IRC log started Wed Feb 17 00:00:01 1999 [msg(TUNES)] permlog 1999.0217 ωνω SignOff Fare: #TUNES (Ping timeout for Fare[esmeralda.enst.fr]) ωνω Fare [rideaufr@esmeralda.enst.fr] has joined #Tunes ωνω Wlad [wlad@profm.arad.ro] has joined #tunes ωνω mode/#tunes [+o Fare] by ChanServ EHLO Wlad: welcome! ωνω Fare has changed the topic on channel #tunes to: Reflective Computing System - http://www.tunes.org/ Fare: why do you think that it's IPv6? Wlad: I'd like the same kind of stuff as you do, only I want the abstraction level of those "vm" to be high-level instead of low-level Wlad: because giving each user a IPv4 address isn't doable Fare: most of big webhosting providers gives you an IP there's no other way to make virtual FTP working in the US. And not one for every sub-user and sub-process but again, I can imagine that it'll be much less of a problem w/ IPv6. Fare: no, this is not what I meant it's not like IBM's VM and by the time your system is ready, you'd better bet on IPv6, anyway. basically it simulates more machines on only one VM can simulate many machines, afaik and it will be a patch for linux kernel, not a new OS (all machines on the same assembly-level model) yes, but VM is not what I want patch to linux? linux isn't modular enough to do that. that'd better be something on top of the flux os kit, then I want to make it look like user has it's own machine Wlad: VM does that Fare: VM is on a process-basis what I basically want to do is to allow the webhosting customer to run it's own daemons I mean, even if we are on the same machine, my sendmail has nothing to do with yours 04:00am do you intend the virtual machine to be a unix-like environment? it seems that HURD is just what you want. it does just that only it virtualizes so much that performance crawls I know personally, what I want is a high-level model by keeping things high-level, you allow the runtime to dynamically optimize and compile, instead of having to inefficiently simulate everything at the bare machine level what about performance loss in such an environment linux really isn't intended for multiplexing and virtualizing in kernel space -- if you do it, it's a *major* undertaking. Wlad: in a high-level environment? performance loss are kept minimal to non-existent: Fare: JVM has proven the contrary... at worse, you interpret, and get the same performance hit as with low-level runtime emulation. JVM is quite low-level that's their problem: a lowlevel virtual machine that you have to emulate, instead of high-level code that you could compile plus the object system design is braindead the worst of both worlds (static typing and dynamic typing) you mean, to dinamically compile the apps? Yes. note that compilation is kept incremental so that you only need recompile what's been changed. and what's the use of this? because the system controls the whole compilation thing, it needn't maintain compatibility with low-level backdoors Wlad: performance. It means that everything that is interpreted in a typical unix system (think all shell, sed, perl, tcl, guile scripts) can be compiled instead! 04:10am Fare: hmmm there's a perl compiler, but not very performant because of eval instruction selfmodifying code perl is not performant because it isn't their primary purpose also, because of their all-dynamic typing strategy a system like OCAML does better. Only their problem is they are all-static (except for a toplevel kluge) I'd like a middle-term between those as of static/dynamic. hmm did you see EROS? I saw their page. Didn't try the reeased code "Extremely Reliable" and no support for SCSI-level disk access commit seems funny IDE gives no guarantee whatsoever for committing data, except if you spin off the disk, which is too slow for anything but system halting. 