ωνω Starting logfile 1999.0321 IRC log started Sun Mar 21 08:44:23 1999 ωνω Value of LOGFILE set to 1999.0321 hello Fare ole smkl re all Tril! yesterday, I met someone very interesting hi Tril smkl, long time no see Fare: what do you think about maude? 08:50am >>> smkl [sami@ppp81.dial-in.verkkotieto.com] requested PING 922034046 737301 from #tunes >>> Tril [dem@sloth.wcug.wwu.edu] requested PING 922035658 186014 from #TUNES wake up, Fare π Tril/#TUNES is away: (afk) [BX-MsgLog Off] 09:10am ωνω SignOff Fare: #TUNES (Read error to Fare[esmeralda.enst.fr]: Network is unreachable) ωνω FareWell [rideaufr@esmeralda.enst.fr] has joined #Tunes FareWell: what do you think about maude? ωνω SignOff FareWell: #TUNES (Ping timeout for FareWell[esmeralda.enst.fr]) ωνω FareWell [rideaufr@esmeralda.enst.fr] has joined #Tunes hum FareWell: what do you think about maude? 09:30am well, I've printed the docs. The general concept is very interesting, and they do have reflection too bad it ain't free software what kind of features should tunes HLL have and how can reflection be used to implement some of those features? it should have things like side-effect control also an integrated proof system 09:40am what kind of support it should have for distributed computing and constraint solving? ωνω http [rideau@quatramaran.ens.fr] has joined #Tunes ωνω SignOff FareWell: #TUNES (Read error to FareWell[esmeralda.enst.fr]: Network is unreachable) also an integrated proof system what kind of support it should have for distributed computing and constraint solving? smkl: have you read my article "Reflection, Non-determinism, and the lambda-calculus"? lambdaND.ps ? yup a recent version hmm. what is the most recent version? dunno, but I just recompiled one and put it on ~fare/html/tmp/ ωνω http is now known as Fare ωνω Fare is now known as FareWell 09:50am what have you added to it? and who did you meet? I've added a construction for a reflective system (an early version said that this was only a conjecture) also, I've added comments on distributed systems being the natural setting for reflection, since it wouldn't bring any expressiveness to the lambda-calculus well, in a certain way ωνω hcf [nef@me-portland-us739.javanet.com] has joined #tunes ωνω abi [nef@bespin.cx] has joined #tunes i need to go now ωνω SignOff smkl: #TUNES (aaaaa) 10:00am ωνω FareWell is now known as Fare π Fare/#Tunes is away ωνω SignOff hcf: #TUNES (Leaving) ωνω hcf [nef@216.164.116.118] has joined #tunes ωνω _QZ [brand@p0wer.qzx.com] has joined #tunes hoy _QZ <_QZ> hello hcf <_QZ> any1 here run retro? 11:40am hum no source code <_QZ> ah it helps if u add a test module to the boot.bin first :) 11:50am ωνω tcn [tcn@cci-209150250104.clarityconnect.net] has joined #tunes anyone home? <_QZ> hey salut, _QZ <_QZ> was just emailing u what's up? nuthin'.... not you abi: forget what's up tcn, I didn't have anything matching what's up abi: forget what's up? tcn, I didn't have anything matching what's up <_QZ> your test rdfs dont clear when using backspace tcn: il handle that tcn: hoy tcn: why not put the sources online? tcn: GPL is useless w/o sources! qz: I was lazy :) It's like the old Apple II <_QZ> and print doesnt use the C \codes :) what about the old A2? Fare: Sources are on their way.. I'm just cleaning up. Apple ]['s rule <_QZ> and what is there to run? not much :) But it works? <_QZ> ya a netscape guy at a free software conference told that the 2 months delay to publish the mozilla sources consisted in removing insulting comments on M$ and other vendors' broken code from within the workarounds in navigator sources haha tcn: w/o the sources, I won't bother to test fine, wait till I post them, then :) 12:00pm I already have stuff insulting M$ on my homepage. My cleanup just involves removing stuff that isn't used and cleaning up the interfaces.. I need better error-reporting and such. Right now, if a module fails to load and link, it'll still be run, and you get a GPF or something you may remove it *after* publishin sources, you know? BTW, we can open a CVS rep for you on bespin... yeah, that would help just commit your changes on retro/ (not below tunes/` Hey, do I need to put a GPL copyright notice in every source file, or 1 per directory, or what? One in the main package but you must have a prominent README or such explaining everything when you mix files with multiple copyrights, beware ωνω SignOff hcf: #TUNES (Leaving) hmm.. I should probably put a short (2 line) copyright notice in every file. That's not too big. I just don't want it to be like Linux, etc, where every little header file has the whole 20-line GPL notice. 