IRC log started Sun Feb 6 00:00:01 2000 [msg(TUNES)] permlog 2000.0206 -:- SignOff hcf: #TUNES (Leaving) -:- hcf [nef@me-portland-us1003.javanet.com] has joined #tunes -:- SignOff rares: #TUNES (KVirc 1.0.0 Millennium BETA 3) -:- SignOff lar1: #TUNES (Ping timeout for lar1[dialup-209.245.133.137.SanJose1.Level3.net]) -:- lar1 [larman@1Cust57.tnt31.sfo3.da.uu.net] has joined #tunes -:- SignOff washort: #TUNES (Ping timeout for washort[d103.narrowgate.net]) -:- SignOff hcf: #TUNES (Leaving) -:- SignOff Downix: #TUNES (Downix has no reason) -:- SignOff lar1: #TUNES (Domo eregato, Mr. Roboto!) whats the best way to save save what? i have a linked list of users kinda like the client list and im trying to save it to a file it stores the correct number of users, and each user structure but it only saves the first letter of each username and password i just have it save/load using fputc/fgetc uh use fprintf () or fputs () i gotta go to sleep cya -:- SignOff eihrul: #TUNES ([x]chat) 03:00am -:- SignOff air: #TUNES (http://www.qzx.com/ :: sleep) -:- SignOff Fare: #TUNES (:Connection reset by pear) -:- Fare [fare@quatramaran.ens.fr] has joined #Tunes -:- smoke [smoke@16dyn128.delft.casema.net] has joined #tunes -:- Fufie [stig@tunnel-44-25.vpn.uib.no] has joined #tunes -:- SignOff smoke: #TUNES (Ping timeout for smoke[16dyn128.delft.casema.net]) -:- smoke [smoke@16dyn128.delft.casema.net] has joined #tunes -:- bineng [Anders@j141.ryd.student.liu.se] has joined #tunes -:- Downix [down@d-gnaps-164.ici.net] has joined #tunes -:- SignOff smoke: #TUNES (One day sheep will rule the world) -:- smoke [smoke@16dyn128.delft.casema.net] has joined #tunes -:- smoke_ [smoke@16dyn128.delft.casema.net] has joined #tunes -:- SignOff smoke_: #TUNES (Read error to smoke_[16dyn128.delft.casema.net]: EOF from client) -:- SignOff smoke: #TUNES (One day sheep will rule the world) -:- smoke [smoke@16dyn128.delft.casema.net] has joined #tunes Smoke! * Downix/#tunes grabs sonme glue, stay put! re all :) downix: did you make up a new name for the company already? Nope * Downix/#tunes is still grumbling 08:00am -:- SignOff Fufie: #TUNES (Ping timeout for Fufie[tunnel-44-25.vpn.uib.no]) do you have an idea what I'm doing now smoke? 08:10am -:- SignOff Downix: #TUNES (Ping timeout for Downix[d-gnaps-164.ici.net]) -:- Downix [down@d-gnaps-164.ici.net] has joined #tunes smoke, know what we're doing now? downix: generally yes downix: i thought of the name `New!' New! ? downix: but it's probably already taken ? probably New! yes. or just New I'll almost guarantee it's taken, but let's see is `See' taken also? :) yes it is 08:20am where do you check, btw? federal trademark commission database url? always best to start there, then look into network solutions oh sure, I close netscape THEN you ask, hold on trademarks.uspto.gov :))) wow. all my ideas are taken :) hehe it's ok, so are all of mine seems like you'd have to settle for some weird word like `baniemoe' (for THAT's not taken yet :) what the heck is a bainemoe? A 3 Stooge with bad breath? hm. baniemoe is the fictive title we once made up for a fake freshmeat.net entry, to see if it got accepted there are SO many weird titles given to useless programs, we thought we could sneak it in. hehe don't think it'll work We have an adertising campaign even worked out, but we need a name for the company the product is Eddas, but no company name for it 08:30am how did you come up with Eddas? Eddas is a collection of norse myths. The word means, litterally, "All that is, was and will be" -:- SignOff smkl: #TUNES (My damn controlling terminal disappeared!) so we made an anacronym for it, "Eddas Doesn't Do Anything Stupidly" oh dear, no :) hm. well probably one of the first corporate recursive anacronyms the first thing that comes to mind (and checked that it didn't exist) is `Sagas' 8) Sagas? reflective word also :) sagS sagaS even what is it? Sagas BTW is open a scandinavian legend .. A Scandinavian legend, or heroic or mythic tradition, among the Norsemen and kindred people; a northern European popular historical or religious tale of olden time. * Downix/#tunes nods it has little to do with your business plan.. the only connection i could think of is with long-term research, instead of hyped fast work let's see if the domain name is free but i don't know if that's a -key- feature of your company -:- smkl [sami@glubimox.yok.utu.fi] has joined #tunes it sounds nice :) it works tho hehe domain name is not availible aaaaah that's too bad let's see if it is owned by a cyber squatter there's no www.sagas.com yes there is, registered let's see who owns it unnamed