IRC log started Fri Jan 8 10:34:36 1999 [msg(TUNES)] permlog 1999.0108 ωνω hcf [nef@escher.sdi.agate.net] has joined #tunes ωνω Tril [dem@bespin.ml.org] has joined #TUNES hi hcf hoy Tril ωνω Tril has changed the topic on channel #TUNES to: http://www2.tunes.org/cgi-bin/TunesWiki?TODO ... Meeting today at 23:00GMT to discuss the web page ωνω Tril has changed the topic on channel #TUNES to: Project to do list: http://www2.tunes.org/cgi-bin/TunesWiki?TODO ... Meeting today at 23:00GMT to discuss the web page what about 'http://members.xoom.com/unios/tunessite.html'? i thot bineng was gonna do real-time summarizing Gakuk! 10:40am Tril: finger nef@bespin, comment the TODO page has everything from tunessite.html that was relevan. i need a tunes blurb, not exceding 48chars hcf: finger dem err, I mean finger dem|less heh, yeah I like yours better I am going to class 1 hour after the meeting starts, so I will try to be on IRC earlier like at 20-21 (It's almost 19 now) in addition to being at the first hour of the meeting a TUNES os blurb, a tunes project blurb, or a #TUNES channel blurb? 10:50am or other? tunes cybersex blurb ;) no, tunes project blurb where is it going? "the new music revolution"? ;) the blurb is going where is in my .plan 48 chars. hmm Reflective Computing Project (28 chars) preferably close to 48chars to further justify the width maybe the above blurb w/ 1 more buzzword only 1? open, dynamic, integrated complete computing system π hcf/#tunes is sick of 'open' caldera has ton it to death that's 53. . take out open put complete first ton? ton/done/ which -centric would best fit? user or idea 11:00am if you use user-centric you can take out dyanmic possibly integrated 'user-centric self-organizing computing project'? what's self-organizing? it can't do that self-maintaining, automated w/ reflection i still don't get it k, n/m self-optimizing, though 'user-centric self-optimizing computing project'? s/project/system/ it's not the MAIN feature of the entire system, which your blurb would indicate main feature being reflection? 11:10am π Tril/#TUNES looks around anxiously yes it is, but we haven't listed clearly the benefits that brings, yet. it's on my .project list 'Specs wrt persistence', whats 'wrt'? abi, wrt is with respect to abi, wrt? it has been said that wrt is with respect to well, more exactly, reflection brings all the benefits we have been clamiing for tunes all along, but we haven't explained how that works, yet. so puting 'reflective' in the blurb would cover most everything? hum I dunno, reflective doesn't IMPLY persistence, what else? 11:20am just say high-level instead, it will make more sense at the expense of being less precise Uhhh maybe not you pick something. it will make more sense after I write this doc on reflection. should the blurb end w/ 'system' or 'project'? Augrh who cares do you "pass" a contract? Fare: in what context 11:30am what kind of contract like, interfaces between modules that you agree upon. no, usually you "make" a contract. the two parties agree? conclude? enter? is it made of "clauses" ? the two parties enter a contract, sure hmm is this where "handshaking" in computers came from? Tril: handshaking is more like price bickering if you are being legalistic, contracts have clauses. I just call them items, informally but you might not want to use that hcf: yeah that's true 11:40am Fare: are you interested in this: We introduce a notion of *Grothendieck logical relation* and use it to characterise the definability of morphisms in *stable* bicartesian closed categories by terms of the simply-typed lambda calculus with finite products and finite sums. Our techniques are based on concepts from topos theory, however our exposition is elementary. do you put a brake "on", or "to" ? categorical logic? you "set the brake", or "put on the brake" "put the brake on" you can also say "braking" if you are actively doing it to slow down putting the brake on, is more like the parking brake the decelerator is more like "using the brake" or "pushing the brake" or more often "hit the brake" I dont know if it's categorical logic, it's from the types forum: Lambda Definability with Sums via Grothendieck Logical Relations Lambda Definability with Sums via Grothendieck Logical Relations by Marcelo Fiore and Alex Simpson if you want it I'll send you the mail, it has urls I have to go π Tril/#TUNES is away: (afk) [BX-MsgLog Off] ωνω Fare [rideaufr@137.194.160.71] has left #Tunes [] 11:50am ωνω _QZ [brand@p0wer.qzx.com] has joined #tunes hoy _QZ 12:40pm <_QZ> hoy 12:50pm ωνω binEng [bineng@dialup208-1-37.swipnet.se] has joined #tunes hi <_QZ> hello hi, _QZ <_QZ> hi what's up, _QZ <_QZ> hmm abi must not like u hi abi <_QZ> hehe abi: HI que tal, binEng now what does that mean... <_QZ> que tal is spansih for whats up well tell that stupid bot I can't Spanish ;) <_QZ> ok i have added some nice additions to give abi a nicer personality how did you do that? changes in the source or just data changes? http://members.xoom.com/unios/tunessite.html <-- Updated list http://members.xoom.com/unios/tunes <-- Updated sample page <_QZ> oh damn http://members.xoom.com/unios/tunes.html <-- Updated sample page <_QZ> we have that meeting today yup.. <_QZ> i might not be here <_QZ> gotta go shopping right now (2pm)and then to the mechanic at 6pm, and the meeting is at 4pm <_QZ> might be able to be back by 4 http://members.xoom.com/unios/docs.html <-- Another sample page <_QZ> might be late QZ: It's not the whole world if you'd miss this... or maybe it is :) 01:20pm <_QZ> its not like i will help out being at the meeting <_QZ> ok i have added some nice additions to give abi a nicer personality _QZ: what did u do? <_QZ> hcf: eh? what did u 'add'? <_QZ> i didnt do naything i know ωνω linuxuser [linuxuser@bjas1-p19.telepac.pt] has joined #tunes Hi!! howdy salut, binEng stop it w/ the double puncuation!! π binEng/#tunes shoots abi in his head binEng: be nice binEng: want me to take out the non-english? <_QZ> hcf: u got gtk installed? _QZ: no <_QZ> this is pissing me off hcf: what? <_QZ> i am trying to install E and gnome <_QZ> but they wont install <_QZ> and i was also gonna upgrade to a 472mhz celeron to make E run better, but the stupid place doesnt have any of hte "good" ones in right now binEng: would u like me to remove the non-english responses from abi? <_QZ> hcf: no dont :) hcf: Never mind... I don't think he's ever saying anything important. <_QZ> abi is a she ic how do you tell? <_QZ> hcf is lonely so he made his bot a she :) <_QZ> whois abi? i guess i am the little girl borg with a big brain yep... girlie _QZ: actually, i prefer gender neutral, but ppl tend to engender bots, so i gave em a head start <_QZ> and does abi sound like a name u would have been able to grow up with, without getting yer face bashed in every week? <_QZ> :) abi: be abi look at me, i'm cute! <_QZ> sex? somebody said sex was not the answer. Sex is the question. 'Yes' is the answer. <_QZ> abi is also a slut okay, _QZ. ωνω NetSplit: sterling.openprojects.net split from forward.openprojects.net [01:27pm] ωνω BitchX: Press Ctrl-F to see who left Ctrl-E to change to [sterling.openprojects.net] <_QZ> damnit ωνω Netjoined: sterling.openprojects.net forward.openprojects.net ωνω jeff [jeff@a1-3h167.neo.rr.com] has joined #TUNES ωνω Tril [dem@bespin.ml.org] has joined #TUNES Tril? Tril is not the one to ask about red hat, not having used it yet <_QZ> haha jesus <_QZ> um abi is not letting me change that slut thing _QZ: what r u trying to change it to? 01:30pm <_QZ> back <_QZ> msg (abi) no, abi is the little girl borg with a big brain <_QZ> MESSAGE (abi nef@bespin.ml.org) sorry... <_QZ> MESSAGE (abi nef@bespin.ml.org) i'm not following you... <_QZ> she cycles between those 2 responces hmm fixed needed to do 'forget abi' 1st binEng: Hi niihau, Tril abi: No Sprecneh Mandarin!!! <_QZ> but "no," shoulda worked Tril: were you working on a "Automated subproject web interface"? yes I am could you explain? it's my main project right now after school related things ok, simply, you create an account which stores your id, name, webpage, and some text info then you are allowed to create a project off the root, or ask to join existing porojects err, I mean you can propose a project off the root. anyone with an id in the system can claim leadership of a proposed project and become some coordinator of that prh? k <_QZ> Tril: yosh on irc.gimp.org#gimp told me what my problem was yesterday so you can propose projects and claim leadership of them yourself or just wait for someone else to does leadership empose any requirements? the project database will only store a short paragraph about the project and a link to a web page. It is intended for this system only to provide a structure for projects, the actual project will be on its web page <_QZ> Tril: the configure wasnt finding the correct glib-config so i had to tell it where it was and gtk compiled qz, oh yeah well -config's sohlud be in the path <_QZ> Tril: now imlib and all the gtk apps wont compile cuz they are all finding the 1.1.12 gtk-config but the 1.0.0 libs <_QZ> Tril: and when i remove the 1.0.0 libs configure totally fails to find any gtk libs, even tho they are right there where the 1.0.0 libs where Tril: My suggestion was that sub-projects are turned into sections in the document page. <_QZ> :) <_QZ> times like this just make me wanna... binEng: in this system the leader is the one people will email about the project. leaders also are responsible for adding/deleting members from their project, disowning subprojects, and claiming disowned subprojects leader can trasnfer leadership to another user id the leeader must keep the webpage and project info blurb up to date s/about the project/to request to join the project/ 01:40pm and the documents? Where will they reside? how about project-@www2.tunes.org aliases to the current leader? any additional info or resources a project hasbeyond a paragraph description and a list of members and subprojects, will go ON THE PROJET'S WEB PAGE. s/aliases/aliased/ This system is not just for tunes, it can be used for any kind oif project. hcf: I'm hoping to have everything related to this program on projects.tunes.org so the Documents for each project are on the project's web page. this system can be used to link to different sections of the tunes web pages, one for each subproject. but they dont have to be on the same server or anything it's just urls I think the different projects should use the same web layout. also we (tcn/I) decided that "jobs" or "tasks" in a project should be listed as proposed subprojects of that project. then when someone wants to use it they claim leadership of it. binEng: sure they can use the same layout, they should because they are part of tunes and on the same site, but that won't have anything to do with my program. π _QZ/#tunes is GONE ( bbl ) ok Eih! Tril, You have any page on theverge.net?? Linuxuser, never heard of it. does it have to do with free os projects or unix administration? 