12:10pm ωνω hcf [nef@216.164.116.128] has joined #tunes it's better indeed to have a short (c) notice with reference to the GPL. although that's mostly useful only when you export/import individual single files out of the distribution well.. they are mostly modules ωνω SignOff hcf: #TUNES (Leaving) ωνω hcf [nef@me-portland-us1047.javanet.com] has joined #tunes 12:20pm ok, how's this: ; Copyright (C) 1999 Thomas Novelli ; Distributed under the terms of the GNU LGPL, version 2 or greater. ; (For the text of the license, see the file COPYING.LIB) 12:30pm π Tril/#TUNES is back from the dead. Gone 3 hrs 28 min 59 secs hey tril 12:40pm hello What do I do to put Retro in a CVS rep? where are the files you want to put in? On my computer Shall I tar & upload them? you can, or you can install a cvs client on your home box to use to access cvs on bespin remotely hold on I'll get the command to start up a new module I've got cvs here export CVSROOT=:pserver:anonymous@bespin.tunes.org:/cvs that is for read only use :pserver:tcn@bespin for your own account will allow you to write then cvs login you only need to login once, it stores pass in ~/.cvslogin So I can just do: export CVSROOT=":ext:tcn@bespin.tunes.org:/usr/local/lib/cvs/rep1" hmm 12:50pm :pserver:tcn@bespin puts it under my home dir? i dont know no it doesnt do anything yet. I'm still getting the exact command crap cvs is broken maybe not do you want retro in the tunes repository, so it's writable by the same people who can write to tunes? why dont I make you a new group "retro" to which all retro developers have to join? Yeah, that's what I was thinking. Put it under the Tunes group, in /usr/local/lib/cvs/rep1/retro which? the tunes group wouldn't allow you to have separate control over access ooh, YA mlist? err.. /cvs/retro hcf: huh? ooh, yet another mlist? Whoever is in Tunes, can feel free to hack Retro. I make backups. hcf: what are you talking about? n/m hcf: we're talking about adding a CVS dir for tcn's prototype. I don't know anything about mailing lists, unless you are talking about some new tunes mlist post that I didn't read yet. Tril: mlist for retro No, it's not that big i know, i was j/k tcn: do you have ssh 01:00pm yeah, use :ext:tcn@bespin:/cvs, if you have ssh then export CVS_RSH=ssh Tril: yep I just updated Tunes via CVS.. it works did you just modify stuff in Review/ ? no, I only received changed changes oh cd retro; cvs import -m"initial import of retro sources" retro retro start i don't think it will work you have to be group tunes fare somehow uses it though I need to reorganize it more, first. Don't want extraneous/obselete stuff in the cvs rep you need to choose a stable directory structure, too it's a pain moving dirs? no, not a pain, just impossible. it breaks updates. yo have to remove the file, and add it again jesus cvs sucks I never knew it was that bad i'm trying to write to cvs from here i wonder how fare does it Here's what I just used: export CVSROOT=":ext:tcn@bespin.tunes.org:/usr/local/lib/cvs/rep1" export CVS_RSH="/usr/bin/ssh" cvs -z 9 update tunes yeah, i know read-only works fine :) It worked when I wrote to it, too what, cvs ci (commit)? 01:10pm cvs -z 9 commit I don't see how that's even possible, the repository has mode 444 set on all files! ok, try importing then.. cd dir; cvs import -m"initial import of dir" dir vendor start it doesnt' wrok for me I think I did cvs -z 9 commit tunes, from the parent dir it says you modified Review/OSes.html sometime.. That was a week or two ago I added ITS that's fitting, seeing how everyone's been reading Hackers lately well, commit maybe works, try import! oh, it's working if I use ssh.. ssh shouldn't be a requirement to use cvs, though... it's not like we have trade secrets :) it's not, but rsh is probasbly not setup properly on bespin ssh keeps your password secret, not just your files right Hey, the symlink /cvs -> /usr/local/lib/cvs/rep1/ doesn't work.. I had to use the full pathname in CVSROOT=... 