IRC log started Fri Apr  9 00:00:00 1999
[msg(TUNES)] permlog 1999.0409
-:- Closing Link: TUNES[bespin.cx] by mccaffrey.openprojects.net (Ping timeout for TUNES[bespin.cx])
-:- 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 norton.openprojects.net)
-:- Your host is norton.openprojects.net[ftp.caus.linpeople.org.], running version u2.10.04.resolv9.nmt.egcs4.tok.pten.tlim4.admin.upper4.whisper3.gipl.modeless7 (from norton.openprojects.net)
-:- This server was cobbled together Mon Jan 25 1999 at 07 12:10 EST(from norton.openprojects.net)
-:- norton.openprojects.net u2.10.04.resolv9.nmt.egcs4.tok.pten.tlim4.admin.upper4.whisper3.gipl.modeless7 dioswkfcg biklmnopstv 
-:- [local users on irc(11)] 5%
-:- [global users on irc(77)] 34%
-:- [invisible users on irc(152)] 66%
-:- [ircops on irc(13)] 6%
-:- [total users on irc(229)]
-:- [unknown connections(0)]
-:- [total servers on irc(34)] (avg. 6 users per server)
-:- [total channels created(61)] (avg. 3 users per channel)
!norton.openprojects.net Highest connection count: 21 (20 clients)
-:- Mode change [+f] for user TUNES
-:- Mode change [+iws] for user TUNES
-:- JOIN activated by "TUNES #tunes tunes@bespin.cx "
-:- TUNES [tunes@bespin.cx] has joined #tunes
-:- Topic for #TUNES: Retro source released! See http://www.tunes.org
-:- topic set by tcn [Tue Apr  6 05:22:51 1999]
-:- [Users(#Tunes:3)] 
[ TUNES     ] [ Fare      ] [ Tril      ] 
-:- Channel #Tunes was created at Sun Feb 28 08:48:06 1999
-:- BitchX+Deb1an: Join to #tunes was synced in 8.217 secs!!
-:- Mode change [-s] for user TUNES
-:- gigi [blah@frc161.upf.es] has joined #Tunes
-:- gigi [blah@frc161.upf.es] has left #Tunes []
-:- Fare has changed the topic on channel #Tunes to: Reflective Computing System - http://www.tunes.org/
-:- LaGgEr [x@pool031-max14.ds12-ca-us.dialup.earthlink.net] has joined #Tunes
-:- LaGgEr [x@pool031-max14.ds12-ca-us.dialup.earthlink.net] has left #Tunes []
-:- binEng [bineng@dialup152-1-39.swipnet.se] has joined #tunes
<Fare> hum
<binEng> hoy
<Fare> manana
11:20am
-:- smkl [sami@MLXVI.rdyn.saunalahti.fi] has joined #tunes
-:- hcf [nef@me-portland-us101.javanet.com] has joined #tunes
-:- Beholder [yea_right@ppp-026.m2-1.sub.ican.net] has joined #tunes
<Beholder> Hello All :)
12:00pm
<Beholder> Tril, I know this sounds stupid, but how do people sign up for the mailing list? (if I'm gone could you e-mail it to me?)  Thanks... BTW: The server works great!
<hcf> hoy Beholder
* binEng/#tunes waves
-:- abi [abi@bespin.cx] has joined #tunes
<binEng> welcome, abi
<abi> :)
12:10pm
-:- bin [patman@142.51.254.20] has joined #tunes
<bin> Take it easy... this is just a test! Don't panic!
<binEng> Testing, testing. 1... 2... 3... *click*
<bin> ack
-:- SignOff bin: #TUNES (Leaving)
<Beholder> Hello hcf :)
<hcf> Hello Beholder :)
<Beholder> Sorry I was in another window... when will they make an IRC program that blinks in the task bar, or beeps or something when the text in the window has changed!
<Beholder> hcf: You got any idea on how to join the UniOS mailing list?
<hcf> u could script ur irc client to do it if u wanted i suppose
<hcf> unios mlist? i dont know
<Beholder> How do you subscribe to the Tunes list?
<Beholder> I'm sure it's the same way
<hcf> using http://www2.tunes.org/cgi-bin/list-control?list=tunes&do=subscribe
12:20pm
<Beholder> Hmm... maybe if I replace tunes with unios it will work, I'll tyr
<Beholder> tyr = try
12:30pm
-:- Pauldoo [PaulRichar@dunvegan1.demon.co.uk] has joined #Tunes
<Pauldoo> Hi binEng
<Pauldoo> What have you been up to recently?  TPP has been silent...
<binEng> hi
<binEng> yeah, I'm like more active here... or maybe not active at all :)
<Pauldoo> I've been making my usual simulation type progies in DJGPP, and also learning linu
<Pauldoo> x
* Pauldoo/#Tunes wonders why his HD is going ape shit for no reason...
<Pauldoo> hmmm...  Must be windoze doing its usual crap
<binEng> I haven't programmed much lately... mostly learning stuff instead
<Pauldoo> Where is this Open Projects server anyway?
<Pauldoo> And what's it about?!?
<binEng> everywhere :)
<binEng> it's about open projects ;)
<Pauldoo> euh?  how can it be everywhere?
<Pauldoo> it's in Hungary...
<binEng> there's many servers... dynamically allocated
<Pauldoo> "provided for and by private users and
<Pauldoo>  organizations.  It is not a public forum."
<hcf> http://www.openprojects.net/ has a list
<Pauldoo> so why are we on it?
<smkl> because we are organization
<hcf> binEng: its only dynamically allocated if u use irc.openprojects.net or something
<binEng> well, you came here because I'm here. I'm here because the rest of 'em is here. They're here because... heck, find out something yourself :)
<binEng> hcf: yes I do
12:40pm
<Pauldoo> hcf?
<abi> hmmm... hcf is mailto:nef@tunes.org, http://www.tunes.org/~nef/
<Pauldoo> and what about this particular room?
<binEng> heh
<hcf> Pauldoo: room? u mean channel
<Pauldoo> yeah
<Beholder> bin: BTW: You and I are the only UniOS members on the Tunes server at the moment.  You have access to edit the web page also I think
<Pauldoo> wait, I found a closer server....
-:- SignOff Pauldoo: #TUNES (Leaving)
<Fare> Hum
<binEng> Beholder: I'll check it out
-:- Pauldoo [PaulRichar@dunvegan1.demon.co.uk] has joined #Tunes
<Pauldoo> I'm back...
<binEng> definitely, Pauldoo
<Pauldoo> so what do you do in this channel then?
<Fare> what's up?
<Pauldoo> (apart from chat)
<Pauldoo> R U using Linux yet, binEng?
<Fare> we meet and synchronize
<Pauldoo> hehe, no I mean - is it a programming channel?
<binEng> Pauldoo: no, but I'm learning UNIX in general. And I do have it installed
<Pauldoo> Me too, I just got Linux installed, but then I got my new chip.  And so it's back to Win95 for QuakeII...
<Fare> Pauldoo: occasionally, we discuss algorithms
<Pauldoo> like math/physics?
<Pauldoo> or programming?
<Fare> programming
<Pauldoo> kewl,
<Pauldoo> I got a problem,
<Fare> well, I have nothing against math or physics
<binEng> :)
<Pauldoo> I've made a top notch program that simulates air currents.  It has convection currents, really cool
* binEng/#tunes is disturbed he doesn't know how to use ssh
<Fare> (I'm not that knowledgeable in more than undergraduate physics)
<Pauldoo> Sorta what they'd use to predict the weather I guess.
<Fare> b: man ssh
<Beholder> ssh? What's it do?
<Pauldoo> anyway, how do the particles of air share their heat between each other?
<Fare> Paul: do you use adaptative grid?
<Fare> P: they do not
<binEng> Beholder: Secure shell, anti-eavesdropping or something
<Pauldoo> I just got them going "towards" the avg temp of all the particles within an x*x box...
<Fare> temp is a macroscopic notion
<Pauldoo> Fare, adaptive grid?
<Fare> for the simulation
12:50pm
<Pauldoo> sorta my own...  I don't know any of the Physics jargon, I'm 15
<Fare> there is the famous equation of heat
<Pauldoo> each particle is free to move about, like it was a single molecule...
<Pauldoo> I can simulate storms and crap, really cool to watch...
<Fare> in macroscopic physics, people rather use fields
<Pauldoo> ah..  But then they can't move about.
<Pauldoo> Isn't that for mapping heat within solids and crap
<Pauldoo> ???
<Fare> the fields evolve, corresponding to macroscopic properties of particules moving about
<Fare> not only solids, but fluids, too
<Pauldoo> yeah, whatever....
<Pauldoo> doesn't sound that good, to look at I mean.  My method is cool to watch.
<Pauldoo> :)
<Beholder> Paul: What did you write it in?
<Pauldoo> C++
<Pauldoo> I've got an EXE here to give you, just to demonstrate...
<Beholder> Paul: Sure :)
<Pauldoo> for DOS but needs DMPI
<Pauldoo> DCC?
<Beholder> ok
<Pauldoo> needs 1024*768*8bit
<Fare> Pauldoo: for number crunching, you might prefer any of Scheme (with Stalin), CommonLISP (with CMUCL), Clean, SISAL, SAC, etc
<Pauldoo> and it'll go mad if it get's too jerky
<Beholder> Fare: How would a person subscribe to the UniOS list?
<Pauldoo> I run it on an AMD K6 200, and it's ok
<Pauldoo> Beholder, what PC do you have/
<Pauldoo> ?
<Fare> Beholder: mail majordomo, with "subscribe unios" as text?
<Fare> (or even APL/J)
<Beholder> Fare: I have a link hfc gave me on the tunes site for subscribing to all the other lists on Tril's server, but UniOS is not there?
<Beholder> Fare: http://www2.tunes.org/cgi-bin/list-control?list=unios&do=subscribe
<hcf> Beholder: s/hfc/hcf/
<Beholder> hcf: Sorry :)
<Beholder> hcf: Didn't catch it until too late
<Fare> Beholder: that, too. Doesn't it work?
<hcf> Beholder: too lazy to use nick completion?
<Beholder> Fare: It looks like it works for the Tunes ones, just no entry for UniOS... hmm.. maybe it hasn't been updated in awhile
<binEng> hcf: would that be lazy?
01:00pm
<Fare> hum.
<hcf> binEng: lazy in the not knowing ur client sort of way
<Beholder> Fare: I'm looking at the HTML source, and it looks like it just needs another option value for the drop down combo box
<Pauldoo> this DCC is slow, does DCC go through the IRC servers?
<binEng> hcf: I wouldn't use nick completion for a three-letter nick either
<hcf> bah
<Beholder> hcf: Is your nick an acronym for somthing?
<binEng> hmpf
<hcf> abi: foldoc for hcf
<abi> hcf may be sought in foldoc at http://wombat.doc.ic.ac.uk/foldoc/foldoc.cgi?query=hcf
<hcf> Beholder: read that
<Pauldoo> Beholder, what spec PC do you have?
