IRC log started Mon Apr 12 00:00:00 1999
[msg(TUNES)] permlog 1999.0412
-:- SignOff _QZ: #TUNES (BRiX [http://www.qzx.com/brix] :: sleep)
-:- SignOff Tril: #TUNES (Ping timeout for Tril[sloth.wcug.wwu.edu])
-:- Tril [dem@sloth.wcug.wwu.edu] has joined #TUNES
-:- Otto [mirc-hisp@193.153.42.212] has joined #Tunes
-:- Otto [mirc-hisp@193.153.42.212] has left #Tunes []
-:- AlonzoTG [Alonzo@client-151-200-121-148.bellatlantic.net] has joined #tunes
-:- Downix [down@d-nh-concord-9.ici.net] has joined #tunes
-:- smkl [smkl@purple.verkkotieto.com] has joined #tunes
-:- SignOff smkl: #TUNES (changing servers)
-:- SignOff Downix: #TUNES (Ping timeout for Downix[d-nh-concord-9.ici.net])
-:- SignOff abi: #TUNES (Ping timeout for abi[bespin.cx])
-:- abi [abi@bespin.cx] has joined #tunes
!koontz.openprojects.net!! Remote CONNECT king.openprojects.net 8005 from lilo
!lackey.openprojects.net!! Remote CONNECT king.openprojects.net 8005 from lilo
-:- hcf [nef@me-portland-us246.javanet.com] has joined #tunes
-:- binEng [bineng@dialup147-1-20.swipnet.se] has joined #tunes
-:- SignOff binEng: #TUNES (I don't really have time to irc now :P)
-:- SignOff AlonzoTG: #TUNES (Have Nice Day :))
-:- Iepos [root@d7.t1-8.tecinfo.com] has joined #TUNES
<Fare> Gakuk!
<Iepos> gakuk!
<Iepos> hey is it GAkuk or gaKUK (accent) ?
<Fare> I pronounce it gaKUK!
<Iepos> okay then. gaKUK!
<Iepos> i'm thinking about starting a prototype soon for the deductive system to boot TUNES ...
<Iepos> but first i need to check out how some other similar systems were implemented ...
<Fare> Iepos: what formal background will you use?
<Fare> a typed system?
<Fare> something like I did in my masters thesis?
<Iepos> ohh...
<Iepos> i haven't read it yet.
<Iepos> formal background?
01:50pm
<Fare> (a _very_ condensed explanation of it is in the lambdanD article)
<Iepos> you mean like which axioms to use?
<Fare> not even, rather, which formal framework to use?
<Iepos> like, the structure of how propositions are expressed?
<Fare> There are lots of differences between things like Coq and things like classical propositional logic!
<Iepos> yeah, Coq has lots of strange features ....
<Fare> I myself know very few systems on any but superficial way
<Fare> Coq is very idiosyncratic;
<Fare> but its very design is IMO enemy to reflection
<Iepos> yes ... i guess so
<Fare> people with a design more like Coq, but where reflection has been stressed is NuPRL
<Iepos> hmm ... haven't heard of it ... maybe i should check it out.
<Iepos> anyway, i was thinking of expressing concepts in the system using only function application trees ...
<Iepos> still there? does that sound reasonable?
<Fare> not sure what you mean. Maybe you mean something like Curry's illative combinatory logic?
* Fare/#Tunes is laaaaggged
<Iepos> yes ... only you can use all kinds of symbols for terms, not just a few constants
<Fare> I haven't seen anything like Curry's work, but on his book (see bibliography for lambdaND)
<Fare> this is definitely possible, and that's what I do in my masters thesis and in lambdaND:
<Iepos> okay ...
<Iepos> where exactly is your masters thesis?
<Fare> ~fare/files/cs/stage_unice.dvi.gz
02:00pm
<Iepos> uh oh... i don't know how to read dvi...
<Iepos> let's see...
<Iepos> what is a good program to read dvi ?
<Iepos> ah ha... there's a ps too...
<Fare> xdvi
<Iepos> oh... i wonder if i have that...
<Iepos> nope ... i'll just get the ps...
<Fare> it's in french, you know...
<Fare> and I didn't know about Curry's stuff at the time
<Iepos> yeah... i've taken almost 2 years :-)
<Iepos> i've got a dictionary ...
<Iepos> i'll be back ... gotta go into X.
-:- SignOff Iepos: #TUNES (Iepos)
-:- Iepos [root@d7.t1-8.tecinfo.com] has joined #TUNES
<Fare> gakuk!