01:50pm ωνω SignOff jeff: #TUNES (Killed (services1.openprojects.net (Too many invalid passwords))) no, only a domain name for a dynamic IP ready !! ωνω jeff [jeff@a1-3h167.neo.rr.com] has joined #tunes Welcome to the TheVerge.network server The making the web work for you. Currently we are building our website please check back at a later date. Thank you. Page Updated(12/24/98). "The making the web work for you"? Bad English there. hi jeff what brings you here..? I'll be back for meeting after I go to school and take care of something (hopefull before 22:00gmt) π Tril/#TUNES is still away There is some place which really offers domains, like dyn.ml.org ??? 02:00pm ?? Eih! check my web page : http://194.65.220.20/~linuxuser/ ωνω Fare [rideaufr@137.194.160.71] has joined #Tunes hi Fare I am adding some things about my OS project ( is under development!!) My PC emulator is still with some bugs, can someone help me with the error msg??? bochs: panic, ESC4: still someone here ?? π binEng/#tunes is here so, the error msg, any one have a answer?? nope :) The Trill, pehraps he knows ?? 02:10pm perhaps someone here works with PC EMULATOR, for start ?? I am write the kernel with gcc,and inline assembler, but... PC emulator? yet another bochs? when I make a flat bin, it says that some variables are not defined, which is not true... free software, this time? pc emulator with inline asm? for what arch??? looks silly enough or else, there's DOSEMU the variables are passed through a function and inside it I have inline assembler with __asm__() in the asm, I put some variables from the param of the func,.. and when I compile to an ELF object, it works well, but... then I link to a flat bin file and it says that, the variables in inline assembler, aren't defined,.... which is not true !! whaddyacall "flat bin"? is it still ELF? are you bitten by _underscore a.out conventions? anyway, variables in inline asm are just *silly* why not write generic C code, anyway? so you get a portable PC emulator... 02:20pm ωνω core [core@core.suntech.fr] has joined #tunes tiens, un core! re tiens, un fare, sans son cheval? bienvenu!! merci :-) π core/#tunes file un e a fare :) <_QZ> core: fare and tril have been whining for months cuz u wont set bespin as primary dns for tunes.org l'autre jour, il a ete suggere ici que je vienne te voir face a face avec un fusil a pompe, vu qu'on vivais "a cote". Il a fallu que j'explique qu'on etait quand meme a deux millisecondes de lumiere de distance... <_QZ> core: could u do that so i dont have to listen to em for another several months? :) qz: i'm spending 10 hours a day at work, and about that much with my future-wife, i am a little short on it :P Fare : it is a com, DOS file!! (ok, you'll say, i still have 4 hours left, that i usually waste sleeping, but..) linux: what compiler are you using? qz: i didn't quite get what Tril wants :/ the 'hidden primary' thing. Fare : gcc why?? lu: if you go through COFF/a.out instead of ELF, between the compiler and the linker, then indeed, you need to underscore variable names... linuxuser: because you suck :) Fare : What I only want is getting the variables in inline asm to work, only that, nothing more, ok ?? core: "hidden primary" means the "official" server is some machine, but the real technical server for DNS is another <_QZ> Fare: u know anything about gtk? QZ: been said that it's good QZ: I don't "do" GUI programming, you know! <_QZ> Fare: do u have gtk installed? so, bye!! ωνω SignOff linuxuser: #TUNES (changing servers) 02:30pm et, j'ai du 1K/s, alors pateince! π binEng/#tunes doesn't speak French, and neither does he want to learn. 02:40pm ωνω gleam [gleam@wfisher.sa.enteract.com] has joined #TUNES ωνω gleam [gleam@wfisher.sa.enteract.com] has left #TUNES [] ding dong... 11 o'clock π binEng/#tunes sighs bE: sigh with me! π core/#tunes humms to himself bE: always look on the bright side of life! 03:00pm like what`? always look on the light side of life! If life seems jolly rotten, there's something you've forgotten :) And that's to laugh and smile and dance and sing. look, I just think we should begin now, since it's time. π Tril/#TUNES is back from the dead. Gone 3 hrs 16 min 41 secs we never managed to get everyone to understand the time of meetings on GGI, so don't worry people can't figure timezones out, apparently :) π Tril/#TUNES will have to depart at 23:50 Hi Core!!! hi david ;) yes, I live π core/#tunes shows :) do you think you could do something aobut the DNS? tril: sorry, was busy as hell as usual; i didn't quite get what you want me to do :( core: see http://www2.tunes.org/cgi-bin/TunesWiki?DNS tril: ok basically, I want bespin to be unlisted primary, your servers + cleese.nas.com + ns1.canuck.gen.nz to be secondaries and listed in Internic canuck is already setup and copying the zone from me, cleese will do that anytime I ask so i would show up as primary on internic records, but you would be primary in the zone file. right? core: right core: yep, that's right ahh.. i see. i'm a bit slow, and not very bright, so bear with me :) AIC? I hate adaptec adapters... fare: better than AIX :) That way I don't have to register a host nic handle. i understand probably way better than A/UX, too ok, i'll send the nic update right now, as soon as i get the acknowledgement from them that it's been changed, and it shows up in their database, i'll setup netnation as secondary and copy the zone from bespin π core/#tunes thinks the french GIX is actually strings and empty yoghurt cans core: ok I'm here, chatting with bineng let us all talk in the channel ;-) sounds good k i read bineng's posts on the ML, they are pretty interesting :) i was just saying to bineng that even though the language is part of the object system, you don't have to use that language 03:10pm it's not a problem, because it's not a language, but a meta-model whichyou can describe any language in, and when you do, that language becomes a first-class part of the system. i read bineng's posts on the ML, they are pretty interesting :) oops binEng: I thought we were going to talk about the web pages well, before you design an object format, i have one ;) Tril: yes, but... :) core: if yours is so good, how come Fare isn't satisfied with it as a base for Tunes? maybe because you won't release the code? :) tril: because he didn't have the code until 5 minutes ago π Fare/#Tunes hasn't really read the ML in weeks tril: he has it now.. as i told him, i'm sorry i'm months late delivering it to him, but i only have 24 hours in a day, and i spend most coding .. :/ and i'm no genius either, so it takes me time Fare: is there a way to get a message to you and put at the front of your to-be-read queue? tril: i just delivered to him, a booting operating system that goes into an interactive debugger and makes use of binary independent components. what if each app could be an OID server...? that would really speed things up (if using a "system language") Tril: send it personally with a URGENT priority ? core: maybe send a copy to Tril, too? Fare:what header field is that? X-Priority? core: Why not share it with everyone publically? π Fare/#Tunes is trnslating the conclusion of his LL99 article... core: irais-tu par hasard a ll99 a Brest? J'y serai conferencier... tril: i wanted a few days of fare's review first binEng: I don't think I understood that part, what is an OID srver..I think I know what it means.. just a container for objects, right? so you're ready to release it? tril: yeah is the nokernel.org website ready? Tril: keeps track of the objects under it in the systree tril: it's not up, but sometime soon yes Tril: like a object database, including what object is using what... are those "stackable?" and when you say stackable, do you mean they can be nested? I.e. decentralized... Tril: How about John Little's model? Tril: yes! that's the point I haven't read Prism stuff in detail yet. I think it sounds interesting but it is definitely on the same track as everyone else maybe resembling the Arrows lang, but easier to grasp :) Tril: unique ID server for global transactions without name clashes there are commit problems for implementing correct OIDs. Fare: yes, that's OID server (not a very good name, but hey...) those servers need to be able to scale upwards also. to add new superdomains Tril: how d'ya mean Another problem is, ATA drives give NO WARRANTY on commit, least you stop spinning the drive everytime you want your commit. why "O" in OID? That's just redundant! π Fare/#Tunes fights "OO" crappiness Fare: for object :) bE: down with "object"! "OO" is a poison in the mind of computer programmers. binEng: if I have a tree of nested oid servers on my box I want to become part of a group of systems .. so we all become nested into one super-oid server. and I want to be able to do that again when we find another cluster, etc. don't spread the virus! binEng: "object" is synonymous with "identity" thus OID is redundant GID, for "global" ID, is much more interesting... 