person in Finland 08:40am ah too bad know anyone that speaks finnish? smkl? but i don't know him :) well, it looks like a cybersquatter so I'll ask 08:50am -:- SignOff Downix: #TUNES (Read error to Downix[d-gnaps-164.ici.net]: EOF from client) -:- ult [noone@user-38lcn5t.dialup.mindspring.com] has joined #Tunes -:- hcf [nef@me-portland-us214.javanet.com] has joined #tunes -:- Downix [down@d-gnaps-265.ici.net] has joined #tunes -:- SignOff smoke: #TUNES (bug in alsa) -:- smoke [smoke@16dyn128.delft.casema.net] has joined #tunes -:- Fufie [stig@tunnel-44-23.vpn.uib.no] has joined #tunes hi fufie hi sup? hullo all -:- SignOff thomas_: #TUNES (Ping timeout for thomas_[193.217.63.152]) 10:40am -:- eihrul [lee@usr5-ppp102.lvdi.net] has joined #tunes -:- thomas_ [thomas@193.217.63.152] has joined #tunes -:- washort [washort@d123.narrowgate.net] has joined #tunes -:- SignOff ult: #TUNES (Read error to ult[user-38lcn5t.dialup.mindspring.com]: Connection reset by peer) -:- lar1 [larman@dialup-209.244.106.35.SanJose1.Level3.net] has joined #tunes -:- water [water@tnt-9-138.tscnet.net] has joined #tunes hey! (full house) Hey, water!! last night was a well-timed good time :) hoy water hey hcf hi 11:40am eih: you here too? yes oh good there's a lot of things i want to work on, but i'm not sure where's appropriate to start (for slate) water: what r the current things to work on? just to start, i'll upload some minor changes to the semantics page hcf: heh. lots for example, i need a definite way of handling data itself sure, i've stated that message-passing wraps around access/update in self, primitives (bytecodes) activate code that handles this ok, i just uploaded the semantics page that i'm currently working with also, it seems (interestingly) that data objects are automatically activation records any thoughts so far? well, they're not... if you wish them to inherit any functionality then you can't overload the parent slot with this dual purpose unless you ensure that the lowest level activation record inherits functionality from an activation record prototype (if its even needed :>) 11:50am hm actually, reading through the self manual ok, so they'd be activation records that couldn't be nested they use a separate link to establish inheritance of scope yes from the parent* link self objects can have multipple parents parents just happen to be shared part-objects in self icuc, http://javalab.informatik.uni-bonn.de/research/darwin/delegation_eng.html ooh eih: so do you think the complications of multiple parents are worth the generality or no? well, it's not a question of multiple parents necessarily... i was planning on having some mechanism allow the same to be done by having each part-object inherit the lookup behavior is the same, but it's used to a different end k so it's not necessarily involving multiple inheritance ok, s/inheritance/delegation water: How about a 'links' page where you can collect lots of links to related stuff? It seems to me that you buid alot on others work, which one then needs to know about. shrug, call it what you will... but it my conceptual model of it is the same :) s/it my/my bin: yeah, i should get to that. i'll start putting it together great 12:00pm * bineng/#tunes is away -:- SignOff lar1: #TUNES (Ping timeout for lar1[dialup-209.244.106.35.SanJose1.Level3.net]) hm delegation vs consultation 'consultation'? check out the link hcf pointed out if i want to program in anything slate-like (but available for practical testing), what language(s) would i best look at? hm self :) or should i just sit back and wait :) it runs on macs, if that helps water: just seems a minor method mop change... to add 'consultation' eih: yes :) 12:10pm if anyone actually wanted it at all :) but we haven't specified Top very well yet atleast, i don't see how its more useful than delegation for atleast 'inheritance' well, it could model a problem better than delegation could in certain settings i.e. better simulation and there's nothing preventing one from adding it if they want it or us providing some mops implementing it :) yep but machanisms like that should be part of the stdlib -:- air [brand@p0wer.qzx.com] has joined #tunes specifying the stdlib might be harder than the language itself, however yeah, because we'll be programming in a language model that hasn't really existed before this is definitely the *last* language i ever work on :) ps.zip... that's original i've seen .ps.bz2 in a few places yes, but ps.zip is violating of something.... 