IRC log started Sat Jul 10 00:00:00 1999
[msg(TUNES)] permlog 1999.0710
12:10am
-:- SignOff _QZ: #TUNES (BRiX [http://www.qzx.com/brix] :: sleep)
-:- hcf [nef@me-portland-us635.javanet.com] has joined #tunes
-:- abi [nef@bespin.dhs.org] has joined #tunes
-:- SignOff hcf: #TUNES (Leaving)
* Fare/#Tunes is back
-:- binEng [Anders@195.84.234.31] has joined #tunes
<Fare> as long as you're not having the argument with your mom...
08:00am
-:- beholder [chaldea@ppp-055.m4-1.osh.ican.net] has joined #tunes
-:- water [water@ppp-tnt-128.tscnet.net] has joined #tunes
<water> hi all
* Tril/#TUNES is back from the dead. Gone 36 hrs 37 min 29 secs
<Tril> hello
-:- SignOff Fare: #TUNES (Ping timeout for Fare[quatramaran.ens.fr])
<beholder> Hello Tril
<water> hey tril
-:- Tril has changed the topic on channel #Tunes to: "...a lot of interesting vaporware" -basile
<Tril> yo
<binEng> hi tril
<Tril> I wrote a section of a new spec. wanna read it?
<water> :P
<binEng> ¨sure
<water> and :P about Fare being the biggest Tunes contributor
<water> sure
<Tril> http://www.tunes.org/~dem/tunes/spec/spec.html
<water> thanks
<water> damn i'm lagged
<beholder> /ctcp water ping
<beholder> dam... I can't enter commands in this damn client
-:- _water [water@ppp-tnt-193.tscnet.net] has joined #tunes
<_water> b
09:40am
<_water> ha! no more lag
<Tril> you solved the lag problem forever? how?
<Tril> for everyone?
-:- SignOff water: #TUNES (Ping timeout for water[ppp-tnt-128.tscnet.net])
<_water> hehe
<beholder> :)
-:- _water is now known as water
<water> can a Tril-set contain itself?
-:- is [ejb@va.debian.org] has joined #tunes
<Tril> i'll skip the complicated answer, and just say yes for now :)
<water> ok
<water> hehe
<Tril> hi, is
<is> hey
<Tril> Debian rocks :) What is your part in it?
<water> what's the point about the membership function being allowed to be undecidable.
<is> I maintain a couple of packages ..
09:50am
<Tril> water: maximum expressibility
<water> Tril: how do you express an undecidable function in your system?
<Tril> water: Um
<water> hehe
<Tril> do lambdas allow it?
<water> hehe no
<water> by definition, no
<Tril> by definition? Hmm
<Tril> then printing out the lambda expression is decidable, too
<water> yes
<Tril> i thought lambdas were about recursion
-:- hcf [nef@me-portland-us923.javanet.com] has joined #tunes
<Tril> and recursion can easily lead to undecidability
<water> remember church's paper?
<Tril> and there is no way to verify whether a function is undecidable or not
<water> you mean non-deterministic recursion
<Tril> i probably didn't read it :(
<Tril> ok, i admit i didnt read it
<Tril> i'm not really into lambdas
<water> the only way a lambda expression could be undecidable is by being more than infinite
<Tril> anyway, all i know is i can write undecidable functions in scheme very easily so i'm not sure your question is that hard, just allowing a function to call itself seems enough.
10:00am
<water> i.e. non-returnable
<water> yes, it's called a bug :)
<Tril> it's also called reflection so tunes must be full of bugs
<water> yes, i've already considered that
<water> the question is how to avoid them
<Tril> i thought you were being sarcastic. I was.
<water> hehe. oh
-:- beholder [chaldea@ppp-055.m4-1.osh.ican.net] has left #tunes [bye]
<water> well, you have to always avoid _actually_ calling a function that is undecidable for the argument you pass it
<Tril> why
<water> umm
<Tril> are you assuiming I can only evaluate one function at a time? 
<water> you like hanging your computer?
<Tril> It won't hang, i can just do something else at the same time
<Tril> and stop it if i want
<water> oh
<water> interruptable
<water> which suggests a kernel ;)
<Tril> undecidable functions are useful
10:10am
<Tril> i didnt reply to that thread about a kernel, but what i always say is in tunes, you can look at it as having no kernel, or as the entire systme being the kernel
<water> well, _possibly_ undecidable outcomes are useful, but trying to evaluate a paradox isn't
<Tril> probably we have different definitions of undecidable?