03:20pm or just ID, if it's not global Fare: Who cares about the drive not guarnteeing anything? there is a few second delay on changes, of course if the power goes out you will lose a few seconds or 10s of seconds of work. No one expects differently. Tril: I don't know if you actually do that... my boss has a nice article being prepared on decentralized global IDs. Tril: because if you communicate ID 100 to another host, you shouldn't reuse ID 100 once you reboot after a failure! Tril, usually you import the remote computers' trees into the network manager Oh, we're talking about the same topic. I thought it was 2 subjects Fare: that problem can be solvd by preallocating ids, and by having a confirmation period for synchronizing new allocated ids ωνω hcf has changed the topic on channel #TUNES to: meeting NOW! topic: web layout. for topic items see http://members.xoom.com/unios/tunessite.html "for the next few hours your ID is tentative, it will be finalized after that" possibly changed at the end of the waiting period. have a look at the main page I did for u... is it OK? binEng: has that page changed since yesterday? I converted that to the Wiki TODO page. http://www2.tunes.org/cgi-bin/TunesWiki?TODO Tril: yes even the preallocated IDs must be committed I don't understand the problem. And we're discussing the site now anyway. timeout is indeed a way to commit. Only ATA doesn't even provides for proper timeouts... besides, using data before a commit is called "optimistic computation", or "speculative computation". binEng: can you put a link to the wiki todo list? http://members.xoom.com/unios/tunes.html Tril: link to what the url I just said ωνω hcf has changed the topic on channel #TUNES to: meeting NOW! take off-topic chatter to private. topic: tunes site web layout. for topic items see http://members.xoom.com/unios/tunessite.html I think especially the "documents" should be integrated into the site.. web pages *are* documents. whatever the layout be, it must be done by a metaprogram the Makefile infrastructure is here already the metaprogram ain't here yet, tho We can use the makefile to put the nav bar on every page. exactly or not do it, depending on people's taste The nav bar changes depending on the page it is on we might even have several versions of the site: for people who like frames, people who like navbars, people who want neither Let's do it with nav bar first. now. which item number is the navbar? oic, #14 So? Tell me what to do! #14??? 03:30pm binEng: you want to volunteer to make these changes to the website? w/ a small navbar at the top like binEng's prototype, theres no need for frames I can make some web pages now. You'll have to fill them yourselves. binEng: it would be easiest if you had a bespin account. I would like to create a directory for the "new" pages, for you to put the pages you make in. Anyone to write the short intro on the first page? well, I have no bespin account msg me the l/p you want, and I'll create it. If you fill out the tunes member form, you may also have a tunes.org mail address. I think the short intro is fine the way it is, but it will be changed. I just copied it off the old intro sorry, girlfriend preempted me - back./ uh, why not just use the CVS? Or did you mean, have an account SO AS to use the CVS commit facilities? what about the links on the navbar (on binEng's prototype) any comments? core: hi-pri IRQ, or NMI? tril: i want to make it public; and i certainly hope Tunes will find it good enough as a basis to ameliorate to reach your goal Fare: We may be creating a new directory tree, or moving subdirs around and renaming files. That's difficult to do if working through CVS. fare: NMI.. :) Tril: ok, ok. I thought the current tree was good enough, when I made it! tril: i'll keep working on it no matter what, but i would be sad to miss the flux of ideas from the tunes people core: what features are working? persistence? reflection? what languages support it? Fare: I told you at the time that I would want to restructure it eventually tril: it's basically a binary components resolver and executor; it boots a pc into protected mode, virtual memory, and connects/executes components tril: it supports any language; persistence would need storage, that i don't have yet; reflection would need a language on your part. Tril: sure :( Tril: I'm not fanatically attached to it. tril: it will do for you what the OS kit does; but it is fully componentized, there is no trick, everything is a component, even the component (dependency) manager Fare: we're not leaving you out tril: it's like KUT, except it does boot and allows for interactively playing with it core: why not allow insertion of OSkit components in it? core: that would be great! tril: but i don't think i am a genius, so if you'll rather go with OS kit i'll understand fare: that's not hard, just someone to write the necessary code core: unless you intend to rewrite linux device drivers from scratch, that is! fare: no, not really :P 03:40pm maybe the entire thing just plain sucks, i don't know. ok everyone talking about the no kernel please join #nokernel. there's a webpage meeting going on in #tunes... sorry, i'll shut up.. well, hcf, what is it you wanted to say I think I'll put bineng's page in a tunes-owned directory, and let everyone mess with them with no CVS moderation i didnt want to say anything specificly π hcf/#tunes had a passing thot that just maybe there supposed to be a meeting going on has anyone objected to bineng's layout? not as far as I've heard except for the colors :) thats just cos noone is talking about the topic except for Fare/hcf, who don't want any color at all. Tril: no, i said i didnt care about the colors π Fare/#Tunes thinks that layout is not lynx-friendly enough... install PHP and use web themes already :) you can have one friendly to lynx, one friendly to me (how tunes.org currently looks like ;) and one for people who really feel annoyed when they don't have their 500 KB of jpegs ;) I don't know what that is..I'm not a web person i haven't used it, but it allows users to select a 'theme' for viewing your pages like WM themes and I think the colors are much too flashy, too i heard it is more than nice, and it works with PHP3; i plan to use it for nokernel.org themes look a good idea to me the pages should be done without any color, that can be added with css with meta tags. (does dsssl do th esame thing?) that's basically what themes do, provide a style like CSS except the user can select it (iirc) dsssl can do anything -- there's Scheme under the hood (and you don't need a css-capable browser, either :) well, I've got to go to class. bineng, mail me that login/password, or tell fare and he'll run adduser frankly, I don't see anything wrong with the current pages. Perhaps there lacks a blurb on the main page 03:50pm and of course, automating the navbar would be cool. bineng thought they didn't have a consistent layout. and many have complained they are hard to understand and navigate. but the general layout looks right and everyone says the content needs to be updated and revised. the Tunes pages are a reference for content, for me, but it's just IMHO π Tril/#TUNES is away: (Class....Bye!) [BX-MsgLog Off] it needs update, but i think the work compiled by Fare is quite rare. ok, ok there are some betterments with respect to the design, but I see no point in having a flashy freedows-like page. fare: opinion shared I want no freedows effect at tunes.org! the colors aren't important right now they seem to irradiate people and force the team to split :P i could probably contribute to the OS review, especially embedded operating systems one, whenever i have some time (ha, ha.) I admit the review subproject is a mess, and the glossary needs be redone properly, but the main pages I did recently are generally good (or so they seem to me), apartf from lacking consistent navbar (for those who like) *shrug* i like your pages. but it's just me. This isn't gonna lead anywhere, is it? π Fare/#Tunes 's primary test borwser for tunes pages is lynx. i can find what i'm looking for in three keystrokes. in lynx, I just /foo to look about foo... binEng: if YOU are in charge, they YOU decide, ok? Only you also get to hear the advice and complaints of everyone else, too! but I dunno what to do now 04:00pm abi: redhat? redhat is 2 CDs abi: dinner? i heard dinner was served.. binEng: is the meeting dead? looks no better And neither Tril nor Beholder is here so what now? did beho ever say he could make it? not really... but he's the one who should. if the meeting is Officially Dead (tm) I'm outta here. u called, u can call it "officially dead" if u want s/called/called the meeting/ --> Meeting killed by bloodthirsty binEng <-- ωνω hcf has changed the topic on channel #TUNES to: ... cya ωνω SignOff binEng: #TUNES (Leaving) RIP the meeting. π Fare/#Tunes calls for one minute of silence... 