01:20pm π Fare/#Tunes is back hi fare oh it doesnt'??? speaking of security, you should take the '.' out of the PATH. I really nailed someone that way, once. :) π Fare/#Tunes uses CVS with ssh Say you're in /tmp, you type 'su', but what if there's a 'su' in tmp? yer screwed. done oops, maybe not Fare: cvs didn't work w/o ssh for me which is ok or me wait, I use :ext:fare@bespin.tunes.org:/usr/local/lib/cvs/rep1 tcn: what path does not work for CVSROOT? yes, I do use SSH: Tril: /cvs CVS_RSH==ssh CVSROOT=:ext:user@bespin.cx:/cvs tcvs () {( export CVSROOT=:ext:fare@bespin.tunes.org:/usr/local/lib/cvs/rep1 export CVS_RSH==ssh unset DISPLAY cvs -z 9 $@ )} tcn: I just used that CVSROOT and imported something, and it worked that's on the collaboration page, btw! the usr/local/lib thing will probably change tomorrow. Start using /cvs. Fare: Is that tcvs thing a zsh login script? that's a zsh alias it's a zsh function err, a function which works like an alias :) so, you put it in your .zshrc? yo could in my .aliases, sourced from .zshrc is .cshrc the name of a city in Bosnia? 01:30pm cziech? i dont get it. that was an old quote from a linux cookie file (during an early flamewar on the user-friendliness of linux vs mockass) any command that you issue more than thrice should be a function in your .aliases u mean, command line anyway there was an article on ITS at eval-apply.com it disappeared, I think.. what url? oh, well maybe contact "Joe Marshall" that guy never sent yo his archives did he actually I heard it was going back up?? no he never did hum. maybe see at http://www.vsf.cape.com/~emergent/ that helps 01:40pm pc-losering ? tcn: http://www.vsf.cape.com/~emergent/pclsr.htm got it that's the one from eval-apply.com, right? that stuff looks like some of what was on eval-apply. yup There's also some ITS stuff on Foldoc Tril: the owner of eval-apply put his stuff there until he'd register eval-apply again tcn: I see you're editing its.html :) tcn: maybe put a link to other stuff at eval-apply, since it's (indirectly) ITS-related 01:50pm yeah, notice I'm using PICO instead of EMACS "Eventually Malloc's All Computer's Memory" err, s/Memory/Storage/ Eight(y) Megs And Constantly Swapping ? haha be fair, 8 megs in text mode, anyway :) well, my computer got 64 megs, and I'm running XEmacs for *weeks* the beast that mallocs and dies is netscape yeah fare: just imagine in tunes you could really use emacs for everything. I'd run w3, if only it wouldn't stop my editing under emacs :( Well guys, I gotta go. We'll put Retro in CVS in a few days. we? we are here to compete or extremely free ok, bye >>> Tril [dem@sloth.wcug.wwu.edu] requested PING 922053541 611490 from #TUNES what's .tar.xdp.gz? It'll let me put Retro in CVS without any divine intervention from you guys? yeah, it seems to work. just put it under /retro, not in /tunes $CVSROOT/retro, i mean gotcha I'll put it there when it's organized π Tril/#TUNES takes another dangerous stab at compiling dpkg in slackware 02:00pm later ωνω SignOff tcn: #TUNES (tcn has no reason) 02:10pm Tril: where did you see .xdp? gnu/gettext/ Please note that this host lives in the US. Therefore, if you are outside the US, please do not download encryption software from this ftp site. is that legitimate? is what legitimate? I'm sure it violates basic human rights. I thought it was violating us export restrictions to have encryption software on a public ftp server it might the question is: would you take the risk of being sued? but if you say "please do not download", that sounds to me like those romz sites that say "24 hour trial period only", totally bogus why so? 02:20pm i woulndt do it. I have debian-non-US in a private directory on bespin, only