<Iepos> gakuk!
<Iepos> oops ... don't have a ps viewer either
<Iepos> i'll have to use ps2pdf
<Iepos> i've got acrobat reader
<Fare> get gv
<Fare> hey, install the .rpm or .deb!
<Iepos> tar
02:10pm
<Fare> slackware?
<Iepos> yes
<Fare> you choose your torment
<Iepos> haha
<Iepos> i do not like dselect
<Fare> neither do I
<Fare> but there's apt and dpkg
<Fare> I *hate* dselect
<Iepos> yeah then you have to worry about dependencies yourself
<Iepos> crazy!
<Iepos> at least with dpkg
<Iepos> what is apt?
<Fare> not with apt
<Fare> it's a frontend to dpkg that manages dependencies
<Iepos> yeah, it's called dselect :-)
<Iepos> oh
<Iepos> oops i misread you
<Iepos> apt is a frontend .... 
02:20pm
!koontz.openprojects.net!! Remote CONNECT jordan.openprojects.net 8005 from jcl
!koontz.openprojects.net!! Remote CONNECT jordan.openprojects.net 6667 from jcl
!koontz.openprojects.net!! Remote CONNECT jordan.openprojects.net 8005 from jcl
-:- SignOff Iepos: #TUNES (Iepos)
-:- veve [h@bay-171-b6.codetel.net.do] has joined #Tunes
-:- veve [h@bay-171-b6.codetel.net.do] has left #Tunes []
-:- FareWell [fare@balance.wiw.org] has joined #Tunes
<FareWell> Hum
<FareWell> the net died to me, and isn't fully resurrected yet :(
02:40pm
-:- SignOff FareWell: #TUNES (Connection reset by pear)
-:- SignOff hcf: #TUNES (Leaving)
-:- AlonzoTG [Alonzo@client-151-200-127-137.bellatlantic.net] has joined #tunes
-:- _QZ [brand@p0wer.qzx.com] has joined #tunes
<Tril> hi
<_QZ> hi
<abi> hello, _QZ
<AlonzoTG> I HATE LINUX!!!!!
<AlonzoTG> Is there *any* way to make decompressions go faster??????
<Tril> kill off some useless daemons
<AlonzoTG> I mean I try to gunzip something and it takes FOREVER!!!!
<Tril> on what, a 386?
<AlonzoTG> I don't know what daemons are running!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!11
<AlonzoTG> THe processor utilization is about 25%
<Tril> what processor?
<AlonzoTG> and the thing should be in high gear at 99% 
<Tril> well, if your cpu utilization is low, that means the program is waiting on the disk.
<Tril> get a faster hard drive
<AlonzoTG> but the thing pretends like it isn't doing jack and just sits there...
<AlonzoTG> the HD light isn't blinking...
<Tril> what CPU do you have, how much RAM, etc?
<AlonzoTG> Its just vegitating...
<AlonzoTG> P100/32 mb
<Tril> how big is the gzip file
<AlonzoTG> PK zip under DOS runs many hundreds of times faster...
<AlonzoTG> 300k
<Tril> it sounds like you probably typed the wrong command
<Tril> what did you type?
<AlonzoTG> oh fuck...
<AlonzoTG> I dunno, I lost my cheatsheet...
<AlonzoTG> thats the only way I could remember that command from hell...
<AlonzoTG> something like: 
<Tril> when it stops like that, it means it's reading from standard input.  To find out, hit ctrl-D and see if it stops.
<AlonzoTG> gzip -cd some/really/horribly/long/path/name | tar xfv -
<AlonzoTG> ?
<Tril> that looks ok
<_QZ> tar zxvf some/really/horribly/long/path/nam
<_QZ> u dont need to run gzip
<AlonzoTG> WTF?
<Tril> you probably just misspelled the path name
<AlonzoTG> but it was working...
<Tril> if gzip can't find the file , it would read stdin
<_QZ> the z will run the file thru gzip for u
<AlonzoTG> just taking forever to decode 2k files...
<Tril> 2000 files?
<AlonzoTG> it DOES work...
<Tril> it works?
<Tril> just slowly?
<AlonzoTG> just takes forever...
<AlonzoTG> brb dinner
04:40pm
<Tril> just tried brix demo in bochs
<Tril> "Unknown 386 3MHz" ... :)
<Tril> qz: have you tried retro?
<_QZ> yes
<Tril> the new one, with forth?