04:10pm π core/#tunes opens a bag of chips.. oops core: not intel chips, I hope? π Fare/#Tunes goes fry an egg... javachips? fare: no, those come in grapes 04:20pm ωνω tcn [tcn@cci-209150250108.clarityconnect.net] has joined #tunes hoy tcn hey, is anyone else here? 04:40pm fare, core jeff.. you guys awake? >>> tcn [tcn@cci-209150250108.clarityconnect.net] requested PING 915841895 from #tunes Hey, did they already have that meeting? kinda I'm snowed in here.. no band practice tonight hmm, reading the irc log.. abi, what's bonsai? bugger all, i dunno, tcn tcn: iv seen something on a band that composed an entire album over the net, try that ;) thats mozilla.org's thing il get u the link bonsai, got it.. is that like a better Wiki? abi: bonsai is also at http://www.mozilla.org/bonsai.html abi: bonsai is also cvs tree control okay, hcf. I see mozilla.org's tools page is at http://www.mozilla.org/tools.html hmm.. I'm working on a v86 monitor.. fun stuff :( 04:50pm dear #tunes lurkers, I've just finished to upload my now english-translated article where is it? submitted to ll99 and accepted. ll99? ll99 is the first french-speaking scientifical conference devoted to free software and free information in general. In Brest, 25-27 jan. 1999 it's in www2.tunes.org/~fare/tmp/ I mean, my article. The translation is *very* rough. Please help improve it! ok, scientifical is just 'scientific' Fare, when did you get the idea for Tunes? 05:00pm tcn: is that an interview? :) π hcf/#tunes likes 'scientifical' better I never got "the idea" for Tunes, for I doubt Tunes is just "an idea" :) fare: I thought I'd write a little thing about the history of Tunes, since it's been going on so long.. I thought it would give the project a better sense of direction if we know where it's been so far tcn: if you read the MOOSE and TUNES archive, and dig through the tunes "releases", you'll know a lot about it :) fare: hehe.. allright, I'll just continue to do that tcn: also, I think my latest article gives some insight on what direction I'm heading towards. tcn: now, if you have a precise question, I'm open and will answer. It's just that you're perhaps being too vague. what happened to MOOSE? Anyway, I've *long* intended to build a better computing system; but my idea of that system has evolved a *lot* in a bit more than ten years... Fare: would u mind converting it to html? fare: well, my idea of it has evolved a *lot* in the past 2 years, since our first discussions if you remember them, so you have the edge hcf: converting what? my article? Fare: yes hcf: go ahead and do it, if you have a latex2html or hevea ready... π Fare/#Tunes hasn't tried hevea yet... hevea? I'm reading it with 'more' :) tcn: uh? less is better than more! fare: less screws up all the accent marks. hmm.. should work for English though :) 05:10pm tcn: uh? there is export LESSCHARSET=latin1 and/or less -r thanks hmm, interesting.. Where you have %%% ??? is that something you're not sure of, or just a latex markup? ωνω jeff [jeff@a1-3h167.neo.rr.com] has left #tunes [] 05:20pm fare: "now, there is no solution to this problem that be reliable for long-lasting, non-superficial effects,"... tcn: something I'm not sure of tcn: particularly in the translation % begins a tex/latex comment fare: "be reliable" should be "is reliable" or "will be reliable" "will be" is right ωνω shinya [shinya@INS36.sapporo-ap2.dti.ne.jp] has joined #TUNES fare, got another one: "throwable %%% jetable ???".. use "throwaway" Hello Hi ωνω SignOff shinya: #TUNES (Leaving) tcn: ok la responsabiliti suit l'ouverture?? what's suit mean? suite? follows? the responsability that follows opening? tcn: that's in a comment, anyway 05:30pm oh, got another one, in the following paragraph.. 'diktat' is 'dictate' tcn: loss of responsibility follows partitioning, responsibility follows opening no, diktat seems good diktat is German it's a german word, but it's permeated in french and english since the Versailles treaty What did you mean by 'diktat' there? A demand, an order? yes, an inadmissible, yet unescapable order. π Fare/#Tunes asks the online webster... hehe normally you say dictature? not the same oops do read the webster that's a dictator a friend of mine has written a script that queries the www site, strips its WWW ugliness, and dumps the result to stdout... the english version is 17 pages; the french version is 18 page. translation is lossy this French dictionary sucks. For 'dictate' it has dicter, dicte'e.. fare: no, you just eliminated redundance :) and the poetry :) haha core: I'd appreciate your advice, too :) le drive? quel drive? floppy? kesako? fare: sure :) You should translate the poetry.. that one example "The loss of responsibility" translates perfectly.. π Fare/#Tunes goes to the "antique" department cpp: command not found kesako? what language is that? 05:40pm brb hey, cool, Tran converted his pmode.asm to NASM tcn: joy. tcn: did he? π core/#tunes watches Fare boot his floppy (?) :) I thought pmode was OLD STUFF. fare: phar lap is still selling :P I can't believe that fare: it's still a pain in the ass! I guess chips rise to their level of incompetency too, as the 80x86 dominates the hell wroke loose! tcn: self quote: As long as software is not free, we'll have hardware compatibility, hence bad, expensive, hardware that has decades-long obsolete design yeah, this is why I'm working on Tunes :) tcn: btw, did you finish reading my article? any more comment? So when I have a big-ass massively parallel computer, I can use it for something fare: I've just read pieces of it so far, give me a little while it reads pretty good though, better than most papers written by Americans and English.. the translation is very rough, though. Half of English derives from French, so it translates well fare, what program translates sgml->html? depends on which sgml that is sgml is a pretty generic name 05:50pm for LDP documents, use sgml-tools 1.0.9, which includes a sgml2html ok, that's what you used for stuff like the FAQ, right? apparemment, y a un programme qui lance cpp depuis le C... oops, wrong channel hehe tcn: sure, since I'm used to Linuxdoc-SGML from the two HOWTOs I wrote 06:00pm fare: s/one can but noticing/one can't help but notice/ k π Tril/#TUNES is back from the dead. Gone 2 hrs 16 min 37 secs fare: What program are you working on in #nokernel? <_QZ> Fare: u know lots about libs right? hey Tril <_QZ> -rdynamic <-- what does that mean on th gcc line qz: man gcc :) <_QZ> -l is lib but what is -r tcn, core's system is almost ready to be released.. cool. <_QZ> core's system??? A no-kernel system? 06:10pm yes, a working system that boots up and has a debugger, it basically is a binary module combiner abi, GRUB? tcn: no idea anyone know? (where it is?) abi: GRUB is GRand Unified Bootloader tcn: it'll be for public consumption in a few days qz, dunno -r, doesn't seem to be in gcc info pages.. maybe egcs specific? abi: GRUB is also http://www.uruk.org/grub/ okay, hcf. tcn: i'm having Tril and Fare review it first -r is a ld option it produces relocatable code. abi, translate #nokernel from French to English, output to #nokernel-english :) haha abi, what do you think of intellectual property? i haven't a clue, tcn abi, what's intellectual property? hmmm... intellectual property is bullshit! abi, where's the 6-bone? tril: i haven't a clue abi, what do you think? wish i knew, core abi, do you think? i don't know, core abi, what's my toafm? tcn, your toafm is Wed/Thu 2100-0300, Fri 2100-0700, Sat 1500-0700, Sun 1500-0300 GMT abi, donde esta la plaza de catalogna? i haven't a clue, core abi, chupa me la poya? i haven't a clue, core haha abi, what's my dog's name? tcn, your dog's name is Simon abi, why doesn't Fare wear panties? wish i knew, core 06:20pm panties? hcf, maybe the negative responses should be customizable in realtime. <_QZ> hehe Tril: maybe PROGRAM ME!!!!! abi: I wish there was a knob on the TV to turn up the intelligence. There's a knob called "brightness", but it doesn't seem to work. -- Gallagher I'm non-sentient. I only do what I'm told. abi, why does a new linux kernel release come out while i am compiling the latest one? core: wish i knew abi, why does the linux kernel take so long to compile? no idea, tcn π Fare/#Tunes still awaits comments on his article... maybe I should post on the list? π tcn/#tunes is still reading it π Tril/#tunes adds commenting on Fare's article to .project π Fare/#Tunes downloads hevea fare: so far so good Fare: whats hevea? hevea? I disagree, though.. C is harder than Assembly! tcn: it's not harder; it's as hard. but you can write and maintain C much faster. even if some of my C, like vmm routines, look like assembly. core: well, harder for the stuff I do. As a high level language, C sucks.. abi, C? rumour has it C is quite low-level, drivers should be rewritten to be high level 06:30pm tcn: what are you doing already; your name is familiar <_QZ> any1 know what "1>&5" does core: I'm doing a prototype Persistent Object Store.. qz: what language? language is pretty much integrated in the whole system, right? tcn: oh, yes. good, my system doesn't have that yet, but all the rest of the base works :) <_QZ> bash π tcn/#tunes smacks abi.. "dumb ass" :) _QZ: redirects stderr tcn: she's just trying to contribute to the conversation. I though "1" was stdout and "2" was stderr.. never heard of "5".. ahem Fare: whats hevea? ca a l'air de marche, apres tous ces efforts! Welcome to clementine. Press F10 to switch keymaps. ouiiiii :)) core: did you do a v86 handler? tcn: not yet, i want to finish user->supervisor gates, usermode, and process context switching first tcn: but everything is setup to have v86 tasks yes tcn: it's not written tho, so your expertise will be welcome π Fare/#Tunes falls literally asleep core: thanks a lot for your patience. fare: what processor does it say as CPU 0? help ^ Unknown command fare: are you kidding? you're welcome.. thanks for yours :) heheh.. well, i'll add '?' to do the same thing as under IOS soon you have PATH problems, it looks like... IOS? type 'show memory caches' (or sh mem c, or s m c :) core: What language are you using? there's no filesystem yet, silly :) 0000200C: 'user' never should be a file system. tcn: C, ugly-but-predictably-compiled-C and assembler tril: i don't need one, at least. slabsize 00001000 objsize 00000040 objcount 59 waste 60 tril: right.. tril: if i can get the persistant store on that, i'll be fine fare: cool. (je connais.. enfin bon, ca gaze:) mask FFFFF000 offs 00000FE8 nextcol 00000008 maxcol 00000038 Fare: whats hevea? hcf: He said it earlier in the logs.. grep hcf: a latex2html rival written in ocaml fare: don't bother typing all, it works.. was just curious ;) fare: what did it tell you as CPU 0? was it right? rpm available from ftp.inria.fr/INRIA/Projects/para/hevea/ Cyrix 6x86 (M1) -- righto fare: cool! what's CPU 1? fare: i'm not so dumb after all tcn: the second CPU in an SMP system? :) 06:40pm well, it's a 6x86L, but you'd have to look at the stepping to know fare: i know. if you can add that... fare: not that it is _vital_ at the moment what? correct CPU detection is the primary feature of an OS. lol I'll upload my thing one of these days well, please add it :) tcn: well, our areas of work seem to overlap I don't want your piece of junk that cannot even detect the latitude of the CPU correctly! tcn: i wouldn't mind persistant store and v86 mode on my current system lol fare yeah, if we can use the same object format.. great.. Alonzo will be very happy. i use one with associated binutils anal raping patches, if you want to improve/modify it it's still time hum, I *really* gotta go to sleep, now fare: good night, and thanks for your trust/patience ;) later fare Fare: Good morning fare: you can reach me on the phone or email if you need bonnuit π Fare/#Tunes is away (sleep: goto sleep) core: yeah, right :) tril: shuddup :) He said you never answer the phone tril: Fare knows how to build the system now. he can educate you. therefore i can go back to being unreachable ;) tril: i'm never home tril: i have a cellphone, but people who give the number away are dead meat core: Where can I fetch a copy of your system? core: you can give it to tcn. He''d be helpful. or, let me give it to him.. lol.. fare just called me! :) tril: i guess tcn: well, nowhere, but i'll give you a copy tcn: same conditions apply than to tril and fare: <_QZ> core: do u have a binary for yer OS? tcn: do not make it public until i do (a matter of a few days, but i want to finish several things), and please have a look at it, don't just put it aside qz: yes. core: do NOT give it to qz I do not trust him. Or, if you do, get a copy of brix out of him in return. <_QZ> core: send it here, i have many many machines to test it on qz: not yet. well, fare just made one, after a lot of effort ;) sure thing, core qz: just wait a few days if you don't mind.. i want it reviewed by a couple people first and a couple of things fixed <_QZ> caore: i can review it :) tcn: bug Fare about how to get it to run ;) s/run/compile/ core: what are the hard/software requirements? tril: any i386+ PC tril: with at least 1 MB of memory tril: if it doesn't match that, it's a bug :) <_QZ> core: how much of it works? tril: although, the virtual memory manager is compile-time configured for i486 individual page flush, you'd just need to edit an header to make it run on i386; that won't be a problem whenever there are i386 and i486 sections in the same object that are choosed by the loader, as planned qz: it boots into an interactive debugger, the components system is working (ie. you can query functionality display, of type text or graphics, which is local, of version >= 2.5 but not 2.3 because it is buggy, and with at least a given set of functions you need) The machine I'd run it on is 486dx2/66, 16meg, and a funky CGA-compatible video card (but mono screen) qz: it manages virtual memory, cpus, there is a proper keyboard driver tril: that's perfect. tril: the entire running system uses 176 KB of runtime memory at this very instant. 06:50pm <_QZ> no one has dcc'd me a binary yet :) (with paging, that is; with a linear memory manager, it would take half i bet) (just write a linear memory manager component and swap it :) core: I was gonna use the virtual memory hardware for peristent storage qz: you'll have to wait a few days <_QZ> core: why??????? qz: i want it reviewed by tril and fare first, and them to tell me what sucks in it; there have to be things. noone has ever seen it except two people. <_QZ> core: i am an OS coder qz: and i want to finish usermode components management, so it doesn't shift under people's feet <_QZ> core: i know more about OS coding then they do qz: i know - have been waiting to see Brix for a while :) ωνω binEng [bineng@dialup45-2-2.swipnet.se] has joined #tunes hi bin qz: i don't doubt that.. please don't think i'm doubting your abilities.. i just want it browsed by a few people i know, before it goes public Just checking by to see if something happened here hello.. hi anything? core's OS oh? more like component set that happens to boot by themselves :) what's that site that gives free POP3 access through the web? tril: france telecom? *duck* tril: or hotmail do you mean? ;) no, they gpl'd the software. oh. hm, no clue there are a couple of programs in GPL to do that already anyway webmailfolder for one π Tril/#tunes tries to remember the address I know, I'll grep my netscape cache. that doesn't help <_QZ> hey did anyone know what 1>&5 did in bash? 07:00pm qz, looking <_QZ> looking where? I'm looking in the bash man page for you. <_QZ> i dont think its there qz: I say the 5 is upside down :) <_QZ> tcn: ya if it was a 2 it would dump the error crap I use tcsh anyway QZ: It redirects standard output and standard error to a file called "5". no, ok, that is wrong. It redirects stderr+stdout to file descriptor 5. ah I found it <_QZ> Tril: that stuff is on page 21 and 22 1>&5 means make 5 the same as stdout. so later you can do 5>filename just as if you had done 1>filename. <_QZ> er 19- 21 tril,tcn: ok, i wrote build instructions for you hopefully it covers all traps Fare fell into, as well cool. <_QZ> well configure is putting that on the end of this gcc line <_QZ> i ran the gcc line in bash and it compiles fine so my gtk libs are fine qz: how's Brix coming, anyway? <_QZ> but configure says it fails and cant find gtk <_QZ> core: if i could get E installed then i could finish brix qz: ahh :/ qz: still working on it alone? <_QZ> yup i have to say that some (small, but useful) bits of what i produced have been written by andreas beck, of ggi fame. it has graphics? 07:10pm tcn: andreas will build the graphics layer RSN. it's not a pet project either. <_QZ> i did make it so it can build either an unprotected non-paging kernel (lite) or a protected paging kernel <_QZ> all in one source qz: what i did can also do that; just have either a virtual memory component, or a linear memory one (unprotected, not paging), both exporting "memory" functionality; just let the user choose what she wants. tcn,tril: i'm counting on your input on what i gave you, ok :) core: ok the way it is now, it is possible for someone to add components to the system <_QZ> core: can u cycle thru them at boot? ωνω NetSplit: asimov.openprojects.net split from devlin.openprojects.net [07:12pm] ωνω BitchX: Press Ctrl-F to see who left Ctrl-E to change to [asimov.openprojects.net] I'm gonna support 8086 real mode :) ωνω Netjoined: asimov.openprojects.net devlin.openprojects.net ωνω Tril [dem@bespin.ml.org] has joined #TUNES <_QZ> tcn: that is dead qz: the code to select it isn't written as of now, but it could be done without trouble qz: and not just for memory; for anything, since anything is a replacable component qz: and "big real mode".. in some cases, that's better than pmode ωνω NetSplit: asimov.openprojects.net split from sterling.openprojects.net [07:13pm] ωνω BitchX: Press Ctrl-F to see who left Ctrl-E to change to [asimov.openprojects.net] ωνω Netjoined: asimov.openprojects.net sterling.openprojects.net ωνω Tril [dem@bespin.ml.org] has joined #TUNES users can redirect console beeps to the parallel port to flick lights on a christmas tree, the system will never know. <_QZ> i was gonna have brix allow the admin to switch kernels while running one but it only adds bloat and brix can boot so fast that the bloat is not worth it you can hotswap components as well; just have the one you want to swap, store its state and the new one gets the stored state and takes over; links from other components are dynamically updated; and there. The idea with Tunes is, you can replace any part of the system while running tcn: i miss persistant store to do that (and have components to replace from :) but the system is ready for that. tcn: Fare heavily influenced by vision of things :P <_QZ> brix doesnt need to hotswap stuf qz: the os doesn't.. the user does. <_QZ> u just unlink it from whatever was using it and its automatically saved core, hey thanks. let's build it once on bespin (pentium) and download the binary. (Sorry I was desynched away) <_QZ> Tril: dcc the