<Fare> Beholder: can you mail Tril about that?
<Beholder> Paul: Celery 400/64m Win95
<Fare> he knows the cgi-scripts better than I do, and doesn't have a bad lag when editing files on bespin
<Pauldoo> Beholder, should be OK then
<Beholder> Fare: Sure, thanks for your help :)
<Pauldoo> anyone here use Xircon?
<Beholder> HCF: There were two entires, I assume it's "Halt and Catch Fire" :)
<Fare> Beholder: I can see the problem in the script, but am not the best person to fix it, and to determine all the implications in changing it.
<hcf> Beholder: yes, sense #2
<Beholder> Paul: Looks cool :)
<Pauldoo> if you leave it for a while it'll settle with the hotter ones at the top...
<Beholder> Fare: NP, I was gonna ask Tril to check it out anyways :)
<Pauldoo> what sensitivity do you get?
<Beholder> Umm... 646
<Beholder> 38 fps
<Pauldoo> EH!!!  Ya what?
01:10pm
-:- NetSplit: koontz.openprojects.net split from norton.openprojects.net [01:12pm]
-:- BitchX+Deb1an: Press Ctrl-F to see who left Ctrl-E to change to [koontz.openprojects.net]
-:- Netjoined: koontz.openprojects.net norton.openprojects.net
-:- Fare [rideau@quatramaran.ens.fr] has joined #Tunes
-:- smkl [sami@MLXVI.rdyn.saunalahti.fi] has joined #Tunes
-:- Tril [dem@sloth.wcug.wwu.edu] has joined #Tunes
-:- abi [abi@bespin.cx] has joined #Tunes
-:- binEng [bineng@dialup152-1-39.swipnet.se] has joined #Tunes
-:- hcf [nef@me-portland-us101.javanet.com] has joined #Tunes
-:- Pauldoo [PaulRichar@dunvegan1.demon.co.uk] has joined #Tunes
<binEng> heh
-lilo(lilo@varley.openprojects.net)- [GlobalNotice] Hi all.  Problems with one of our hubs, recovering now
!doogie:*! if only real life recovery were so easy.  :|
-:- SignOff hcf: #TUNES (Leaving)
!xing.openprojects.net!! Remote CONNECT lu.openprojects.net 8005 from lilo
!asimov.openprojects.net!! Remote CONNECT lu.openprojects.net 8005 from lilo
<Pauldoo> DCC is a bit slow...
!norton.openprojects.net!! Remote CONNECT vinge.openprojects.net 8005 from lilo
<binEng> well, yes
<Pauldoo> accept it again, I've increased the frame size or whatever it is.
<Pauldoo> block size
01:20pm
<Pauldoo> binEng?  R U there?
<Pauldoo> :)
<binEng> I'm here
<binEng> had to fight a ton of web popups :p
<binEng> send again
<Pauldoo> :)
-lilo(lilo@varley.openprojects.net)- [GlobalNotice] Hi all.  We're back.  Apparently one of our hubs had an unscheduled reboot.  Apologies for the inconvenience!
<Pauldoo> accept ! ! !
<binEng> It doesn't even show up
<Pauldoo> get ready again...
<Pauldoo> whey!  Don't ya just love DCC?
<binEng> why should I?
<Pauldoo> never mind
01:30pm
-:- SignOff Pauldoo: #TUNES (Leaving)
-:- Downix [down@P191.bit-net.com] has joined #Tunes
<Downix> Hello all
<smkl> how easy it is to analyze nondeterministic programs like in lambdaND?
<smkl> hi Downix
<Downix> Hello
<binEng> hi
* Downix/#Tunes hasn't looked into the tunes project for awhile, how is it going?
<binEng> how long is 'a while'?
<Downix> um... almost a year
<binEng> :)
<binEng> I'm afraid you'll find that not as much as you're hoping has happened
<Downix> Sounds like my own hardware project.  8)  In a year, all we've gotten done is some basic paragrams and selected our distribution mode.
<smkl> Fare has invented lambdand, and tcn has written new tunes prototype
<Downix> hmm
01:40pm
<Downix> fun
* Downix/#Tunes shakes his head.  Do you know where you are headed at this rate?
* Downix/#Tunes points to HURD, 15 years in development, and it was outdone by an upstart within a year of the upstarts conception.
* Downix/#Tunes sighs
<binEng> explain that with Hurd...
* Downix/#Tunes has used Hurd, it's ok, but Linux still outdoes it.  BSD does as well, but BSD is older.
<binEng> Hurd has been in development for 15 YEARS??
<Downix> binEng:  According to GNU, Hurd's first paragram was made 15 years ago
<binEng> geez...
<binEng> what a maredream scenario <shrug>
<Downix> Linux's first part was in 1988
<Downix> hehe, of course BSD's first part was in 1969, with UNIX.
<smkl> if you have committed choice nondeterminism, what use are linear types or monads for IO?
<binEng> well, that's right...
<Downix> smkl:  This is true.
<Downix> smkl:  In the end, I'm just saying, don't let things stagnate too badly.  I'm trying to stop stagnating with my hardware design, and it's hard
<smkl> well, we can always change the project name ...
* Downix/#Tunes sighs.
* Downix/#Tunes looks at all of the rejected designs he's had so far
* Downix/#Tunes likes the concept behind Tunes, would not mind if it would be on his hardware design.
<smkl> we really need to start implementing a language with good reflection features
01:50pm
* Downix/#Tunes nods.
<binEng> I'll wait with anticipation...
<Fare> hum
<smkl> how easy it is to analyze distributed programs like join-calculus?
<Downix> Some thought must always be given to the final implimentation, what is it for, how shallit look, etc.  
<smkl> Fare: can you answer my 3 questions?
<Fare> smkl: can you repeat them?
<Fare> what do you call "analyse"?
<Fare> for instance, jocaml has a type system that extends the one of ocaml
-:- SignOff Downix: #TUNES (Read error to Downix[P191.bit-net.com]: EOF from client)
<Fare> (except that compatibility with the object system wasn't currently implemented)
-:- down [down@Q082.bit-net.com] has joined #tunes
<Fare> gakuk, down!
<down> huh?
-:- down is now known as Downix
<binEng> that was a new use for 'gakuk', not?
<Fare> nope, the same.
<Fare> I said "gakuk" specifically to down.
<smkl> well, check if a program implements it's specifications
<Fare> smkl: what language is the specs written in?
<binEng> ok I thought you said it before he rejoined
<smkl> also analyze the program to optimize it
* Fare/#Tunes is writing the scenario for a novel about matter transmission
<binEng> Fare: over or under c?
02:00pm
<smkl> Fare: any language that is good for writing specs
* Downix/#tunes is gone (04:25PM): (AutoGone - 5 mins)
* Downix/#tunes has returned...  [atl]
<Downix> UniOS?
<abi> rumour has it UniOS is at http://members.xoom.com/unios/
<Fare> uh, "over or under c?"?
<Fare> abi: abi?
<abi> i am the little girl borg with a big brain or #tunes' personal little whore or an infoslut or gaklosmontic or flurivostuginuous or xyvarestoplik
<binEng> light speed, that is
* Downix/#tunes bitchslaps api
<binEng> abi: forget UniOS
<abi> binEng: I forgot unios
<Fare> smkl: know any good such language? ;->
* Downix/#tunes just noticed the UniOS link from your page, looks interesting
<binEng> abi: UniOS is at http://unios.dhs.org
<smkl> for example coq's specification language
<Fare> bE: did you update the OSes.phtml page?
<Fare> and the index.phtml page, too?
<binEng> should I?
<Fare> smkl: coq is quite interesting; there are still many things I dislike about it.
* Downix/#tunes did not notice those
<Fare> It's not reflective, for one.
<Downix> coq?
<abi> it has been said that coq is quite interesting; there are still many things I dislike about it.
<smkl> abi: forget coq
<abi> smkl: I forgot coq
* Downix/#tunes noticed the OS layout ideas, one of them looks suspiciously like AmigaOS
-:- down [down@P224.bit-net.com] has joined #tunes
-:- down is now known as downix2
-:- downix2 is now known as downix030
<smkl> abi: coq is a proof assistant, http://coq.inria.fr/
02:10pm
<Fare> smkl: faster than me!
-:- downix030 [down@P224.bit-net.com] has left #tunes []
-:- SignOff binEng: #TUNES (Ping timeout for binEng[dialup152-1-39.swipnet.se])
-:- downix030 [down@P224.bit-net.com] has joined #tunes
<downix030> ok
-:- downix030 is now known as Downix2
<Downix2> Better
<Downix2> Did anyone see what I sent?
-:- SignOff Downix: #TUNES (Ping timeout for Downix[Q082.bit-net.com])
<smkl> Downix2: well what?
-:- SignOff Downix2: #TUNES (Ping timeout for Downix2[P224.bit-net.com])
-:- DOwnix [down@Q029.bit-net.com] has joined #tunes
<DOwnix> AAARRGGHHH!
<DOwnix> I'm going to kill my brother!
-:- DOwnix is now known as Downix
<Downix> He's annoying me because his girlfriend got so annoyed/scared of him that he can't call her anymore, so now he's askimg ME to call her!
02:20pm
<Downix> Ok, so has anyone made any decisions for Tunes yet?
* Downix/#tunes has been looking at the Tunes page
* Downix/#tunes notes the link to Flux, and looks
<Downix> What the?
<Downix> It's availible now?
>>> Downix [down@Q029.bit-net.com] requested VERSION  from #tunes
02:30pm
<Fare> hum
<Fare> Tunes is public
<Fare> you can see everything that's *done*
<Fare> tcn is building a basis on top of which we intend to write a forth machine and a scheme implementation
<Downix> This is true
<Downix> hmm
* Downix/#tunes has a dozen schemes he came up with for his own system back when he was considering writing his own OS
<Downix> But doing the hardware is difficult enough
<smkl> i would like to design HLL as fast as possible
<Fare> smkl: go ahead!
* Downix/#tunes nods.  I liked the Amiga-like concept that was proposed.  it would be very fast and very flexible for the user
<Fare> smkl: start from scheme, add and remove features
<Downix> HLL?
<Fare> high-level language
<Downix> DOH!
<Downix> Hmm
* Downix/#tunes ponders
-:- Pauldoo [PaulRichar@dunvegan1.demon.co.uk] has joined #Tunes
<Pauldoo> howdy
<Downix> Hey Paul
<Downix> Anyone here know what they want for a computer system?
<Pauldoo> Cray
<Pauldoo> or other supercomputer
<Pauldoo> Anyone know how particles of air share out heat>
<Downix> seriously
<Pauldoo> ?
<Pauldoo> yeah, I'd want a cray..
<Downix> Ok, someone else, what options do you want (not system, options)
<smkl> i want quantum computing system with TUNES
<Downix> Uh huh.
02:40pm
* Downix/#tunes doesn't think he'll get much info here then
<Fare> smkl: sure, why not?