<_QZ> yup
<_QZ> it was made yesterday
04:50pm
<_QZ> and that 3mhz for bochs is correct
<Tril> well what does it mean then
<_QZ> it means bochs has the same performance as a 3mhz 386
<Tril> not noticable in brix demo.!!!
<Tril> or in retro, for that matter.
<_QZ> what isnt noticable?
<Tril> I can't tell that bochs is slow, since neither application is CPU intensive. 
<_QZ> heh
<Tril> In retro it's just a forth interpreter, so simple stack ops are plenty fast
<Tril> that brix demo would look the same on a C=64
<_QZ> :)
<_QZ> it wasnt a demo
<Tril> also, I can't tell what lag is caused by bochs, and what is caused by netlag. since I run bochs over remote X sessions.
<_QZ> it was a test for yer floppy drive
<Tril> my bochs floppy drive works fine :)
<_QZ> oh and if u try it on a real floppy it will be slowwwwww
<Tril> I told u it worked fine on my floppy drive. but failed for some other reason.
<_QZ> did u notcie how long it took to display the number of objects it loaded?
<_QZ> on yer floppy
<_QZ> the boot sector would load the kernel just fine but then the kernel would fail to load the objects. so i stuck in a disk reset before each read (7 of them)
<_QZ> i now know that it was failing cuz i was turning OFF the floppy after each read :)
05:00pm
<_QZ> oh piss
<_QZ> i just got some email from some guy that ran that test on several machines
<_QZ> appears that the speed reported for a 386 is half what it should be
<_QZ> so bochs is a 6mhz 386
<_QZ> and none of his machine, nor any of my machines, failed on the cpuid
<AlonzoTG> om
<_QZ> i think u should just throw that machine of yers away :)
<AlonzoTG> =P
<Tril> sorry, that machine locked up I had to go to another lab
<_QZ> hmm he tried to run brix in dosemu :)
<Tril> I don't remember how long it took to display the number of objects.
<Tril> half a second ?
<_QZ> it was more like 1.5-2 seconds
<_QZ> u know a joshua mayo?
<Tril> no...
<_QZ> name is familiar
<Tril> do a dejanews profile search
<_QZ> maybe he keeps emailing me to release brix :)
<Tril> so what about it hanging on my box
<_QZ> throw yer box away
<Tril> yeah, i should get a new one
<_QZ> maybe the floppy was corrupt
<Tril> we'll see after I pay tuition
<Tril> that's possible
<Tril> I only tried it on one disk
05:10pm
<_QZ> the 8086 only used the low 8bits of the flags register right?
<Tril> Uhh... 
<_QZ> isnt that the test for an 8086?
<Tril> oh, maybe
<Tril> trying to set those bits fails?
<Tril> are you trying to distinguish an 8086 from 8088, 80186, and 80188?
<_QZ> no u clear them and then reread and if they are still set then its an 8086, i think
<_QZ> no, this guy attempted to run brix on an 8086
<Tril> brix requires a 386 tho
<_QZ> i know
<Tril> you just wanwt it to print a message?
<_QZ> he complained cuz it didnt print an error message, it just hung
<_QZ> u think i should add the code?
<Tril> no.
<_QZ> ki wont then :)
<Tril> I think u should release the source :)
<_QZ> fine
<AlonzoTG> om
<AlonzoTG> munch munch
05:20pm
-:- ChanServ has changed the topic on channel #tunes to: Millenia Year Application Software System -- www.qzx.com/myass.txt
-:- mode/#tunes [+o Tril] by ChanServ
<Tril> Oh crap
<Tril> chanserv has gone berserk
-:- mode/#tunes [-o Tril] by Tril
-:- Tril [dem@sloth.wcug.wwu.edu] has left #TUNES []
-:- Tril [dem@sloth.wcug.wwu.edu] has joined #TUNES
-:- mode/#tunes [+o Tril] by ChanServ
<Tril> grr
-:- Tril [dem@sloth.wcug.wwu.edu] has left #TUNES []
-:- Tril [dem@sloth.wcug.wwu.edu] has joined #TUNES
-:- mode/#tunes [+o Tril] by ChanServ
05:40pm
-:- ChanServ has changed the topic on channel #tunes to: Millenia Year Application Software System -- www.qzx.com/myass.txt
-lilo(lilo@varley.openprojects.net)- [GlobalNotice] Hi all.  Please bear with us as we debug a small enhancement to services :)
<_QZ> Tril: did u bitch to lilo? :)
<Tril> no. all I did was join the #openprojects channel!
<Tril> i didn't say anything.