binary here same rules apply to binary :P core: So the faster I give feedback the faster you will release it? where are instructions? tril: right :) tril: feedback, patches, "your design is stupid, here is how to do it", things like that ;) <_QZ> Tril: i gave u a copy of brix. its not my fault yer floppy drive was broke qz: we aren't trading systems behind people's back. my work will be available to everyone to tear apart soon, i want it hammered by people i know first, just be a little patient ok? 07:20pm qz: Yeah I'm gonna try it on my oehter 2 floppy drives .. qz: i'll happily get your ideas and feedback in a couple days.. core: Where are the build instructions, again? <_QZ> :) <_QZ> core: yer OS was called what? qz: it's not an OS, it's more like a set of components put together that boot into a working system :) but it's called clementine, because it has no kernel hahah core: binutils has XCOM support? how did you get that? didn't Billy Tanksley invent XCOM? or was it you? tril: come again? tril: no, i came up with XCOM tril: and i wrote a 100K+ patch to binutils, that's how :) oh, missed that line :) tril: but tanskley and other Dolphin people asked if they could use it, so i said sure :) what is XCOM? tcn: executable component format is this what Alonzo wants, as a "universal linker"? tril: well, it's not OS specific at all. it is designed to interoperate as much as feasible tril: with XCOM you can define an interface for a component to Tunes, and another to Brix, as long as both systems use XCOM (or have a loader for it, at least) <_QZ> brix doesnt use components tcn: it's an executable format with dependency information so that components can connect and call functionality in eachother tcn: like, i export that, i import that, etc :) qz: ah well, you get the picture :) qz: then what is Brix? i thought brix stood for bricks, as in components :) <_QZ> smaller bricks :) me too qz: like smalltalk? qz: 3 lines of code can be a component <_QZ> brix is just like tunes <_QZ> but it uses protection qz: brix is reflexive? <_QZ> ya qz: so does tunes, at a different time ωνω Tril_ [dem@hh154ws03.hh154lab.wwu.edu] has joined #TUNES tril, I got Squeak.. hmm, looks like a Mac :) qz: you can write code with code? <_QZ> tunes uses its compiler to add protection Core: so every component in the system is encapsulated in an XCOM module? <_QZ> brix adds it with the kernel tril: correct tril: well, it can be something else, as long as it provides component semantics <_QZ> so the programmer is free to code how he wants tcn: Incorrect. Macs look like Smalltalk :) qz: what is brix coded in? <_QZ> its all based on fucntions pure asm ωνω NetSplit: devlin.openprojects.net split from varley.openprojects.net [07:29pm] ωνω BitchX: Press Ctrl-F to see who left Ctrl-E to change to [devlin.openprojects.net] <_QZ> asm ωνω Netjoined: devlin.openprojects.net varley.openprojects.net ωνω Fare [rideaufr@137.194.160.71] has joined #Tunes Tril: I'm kinda curious about that abstraction system of yours... tril: all you need is a component exporting loader functionality for your given object format, and that the object format can store component dependency information; XCOM does that, but you can use another 07:30pm ωνω lar1 [larman@1Cust21.tnt20.sfo3.da.uu.net] has joined #TUNES I guess that means QZ and I are the only ones in here who like asm.. Is TUNES persistant? bineng: yeah I'm going to go home now, and try out all these OSes. tril :) ωνω lar2 [larman@1Cust21.tnt20.sfo3.da.uu.net] has joined #tunes hey lar Hi what's up, lar2 although I wonder if it is possible to boot off zip disk in this computer lab... tcn: no, i love assembler Sorry I wasn't here for the meeting... binEng: Some other time, sorry :( k bye ωνω lar1 [larman@1Cust21.tnt20.sfo3.da.uu.net] has left #tunes [] core, what's IOS? ωνω SignOff Tril_: #TUNES (Read error to Tril_[hh154ws03.hh154lab.wwu.edu]: EOF from client) ωνω SignOff binEng: #TUNES (Leaving) ωνω lar2 is now known as lar1 tcn: but i started writing this in may 98, i couldn't achieve that pace just with it tcn: and i don't want to be tied to one platform <_QZ> core: how come tcn gets a copy? u dont know him tcn: there are assembly parts, and some C that is specifically written to compile as fast assembler, and the rest is C so far What did I miss at the meeting? core: Actually my goal is to use languages for which a compiler/interpreter can be easily written in assembly <_QZ> lar1: the beer, the girls, the drugs.. Hehe qz: what he worked on is very interesting to me, and i trust tril who told me it'd be a good thing; again, please don't feel bad, it's a matter of a few days tcn: forth? ωνω SignOff Fare: #TUNES (Ping timeout for Fare[137.194.160.71]) I think the meeting took place a few hours early <_QZ> tril: was just kidding about not trusting me <_QZ> s/:// core: Yeah, forth, lisp, things like that. core: So far I only wrote a forth intrepreter ωνω Tril [dem@hh154ws03.hh154lab.wwu.edu] has joined #TUNES tcn: it'd be easy to run a forth interpreter on top of what i have now tcn: heh, i didn't, but i have the rest ;) I'm back, missed the bus. next one in 40mins tril: ack hhehe tril: Fare called me before going to bed :) he wanted to see if my phone really existed i guess ;) If I was at school at 7:30 I'd be the only one core :hehe there is one other person in the lab. there were more at this time yesterday, but today's Friday oh yeah Did we have a prob with external msgs? i would stay late at work if i was alone :P I don't start school till the 19th.. core: I wrote the forth interpreter to run in realmode and pmode, so I should have no trouble adding it in tcn: that could make something quite useful ωνω Fare [rideaufr@137.194.160.71] has joined #Tunes i hope you don't think my work stinks too much.. i work exclusively on that, therefore i don't have a very objective opinion :P exclusively on what? on this system hehe not on stinking - hopefully :P I thought maybe you had a job programming Forth, actually there is probably *much* to be changed in the whole thing, but it has the big advantage of working. 07:40pm working is always a plus :) tcn: lol.. no. i'm not that eager to get money :) i'm paid to do OS research I think core has an internet company, right? tril: it's doing so many things it's not even funny IOS? what's that? tcn: oh, sorry; it's cisco routers operating system; i like their command line, as you guessed by now :) oh, is that like DECSERVER, VMS, TOPS, etc? core: does that mean your nokernel is property of some corporation? tril: yeah, we have two sides; suntech both are a large IAP - hosting two backbones nodes in our offices - and software development company tril: no >>> Tril [dem@hh154ws03.hh154lab.wwu.edu] requested PING 375626512 from #TUNES tril: they pay me to work on it and related stuff, but it'll be GPLed That's cool. I'm paying to do OS research and get college credit for it :) OK, actually the government and my grandmother are paying me to do OS research and get credit for it. tril: heheh but I don't think either of them care what I do, as long as I go to school :) π tcn/#tunes has just about had enough of school!! tril: the company cares very much, they want to make lots of intelligent devices with it. tril: i don't really care what they do, and they know it; if they stopped now, i did the hardest part, and the nokernel thing will only go forward now that it is publicly released tril: there is enough code to gather interest and be useful to people for playing with it and changing it.. apparently just talking to a few people about it here sparked interest :) i'll be very happy if you hammer it until it fits into the shape of tunes :) π core/#tunes shuts up :) Well, people are sick of kernels.. they get in the way :) sounds like a great company tril: they're the best i've been at so far, and i've been at some rather known ones like sony or ubisoft. Anyone know much about ITS? tril: except i have little time for anything outside, as you noticed :-) abi, its? its is probably Incompatible Timesharing System, written at MIT around 1970.. abi, tunes? tunes is in my head or tunes is a useful nevertheless expedient system. abi, clementine? i don't know, core abi, you should go out more often. hmm, i think i killed kenny. core: that an offer? I can just see it: instead of a logo, just write TUNES IS IN YOUR HEAD across the top of www.tunes.org ;) 07:50pm abi, sure, just wear that tank top of yours abi, kenny? tcn: i haven't a clue ωνω SignOff _QZ: #TUNES (Ping timeout for _QZ[p0wer.qzx.com]) is it just me or Qz is a little touchy feely. π tcn/#tunes rofl abi, what's qz? i heard qz was the two least used letters of the alphabet what's _qz? rumour has it _qz is the creator of BRiX, and webmaster of http://www.qzx.com He is? abi, what's core? i don't know, core doh! abi bzip2? i guess bzip2 is http://www.muraroa.demon.co.uk/ Thank you abi lar1: no worries core is a file you don't wanna see in your home dir Hehe tcn: well, i have a funnie about that. my email account is called core, as well Hey, we have 16k of iron core in a drawer at school :) once, i was in /var/spool/mail after doing some crap in it, and i ran a program that dumped core i really hated myself. Do you still use that iron core? haha π core/#tunes ulimit -c 0's on that server since then. we might hook it up, if we get around to it ;) 16k is actually not bad, for 30 years ago i know someone who was selling 64K of ram extension to ibm 30/70's computers for $50.000 not so long ago. I have an original IBM XT... Looks like a bunch of miniature beads on threads.. like 40 gauge wire lar1: I told QZ some people still used 8086 real mode :) (my OS can run in real mode) 8086 real mode eh? tcn: cool, i can run your OS as a process of mine ;-) I hate bloat, so I decided it should run on any computer that still runs tcn: i can't bother with 8086 assembly, i come from the M68K world. 