12:20pm s/violating/a violation ahem slate :) eh, i'm waiting for a paper to download ok... this is a first the paper is UPSIDE DOWN lol 12:30pm water: "root" and "top" r the same thing right? nope oh top happens to be the language's primitive behavioral meta-object hm. i'll list all the stuff i know so far on the semantics page -:- ult [noone@user-38lcn5c.dialup.mindspring.com] has joined #Tunes 12:40pm water: hmm, consultation looks like it could be useful for providing different interfaces to the same object... or "roles" as the paper terms them right which is what part-objects will be used for 12:50pm hm eihrul: how do i read a string with spaces from a file? well, firstly, you need a delimter... commonly '\n' in which case fgets () will suffice so fscanf wont work? not really crap 01:10pm the code is so much cleaner and smaller with fscanf uhh fgets doesnt say anything about a delimiter yes it does read ckiseky ah it stops at newline er closely off-by-one-key error i have multiple strings on one line each starting with a = one-key eh? :) so worst you can do is fgetc () until you hit a '=' i will just have to settle for username/passwords that dont contain spaces then cuz im not going to the trouble of doing that heh sounds like someone's using a lame language :) uh, it's not hard... i know its not hard but i just dont wanna do it right now dude you already have code to do it :P bfe, string_read () ya i know damn, where'd core disappear to? car accident 01:20pm uh, seriously? uh, no :) well, it's plausable... it's been a month email him i have! hmm 22 days ago did u email him multiple times? -:- SignOff hcf: #TUNES (Ping timeout for hcf[me-portland-us214.javanet.com]) nothing says "respond" like 1000 identical emails in yer inbox :) nothing says 'asshole' like a 1000-email mail bomb hmm heh it says "justifiable homicide" to me it might be a mail bomb to us but remember core has an OC100000 connection sure, with 30 day lag times... maybe he missed yer email if u accidentally click too many times while deleting an email u just read u might nuke an unread email 01:30pm if i do a, #define name(x), how do i use the x in the define? you just use it... air: you can use ##x to append it to variables, or just `x' to substitute or #x ro stringify thanks 01:40pm eih: can you think of any reason why behavioral meta-objects should not all descend from Top? * eihrul/#tunes ponders. descend along the meta-object hierarchy or inheritance hierarchy? inheritance hm i don't know if inheriting from top is useful that (allowing any bmo) could allow for some really weird (and/or cool) stuff unless it provides any functionality for implementing mo's above and beyond lookup-apply it does, though but mostly i see mo's being meta'd by Top, not parent'd i mean, any time you clone and modify Top, the new bmo inherits from Top that's how you get other-than-default meta-behavior i suppose Top's 'meta' should be immutable yeah it'd be disastrous otherwise :) hm lest ye reinstate unending meta-regression again.... actually it wouldn't because a clone of Top would just refer to Top for it's un-implemented protocol otoh, if you modified those messages (like lookup-apply) then it would ok you're right 01:50pm say good bye to "default object format" under semantics i never said hello heh if color1 is a structure of ints cant i do this: tmp->color1={1,2,3}; 02:00pm no... { ... } is an initializer list and hence, can only be used to initialize an array or structure during its declaration... hm... basic language policy: "really important stuff" should always be immutable another thought: where should the mutability / reference-handling protocol go? maybe it should go in Root, but otoh a separate Reference or Slot object would be nice 02:10pm eih: any thoughts? well, mking slot objects could potentially complicate things... they'd just be reifiers hm which means that Root definitely has to handle mutability well, you can provide different versions of Root :) hm note: i just realized that applyTo_withArgs_withCont could/should be drastically simplified well, continuations complicate things a bit... and the withArgs part as well 02:20pm the args part is just binding the slots of the method('s clone) * eihrul/#tunes ponders a way of getting rid of irix on his indigo2. how the hell'd you get an indigo? used it's been a while since i had a used workstation i got it for playing around with language development "playing around with language development"? compilers :) for whom? which lisp do you run on the indigo? (irix was nice as Un*ces go) i'm still trying to get rid of irix :) i just got it irix is ok if you got the native compiler as well i don't, unfortunately IRIX is never ok! -:- SignOff ult: #TUNES (Leaving) you have no compiler? and thus begins my problem * eihrul/#tunes nods. augh and the compiler costs $2000? expensive thing yes... so i'm thinking sgi-linux... I think sgi ships freeware cds with gcc at leasts, but I can't remember if you need the development system to get the necessary header files because none of the *bsds seem to support it * Fufie/#tunes liked irix but he had all cds he wanted when he used irix yes a pile of support software for an incomplete product definitely masks the usual deficiencies but with the support software I prefer irix over other commercial Unix-systems i don't doubt it slowaris is clunky to develop on digital unix was a pain old sun4 was nice to develop on.. most stuff worked on sun4 hp sux 02:30pm :) anyway i believe we were discussing mutability of references and coming up with a good protocol for handling this issue -:- SignOff bineng: #TUNES (bye ) you guys heard of the Vapour project? http://ftp.rook.com.au/ yep -:- SignOff smoke: #TUNES (z) water: going to go buy a network card so i can get rid of irix, bbl ok 02:40pm fufie: do macros apply well to clos? umm.. what do you mean? are they useful? yes within clos? most of CLOS is done with macros. :) remember that clos is written with cl macros, functions and variables well sure hm even within clos yes.. I use macros heavily to define various methods and generate classes ok s/clos is written/clos can be written/ there are a few native implementations of clos 02:50pm AMOP which you should read if you plan on doing _anything_ related to meta-objects and meta-object protocols is basically an implementation of CLOS AMOP rocks. =) yes even though i didn't understand half of it lol the half that i *DID* understand was really cool though :) well i'm pretty sure i've got most of the ideas i liked out of amop lisp is a language one grows into.. it is very seldom one gets or understands a heavy lisp book on first attempt (I think I needed two attempts on AMOP and I might need a third :-) "If you give a programmer Fortran, he's got Fortran. If you give a programmer Lisp, he has any number of languages." water: umm.. to get amop you must also understand the implementation.. that's a very important half.. i know unfortunately, the last time i read amop was in college when i didn't understand very much and it wasn't my copy your local library should have it yeah, right seriously, they do here at least right next to a major cs lab, i'm sure yes exactly actually, i could run over to uwash any university library should have a copy *should* :) yes.. normative sentences are good :) normative sentences should not be uttered. heh would be pretty quiet then :) declarative sentences are false how goes it, fare? 03:00pm badly I'm such a lazy bum heh sorry have you had the presentation or is it this week? haven't done anything worthwhile in years did it how did it fare? I have slides in french, if you like my french is poor unfortunately, I know just enough french to understand english literature :) well, the listeners listened, and were interested; the leaders were not convinced. and the hackers were not moved but they listened with respect, at least what presentation was this? the CAML guys don't respect, on the other hand; but that's as usual a presentation on a reflective execution framework how about Ian and the vm-guys? aka the TUNES LLL oh Ian was vaguely interested, but he what he wants is practical stuff. all in all, I think they all felt there was worth in it, but didn't relate exactly to what they were doing so, any ideas on how to proceed now? the leader, Bertil Folliot, would still wonder why being reflective was better than just being meta Ian? Ian Piumarta? yes nifty. =) Ian explained to him that being fully reflective was easy if you could be meta but that still wouldn't do. hm what does Bertil lead? a lab at lip6, which leads the VVM work see around www.lip6.fr see also around www-sor.inria.fr ok 03:10pm -:- ree [js@twisted.goodnet.com] has joined #tunes they were made aware of the issues of reflection awareness is the first step umm.. how long has the tunes project been around? 95? under other names, 13 years four score, .... decades ago under the TUNES name, 5 years but no implementation yet? ree: four score? * Fare/#Tunes hides what came before moose (if any)? heh water: there was OSL osl? my project of an "object&stack language", inspired by the HP28S Fufie: the official strategy is to hack from the bottom up, as Ian does OSL was not reflective, except at the very end at that time, I was trying to understand the concepts behind "OO" until I found that there were none