<Fare> Downix: considering hardware, software, UI, what?
<Fare> I want voice-based computing
<Fare> I want music recognition
02:50pm
-:- Connection closed from irc.us.openprojects.net: Success
-:- Use /Server to connect to a server
-:- Connecting to port 6667 of server clarke.openprojects.net [refnum 1]
-:- BitchX+Deb1an: For more information about BitchX type /about
-:- Welcome to the Internet Relay Network TUNES 
-:- Your host is clarke.openprojects.net[blueberry.inwa.net], running version u2.10.04.resolv9.nmt.egcs4.tok.pten.tlim4.admin.upper4.whisper3.gipl.modeless7 
-:- This server was cobbled together Mon Jan 25 1999 at 07 12:10 EST
-:- clarke.openprojects.net u2.10.04.resolv9.nmt.egcs4.tok.pten.tlim4.admin.upper4.whisper3.gipl.modeless7 dioswkfcg biklmnopstv 
-:- [local users on irc(12)] 5%
-:- [global users on irc(84)] 36%
-:- [invisible users on irc(150)] 64%
-:- [ircops on irc(12)] 5%
-:- [total users on irc(234)]
-:- [unknown connections(0)]
-:- [total servers on irc(33)] (avg. 7 users per server)
-:- [total channels created(69)] (avg. 3 users per channel)
!clarke.openprojects.net Highest connection count: 27 (26 clients)
-:- Mode change [+f] for user TUNES
-:- Mode change [+iws] for user TUNES
-:- JOIN activated by "TUNES #tunes tunes@bespin.cx "
-:- TUNES [tunes@bespin.cx] has joined #tunes
-:- Topic for #TUNES: Reflective Computing System - http://www.tunes.org/
-:- topic set by Fare [Fri Apr  9 10:36:05 1999]
-:- [Users(#tunes:7)] 
[ TUNES     ] [ Downix    ] [ binEng    ] [ Fare      ] [ smkl      ] 
[ abi       ] [ Tril      ] 
-:- Channel #tunes was created at Sun Feb 28 08:48:06 1999
-:- BitchX+Deb1an: Join to #tunes was synced in 7.706 secs!!
<Downix> Fare:  That's the Palmtop
-:- Mode change [-s] for user TUNES
<Downix> B/W screen tho
<Fare> for $300? When can I buy it?
<Fare> oh, how much battery autonomy?
<Downix> Fare:  you can't yet.  We're still developing it.
<Fare> what's a RM5260?
<Downix> Fare:  6 hours in two AA
<Downix> Fare:  it's a 64-bit MIPS CPU availible from QED (www.qedinc.com) for $45
<Fare> 6 hours is not _that_ much, you know.
<Fare> what about $100 more and Li-Ion or better batteries?
<Downix> Fare:  it is when I'm considering active mode.  If idling, it stretches out for 48 hours.
<Fare> hum, maybe.
<Downix> We're still working on the design.  Right now the chips are our priority, once they are done we'll finalize what battery we'll use.
<Downix> thank goodness for HDL
<Fare> what about caches?
<Fare> better a 32-bit w/ cache than a 64-bit w/o
<Downix> The 5260 has a 64kb L1 cache and 256kb L2 cache.
<Fare> wow!
<Fare> looks mighty.
<Downix> It's a MIPS, you won't find them w/o a cache, in fact many have L3 cache's in the 4-12 meg range
<Downix> The R12000 has a 32 meg L3 cache
<Downix> We're making a scalable design, and then open licencing it.  We produce the chips, whoever wants to can produce the systems
<Tril> Downix: What company do you work for?
<Downix> By using the RM7000 CPU over the 5260, you turn the unit into an ideal workstation
<Downix> Tril:  carrad Multimedia
<Downix> BTW Fare, if you wanna see a similar idea to the MISC, look at the MIPS EE from Sony, the core to the next-gen Playstation
<Downix> Same idea, but 128-bits & 3D acceleration
* binEng/#tunes smells the word MISC
03:20pm
<Fare> Downix: any URL?
<Downix> Fare: I wish, our domain name needs a host for it
* Fare/#Tunes loathes 3D personally
* Downix/#tunes doesn't care, he has a 3D add-on designed for his system, but it's not necessary.  And removes a chip, turning it into a 3 chip solution
<Fare> I wish people would spend more time on the contents rather than on the looks.
* Fare/#Tunes wants a free hand-held design to program Tunes on.
<Downix> Our biggest worry now is the OS design.  We've got the hardware 40% done now
<Fare> take linux, to begin with.
<Fare> with 32meg RAM and 144meg HD, that's enough.
<Fare> although I'd like Tunes to run on such a beast!
* Downix/#tunes was considering, but booting Linux off of a ROM proved problematic
<smkl> TUNES will be ready soon (year 2020)
<Downix> I've had better luck with BSD and XOK
<binEng> smkl: then you'll be too late. UniOS is released already by 2017
<Downix> but Linux 2.2 *may* work.
* Downix/#tunes is trying it
<Fare> maybe OpenBSD?
<Fare> or eCos, with a Scheme interpreter/compiler on top?
<Tril> are you guys talking about a portable system?
* Downix/#tunes has OpenBSD already done.
<Downix> Tril:  Portable, or rather, the minimalist implimentation of it would be portable
<Downix> The same chips can scale up to supercomputer stage if you want it to
* Fare/#Tunes is more interested in source-level compatibility than in hardware-level compatibility
<Fare> that's the beauty of free software: you needn't have bug-compatible hardware anymore.
* Downix/#tunes is more interested in having a working system that can do what he wants it to w/o annoying me.
* Downix/#tunes nods to Fare
<Tril> what does "working" have to do with hardware?
<Fare> No more 32-bit extension to 16-bit revamp of a 8-bit computer successor to a 4-bit design by a 2-bit company that can't stand 1-bit of competition.
<Downix> yup
<Downix> Tril:  I mean right now my hardware isn't working, so obviously my system is not working
03:30pm
<Fare> Downix: I admit I know little about hardware design; all I know was read in old Rodnay Zacks books.
<Fare> (sp?)
<binEng> Tril: What kind of projects can be hosted by bespin?
<Downix> Beasides, a system is more than just software.
* Downix/#tunes nods.
<Tril> bineng: i publically offer free hosting for free OS or language projects. Others are possible by request...
* smkl/#tunes can't design HLL because he doesn't know how to add reflection to it
<Fare> Tril: what about hosting the Closures project for a web browser in CommonLISP?
<binEng> I have a guy here who's just (2 h's ago!) started a modest Linux programming group...
<Tril> smkl: do you know what a MOP is?
<Fare> smkl: to begin with, reflection can be just reification of source.
<smkl> Fare: like an AST?
<Fare> but actually, there needs be some kind of modal logic
-:- hcf [nef@me-portland-us316.javanet.com] has joined #tunes
<smkl> Tril: yes
<Tril> Fare: I don't mind, go ahead
<Fare> (something that integrates the module system as a way to do reflection)
<Tril> bineng: sure he can have an acct
<binEng> Tril: nice! I'll ask him ;)
<Fare> smkl: like an AST indeed.
<Fare> Btw, SEX are generic AST.
<Tril> smkl: i'd start reflection with adding a MOP to the language
<Tril> fare: not generic enough, IMHO
<Fare> wait ain't generic enough? SEX?
<Fare> the problem with lisp is that the module system sucks.
<Tril> that's one problem :)
<binEng> Tril: I'm calling him in
<Fare> T (a Scheme dialect) does better already.
* Downix/#tunes is gone (05:59PM): (AutoGone - 5 mins)
<Fare> MOP means "Machin Oriente Pipo"
<Tril> pipo?
<Fare> machin = thingy; pipo = hype
<Fare> hype-oriented thingy
-:- ProGuy [ProGuy@p128-14.ppp.get2net.dk] has joined #tunes
<Tril> Fare: Are you dissing the MOP?
<smkl> hello ProGuy
<binEng> hi ProGuy
<ProGuy> hey
<Tril> greets
<Fare> yes: people have come to call everything and its contrary a "MOP".
<Fare> great things have been done, but not much was understood.
<binEng> everything and its contrary? ;)
<Fare> it's time to clean up the issue :)
<Tril> Fare: the idea seems very strictly defined to me, a well-defined interface between the language and the programs written in it.
<Tril> yeah let's clean it up, sure
<Fare> Tril: I've seen "MOP" used to mean quite anything.
<binEng> ...and its contrary :)
<Fare> that, too.
03:40pm
* ProGuy/#tunes is lost
<Tril> Fare: that's because they're so cool :)
<Fare> ProGuy: do you know the Tunes project already?
<smkl> ProGuy: welcome to club
* Tril/#TUNES hands smkl a map
<ProGuy> I am familar with it, (I think)
<binEng> ProGuy: you are? I didn't know that
<ProGuy> smkl, maybe we can share that map ;)
<Tril> ProGuy: binEng told you it was the lousy second competitor to UniOS?
<Tril> abi, AMOP?
<abi> AMOP is quite interesting. or _The Art of the Metaobject Protocol_, by Kiczales, des Rivieres, and Bobrow
<ProGuy> Well, binEng I've been here before
* binEng/#tunes leans over to glance at that map
<binEng> ProGuy, ah ok
<ProGuy> Tril, No he didn't tell me
<binEng> That doesn't make you familiar with Tunes, though ;)
<ProGuy> But I can't remember what it was about at all ;)
<ProGuy> I think somebody tried to explain it to me, but I am not sure!
<binEng> abi: Tunes?
<abi> i think Tunes is in my head or tunes is a useful nevertheless expedient system. or at http://tunes.org/ or a free reflective computing system
<binEng> Beware, it's really messy
<Tril> tunes is to programming languages what the internet is to computer networks.
* ProGuy/#tunes thinks he will be lost for ever in tunes
<binEng> Tril: Nobody told ME that
<Tril> well, I just thought of that last week. I haven't told anybody.
<binEng> ProGuy, no need to get involved... just get even with Tril
<smkl> ProGuy: what programming languages and operating systems do you know?
<Tril> "get even with Tril"?
<Tril> I don't like the sound of that
<binEng> ;)
<ProGuy> smkl, Well primary C/C++ for dos/win, but I have started C/C++ Linux programming
<Tril> what do you mean by "get even"?
<smkl> smurf Tril
<binEng> what I really meant was that you two could set up that project...
!mccaffrey.openprojects.net!! SERVER Numeric Collision: varley.openprojects.net != drexelhill.pa.us.opirc.nu
<Tril> what is it?
* binEng/#tunes turns to ProGuy
<Tril> I recommend a DHS.ORG address for it.
<Tril> or other domain if you can afford it.
<ProGuy> Sorry, I have been browsing!!
<Tril> well, I'm doing homework.
<Tril> so don't worry
* binEng/#tunes hits ProGuy with a plastic hammer
<ProGuy> Oh, ok
* ProGuy/#tunes is ashamed
<Fare> quick, learn something else than C++ to decorrupt your mind
<ProGuy> Fare, eh?? That doesn't make sence!!