80386+ is just acceptable. 68K!! AHHH! core: oh yeah, the 68k is great! lar1: what with is? s/is/it/ I don't like it lar1: why not? x86 makes more sense... 64k is more coplicated lar1: umm.. sure..? ωνω _QZ [brand@p0wer.qzx.com] has joined #tunes Hi _QZ well, I know someone at college who thinks x86 is better than 68k, but nobody agrees with him on anything ;) no shit, going from an orthogonally designed cpu with a flat address space, 16 general-purpose registers, equal length instructions, to a 8-bit spinoff with 8 specialized registers, variable length instruction opcodes and tricks all around, wasn't much fun. oh yeah, and an illogical operand ordering too. one question, core: Why did you bother? tcn: when i started working at ubisoft, i had little choice.. games on amiga weren't too hot then (anymore) 08:00pm and well, the m68K world has stagnated since then I've seen some pretty cool Amiga games, even on the old ones. tcn: me too; but they are old. tcn: be sure the system will run on m68k sometime though. What I don't get is, Motorola says they're not making a successor to the 68k, but Intel is still making x86's.. core: I'll work on the 68k code with ya tcn: that's what i meant by stagnation tcn: that'd be cool, as my days are still only made of 24 hours :) What is so good about 68k? i have an a4000/040 on my left to test it. it hasn't been switched on for maybe 6 months tho, last time i listened to modules. <_QZ> lar1: it was a 32bit chip from the beginning core: The 68k is superior, and Motorola deemed it "a thing of the past".. and still Intel thinks their piece of shit x86 has a future.. lar1: registers aren't specialized; it was a 32-bit chip in 1978; it has no special I/O bus, only memory mapped devices that do substractive decoding like normal devices.. <_QZ> tcn: intels "piece of shit x86" is making money, motorola's wasnt qz: right, that's why we all work on x86 now <_QZ> and the ppc replaced the 68k core: So why didn't the XT run on a 68k chip? lar1: It can have 8 stacks at once, and it can have position-independent code x86 is not _so_ bad since the 386 came out, it's just a pity its design is ruling the computing world qz: the ppc isn't a design of motorola lar1: IBM FUBAR hehe lar1: actually, it did. My XT is an 8088... lar1: Hey, I live in IBM's hometown. I can assure you, their management sucks. <_QZ> core: motorola, ibm and apple designed the ppc qz: not really qz: ibm gave their 'power' cpu line design, motorola trimmed it, it became the ppc, apple used it. qz: the powerpc is a mass version of the ibm power cpu; now it's faster than them because more money was put into it to make it do that. <_QZ> ya, that makes it motoroal's Amiga uses the PPC now too qz: well, if i took Brix and trashed code until it did only what i wanted, whose work would you call it? umm, if you had to write a lot of code to make it do what you want, it would be yours tcn: no, just removing code <_QZ> ya Or, information is free motorola didn't add, they removed. IBM gave it to them Just like Xerox gave Apple smalltalk yeah, but the point is, it wasn't designed by the motorola geniuses who came up with the m68k it wasn't a cleanroom design as they sold it. So the 68k is nicer? I haven't looked too hard at the PPC design. tcn: it was designed in 1978, so it lacks some features of modern cpus, but from a design point of view it was a gem. of course, we aren't living in the past, so today we have x86 and we live with it. <_QZ> its those who used the 68k in machines that killed it there's also so much documentation on them that anyone can write an OS for it. 08:10pm qz: what do you mean? <_QZ> apple and amiga I'm interested in designing new CPU's.. qz: oh, yeah <_QZ> the pc world had clones <_QZ> "cheap" clones tcn: i know zilch about hardware tcn: i have internal contact at cyrix (i wrote their mediagx drivers) but that's it. And clones gave us 50000 different components, all incompatible, which we have to find/write drivers for <_QZ> tcn: hehe The IBM PC was a cob-job to begin with. If they'd had some standards in place, we could have had cheap AND compatible clones <_QZ> agreed Well, if we get something like Tunes out there, there will be lots more room for competition with Intel tcn: the point of Tunes being free, is that you aren't tied to intel anymore. for all we know, we could be running Tunes on the pinky processor someday. core: Not just free (Linux is free) but trivial to port tcn: linux is trivial to port :) tcn: but i know what you mean, and i agree :) <_QZ> is thread support just a lib on linux? hehe.. Forth is trivial to port. I wrote that interpreter in a month, rewrote it in 2 days, and that was on the x86.. imagine how quick it would be on the 68k.. qz: no, there is kernel support for creating threads too <_QZ> what kernel qz: linux 2.0? tcn: yeah; i think it was introduced in 1.31 though s/1.31/1.1/ <_QZ> i have 2.0.34 and dont remember seeing thread support qz: you haven't looked very far then <_QZ> well the word thread was not found in my .config _qz: what would it be doing ther? s/ther/there/ _qz: /usr/include/asm/unistd.h _qz: at the end. the kernel_thread() code. <_QZ> .config is where all options are set _qz: it's not an option. _qz: it's a system call. <_QZ> so why did glib say i didnt have thread support linuxthreads.tar.gz because it didn't have the necessary userspace functions to call that system call, among others and it wasn't thread-safe. that's a package that provides thread support and interface to the kernel, as per posix specs. qz: but you can call kernel_thread() even without that. 08:20pm qz: libc is not the kernel, and is not affected by .config; libc is here to wrap system calls around userspace functions (among a lot of others) <_QZ> i know that <_QZ> what is pthreads? qz: posix threads. <_QZ> same thing as linuxthreads what you're looking for is in unistd.h, it shows how to call the linux syscall creating a thread core: is it necessary to compile binutils before compiling nasm? tril: no ok, I'm going to do both at the same time. tril: it can be done in any order, but you need both to compile the system after that tril: sure :) how old is binutils 2.9.1.0.14? <_QZ> u only need nasm to build brix :) tcn: a couple months QZ: Same for my OS. we're cool :) <_QZ> well nasm and all the other programs i wrote qz: binutils spared me from writing a lot of 'other programs'. <_QZ> core: u got an fs? qz: stuff for packaging objects, etc? How do I find what version of binutils my 2.9.1 is? (that I just downloaded from cdrom.com) qz: no, not needed yet; if you want to put one, you're welcome to <_QZ> core: brix has a working fs Tril: That's what I have.. I think it's 2.9.1.0.0 qz: um, some readme i guess _qz: i've started working on it in may of 1998. qz: what does it have to do with it? <_QZ> eh? tcn,tril: oh, that'll work too (or you can get patches to bring it up to latest), just patch against binutils-2.9.0-xcom.diff, not 9.0.14 qz: i only started this in may. <_QZ> i started in july I'm installing debian 2.1 one of these days any reason to prefer the newer binutils? (I don't know where to get the patches) I'll get binutils then qz: but not of 1998, i've seen brix up for much longer. Hmmm be back later ωνω SignOff lar1: #TUNES (Leaving) <_QZ> ya but i wasnt working on it it was vaporware :) qz: there was supposed to be a working version in april 1998 <_QZ> ya still is vaporware in a sense, since none of us have seen the whole thing <_QZ> core: ya but i never did any coding qz: *shrug* well, who knows what you did. i started working on it as a whole, even writing things down, in may; we're in january, i'm quite happy of what i have.. again if you want to plug an fs on top of it, you're welcome <_QZ> core: ok send me the source I think QZ has an object fs <_QZ> :) <_QZ> tcn: ya qz: you'll have it, no worries :) persistent storage? <_QZ> i have cancer and will be dead by sunday morning brix has persistent storage <_QZ> tcn: yes qz: then you can't work on the fs anyway :) <_QZ> core: i can work fast :) qz: heheh qz: well, tell tril to hurry then :) qz: what if you're still alive monday? <_QZ> tcn: miracle 08:30pm qz: ok, he could give it to ya, but then he'd have to kill ya on monday if we get rid of the 'not invented here' syndrome and put everything together, it'd be a shame if we don't get a nicely working thing :P hahaha <_QZ> YES YeS YES YES YES <_QZ> i had compiled glib with thread support (default) and that is why apps couldnt use gtk core: you're right, we all have our own pet OS's.. :) tcn are you gonna compile it yourself or wait for me to tcn: yeah tril: Go ahead and send me the binary.. I won't get the new ld today i'll head to bed <_QZ> i have the newest binutils see ya my girlfriend has been asleep for 6 hours :) hahah <_QZ> i have the newest everything so does prep.ai.mit.edu :-) where? tril: /pub/gnu ? :) oh, hum, i think Fare found .19a on sunsite^Wmetalab anyway, be sure to email me at core@suntech.fr, if it's not from a ML i'll read it before next year ;) 'night, have fun :) ωνω SignOff core: #TUNES (this is *not* the default quit message) ωνω luxury [lux@client216-174.raleigh.intrex.net] has joined #TUNES hi, lux. hello hey what's up, tcn sup? awwwwwwwww yeahhh... ωνω luxury [lux@client216-174.raleigh.intrex.net] has left #TUNES [] abi scared lux away <_QZ> gee i sure wish they