<_QZ> hehe
<Tril> but I think it was changing topics on ALL the channels
06:00pm
<_QZ> Tril: did it report a mhz when it crashed on yer machine?
<_QZ> > 6x86L-PR200+ @ 150MHz
<_QZ> > Above machine with Linux DosEmu 0.66.7.0 [crashes at speed test]
<_QZ> > reports: CyrixInstead ??-P5-?? FPUgeneral protection at 0x9d66: f
<Tril> I don't remember; I can find out at home by looking at my sent-mail, do you still have the mail I sent you?
!lilo:*! I think the services changes are relatively stable.  I'll probably have to tweak them later. 8)
<_QZ> most likely
!doogie:*! it is later
<Tril> ..
<_QZ> > > It loaded 2 objects, turned off the floppy drive light, printed
<_QZ> > > "Found: GenuineIntel " and then locked up.
<_QZ> so it didnt even report a cpu?
<Tril> I wrote exactly what it said
<Tril> right, it didn't say what Cpu, just that it was an intel
<Tril> I guess it was going to be the next thing it printed?
<_QZ> i have no idea why it would fail
<Tril> there was a space after GenuineIntel and the cursor was after the space
<_QZ> cuz it works fine on my 486
<_QZ> want the code?
<Tril> i guess
<Tril> this is the code run right after printing "GenuineIntel "?
<_QZ> this is the code that prints that entire line
06:10pm
<_QZ> except for the "found: "
<Tril> why can't I find the string GenuineIntel in it then?
<_QZ> that is from the cpu
<_QZ> the cpuid instruction returns that string in ebx,ecx,edx
<Tril> what's pushfd?
<Tril> doubleflags?
<_QZ> push flags 32bit
<_QZ> pushf pushes 16bits
<Tril> i'll take your word for it that's supported on a 386
<_QZ> yes it is
<_QZ> its how u push the 32bit flags on a 386
<_QZ> it also works on a 286
<Tril> what code is used for print_ ?
<_QZ> %MACRO print_ 1
<_QZ>         mov si,%1
<_QZ>         call print
<_QZ> %ENDM
<Tril> ..
<Tril> and print?
<_QZ> print:
<_QZ>         cld
<_QZ>         xor bx,bx
<_QZ>         mov ah,0eh
<_QZ> .nch:   lodsb
<_QZ>         int 10h
<_QZ>         cmp byte [si],00h
<_QZ>         jne .nch
<_QZ>         ret
<Tril> int 10h guarantees to preserv EAX? HA ha ha ha ha...
06:20pm
<_QZ> it does on every other computer :)
<Tril> never trust a BIOS.
<Tril> wait a minute
<Tril> you moved crap into AH
<_QZ> tell me how it printed the other strings then?????????
<Tril> no, look at line 99, it says print_ msgVendor
<Tril> 2 lines down it depends on a value in EAX being preserved from the result of CPUID
<Tril> but you yourself put crap in AH during the print_ call
<Tril> my first guess is it is incorrectly detecting that my 486 has a full cpuid
<Tril> since there's much more code executed for the full cpuid than the partial one, the partial cpuid path looks safer :)
<_QZ> hmm
<Tril> either that, or fninit
<Tril> is fninit a 80287 command?
<_QZ> i dont know
<Tril> how about 487
<Tril> did pentium add any fpu instructions?
<_QZ> that is from intels code to detect an fpu when the cpu doesnt have full cpuid
* Tril/#TUNES  is away: (afk) [BX-MsgLog Off]
<_QZ> i just saved eax before the print
<_QZ> i would let u test it but the source can not be assembled right now
06:30pm
-:- SignOff AlonzoTG: #TUNES (Have Nice Day :))
-:- limon [romanl1@nog1-42.telmex.net.mx] has joined #Tunes
<limon> hola
<limon> alguien habla espaņol
07:50pm
-:- limon [romanl1@nog1-42.telmex.net.mx] has left #Tunes []
!lilo:*! that was requested by the user, btw 8)
!lilo:*! I just wanted people to know we have not turned into your favorite /kill crazy net :)
!lilo:*! isho 4
-:- KeLp_ [kelp@xws107.xtrn03.wwu.edu] has joined #tunes
-:- SignOff KeLp_: #TUNES (rebooting)
<_QZ> anyone here?
10:00pm
-:- SignOff _QZ: #TUNES (BRiX [http://www.qzx.com/brix] :: sleep)
[msg(TUNES)] newlog 1999.0413
IRC log ended Tue Apr 13 00:00:00 1999