<Tril> sure it does
<Tril> Just like BASIC
<ProGuy> Eh, I am still confused!
<binEng> don't listen to them... they'll advice you to learn perl or scheme or who knows what
03:50pm
<smkl> do you need something else but source reification for reflection?
<Fare> The use of COBOL cripples the mind; its teaching should, therefore, be
<Fare> regarded as a criminal offence.
<Fare>                 -- Edsger W. Dijkstra, SIGPLAN Notices, Volume 17, Number 5
<smkl> learn Haskell
<Fare> same could be said of BASIC, C++, and more.
<smkl> haskell?
<abi> rumour has it haskell is lazy purely functional programming language at http://www.haskell.org/
<Fare> smkl: yes and no.
<ProGuy> I stick to C++, I like it!
<Fare> smkl: did you read the conclusion to lambdaND?
<binEng> ProGuy, don't listen too much to them, THAT if something is not good
<Fare> PG: your mind is sure crippled.
<Tril> binEng: You don't like haskell?
<ProGuy> Wow, dangerous people, eh? ;)
<Tril> if you use C++ in a consistent way it's not as bad..
<Fare> PG: go learn COBOL and VISUAL BASIC!
<ProGuy> Visual Basic, yeah right!
<Tril> Fare: Don't forget Javascript
<Tril> TCL
<binEng> Tril, To be honest I've not spent too much time on it... can hardly remember it
<Tril> ProGuy: anyway, what is your project?
<ProGuy> http://prog.virtualave.net
* Fare/#Tunes is away (shower)
<ProGuy> That is my project
<Tril> stupid term client doesn't have copy/paste
<binEng> prff haha
<Tril> popup ad boxes, die!!!!!
<Fare> Tril: screen has copy/paste
<ProGuy> In that case it is good that it isn't a members.wierddomainname.net/members/archieve/~zone51/user/prog address
<Tril> fare: I'm in NT
<Fare> Tril: sucks to be you!
<Fare> Tril: download the Tuxissa virus :)
<binEng> yeah :)
<ProGuy> What's wrong with NT, besides that it sux?
<Fare> what's wrong with FOO, besides that it's wrong?
<ProGuy> hmm... No idea
<Fare> what's FOO with BAR, besides that it's FOO?
<binEng> :)
-:- timestamp [dancer2@207.16.36.12] has joined #tunes
-:- timestamp [dancer2@207.16.36.12] has left #tunes []
<ProGuy> Could somebody give me that map again ??
<Tril> of course, the only REASON I'm in NT is because I'm downloading CD images of Debian GNU/Linux for my friend to burn off of the NT server's temp volume
<binEng> now you're getting to the hard philosophical issues, Fare
<Fare> ProGuy: left door in the bottom of the corridor
* smkl/#tunes gives ProGuy a top secret army map
-:- timestamp [dancer2@207.16.36.12] has joined #tunes
<ProGuy> Thanks smkl, now the only problem is that I am too stupid to read a top secret map ;)
<Fare> Tril: telnet into bespin, and use screen there
<smkl> hello timestamp
<binEng> ProGuy: then right at the second stairs, over the roofs and down in the pool, into the grate, and you're close
>>> Tril [dem@sloth.wcug.wwu.edu] requested PING 923698897 587249 from #TUNES
<ProGuy> thanks binEng ;)
<Tril> Fare: Internet explorer won't run inside screen
<Fare> tril: lynx will!
04:00pm
<binEng> long live lynx... but soon its time for it to DIE DIE DIE
<Fare> well, emacs w3, then!
<ProGuy>  binEng, eh?? behave yourself 
<Fare> long live lynx!
<ProGuy> YEAH Spice Girls, in the radio, YAHOOO!
-:- tcn [tcn@cci-209150250085.clarityconnect.net] has joined #tunes
* ProGuy/#tunes hates them
<Fare> mu. they don't deserve hate.
<binEng> yep
<binEng> want an mp3 with 'em? ;)
<Fare> binEng: /vmlinux
<ProGuy> Well, that would be waste of bandwidth, binEng!
<binEng> eh?
<tcn> hey.. what's up?
<binEng> ProGuy, agreed
<Tril> what is *~list?
<Fare> tcn: I'm Francois-Rene, not Francois-Fare!
<binEng> Fare: whadda mean with '/vmlinux'?
<Fare> Tril: everything but list.
<Fare> binEng: good mp3-alike of the spice girl?
<tcn> yeah
<Fare> tcn: I think the problem is with the keyboard code.
<binEng> ah good substitute
<Tril> they need bochs on these computers
<Fare> tcn: could you explain how to debug for dichotomy-boot?
-:- _QZ [brand@p0wer.qzx.com] has joined #tunes
<Fare> (hey, there could be easy dichotomy-boot by having a counter in one object file, without having to reassemble the other ones)
<tcn> Fare: hang on..
<Fare> tcn: also, consider having only the barest FORTH system, w/o a FORTH compiler in FORTH, having the FORTH compiler in Scheme instead.
* Fare/#Tunes is away (really need a shower)
<_QZ> hmm, i heard that the french didnt shower
<tcn> hehe
<binEng> heh
* Downix/#tunes has returned...  [atl]
04:10pm
-:- SignOff ProGuy: #TUNES (Ping timeout for ProGuy[p128-14.ppp.get2net.dk])
<_QZ> wow, with all these ppl here tonight is there any good conversations?
<tcn> yeah I can't keep up with it
<tcn> anyway, to answer fare's questions..
<Downix> Ok, I need to get to a biz meeting, bbl
<tcn> 1) what's a dichotomy-boot?
<Downix> email me at down@ici.net if you'd like, I could use some ideas for sample designs to give out with my licence
<binEng> I was hoping that ProGuy would get an account set up tonight, but he even left...
<tcn> 2) A forth compiler is easy, actually, it's *necessary* if you want to define words/functions
-:- SignOff Downix: #TUNES (*wave*)
<_QZ> crap, i would love to stay and chat but im going paintballing right now :(
<smkl> why would somebody like paintball? isn't that bit like war?
<tcn> smkl: but you don't die
<binEng> I'd like it, I think
<Tril> 1) dual boot?
<Tril> 2) you don't need to add new forth words, if the only thing written in Forth is the Scheme interpreter
<smkl> you should go to millitary service instead of me
* Tril/#TUNES doesnt paintball because it costs money
<_QZ> smkl: yes ;)
<_QZ> smkl: i was in the us army
<Tril> Q all you can Z
<_QZ> smkl: i go hunting and kill animals with a bow
<_QZ> tcn: u can die
<_QZ> tcn: i had a hole shot in my forearm
<tcn> how?
-:- ProGuy [ProGuy@p412-109.ppp.get2net.dk] has joined #tunes
<ProGuy> heja!
<_QZ> a 68 caliber paintball travelling at 240fps from a distance of 1 foot
<binEng> Hey ProGuy, still not paid your ISP? :)
<tcn> ow
<binEng> QZ: how much is a fps?
<_QZ> feet per second
<binEng> about the same in km/h, then?
<tcn> Tril: (2) The Scheme interpreter would be considered a Forth word.  Oh yeah, I'll write most of Forth in itself, so I gotta have compilation. That settles it :)
-:- ProGuy_DK [ProGuy@p131-46.ppp.get2net.dk] has joined #tunes
<ProGuy_DK> Hiya!
<_QZ> i have seen paint go thru the cracks in the mask and cut up the eye
<binEng> Hey ProGuy, still not paid your ISP?
<smkl> do they have paint bazookas?
<ProGuy_DK> binEng, how could you guess that??
<_QZ> smkl: no, but they have paint grenades
<binEng> a wild guess...
<binEng> air strikes, too? missiles? robots??
<binEng> some napalm variant, maybe? :)
<tcn> robot wars would be cool
-:- SignOff ProGuy: #TUNES (Ping timeout for ProGuy[p412-109.ppp.get2net.dk])
-:- SignOff Tril: #TUNES (Ping timeout for Tril[sloth.wcug.wwu.edu])
<binEng> tramp mines :)
<smkl> they could have real napalm easily
<ProGuy_DK> Wow, finally I quit!
-:- ProGuy_DK is now known as ProGuy
04:20pm
<smkl> and tear gas
<binEng> heay
<binEng> yeah
<_QZ> This memo is to announce the development of a new software system which will
<_QZ> be Year 2000 compliant. known as:
<_QZ> "Millennia Year Application Software System" (MYASS).
<_QZ> Next Monday there will be a meeting in which I will show MYASS to everyone.
<binEng> I can't wait...
<_QZ> :)
<tcn> haha
<_QZ> any1 want the rest?
<tcn> Windows-2000 is MYASS-Compliant
<ProGuy> Windows-2000 looks promising! It runs pretty good on a p150 ;)
<_QZ> www.qzx.com/myass.txt
<binEng> Anyway, how about a weekly time when people could come and find other here and discuss, etc?
-:- Tril [dem@sloth.wcug.wwu.edu] has joined #TUNES
<tcn> Isn't that what we usually do on Friday at this time?
-:- _QZ has changed the topic on channel #tunes to: www.qzx.com/myass.txt 
<smkl> all the time
<binEng> it is? officially?
<Tril> anyone seen the Matrix yet?
<tcn> oh yeah, I no longer have band practice on friday night.. I should change my toafm..
<Tril> get YOURASS over there now!
<tcn> matrix, nope
<binEng> Tril, is the Matrix a new movie?
* Fare/#Tunes feels fresher
<tcn> yeah what's it about (don't spoil it)
<Tril> bineng: yeah
<ProGuy> I haven't even heard about Matrix
<binEng> then it'll probably come here in a month or so. I guess.
<Tril> tcn: it's about Reality
<tcn> Tril: Is it realistic? (I typed that before you said it's about reality)
<Tril> it's about the future. I don't personally believe the future could be like that, but...
<Fare> tcn: 1) it's a debug technique to determine where system code fails, by inserting foo: halt;jmp foo statements at dichotomy-varying places in the code to find the precise moment where it fails
04:30pm
<Tril> it was an interesting one though.
<Fare> 2) it'll be easier and cleaner to have the compiler in Scheme
<Tril> i would pay for you all to see it
<tcn> Fare: We could just write the Scheme compiler in assembly too
<Fare> or rather, write a simple Scheme (subset) compiler in Scheme: that way, we're bootstrapped.
<tcn> s/compiler/interpreter
<tcn> I'm gonna get Forth working (almost there) then worry about scheme
<_QZ> Tril: ive been waiting for some1 to tell me it was a good movie, now i will go see it
<Fare> ok, then I'll get scheme working meanwhile
<Tril> qz , yes
<binEng> Does anybody here know of any unix program that lies in the background and tries to fetch files from the net, not giving up the first time it fails?
<Fare> tcn: we do need dichotomy boot do deter hardware problems.