had had that oc'able 300A today. compiling all this crap would go SOOO much quicker ωνω SignOff tcn: #TUNES (Read error to tcn[cci-209150250108.clarityconnect.net]: EOF from client) qz, my friend got 46th week santa somewhere or other, a monster fan/heat sink ($35) and heat goop and his is working at 450 ok ωνω tcn [tcn@cci-209150250116.clarityconnect.net] has joined #tunes <_QZ> 450??? <_QZ> it can do 472 maybe my ISP boots people who've been on for 4 hours 472 what? is his an old one? tcn, MHz <_QZ> with the dinky cerlon fan do you know the steppings? <_QZ> my isp boots at 8hrs my isp doesnt boot <_QZ> if the chip can be oc'd then it can go all the way 08:40pm not cos they dont want to, but they cant get the cfg rite <_QZ> there are 2 300A's: the one that does only 300 and the one that does 472 <_QZ> i have heard the 300 wont do 472 in the abit board <_QZ> i have it running at 472 in an asus p2b <_QZ> s/300/300A/ <_QZ> u need to use a 103 bus and enable turbo to make the bus 105 <_QZ> turbo is not the turbo switch <_QZ> its a setting in the bios for testing the stability of the hardware at a given speed <_QZ> if it can run fine being oc'd 2% than it will run fine at its normal speed under most conditions <_QZ> but if it can run fine oc'd 2% then leave it there :) <_QZ> hcf: hehe π tcn/#tunes is reading a smalltalk tutorial.. looks kinda like OO Forth :) tril.. I'm not sure I like the 'message passing' concept (I'll have to read more though). I should talk to BRice about this "objects as functions" concept.. you still at school? <_QZ> message passing in brix is the best tcn: fare has convinced me msg passing is bad, too. <_QZ> cuz the system creates threads for each message I was originally using it in my OS model. but tunes is better <_QZ> so the system is narturally multithreading without the programmer needed to know about threads I am at the university computer lab, yes.. I should leave in 15 min I'll let you know when it's midnight here <_QZ> Tril: dont bother, u missed the bus. trust me i know ;) yer walkin' :) <_QZ> Tril: u smoke? qz, nah <_QZ> Tril: u got a warm jacket? 08:50pm yup that I do... there are 2 more busses to my house from here yet tonight. anyway, what I'm after is a system that works for interprocess communication AND the "flow of data" within a program how about a system where there isn't a difference between the two? i.e. eliminate the concept of isolated programs altogether ωνω Tril_ [dem@hh154ws03.hh154lab.wwu.edu] has joined #TUNES Yes, in a parallel system, that should happen.. >>> Tril [dem@hh154ws03.hh154lab.wwu.edu] requested PING 380031115 from #TUNES (you didn't miss anything) was I split? >>> Tril [dem@hh154ws03.hh154lab.wwu.edu] requested PING 380102698 from #TUNES nevermind ;) ok I am going to leave early to make sure I find the way out of this building. take it easy I should be able to test clementine as soon as I get home. (and brix) ωνω SignOff Tril: #TUNES (Read error to Tril[hh154ws03.hh154lab.wwu.edu]: EOF from client) ωνω SignOff Tril_: #TUNES (Read error to Tril_[hh154ws03.hh154lab.wwu.edu]: EOF from client) <_QZ> dcc me the binary hehe <_QZ> well <_QZ> we will see how many of u get brix I don't have the binary <_QZ> u have the source tho 09:00pm yeah, it's prolly better than Brix too :P <_QZ> no jk <_QZ> brix has gfx, floppy, fs, and brix is all NASM code <_QZ> speed baby are the helper programs in ASM, C, or what? <_QZ> everything is in asm cool <_QZ> once i get the language working then i will write everything in that what kinda language? <_QZ> but the language will be close to asm <_QZ> so it will compile to near asm sounds like Forth <_QZ> well gee asm and forth are my 2 favorite languages <_QZ> so u figure out what it will be like You haven't written your forth interpreter yet? <_QZ> and brix is word based like forth I'll trade ya mine for Brix :) <_QZ> yers wor not work <_QZ> s/wor/would/ how come? <_QZ> have u read the brx page? hmm.. maybe 2 weeks ago.. anything new? <_QZ> no <_QZ> brix uses actions, which are like forth words so that's what actions are :) <_QZ> but an action works on an object You have an object stack? <_QZ> an action is the message part of the message u send to objects <_QZ> object.action() <_QZ> the entire system works by sending messages <_QZ> and the system creates threads to handle each message a thread for every message? that doesn't sound fast.. <_QZ> brix has synchronous and asynchronous messages <_QZ> its called pervasive multithreading <_QZ> it makes the system extremely responsive and fast <_QZ> and when u have smp it screams Do you have a kernel per se? Or is it just an exception handler? <_QZ> kernel abi, brix? rumour has it brix is a free portable protected preemptive pervasive-multithreading multiuser object-oriented high-speed distributed operating system located at http://www.qzx.com/brix <_QZ> checking for perl... /usr/bin/perl <_QZ> checking for mass_quantities_of_bass_ale in -lFridge... no <_QZ> checking for mass_quantities_of_any_ale in -lFridge... no <_QZ> Warning: No ales were found in your refridgerator. <_QZ> We highly suggest that you rectify this situation immediately. distributed? how far along is that part of it? <_QZ> hehehe haha <_QZ> atleast rasterman hasa sense of humor <_QZ> thats in teh configure for E yeah, even if he can't code worth a shit.. :) abi, rasterman? tcn: i haven't a clue abi, E? E is probably Enlightenment Window Manager @ http://www.rasterman.com 09:10pm <_QZ> uhh what's abi? hmmm... i am the little girl borg with a big brain <_QZ> E is at www.enlightenment.org what's abi? i am the little girl borg with a big brain or a little ho abi is also #tunes personal little ho okay, hcf. what's abi? i am the little girl borg with a big brain or #tunes' personal little whore hcf comes on here all of a sudden, just to dis his bot :) abi: abi is also an infoslut okay, hcf. haha QZ: is Brix capable of distributed computation yet? or what did you have in mind for that? <_QZ> it has no networking yet in my project timetable from September I said I'd have tcp/ip done by now :) college intervened. You have like 7 computers on a LAN, right? 09:20pm <_QZ> soemthing like tht <_QZ> 8 Must be nice get Brix running on all them at once.. <_QZ> well i plan on having tcp/ip and ethernet suppor before any modem support <_QZ> i might leave modem support upto someone else yeah, not too hard.. serial + PPP <_QZ> cuz i will always have my server running linux There's a RFC for PPP <_QZ> i know howto do ppp just takes time well, I'm taking off.. <_QZ> cya _QZ: what do u hav left to do on brix? <_QZ> for? <_QZ> for me brix will be finished when it can do everything i do in linux on this box <_QZ> mp3 player, cd player, browser, etc.. i ment b4 a release 09:30pm <_QZ> a release will happen the day i feel like letting everyone see it <_QZ> cept for tril,fare,core and tcn <_QZ> :) and the #asm crowd <_QZ> heh <_QZ> *** Warning: This library needs some functionality provided by -lX11. <_QZ> *** I have the capability to make that library automatically link in when <_QZ> *** you link to this library. But I can only do this if you have a <_QZ> *** shared version of the library, which you do not appear to have. <_QZ> <_QZ> u know what that means? npe <_QZ> u know how big bashes buffer is heh, thers a perl debug module called, D'oh <_QZ> like how much command line crap i can type while its compiling? <_QZ> hehe ωνω SignOff tcn: #TUNES (Leaving) Richard's bio-munition: Yes, folks... Richard White has a tatoo of the Perl program which implements RSA encryption and decryption on his upper arm. <_QZ> HHHAHAHAHAHA <_QZ> creating redmond95/Makefile <_QZ> creating redmond95/Theme/Makefile <_QZ> creating redmond95/Theme/gtk/Makefile <_QZ> im gonna get a tatoo of the brix kernel source on my arm 09:40pm ωνω SignOff hcf: #TUNES (Leaving) ωνω Tril [dem@ppp435.whatcom.pacificrim.net] has joined #tunes _QZ you here? I tried brix on my mom's computer and it got farther I also tried a different floppy drive on my computer and it did not work Same disk 10:20pm <_QZ> bummer <_QZ> as u can see it does work <_QZ> when the floppy drive does I tried 2 working floppy drives with brix. It's not the drive. <_QZ> all my drives work and everyone that i have given that too has gotten it to work my mom has an Unknown 486 with FPU [0053] and after that it hangs with the floppy light on for some reason it turns the text red <_QZ> the text should be red <_QZ> red on black Well I don't have color on my computer so I wouldn't be able to tell now I am overwriting my brix disk with clementine ωνω SignOff Tril: #TUNES (Tril has no reason) <_QZ> ok dcc that here 10:30pm ωνω Tril [dem@xws128.xtrn03.wwu.edu] has joined #tunes Clementine didn't work either I don't know what's going on. I can boot Win95 CLI off a floppy just fine. or Linux abi, brix? rumour has it brix is a free portable protected preemptive pervasive-multithreading multiuser object-oriented high-speed distributed operating system located at http://www.qzx.com/brix 11:00pm abi, date? well, date is Fri Jan 8 23:05:56 1999 abi, dolphin? tril: i haven't a clue 11:10pm ωνω Tril has changed the topic on channel #TUNES to: http://www2.tunes.org/ - tunes project ping time: 666.666 ms <_QZ> hahah <_QZ> told ya brix was FINE 11:20pm hackernews.com- thank God lou confessed the declaration of war was a mistake. gotta go...l8r ωνω SignOff Tril: #TUNES (Tril has no reason) 11:40pm ωνω SignOff _QZ: #TUNES (BRiX [http://www.qzx.com/brix] :: sleep) !lackey.openprojects.net!! Remote CONNECT king.openprojects.net 8005 from ^lilo [msg(TUNES)] newlog 1999.0109 IRC log ended Sat Jan 9 00:00:01 1999