04:20am it's very alpha sure I have to go now bye bye! ωνω SignOff Wlad: #TUNES (Wlad has no reason) 04:30am ωνω SignOff yery: #TUNES (Ping timeout for yery[benetnash.ffke-campus.mipt.ru]) ωνω SignOff Fare: #TUNES (Connection reset by pear) ωνω Fare [rideaufr@esmeralda.enst.fr] has joined #Tunes ωνω yery [ost@benetnash.ffke-campus.mipt.ru] has joined #tunes ωνω hcf [nef@escher.sdi.agate.net] has joined #tunes ωνω SignOff yery: #TUNES (Ping timeout for yery[benetnash.ffke-campus.mipt.ru]) ωνω NetSplit: sterling.openprojects.net split from koontz.openprojects.net [07:38am] ωνω BitchX: Press Ctrl-F to see who left Ctrl-E to change to [sterling.openprojects.net] ωνω Netjoined: sterling.openprojects.net koontz.openprojects.net ωνω Fare [rideaufr@esmeralda.enst.fr] has joined #tunes ωνω yery [ost@benetnash.ffke-campus.mipt.ru] has joined #tunes !tolkien.openprojects.net!! Remote CONNECT sterling.* 8005 from chrome ωνω SignOff yery: #TUNES (Ping timeout for yery[benetnash.ffke-campus.mipt.ru]) ωνω yery [ost@benetnash.ffke-campus.mipt.ru] has joined #tunes ωνω SignOff yery: #TUNES (Ping timeout for yery[benetnash.ffke-campus.mipt.ru]) ωνω yery [ost@benetnash.ffke-campus.mipt.ru] has joined #tunes ωνω NetSplit: koontz.openprojects.net split from norton.openprojects.net [08:38am] ωνω BitchX: Press Ctrl-F to see who left Ctrl-E to change to [koontz.openprojects.net] ωνω abi [nef@bespin.cx] has joined #tunes ωνω ^lilo [lilo@varley.openprojects.net] has joined #tunes ωνω yery [ost@benetnash.ffke-campus.mipt.ru] has joined #tunes ωνω Fare [rideaufr@esmeralda.enst.fr] has joined #tunes ωνω hcf [nef@escher.sdi.agate.net] has joined #tunes ωνω SignOff hcf: #TUNES (Leaving) ωνω NetSplit: zsoldos.openprojects.net split from koontz.openprojects.net [01:05pm] ωνω BitchX: Press Ctrl-F to see who left Ctrl-E to change to [zsoldos.openprojects.net] ωνω Netjoined: zsoldos.openprojects.net koontz.openprojects.net ωνω yery [ost@benetnash.ffke-campus.mipt.ru] has joined #tunes privet! privet! yery: what are your specialties that you'd use in the Tunes project? Fare: nothing interesting :) what is your current job? Fare: I am translating the GNU Emacs manual Fare: I work for FSF as a translator that's not a full-time job, is it? I mean, it doesn't pay your bills, does it? 01:20pm Fare: that is a full-time job without a payment Fare: am I a hero? :) yes you are, but soon, a hungry hero :( Fare: I am really hungry, now. good appetite. π yery/#tunes is trieng to find something to eat Fare: thank you. it will be an apple 01:30pm Do you believe in GUD? that's RMS' note in the Emacs manual :) I've never used GUD. The last debugger I used was gdb for my eforth-1.0c on linux before that, it was Turbo C 01:50pm Fare: you use gdb as it is? Fare: I mean, without nicer interface? long ago, and for a small asm program ωνω tcn [tcn@cci-209150250099.clarityconnect.net] has joined #tunes besides, the FORTH threading model made a C-oriented debugger useless the only thing usable was low-level breakpoints. break-on-return was impossible using gdb :( tcn: Gakuk! You wrote Eforth? no, I only ported it to Linux. Fare: in fact, this note shows that RMS' and ESR's conflict isn't that new strong personalities have strong divergences but they might also strongly appreciate each other, up to their convergences I think I missed something tcn: read the Tunes log :) tcn: yes, rtfl :) Proprietary software is dead software. Using it is akin to necrophilia. haha Like that starbridgesystems thing.. Instead of running Tunes, they came up with some lame proprietary OS for their new supercomputer That's ok. My supercomputer will run Tunes, and