<Tril> bineng: any unix program can go in the background. but wget does that
<Fare> binEng: lftp
<_QZ> binEng: wget
<binEng> which?
<abi> which is what
<Fare> wget, too, although I prefer lftp for ftp.
<binEng> abi: what is which
<abi> which is what
<Fare> and w3mir for the web
<Fare> btw, screen is your friend
<ProGuy> Well binEng, so much for that idea ;)
<binEng> shutup
<tcn> Fare: so a dichotomy boot is just inserting 'jmp short $' and trying to find the place before it crashes?
<Fare> tcn: dichotomy boot is trying to find the place where a program crashes and reboots, by having it halt precisely at the moment before which it reboots.
-:- SignOff hcf: #TUNES (Leaving)
<tcn> Fare: ok, I gotcha
<Fare> retro reboots early on on my laptop; dichotomy boot may tell me when exactly.
<Tril> how the heck do you do that
<tcn> Fare: Start by putting 'jmp short $' at various places in boot.asm, then init.inc
<Fare> tcn: ok
<tcn> laptop keyboards tend to be weird :)
<Fare> the usual suspect is the keyboard driver.
<tcn> Does it say anything on the screen before it dies?
<Fare> it is *very* difficult to get the keyboard right on all computers.
<Tril> Why can't you add more verbose output? The boot image can't be made any larger?
<Fare> I wonder how linux does: better do it the same.
<tcn> Tril: Now that I rewrote the loader, it can be as big as necessary
<Fare> Tril: because when it hard reboots, you don't get to read the output
<Tril> so add delays, too!
<smkl> Fare: what do you think about using elisp for teaching functional programming?
<Fare> that leads to as much code or more than dichotomy boot for the same result.
04:40pm
<Tril> Fare: but saves considerable human time.
<Fare> smkl: not perfect.
<Fare> smkl: to whom?
<tcn> We need a scrollback buffer or 'more' pretty soon :)
<Fare> Tril: verbose output is complementary to dichotomy boot, not opposite to it
<Tril> can Bochs be installed on NT without modifying the registry or putting any files in WINDOWS\SYSTEM?
<smkl> Fare: some university
<Fare> Tril: I doubt it
<Fare> but who knows?
<Tril> does it use installshield?
<Fare> smkl: I'd go for Haskell;
<binEng> Can I logout and still have screen running?
<Fare> elisp is klugy, and it's doomed;
<Tril> bineng, ya, that's the idea
<ProGuy> Tril, usr Cleansweep ;)
<binEng> just checking
<Fare> and it is inefficient -- might lead to a false impression of FP
<tcn> Delay: Do 'extern delay' then 'callproc delay, <milliseconds>'
<Fare> tcn: what about verbose output?
<Tril> nasm has macros now?
<tcn> I also use 'callproc print_hex, <value>' alot
<tcn> Tril: where have you been the last year?
<binEng> How did you say I start a new screen, now again?
<Fare> considering my experience with macros in TASM and as86+m4, I far recommend an open Scheme compiler over any pseudo macro language
<Tril> ^a c
<binEng> ahh tnx
<Fare> man screen
<Fare> ^A ?
<binEng> yeah yeah I've done that... I just hadn't it availible
<Fare> <Fare> tcn: what about verbose output?
<tcn> Fare: 'callproc print, <string ptr' or 'callproc putc, "*"' (or whatever character you want)
<tcn> Strings are a pain in the ass in assembly. When we write our assembler it should have an easy way to insert strings inline
<tcn> (example: push "hello world" ; call print)
<binEng> jesus, 'man screen' is 46 pages??
<_QZ> tcn: why the hell do u push yer strings?
<tcn> by the way, I made a shitload of changes to Retro since I last committed it yesterday.
<Fare> tcn: well, time to commit again, then!
<tcn> qz: I'd *like* to be able to push strings that way (have the assembler put the string elsewhere, and generate a 'push <pointer to string>'
<Fare> tcn: I did it with a nasm macro, long ago.
<Fare> ok, maybe time to resurrect that asm-in-scheme
<tcn> Fare: I'm almost done.. just changing the readline function
04:50pm
<Fare> (or maybe just nasm-outputing scheme?)
* _QZ/#tunes is GONE ( paintballing )
* ProGuy/#tunes will be leaving soon (VERY TIRED)
<tcn> nasm ain't great
<tcn> A 'high level assembler' would be great
<Tril> tcn: you should really talk to the 2nd-to last new member
<Fare> tcn: you can unearth the asm-in-scheme files from the tunes project...
<Tril> he';s working on a reflective assembler
<Tril> ops, I guess he didnt join holdon
<tcn> Fare: I've seen asm-in-scheme; seems like a good idea
<tcn> asm-in-forth looks useful, too
<tcn> Tril: Who is he?
<Tril> mate at c3.hu
<Fare> c3.hu?
<Fare> tcn: win32forth has asm-in-forth
<Tril> he wrote me some private mails about his project, it is a dynamic assembler, in assembly
<tcn> Fare: yeah, I was thinking of stealing that one :)  It does 2 or 3 syntaxes, very nice
<Fare> the sources are very cryptic, tho
<tcn> I think that's how Forth is :)
<tcn> cryptic
* ProGuy/#tunes found a win16/32 emulator for Linux
<tcn> But it Forth gives you a lot of power from such a small compiler
<Fare> sure
<Fare> Scheme isn't that large.
<Fare> and ANS forth has a *lot* of legacy
<tcn> ANS forth sucks
<Fare> (not ANS is pointless)
<Fare> so all in all, forth has problems
<Fare> unless seen as a vm under a higher-level system
<Fare> (see the OCAML bytecode -- it's kinda forth)
05:00pm
<ProGuy> win95 programs with win 3.1 gui, cool!
<Fare> (well, not exactly)
<Tril> postscript
<tcn> Fare: Have you read hbaker's article "the forth shall be first" or something like that? It talks about "linear lisp" which looks like a cross between forth and lisp..
<Tril> how much is that like forth
<Fare> tcn: I love hbaker.
<Fare> I've read all his stuff -- great
<Fare> and yes, I want a LISP with linear types, too
<tcn> Tril: why, I never noticed that! postscript is forth, more or less ;)
<binEng> Is the source of wget availible? where?
<tcn> Fare: I'll have to read the whole thing.. it's been sitting on my HD since I got into Forth last year
<binEng> abi: wget?
<abi> wish i knew, bineng
<smkl> if you have committed choice nondeterminism, what use are linear types or monads for IO?
<tcn> As for Scheme/LISP, it can't be that hard to implement. They were using it in the 50's.
<smkl> modern lisps are much bigger and better than original lisp
<Fare> tcn: 1958 for the earliest version
<Fare> smkl: we're not forced to include full CommonLISP in the core
<Tril> yes, you are! :)
<Fare> the goal is precisely to have an open system so that starting from a small core, we can expand
<Tril> common lisp must run in ring 0!
<tcn> We want to write as little as possible in assembly
<Fare> smkl: if you read lambdaND, you'll see that cc-nd is just call-by-value in a nd environment
<tcn> write the core, then bootstrap the rest of the language by writing it in itself
<Fare> smkl: it doesn't quite remove the interest of linearity.
<tcn> I did that with my old 16-bit forth
<tcn> Tril: GCL doesn't run as root
<Tril> i'm kidding
<tcn> Should I add some shorter macros for verbose debugging output?
<smkl> oh yeah ... order of IO operations must be specified
<tcn> ('x <value>' to print hex, 'p <value>' for a string.. ?)
<Tril> can someone sum up what linear lisp means, for me who did not read the article?
<tcn> fare?
<abi> it has been said that fare is asleep
05:10pm
<tcn> allright guys, Forth is about to work
<Tril> go go go
<Tril> are we forthing at the mouth?
* tcn/#tunes crosses his fingers
05:20pm
<tcn> dooough!
<tcn> you can make alot of stupid mistakes in asm :)
<tcn> well, it pseudo-works
<tcn> I mean, it works!
<binEng> ok. So how DO I logout with screen started?
<tcn> ask tril, he uses it
<binEng> I ask anybody who cares to answer
<tcn> beep him
<Tril> ^a d to put screen in the background
<Tril> then type 'screen -r' to get back to it
<binEng> oh
05:30pm
<tcn> heh, it would help if Forth could parse numbers :)
<tcn> I seem to have forgotten that important item
* Fare/#Tunes is back
<Fare> Tril: just go and do read hbaker's articles :)
<tcn> fare, have you written anything for retro yet?
<Fare> tcn: not yet. I'm rather thinking about the Scheme part currently
<Fare> what's the execution/memory model you chose for retro?
<tcn> Fare: good :)  that means I can't have broken it ;)
<tcn> I just changed the memory layout
<tcn> First 4 megs mapped linear->physical, last 4 megs are for the kernel & modules, allocated as needed
05:40pm
<tcn> There are CODE, DATA, and STACK areas
<Tril> tcn: hey, you copied the NT memory model exactly! :P
<Tril> uh, don't you mean gigs?
<tcn> no, megs.
<binEng> How can I kill a screen that's stopped to respond?
<tcn> You can always map more memory
<Tril> you mean a window inside the screen, or all windows?
<tcn> Fare: need to know more?
<binEng> just a window
<Tril> ^a k
<binEng> ahhh
<tcn> damn, forth compilers are so easy to write!
<Tril> so what's up?
<Tril> retro has forth in it now?
<tcn> yeah, I'm just adding words to the dictionary now
<tcn> and debugging
<tcn> in 5 minutes it'll be doing basic math, at least :)
<Tril> unbounded precision numbers?
<tcn> yeah ;)
05:50pm
<tcn> as long as you keep them under 2^32
<tcn> hmm..
<Tril> uhh...
<Tril> i guess not.
<tcn> Scheme or LISP does that alot, I've noticed
<tcn> arbitrary precision, i mean
<Tril> It's called robustness
<tcn> who ever said forth is robust?
<ProGuy> I am leaving now, cya later
<Tril> k, bye!
<Tril> i dunno tcn.
<ProGuy> and thnx Tril
<binEng> Tril: were you spying on me now again?
<Tril> no
<Tril> unless running 'w' counts as spying
-:- SignOff ProGuy: #TUNES (Read error to ProGuy[p131-46.ppp.get2net.dk]: Connection reset by peer)
<binEng> so why did it say "write: dem is logged in more than once; writing to ttya7" when I logged out?
<binEng> and what is 'w'?
<Tril> that's my notify program
<binEng> which...?
<abi> which is what
<tcn> haha
<Tril> every time you log off, it looks in the ~/.notify of every user to see if they want to be informed of your leaving/coming
<binEng> ok
<Tril> also when you logon
<Tril> you can put names in your .notify if you want to know when people sign on or off
06:00pm
<Fare> tcn: I'm more interested in the register layout
<Fare> and with stuff whose address is possibly hardcoded.
<Fare> the rest is resolved at link times
<Tril> fare, are you actually going to do some work on tunes now? with scheme ?
