =_QZ
11:00pm
-:- SignOff lar1: #TUNES (Ping timeout for lar1[dialup-209.245.142.131.SanJose1.Level3.net])
-:- lar1 [LARMAN@dialup-209.245.138.19.SanJose1.Level3.net] has joined #tunes
<water> wb
<ruiner> later all
<liar> lar1: qz gave me the source code to brix and told me to take it over
-:- SignOff ruiner: #TUNES (Leaving)
<water> oh. my bad
<lar1> qz gave up?
<liar> ya
<lar1> Why?
<liar> he has cancer and was too weak 
<lar1> Really?  My god... wow
<water> lol
<liar> i liked all of qz's ways of doing stuff in brix so he chose me to finish it
<liar> but i wont be giving it away for free like he was going to do
<liar> it will cost u $100 to buy it from me
11:10pm
<water> it seems that soon Squeak will support an entire office suite
<water> a lightweight, powerful one, of course :)
<liar> brix will soon support squeak
<liar> NOT
<water> they haven't added postscript support yet, though. only rudimentary TeX output so far
<water> liar: that's ok, it'd be really easy to port squeak to brix
<water> :)
11:20pm
<liar> ok events are fixed
<liar> now brix is 100% stable
<liar> u cant crash it
<water> hehe
<liar> if u can crash it ill give u this machine
<water> that sounds like a challenge
<water> make a web server for brix and test it that way ;)
<liar> hehe
<liar> well i must sleep now
<lar1> g'night
<liar> cya
<water> bye
-:- SignOff liar: #TUNES (BRiX [http://www.qzx.com/brix] :: sleep)
<lar1> So does qz hang out here anymore?
<water> lar1: liar = _QZ
<water> liar was lying ;)
* lar1/#tunes feels _very_ stupid right about now...
<lar1> hehe
<lar1> Hmm, every time I look at the BRiX docs I think... uh oh... I don't understand 1/2 this stuff
<water> os-coding _is pretty arcane
11:30pm
<lar1> Well I have redesigned my OS more times then I want to count... thats why there is very little code
<water> what's the current idea?
<lar1> What exactly do you want to know about it?
<lar1> I am not sure how to describe the whole thing
<water> just what kind of things you want to code to make the os the Right Thing (TM)
<water> i.e. the cool or good features or properties that you want
<lar1> Well I am looking for small
<lar1> like Qnx
<water> ah... real-time, too?
<lar1> yes
<water> but not posix, i'm guessing
<lar1> No way
<lar1> brb
<water> k
11:40pm
<lar1> back
<water> k
<water> so what kind of programming interface do you want to provide?
<lar1> What do you mean by programming interface?  Like kernel functions?
<water> yeah, the equivalent of posix api's
<lar1> I guess some call gates
<lar1> Or interupts mabye
<water> ok, but keep in mind that int-based syscalls can really slow things down
<lar1> What is a better way?
<water> dunno. maybe inter-process comm's using messages or something
<eihrul> water: hrmm, how so?
<lar1> How do messages work
<water> eihrul: they tend to cause lots of cache-flushing
<water> lar1: look at a uK design
<lar1> Wheres that?
<abi> i heard that was current
<eihrul> i assume this is associated with call gates as well?
<water> uK?
<abi> it has been said that uK is monolithic right?
<water> uK is a microkernel
<water> abi: no, uK is microkernel
<abi> okay, water.
<eihrul> or is the cache-flushing only a symptom of trap gates?
<lar1> Whats the difference between a microkernel and kernel... I never quite understood that
<water> uK's allow os services to be mostly performed by app-level server processes
<water> like file systems or virtual memory schemes
<water> sort of like a modular kernel, but uK's almost always use threads and messages
<lar1> So programs write their own files and the like instead of the kernel?
<water> err
<water> not quite
<eihrul> microkernels tend to provide enough functionality so that services usually provided by the kernel can be easily implemented by other applications
<eihrul> so you might have one program specifically managing files
<eihrul> which the program wanting to access the file would communicate with
11:50pm
<lar1> Whats the point of that?  Isn't that more overhead?
<eihrul> but a program could just as well implement it's own file system services within the process
<eihrul> yep, context/address space switching
<lar1> So not using a uK is faster?  Or no
<eihrul> it's not really faster at all
<lar1> Why is it better?
<eihrul> some say it offers security since the various functionalities are separated by address spaces
<eihrul> some say it's modularity, which need not really be taken advantage of at all... you can have a monokernel on top of a microkernel
<lar1> a mono kernel is just a plain kernel?
<eihrul> yep, most hardware services are integrated into a monolithic kernel
<lar1> So a monolithic kernel would preform better on slower machines?
<eihrul> it will generally perform better in most cases
<eihrul> i don't think there is an exception to that thus far
<eihrul> it's just microkernels are way easier to make :)
<water> <eihrul> or is the cache-flushing only a symptom of trap gates?
<lar1> Ohh
<eihrul> water: well, may i have an answer? :)
<water> cache-flushing often occurs for interrupts because the code is arranged for the case of not being interrupted
<eihrul> you could probably get all the benefits of a microkernel with a small kernel that just dynamically loaded modules
<water> a.k.a. linux
<eihrul> yep
<eihrul> which i've been thinking about doing myself lately
<water> no thanks
<eihrul> am i telling you to do it?
<water> hehe
<water> no, of course not
[msg(TUNES)] newlog 1999.0819
IRC log ended Thu Aug 19 00:00:00 1999