SBS will go outta business :) Is GUD really a debugger? tcn: GUD is Grand Unified Debugger tcn: it is an interface to gdb, Perl debugger, jbd ... And RMS and ESR got in a fight over that? tcn: no, ESR wrote the GDB, RMS wrote that note 02:00pm did ESR go to MIT too? ESR wrote the GUD of course, sorry tcn: afaik, no ωνω moonbeam^ [chg@152.205.224.192] has joined #TUNES Hello room howdy hello hi, yery Anyone home? Yeah, 3 of us.. what's up? later, then abi's a borg bot, whatever :) tcn: robot abi is a stray biorobot Hey Fare, I finally tried out ML.. nice language.. 02:10pm so, who is real in here? besides me, i mean moonbeam: I, real oh, good Me too i was starting to get spooked I'm new to mirc there's nobody home in newbies seems kind of quiet around here oh well, nice chattin with ya see ya ωνω moonbeam^ [chg@152.205.224.192] has left #TUNES [] hehe.. OS design isn't for newbies.. :) Well, they could help us figure out what "easy to use" means, but they ain't gonna get far without people like us.. tcn: the problem with ML is its type system: either too powerful or not enough, but not flexible. hahaha.. the military is retarded.. http://www.national.com/appinfo/milaero/napa1000/DARPA.htm 02:20pm doing stuff by hand that should be automatic, using C, ... C people suck. Doing stuff by hand that should be automatic, using LISP, ... seen Tril, lately? btw, happy new year of the Cat! hehe I think I saw him on here Friday, for about 2 seconds busy, I guess 02:30pm fare: Have you thought about how Tunes will deal with reconfigurable hardware (i.e., DPGA's/FPGA's).. ??? >>> tcn [tcn@cci-209150250099.clarityconnect.net] requested PING 919275547 from #tunes 02:40pm ωνω SignOff tcn: #TUNES (Leaving) tcn: I've dreamt about it, but nothing precise. this is one of the many engineering problem that Tunes might reveal, but that are quite independent from Tunes itself. Tunes doesn't target any particular engineering problem as much as it tackles an essential design problem. duh, he's gone. 03:00pm ωνω NetSplit: lucas.openprojects.net split from sterling.openprojects.net [03:15pm] ωνω BitchX: Press Ctrl-F to see who left Ctrl-E to change to [lucas.openprojects.net] ωνω Netjoined: lucas.openprojects.net sterling.openprojects.net ωνω Fare [rideaufr@esmeralda.enst.fr] has joined #tunes hehe. New TUNES server: http://d013.paris-135.cybercable.fr Fare: what's new? new is better (tho u didnt ask me) yery: nothing. Just that I now have my own baux on the 'net. Fare: um my home has a world-visible web-server (nothing on it, dhcp, but still)! abi: what are you dreaming about? i don't know, yery Fare: I have a web-server too. for testing, usually yeah. I put the web server just so as to be able to identify the machine easily from anywhere (i.e. check whether the dyn ip has changed, and if so, scan among possible addresses) Also as a gate to the internet for my family :) 05:50pm ωνω SignOff Fare: #TUNES (Back to the Tunes OS project -- /whowas Fare) ωνω _QZ [brand@p0wer.qzx.com] has joined #tunes ωνω iStormy [stormy@rain.futuresouth.com] has joined #Tunes trying out linux 2.2 smp...who was curious about that 06:50pm ωνω SignOff iStormy: #TUNES (bah, back to freebsd) ωνω SignOff _QZ: #TUNES (Ping timeout for _QZ[p0wer.qzx.com]) ωνω _QZ [brand@p0wer.qzx.com] has joined #tunes ωνω hcf [nef@escher.sdi.agate.net] has joined #tunes ωνω SignOff hcf: #TUNES (Leaving) ωνω _QZ has changed the topic on channel #tunes to: The day we met, my little pet, -- I knew with just one look -- You'd bear a son, and now that's done, -- So shut your mouth and cook! ωνω SignOff _QZ: #TUNES (BRiX [http://www.qzx.com/brix] :: sleep) [msg(TUNES)] newlog 1999.0218 IRC log ended Thu Feb 18 00:00:00 1999