<Fare> I hope so.
<Tril> I'm revising my spec
<Tril> It's kinda incoherent
<Tril> I wrote it in a rush last quarter
<tcn> Fare: Registers? They're all available, except for ESP
-:- SignOff binEng: #TUNES (Ping timeout for binEng[dialup116-4-40.swipnet.se])
<tcn> Tril: spec for tunes? persistence? ..?
<Tril> yes, that's it :)
<Tril> lots about persistence
<Fare> tcn: I mean, the calling convention
<Tril> I have invented persistence with no GUIDs
<Fare> mind you, scheme is quite different from C.
<Fare> It doesn't even have linearity guarantees on continuations!
<Fare> Tril: my boss has already invented distributed programming w/o GUIDs
<Tril> i'm sure it's not a new idea.
<Fare> he's a very bright guy, but lost in administratrivia so he cannot do much research anymore, only lead other people's research!
<tcn> Fare: The kernel functions mostly use stack frames.. when I commit, look at 'odb.inc' and 'loader.inc' for some good examples.
<tcn> hey, what's a GUID stand for?
<Fare> Globally Unique IDentificator
<Tril> my idea is that since all references to objects are in the same persistent store as the object they refer to, that having a GUID is unnecessary: all links are internal.   I guess it's standard persistent hyperprogramming (and hyper-data-structure-building)
* Fare/#Tunes needs food to learn
06:10pm
<tcn> ah, I've been reinventing what your boss invented.. I'd like to see his ideas sometime..
<Tril> much like Bill gates.. but does Linus do much coding anymore?
<tcn> he works for transmeta
<tcn> = paul allen
<Tril> i mean does he program the kernel, or just modify and approve stuff
<tcn> I read somewhere that he still hacks alot at night
<Tril> what did you reinvent tcn?
<Fare> Tril: probelm is with *distributed* computing
<Fare> Tril: persistence on monoprocessors is long known (see Eumel, or machines at Xerox)
<Tril> that's easy, you can easily convert a persistent store into a sub-store of another, just add precision to the implicit GUID.
<tcn> I haven't reinvented it yet, I'm just trying.. in designing a persistent store, it's easy to start thinking about distributed persistence
<Fare> but having a global pool w/o external links sucks wrt distributed computing (including cooperative computing and even partial backups!)
<Tril> types are sets, you can do set union on two types
<Fare> he does program the kernel, mostly SMP and mem mgt
<Tril> fare: There are no "external" links. If it's distributed, that means remote and local are unified into one store.
<Fare> you must think about 0/1) read and write 1/2) cache and invalidation
<Fare> s,1/2,2/3,
<Tril> the whole way to do distribution is to provide a way to allow links between the places , as if they were the same
<Tril> fare: i do that by 2 way links: the object referred to has the authority to modify all references to it, anytime it moves or changes.
<Fare> Tril: when you implement distribution, all the more in presence of various failures, you must differentiate local and remote
<tcn> Fare: In forth, calling conventions are simple: ESP is the return stack pointer, EBP is the data stack pointer, and you can still call C-style functions from Forth functions, no problem
<Fare> tcn: why not reuse XCOM?
<Fare> the author (core) is willing to adapt it to your needs if you want
<Fare> and it works with NASM, too
<Tril> they are differentiated at one level, unified at the other, isn't that the whole point of tunes?
<Fare> anyway, we'll use our own scheme-programmed format, anyway.
<tcn> right
<tcn> XCOM support can come later.. I did mention that I wanted to support it at some point
<Fare> Tril: sure
<Fare> Tril: as implementers, we must take into account the difficult part, the one where they are differentiated
<Tril> Well, I'm just writing the spec, so I'm not worrying about that, yet!
<Fare> (or rather, the difficult transformation, implementing unified  differentiated)
<Tril> I am not the implementer yet.
<Fare> in mpfas, I describe a generic model for metaprogramming;
<Fare> a very important thing is the concept of caches.
06:20pm
<Fare> tcn: can you explain odb.inc ?
<Fare> doesn't look like it works for circular data structures...
<tcn> Fare: wait for me to commit it.. I rewrote it
<tcn> Fare: btw, ODB = object database... I'm using that term to describe the structure that holds all the object headers.. but the actual structure of it could change
<Fare> <subliminal>commit commit commit</subliminal>
<Fare> tcn: you should let that be written in HLL
<Fare> this is no job to do in asm
<tcn> Fare: right now it's a linked tree
<tcn> asm beats C any day
<Fare> the LLL should be minimal
<tcn> agreed
<Fare> tcn: for the odb, we need flexibility; asm won't give us that
<tcn> asm -> core lll -> rest of lll -> core hll -> full hll
<Fare> (a few spots of inline asm will be welcome, of course)
<Tril> Fare: do you mean section 5.1?
<Fare> asm -> core lll -> core hll -> rest of lll -> full hll
* tcn/#tunes reminds everyone that this is a prototype
<Fare> 5.1?
* Fare/#Tunes is confused
<Tril> Reflective Programming / Reflective Systems : "Let us examine the properties of a reflectivee development framework,..."
<tcn> Fare: order doesn't matter there
<tcn> hey
<Tril> i'm trying to find where your "generic model for metaprogramming" is you just said
<Fare> rather 2.2
<tcn> the HLL should encompass the LLL
<Tril> a contents or index would be niec
<Tril> or is there one? I'm looking at the html version.
<Fare> Eat flaming death, non-reflective mongrels!
<Fare> the html is translated from latex that didn't have toc
<Fare> <subliminal>commit commit commit</subliminal>
<Fare> Tril: it's quite condensed
<Tril> am I a non reflective mongrel? why?
<Fare> not you!
* Fare/#Tunes looks up the webster at "mongrel"
<Tril> did you mean tcn then?
<Fare> nope
06:30pm
<Fare> just preparing an assault against non-meta systems
<Tril> Ok just checking. i thought one of us may have been being non-reflective, but i couldnt' see where or why
<Fare> :)
<Fare> have you seen the stuff on polycontextural logics?
<Tril> not me
<Fare> anyone here reads german?
<Fare> my, why didn't I learn german?
<Tril> my sister is learning german in high school..
<Tril> because french is better, probably
<Tril> even if it's not
* tcn/#tunes preparing to commit..
<Fare> my sister is, too. But she ain't making much efforts.
<Fare> French is indeed, better. Why don't everybody speak french? ;->
<tcn> english includes french :)
<tcn> ok, here goes
<Fare> did you s/Francois-Fare/Francois-Rene/ ?
<tcn> hehe
* Fare/#Tunes tries to find people in Paris to work closely on Tunes.
06:40pm
<tcn> gotcha
<Fare> you mean, you committed?
<tcn> un momento
<Fare> maybe time to wake core and tell him to really do publish code!?
<tcn> has he worked on it since he gave us copies?
<Tril> of course.
<Tril> he works on it all the time
<Tril> maybe he's afraid if he releases it, his company will be disappointed about the response, and cancel him
<Fare> but he wants the GPL to be legally granted
<Tril> ?
<Fare> you can't start big
<Fare> you must start small.
<Fare> if he's afraid to start small, he'll never manage
* tcn/#tunes has been committed
06:50pm
<tcn> So what else do we need now that we have Forth?
<Tril> to completely test forth?
<Tril> debug everything
<Tril> rigorously
<tcn> feature-wise
<tcn> obviously we need to compile Forth programs from disk
<Tril> what forth features does it have (note I'm not all that familiar with forths)
<tcn> not much yet :)
<tcn> It has stack operations, memory read/writes, some text & number I/O..
<tcn> the hard part was the object database, which is useful to the whole OS not just Forth.
<Fare> tcn: we need get the thing to work on my laptop :)
<Fare> tcn: a disk driver?
<Fare> and/or a VM86 or RM86 mode
<tcn> I wrote a VM86 handler but I took it out. I'll put it back when I add multitasking.
<tcn> Have you looked at floppy.asm?
<tcn> it's sloppy but it works
<tcn> want IDE or SCSI support?
<Fare> IDE is simpler -- only one "chipset"
<Fare> (well, not so, after all)
07:00pm
<Fare> SCSI is really one completely different driver per chipset.
<tcn> ack
<Fare> SCSI is mucho better designed, however.
<tcn> well I only have IDE
<Fare> yeah, IDE is everywhere
<Fare> maybe reuse the drivers from oskit?
<tcn> yeah
<Fare> that should be a goal for us: reuse
<tcn> what language are they? mostly C and a little asm?
<tcn> we might be better off rewriting drivers, using OSkit as a guide. Alot of interfaces are different in Retro/Tunes
<Fare> C mostly, with gcc inline asm at spots
<Fare> tcn: the one who does decides.
<Fare> you might inspire from MMURTL, too
<tcn> hmm.. this is looking like a real possibility now: Compile to XCOM using linux gcc, and import the object into Retro
<tcn> MMURTL?
<tcn> We could have a "Unix sub-environment" with a C library and all..
<Fare> tcn: available by ftp from cygnus; there's a book about it, too
<Fare> tcn: we will have a linuxemu, but later
<tcn> what's mmurtl do though?
07:10pm
<Fare> yeah, importing C-compiled XCOM might be a good thing
<Fare> until we have a C compiler or a linux emulator (not before a loong time)
<tcn> is there an XCOM patch for nasm?
<Fare> core had one with clementine
<tcn> ok
<Fare> mail him about it. Or maybe I have an archive of it?
<Fare> hum, not here
<tcn> When we write an ML compiler, it'll be similar to C, syntax-wise
<Tril> i have it
<Fare> tcn: the difficult part is not the syntax
<Fare> (although in the case of C, the syntax is *quite* annoying)
<tcn> oh C's not that hard
<tcn> I hate lex/yacc
<Fare> C has its tricky parts;
<Tril> the patch is in os/tools/patches
<Fare> but the most horrible is dealing with CPP in an intelligent way
<Fare> (and dealing with GCC extensions, too)
<tcn> Isn't there a better way to parse infix than with LALR(1) grammars?
<Fare> <shudder>
<Fare> I hate parsing.
<tcn> haha
<Fare> Which is why I like LISP
<tcn> and I like forth
<Fare> I think parsing is so stupid a task!
<Tril> I wrote a parser for my OS before I joined tunes
<tcn> that's why GCC is so slooow
<Tril> it worked well
<tcn> hey fare, tried the new retro yet?
<Fare> not yet, would have to deconnect for that.
<Fare> compiled it only
<tcn> unless it worked on your laptop :)
<tcn> read through kernel/Notes.txt, odb.inc, forth.asm...
<Fare> if you really like FORTH, maybe write an openboot interpreter ;->
<Tril> tcn, can you explain how to use bochs so I can try retro?
<Tril> I got bochs compiled, but it wants a hard drive image file. I've never used it before, how do I make it load a retro image?