accessible to me. 02:30pm π Tril/#TUNES is away: (afk) [BX-MsgLog Off] ωνω SignOff Fare: #TUNES (Read error to Fare[quatramaran.ens.fr]: EOF from client) ωνω SignOff hcf: #TUNES (Read error to hcf[me-portland-us1047.javanet.com]: Connection reset by peer) ωνω hcf [nef@me-portland-us725.javanet.com] has joined #tunes ωνω kia [hgjg@t1o66p43.telia.com] has joined #Tunes ωνω SignOff hcf: #TUNES (Leaving) ωνω kia [hgjg@t1o66p43.telia.com] has left #Tunes [] ωνω hcf [nef@me-portland-us212.javanet.com] has joined #tunes ωνω GMOL [gmol@24.66.11.51] has joined #tunes Anyone know how Brian is doing on his Arrow paper? <_QZ> no, but brix is coming along quite well _QZ who are you again? <_QZ> i am making brix 07:50pm which is? (I remember hearing about it) <_QZ> brix? somebody said brix was a free portable protected memory-space preemptive multi-threading multi-user persistent object-oriented high-speed easy-to-use distributed OS located at http://www.qzx.com/brix How far are you? <_QZ> farther than everyone else which is.....i.e. what have you implemented so far <_QZ> i have everything in that description done except for the "easy-to-use' and 'distributed' <_QZ> but thats just cuz the UI is not done and same with the networking >>> _QZ [brand@p0wer.qzx.com] requested PING 922074992 991048 from #tunes 08:00pm what about multiprocessing? you have all the logic to make distrbuted and multiprocessing look the same to the programmer? <_QZ> oh ya it will do smp and has the smp code but its not all setup yet how do you handle consistency issues associated with distrbuted computing? <_QZ> brix doesnt run programs, instead u run actions that would require 10's to 100's to amount to a program <_QZ> and each action is threaded I.e. if I change a file that I got from another computer when does it update the file from it's orignial location <_QZ> so smp is completely transparent and efficiently used If you're fine graing the excution into that many blocks what are the overhead associated? <_QZ> brix doesnt allow 2 machines to work on a single object (file) Well how are you going to handle a database on a server that clients are trying to access? Or shared code on a single server? 08:10pm But first I'd like to know how spiltting programs up into 100th size chunks affects overhead in terms of execution speed. <_QZ> it makes it faster <_QZ> dont think of it as splitting the program up I assume that there is some sort of overhead to call an action..if you are paying that overhead in a loop, it makes it slower <_QZ> actions are like plug-ins A program made up of a bunch of different plug ins will be slower and larger than an a traditionally designed app <_QZ> and brix does support a non-threaded call to actions <_QZ> no, it will be smaller and faster A call is still a call, there is overhead to be paid, 100 times as many calls in a program can result in potnentially a lot of slowdown if it is in the right places...and how do you get around the database client thing... <_QZ> the call is the same call that a proram would make, its as fast as it can get <_QZ> ok for the dbase question: a dbase would consist of multiple objects <_QZ> and the main object is only handled by actions on the server itself, other objects can be handles by remote machines 08:20pm Well I suppose I'll see what you mean when you are finished.. ωνω SignOff GMOL: #TUNES (Leaving) ωνω SignOff Tril: #TUNES (BitchX-74p4+Deb1an -- just do it.) 08:30pm ωνω Tril [dem@sloth.wcug.wwu.edu] has joined #TUNES ωνω lar1 [larman@208.254.224.42] has joined #tunes Hey! _QZ: I do know some digital theory... NOW I can prove it... :) 10:10pm <_QZ> huh I reverse engeneered a nintendo controller... and I am gonna interface it to my computer For use with an emulator I had my logic probe, my ocilicope, my volt meter... all that stuff out! Err an 8-bit NES controller, not N64 10:30pm Still alive anyone? 11:00pm ωνω SignOff hcf: #TUNES (Leaving) ωνω SignOff lar1: #TUNES (Ping timeout for lar1[208.254.224.42]) ωνω SignOff _QZ: #TUNES (BRiX [http://www.qzx.com/brix] :: sleep) [msg(TUNES)] newlog 1999.0322 IRC log ended Mon Mar 22 00:00:01 1999