<water> maybe you mean non-deterministic
<water> undecidability is usually a term reserved for predicates
<water> abi forget undecidability
<abi> water: I forgot undecidability
<water> abi forget undecidable functions
<abi> water: I forgot undecidable functions
<water> abi forget the question
<abi> water: I forgot question
<Tril> "Any problem that can be stated as a question with a yes or no answer is called a _decision problem_. ...A decision problem is called _decidable_ if there is an algorithm that halts with the answer to the problem. Otherwise, the problem is called _undecidable_.  A decision problem is called _partially decidable_ if there is an algorithm that halts with the answer yes if the problem has a yes answer but may not halt if the problem has a no answer.
<Tril> Ok, i meant partially decidable
<Tril> there is an end quote on that line BTW :)
<water> "may not halt"?
<Tril> Later on... "Although the validity problem for the first-order predicate calculus is undecidable, it is partially decidable."
<water> right. ok
<water> no wait
<Tril> implying that "undecidable" means that IT IS NOT GUARANTEED TO TERMINATE.  That doesn't mean it MIGHT NOT terminate.
<water> that's not quite right
<water> if it doesn't terminate, then the answer (whatever you assume to be the answer) means nothing
<Tril> it's interesting how they refer to the problem as un/decidable, not the algorithm!
<water> the answer would not of necessity be "no"
<water> right
<water> i usually call membership a predicate
<water> but that's me :)
-:- is is now known as isaway
<Tril> i just consider predicate a particular case of function
<Tril> a function that returns boolean
<water> i consider it the opposite :)
<water> a predicate that is left-deterministic
-:- Fare [rideau@quatramaran.ens.fr] has joined #Tunes
<water> wb
<Tril> no, functions can return things other than boolean. predicates always return booelan
<Fare> Gakuk!
<water> yablutsk!
<Fare> Tril: who could serve as secondary DNS for scwm.* ?
-:- isaway is now known as is
<Tril> Fare: is it open source?
<water> tril: that's not what i said
<Tril> water: you didn't say anything you said "opposite" maybe clarify?
<Fare> Tril: sure, it's the Scheme and Constraints Window Manager, built atop GUILE
<Fare> Tril: I wouldn't ask you if it weren't!
<water> tril: functions are left-deterministic predicates
<Fare> water: what's "left-deterministic" ???
<water> ugghh
<Tril> abi: left-deterministic is injective
<water> no!
<water> abi forget left-deterministic
<abi> water: I forgot left-deterministic
<Fare> abi: forget water!
<abi> Fare: I forgot water
<Tril> ok, water, i guess left-detrerministic is confusing me, lets fix that before worrying about the predicate part
10:20am
<Tril> Fare: oh, I think it would be good to set up a network of DNS servers who will provide free service for this kind of project
<water> P(x,y) and P(x,z) => y=z
<Tril> implicit "sub-P" on that = sign
<Fare> Tril: I guess he first assimilates partial non-deterministic functions to relations, relations to predicates, and takes "functions" as the degenerate deterministic case
<Tril> "equality under P"
<Fare> Tril: I explained that in my paper rndlc
<water> well, i'm talking about first-order logic
<water> _mathematical_ logic
<is> Fare: I can hsot dns ..
<Fare> water: logic has been mathematical ever since George Boole (praise him!).
<Fare> is: good! Send mail to Maciej Stachowiak <mstachow@alum.mit.edu>
<water> shut up, Fare
* Fare/#Tunes shuts water down
<is> heh .. find me on irc in the next couple of days ..
<is> i won't remember
<Fare> Tril: we should replace water's program. It's even buggier than abi!
<hcf> Fare: abi is buggy? ;)
<Fare> is: just send a mail now using paste&copy
* water/#tunes regrets working with programmers. they're such control freaks and absolutists
* Tril/#TUNES likes working with water
<Fare> LOAD WATER\n0 END\nSAVE WATER
<Tril> we have too many different terms. why don't we make some rules about what to call things in duscssion?