<Tril>          iodev/          pc_system.cc
<Tril> LICENSE         config.cache    debug/          macintosh.txt   pc_system.h
<Tril> Makefile        config.h        disasm/         macutils.c      pc_system.o
<Tril> Makefile.in     config.h.in     docs-html/      macutils.h      state_file.cc
<Tril> README          config.log      dynamic/        main.cc         state_file.h
<Tril> bios/           config.status*  font/           main.o          state_file.o
<Tril> bochs*          configure*      gui/            memory/         win32.txt
<Tril> bochs.h         configure.in    instrument/     misc/
<Tril> dem@bespin:~/src/bochs-990219$
<tcn> uhh.. use dd
<Tril> whoa, I can paste! now I have to figure out how.
<Fare> hum. I haven't recompiled bochs since a few months ago. Only two tries per year, the license says.
<tcn> hang on..
<Tril> this is my first try ever.
<Tril> tcn: dd ? Why? I want to load the image off a hard disk. I have no floppy drive
<Tril> (on bespin)
<tcn> tril, go to retro/kernel
<Fare> tcn: multitasking can be programmed in FORTH/Scheme
<tcn> there's a .bochsrc there, and a 'make bochs' target in Makefile
07:20pm
<tcn> Fare: I want it to be as efficient as possible: ASM
<tcn> Fare: Some things are easiest in asm, anyway
<Fare> FORTH and Scheme can eventually do inline ASM...
<tcn> yeah
<tcn> too bad 80x86 machine code is so convoluted
<Tril> now I need nasm. Is there a .deb for the nasm I need?
<Fare> you really only need do the actual switching in ASM
<Fare> Tril: you need recompile the latest version
<tcn> Tril: see http://bespin.cx/~tcn/retro.html
<tcn> I've got a link to it there
<Tril> that link has a typo btw
<tcn> oh yeah, thanks
<tcn> hcf said I had a shitload of typos :)
<tcn> Tril: You'll need to edit the vgaromimage: line near the top of .bochsrc
<tcn> I wish you luck
<Tril> problems compiling nasm
<Tril> getting autoconf and automake
<tcn> hehe
<Tril> configure in nasm isn't working
<Tril> maybe I'll read the instructions
* Fare/#Tunes is away
<Tril> or you could explain it ...
<Tril> why do I need this version of nasm?
07:30pm
<tcn> nasm-0.97 has RDOFF 1.1 (which had some oversights), nasm-0.98 has RDOFF 2.0 (which is what Retro uses for modules)
<tcn> What's configure choking on?
<Tril> 0.97 won't work at all?
<Tril> it's choking on me not knowing how to compile it
<Fare> ./configure ; make ; make install
<Tril> the ./configure that comes with that nasm file, is corrupt, it doesn't run.
<Fare> sh configure ; make ; make install
<Tril> : command not found
<Tril> : command not found
<Tril> : command not found
<Tril> : command not found
<Tril> : command not found
<Tril> '/configure: line 57: syntax error near unexpected token `do
<tcn> does 'make' by itself work?
<Tril> '/configure: line 57: `do
<Fare> oh, I remember:
<Fare> you must pkunzip -t
<Fare> to convert CRLF into LF
<Tril> what's the info-unzip equivalent, then?
<Fare> unzip -a
<Tril> it works, thanks!
<tcn> haha
<Tril> I am now the only person in this lab with 40 computers
<Tril> and I can't access the cd burners :(
<tcn> If you were an MIT student you would figure out how to unlock them :)
<Tril> hacking NT profiles isn't my idea of fun.
<Fare> install Linux on that machine!
<Fare> Tuxissa, where are you?
<Tril> actually, they ARE installing Linux on these PCs as soon as the Linux server arrives.
<tcn> oh my god
<tcn> a college, not using linux?
<tcn> er, a college, using linux?
<Fare> er, a college?
07:40pm
<Fare> f00d!
<tcn> my college repossessed the linux servers we had in our engineering lab
<Tril> ok, "make bochs" produced an image
<tcn> allright, try 'bochs'
<Tril> wait you said I had to edit .bochsrc
<tcn> oh yeah, about the 10th line down
<tcn> you need to change it to your bochs directory
<Fare> tcn: scheme has a heap-based model, not quite the same as FORTH
<Fare> what to do about that?
<tcn> Fare: that's close though
<Fare> it also has a precise memory model
<tcn> where's some good (short) documentation on the internals of scheme?
<Fare> Rscheme.org
<Tril> bochs: panic, could not open hard drive image file '10M'!
<Tril> hmm, that's what it said before I went through all this trouble! :)
<Tril> does .bochsrc have to be in ~ ?
<tcn> maybe it's 'bochs -a'
<Tril> no, that's the dosemu command
<tcn> bochs looks for .bochsrc in . then ~ then some bochs dir
<tcn> uhh
<Tril> it didn't work in bochs
<tcn> ok, I have a 10M hd image in there
<tcn> I think bochs doesn't let you run w/o a HD.. 
<Tril> ?
<Tril> so I make a msdos hd image called "10M"?
<tcn> you could try commenting out the 'diskc: ...' line in .bochsrc... see if your version of bochs will work with it
<tcn> s/with/without
<Tril> bochs: panic, could not open hard drive image file ''!
<Tril> does it have to be 10M in size, or just 10M in name?
07:50pm
<tcn> it has to be whatever size the cylinders/heads/sectors specified in .bochsrc would be
<Fare> "even the blue jelly fish  do it"
<Fare> 10M in name
<Fare> until actually read from
<tcn> I guess you could say "8 cylinders, 2 heads, 8 sectors" = 8*2*8*512 = 64k
<Tril> hmm that dd command didn't work
<tcn> what command?
<Tril> dd if=/dev/z<tcn> You see now why I don't use bochs? :)
<Fare> tril: try bs=1k count=1k
<Fare> default bs=512
<Tril> now I should format that file..
<tcn> you don't have to
<Tril> I don't>?
<tcn> retro won't access it
<Tril> bochs won'?
<tcn> bochs just insists on it being there :)
<Tril> yes! a new error!
<Fare> UHE
<Tril> I got a bochs window!
<Fare> User Headspace Error?
<Tril> Retro is loading!
<tcn> I bet it's slower than my 486
<Tril> why is this taking so long
<tcn> or my 8086 (now R.I.P.)
<Tril> is bochs a piece of crap?
<tcn> oui
<Fare> nope
<Fare> bochs is great
<Tril> it says, "RETRO #8 bootloader/Protected mode startup successful./Loading modules.../Loading... FFE00000/Loading... ~"
<Fare> but it *emulates* a piece of crap!
<tcn> bah
<tcn> Maybe if you only have a Cray, and you need to run some PC program
<Tril> who's running bzip2
<Fare> on bespin? I am
<Tril> is this the backup script?
<Fare> yup
<Fare> 18M currently
<Tril> now I know why bochs is going so slow, it is getting only 45.8% of the CPU
<Tril> your backup script may have an 18M file, but it only takes 6M of RAM. Bochs is using 18M of ram
<tcn> haha
08:00pm
<Tril> I do like X-Win32 :)
<tcn> Retro is using less than my pocket calculator
<Fare> a hp48?
<tcn> no, ti85 piece of crap
<tcn> 48's are nice
<Tril> bzip2's almost done
<Fare> no, I renice +20 'd it
<Tril> oh you reniced it, thanks
-:- tcn2 [tcn@cci-209150250113.clarityconnect.net] has joined #tunes
<Fare> oh, no! one of them was enough!
<Fare> just when I had been writing a novel on matter duplication!
<tcn2> that was my isp telling me it's time to rest my eyes :)
<Fare> (actually, writing a *plan* of such a novel)
<tcn2> anyone know Downix's email?
<Fare> (either you're meta, or you're not)
-:- SignOff tcn: #TUNES (Ping timeout for tcn[cci-209150250085.clarityconnect.net])
<tcn2> oh, prolly down@bit-net.com
<Fare> so you know how to read irclogs, too?
<Tril> come on, if it's going to go this slow, it should show me some nifty blinkenlights
<Fare> tcn: what about the memory model?
<Fare> das ist nicht fur den dumbkopfen!
<tcn2> Tril: es muerte?
<Fare> tcn: the worst about the scheme memory model is *interrupts*
<Tril> no, still very running.
<Fare> (uh, no, SMP is actually _even worse_)
<tcn2> Fare: ut oh
<Fare> processes using the heap must cooperate
<Fare> (although this may be compiler-enforced)
<Tril> does anyone know what loads mean?
<Fare> shitloads?
<Tril> load averages
<tcn2> loads: (low-d\Z) puts on
<tcn2> :)
<Tril> what do the numbers mean?
<Fare> they mean that you are confused
<tcn2> oh
<tcn2> Fare: We could have separate heaps for each process
<Tril> is it a ratio?
<Fare> tcn: define "process"
<tcn2> task
<Fare> tcn: at first, we won't have that notion
<tcn2> anything that has a TSS is a process
<Tril> i guess no one knows
<Fare> tcn: at first, there will be only one such
08:10pm
<Fare> Tril: dunno
<Fare> tcn: in "small system" targets, there will be only one global "process"
<Fare> tcn: as in handheld devices, too
<tcn2> yeah, and they can cooperate
-:- tcn2 is now known as tcn
<Fare> let's have it this way:
<Fare> scheme-level threads will have "safe points"
<Fare> at safe points, the forth stacks must be empty (or filled with valid tagged data), and the heap must be ok
<tcn> hey, anyone know how to get an old Mac into it's ROM Forth?
<Tril> not i
<Fare> a safe point must always be reachable in (small) finite time from anywhere in the code, so the interrupt may resume it until it's reached.
<Fare> high-reactivity code must not use the heap
<Fare> low-reactivity interrupt code (or context switch) may resume computation until it reaches a safe point
<Fare> special-resume may be declared for some tight loops that want to avoid overhead of normal safe points
<Fare> tcn: only newer powermacs have a ROM Forth
<Tril> openbios?
<Fare> older Macs had a monitor, tho
<Fare> Tril: indeed
<Fare> no info.html for tunes-specs
<Tril> lists are irrelevant
<Tril> did you notice that the review list is not even on the collaboration page
<Fare> are they?
<Tril> not to mention tunes-specs, and tunes-all.
<Tril> well, there was no decision to which lists we should have, so I'm still not sure
<Tril> the traffic does not necessitate more than one list.
<Tril> yet, you think lll should be kept around.
08:20pm
<Fare> if retro is successful, lll might get full, quick
<Fare> oops, forgot to add /serv/w3 to backup.sh
<tcn> yeah, I told people to use lll and specs for Retro
<tcn> I'm surprised at how fast Retro is coming along.. pretty soon we should call it Tunes
<tcn> this summer, I guess
<Fare> hey, we're not there, yet :)
<Fare> but there's hope... :) :)
<tcn> fare, I'm writing a Retro Hackers' FAQ for ya
<Fare> tcn: thanks
<Fare> I'll seriously will have to implement a stupid retroscheme->retroforth compiler
<Fare> (written in the retroscheme subset of scheme)
<Fare> well, sleep time
<Fare> bye!
<Tril> can't I have emulated RETRO in my head faster than bochs?
<tcn> later
<Tril> bye fare
* Fare/#Tunes is away (rrrrr zzzzzzz)
<tcn> it's still not done? haha
<tcn> what's it said so far?
<Tril> it hasn't moved
<Tril> i told you
<tcn> how many "Loading..."'s?
<Tril> it's still on the second Loading... line.
<Fare> hung!
<Tril> it's not hung, bochs is using all the cpu.
<tcn> ok, that's the ints module
<Tril> is there some way for me to trace it
<Tril> what do I hit to get a dump of the cpu contents
<tcn> I dunno
<Tril> or bring up the bochs control interface, or whatever?
<tcn> haven't used bochs in 4 months
<tcn> all ints.asm does at load time is generate default interrupt handlers, and install handlers for 3 common exceptions.. it shouldn't take this long
<Tril> what does ~ mean?
<tcn> I'm guessing it's in an infinite loop.. but you can leave it running if you like :)
<tcn> ~
<tcn> oh, 'done'
<tcn> Did you get an error message?
<Tril> done what?
<Tril> I can't see an error
<Tril> oh, maybe...
<Tril> floppy_command: attempt to read/write sector 36, sectors/track=18
<Tril> floppy: DMA and interrupt capabilities disabled
<tcn> Well, the last module (ddr2) should never return control to the bootloader.. but if it does, it prints a ~ and does a 'jmp short $'
08:30pm
<tcn> how many Loading..?
<Tril> Loading modules...
<Tril> Loading... FFE00000
<Tril> Loading... ~ <still here>
<Tril> I got some messages about vga, then those ones about floppy, then some about pic
<tcn> those are normal, I think
<tcn> ok, that means it failed to load the second module
<Tril> was I supposed to do something with the modules?
<tcn> Normally there should be an address shown after Loading... it CALLs that address..
<tcn> I know!
<Tril> I just checked out retro, cd kernel/, and typed 'make bochs', was i supposed to build some modules?
<tcn> hang on
<tcn> edit Makefile
<tcn> Find the 'bochs:' target
<tcn> comment out the 'dd' command and use this instead: ../util/writeboot $< $@
-:- Iepos [root@d13.t1-3.tecinfo.com] has joined #TUNES
<tcn> howdy
<abi> hey, tcn
<Iepos> howdy
<Tril> hi 
<abi> what's up, Tril
<tcn> haha
<Tril> hi abi
<Iepos> hey Tril, did you ever find out what was wrong with the new-member-application thing?
<Tril> no, i didn't, do you remember what you wrote?
<Iepos> pretty much
08:40pm
<Iepos> do you want me to give it to you manually?
<Tril> do you know CVS?
<abi> CVS is Concurrent Version System at http://www.cyclic.com
<Iepos> nope
<Tril> ok yeah, then send me the text in email
<Iepos> all right
<Iepos> i'm going to try to get CVS to work sometime, though...
<tcn> Tril, any luck?
<Tril> I closed bochs, if that's what you mean
<Tril> you told me to holdon.
<Tril> I thought you had something for me to try.
-:- NetSplit: asimov.openprojects.net split from devlin.openprojects.net [08:44pm]
-:- BitchX+Deb1an: Press Ctrl-F to see who left Ctrl-E to change to [asimov.openprojects.net]
<tcn> yeah.. go to the kernel dir
<tcn> <tcn> edit Makefile
<tcn> <tcn> Find the 'bochs:' target
<tcn> <tcn> comment out the 'dd' command and use this instead: ../util/writeboot $< $@
<tcn> It has to tell the bootloader how big the kernel is, with the modules tacked on
<Tril> now 1.44M_test is not 1.44 M in size
<Tril> hey,...it worked... 
<tcn> oops
<tcn> haha
<Iepos> Are you guys trying Retro now?
<tcn> so you got a prompt?
<Tril> i got a prompt!
<Iepos> (wow)!
<Iepos> :-)
<Tril> I typed 'forth', then I typed '+ 1 2', now it is writing infinite capital 'S' on my screen
<tcn> reboot :)
<tcn> there's no interrupt key yet
<tcn> you could hit PAUSE but that might not work
<Tril> what can I do in Forth?
<tcn> in bochs
<Tril> the pause key? 
<tcn> it's '1 2 +' .. you're thinking lisp
08:50pm
<Tril> ok, it didnt loop this time. now it says "  " undefined
<tcn> actually it's '1 2 + .'  The . is to pop the number off the top of stack and print it.
<tcn> Oh yeah, I didn't add + yet.
<tcn> try 'here .'
<Tril> 2162688 ok
<tcn> that's forth's heap
<tcn> try '234 here !' then 'here @ .'
<Tril> the first one said ok but the second one did the S thing again
<Iepos> heh heh I wanna try too
<Iepos> let me get into x so i can use bochs
<Iepos> i'll be back
-:- SignOff Iepos: #TUNES (Iepos)
<tcn> try '234 920 0983 8' then '.s'
<tcn> that does a stack trace
<tcn> then try 'dup' '2dup' 'over' 'swap' 'rot' '-rot' 'depth .'
<Tril> < 4 > 234 920 983 8
-:- Iepos [root@d13.t1-3.tecinfo.com] has joined #TUNES
<Tril> what's all this numeric stuff
<tcn> random numbres :)
<Tril> 8
<tcn> iepos, try this:
<tcn> <tcn> try '234 920 0983 8' then '.s'
<tcn> <tcn> that does a stack trace
<tcn> <tcn> then try 'dup' '2dup' 'over' 'swap' 'rot' '-rot' 'depth .'
<tcn> do a .s after each of those commands
<Iepos> ??? I need the disk image first, don't I?
<tcn> '.' pops a number off the stack, but '.s' leaves the stack untouched
<Tril> can he just copy mine?
<tcn> I just put it on bespin
<tcn> http://bespin.cx/~tcn/retro.html
<Tril> it's on the web page? good.
<Iepos> is that daily snapshot the newest version?
<tcn> no, http://bespin.cx/~tcn
<tcn> then go to 'retro files'
<Tril> you mean the kernel directory
<tcn> yeah
<Tril> the image sizes are differnet
<tcn> get 'writeboot' from the util directory too
<Iepos> okay... getting
<tcn> I changed a few things
<tcn> nothing material :)
<Tril> gotta go...
<tcn> later
<Tril> my work here is done , for now
<Iepos> bye
* Tril/#TUNES  is away: (AFk) [BX-MsgLog Off]
09:00pm
<Iepos> hmmm...
<tcn> you have the snapshot, right?
<Iepos> yeah, i'm playing with bochs
<Iepos> now
<tcn> actually it would be easier to just 'writeboot image /dev/fd0' and reboot :)
<Iepos> true
<Iepos> but my linux systems takes like a couple minutes to boot back up
<Iepos> and then another couple for X
<tcn> there's a help screen for the debugger, but nothing for Forth yet
<Iepos> :-(
<tcn> do it before you hit the sack :)
<Iepos> okay
<Iepos> hmm ...
<Iepos> i am having a problem
<Iepos> ah ha
<Iepos> it seems to be working
<tcn> you got the prompt?
<Iepos> yes
<Iepos> hmm ... how does the forth thing work?
<Iepos> can't check out but can check in !
-:- hcf [nef@me-portland-us919.javanet.com] has joined #tunes
<tcn> haha
<Iepos> how does the forth thing work?
<tcn> I didn't bother figuring out how to exit forth :)
<tcn> gotta do some stack sleight-of-hand
<Iepos> hi, hcf
<tcn> hey
<Iepos> stack sleight-of-hand, huh?
<hcf> hoy Iepos
<Iepos> can it add, substract, multiply, divide?
<Iepos> weird RPL syntax?
<tcn> didn't add the math operators yet (they're in my old version, just gotta cut&paste, and change a few things)
<Iepos> RPN or whatever...
<tcn> yeah
<Iepos> oh okay.
<Iepos> so what should i do...
<tcn> .s is your best friend
<Iepos> what does .s do again?
<tcn> type a few numbers (separated by spaces) then type .s
<tcn> you'll see
<Iepos> oh i see ... it shows the stack
<Iepos> . pops one off like you were saying earlier
<tcn> right, and prints it
<tcn> . is pronounced 'print'
<tcn> , pops one and writes it to the address 'here' points to, then increments 'here'.
09:10pm
<tcn> 'here' is the heap pointer
<Iepos> how do you decrement here?
<Iepos> can you get here to point to here?
<tcn> I forget if you can
<Iepos> hmmm ... 
<Iepos> so how do you get something out of the heap ... ?
<Iepos> once you put it in ?
<tcn> haven't written any forth programs in like 6 months
<tcn> oh, you have to keep track of what 'here' was when you created the thing.
<tcn> either by keeping it on the stack, or by putting it in a variable (there are no variables yet :)
<Iepos> oh
<Iepos> oh i see
<tcn> you could type 'here dup 4323 , @ .'
<Iepos> what is @ ?
<tcn> 'fetch'
<Iepos> oh
<Iepos> is there a 'store' or do you just use the heap
<tcn> ! is store
<Iepos> oh
<tcn> used like this: 'x addr !'
<Iepos> yeah
<tcn> I'll add the rest of Forth in the next few days
<tcn> It really needs a 'words' command to show you what the commands are
<Iepos> okay i get it
<tcn> 'help' would be nice
<Iepos> heh heh, but i want to be able to change the heap pointer!
<tcn> PFE (portable forth environment) has a nice help system
<Iepos> i've actually never used forth before
<tcn> after I add variables I'll remember how to change 'here'
<Iepos> oh is "here" a normal forth thing, then?
<Iepos> not just a os debugging thing ... ?
<tcn> yeah, here is just a variable.. but I implemented it in assembly instead of forth
<tcn> it's a commonly used word in forth
<Iepos> oh
<Iepos> all right
<Iepos> i'm going to bed
<tcn> there's the stack control words: dup drop swap over 2dup rot -rot
<Iepos> but i'll write this to a disk first :-)
<Iepos> what is over?
<Iepos> rot?
<Iepos> -rot?
<tcn> fuck, i gotta get outta here too
<tcn> just type '1 2 3 4' and try those words one at a time, with .s in between
<Iepos> okay
<Iepos> well good night
09:20pm
<tcn> goodnight
<Iepos> cya
-:- SignOff Iepos: #TUNES (Iepos)
-:- SignOff hcf: #TUNES (Leaving)
-:- SignOff tcn: #TUNES (tcn has no reason)
09:30pm
-:- GMOL [gmol@24.66.11.51] has joined #tunes
<GMOL> Does anybody know how Brian's arrow paper is going..it's been ahwile since I've heard from him
10:20pm
-:- GMOL [gmol@24.66.11.51] has left #tunes []
[msg(TUNES)] newlog 1999.0410
IRC log ended Sat Apr 10 00:00:00 1999