<Fare> Tril: water hates to call things the same as other people ;->
* Fare/#Tunes is in teasing mood today
<Tril> Fare calls non-deterministic functions functions, and i'm trying rto figure out what water is calling functions
<Fare> deterministic functions; perhaps even total deterministic functions
<Fare> Tril: most people only call functions the deterministic ones
-:- water has changed the topic on channel #Tunes to: Fare's private totalitarian kingdom
<Fare> Tril: and most mathematicians even only consider mappings (total functions)
<Tril> yes... I call functions deterministic (one association per key) and total (every elemnt in domain is a key) 
<Fare> my kingdom should be dry as a desert! Let's remove the water!
* water/#tunes wonders what keys have to do with functions
<Tril> watch out i'm experiecing modem lag. i think my modem is trying to use a faster speed than the line can handle
<Fare> Tril: if you're total, "mappings" is a more precise word in mathematical context, although lots of people also only say "function" and assume totality
<Tril> i'm going to have to make a table
<Fare> Tril: force the modem to be slower by downsizing the modem<->computer speed
<Fare> water: asociation tables are conceptually but functions with a finite codomain
<water> Tril: oak or maple?
<Tril> water: html
<water> Fare: don't care
<Fare> maple sucks, and it's proprietary
<Fare> oak was renamed java, and sucks, too
* water/#tunes hasn't used oak
10:30am
<Fare> oh, there was an famous "oak lisp" of old. Maybe still available somewhere by ftp
<water> Fare: could you campare Maple and Mathematica?
<water> oh that
<Fare> water: mathematica is more expensive
<water> hehe
<Fare> maple isn't designed
<water> Fare: name a lisp environment that is better
<Fare> mathematica is designed, albeit with few flaws
* Tril/#TUNES tries to read the FOL syntax for water's meaning of function
<Fare> water: EMACS
<Fare> a/with few/with quite a few/
<water> Fare: well, emacs doesn't handle mathematical symbols
* Fare/#Tunes tries to convince LispWork people to go free software
* Fare/#Tunes had no luck with ILOG Talk
<water> Fare: does any actually have full Latex typesetting?
<Fare> water: whaddyamean?
<Tril> water and fare: left-deterministic means a total deterministic function right?
<Tril> err, left
<Fare> water: want curl?
<water> Fare: i mean wysiwyg mathematical typesetting and interaction
<Fare> water: want curl? get curl!
<water> Fare: also real-time
<water> Fare: what's curl?
<abi> curl is at http://curl.lcs.mit.edu/curl/wwwpaper.html
<water> ahh
<Fare> water: tex is not real-time, nor is it wysiwyg!
<water> Fare: looked at Lyx?
<Fare> LyX is not TeX
<water> Fare: what's the diff?
<Fare> LyX is LyX
<abi> it has been said that the diff is what you get when you run 'diff' on the files, not?
<Tril> water and fare: left-deterministic means a total deterministic relation in the left arguemnt??
<water> abi forget diff
<abi> water: I forgot diff
<water> sure
<Fare> LyX?
<abi> LyX is not TeX
<Tril> ok then what's a predicate?
<water> hehe
<Fare> abi: you stupid gaklosmontic little whore!
<Tril> abi: gaklosmontic?
<abi> gaklosmontic is probably the opposite of flurivotuginuous
<water> Tril: a function that returns a boolean
<Tril> no!!
<Fare> a predicate is a kind of a function that returns a boolean
<Tril> water: now you're just saying that to end the conversation. I said that earlier, but you disagreed with me!
<Fare> abi: and the spelling is flurivostuginuous, you stupid!
<water> i didn't disagree, i said that i think of it differently
<Tril> ok.. how do you think of it
<water> Tril: the definitions are definitely circular and arbitrary
-:- s_r [sr@phila-dialup567.nni.com] has joined #tunes
<Tril> I don't see how you can define a function in term of a predicate
<water> sighh...
<Fare> in FOL, the "boolean" returned by a predicate is not within the logic, but outside of it
<water> the predicate holds if the function takes the first argument and returns the second
<Tril> i dont mind circular definitions, just dont see one here
<water> s/if/iff
-:- is [ejb@va.debian.org] has left #tunes []
<water> hi sr
<Fare> a predicate is thus a logical construct that you can apply to first-order objects, and that you can externally observe to be true or not
<Tril> water you can't turn that into "application" without pattern matching, though.
<Fare> in intuitionnistic logic, a predicate may be neither true nor false
<s_r> why was someone from debian.org here?
<Fare> s_r: why not?
<s_r> heh
<Tril> (or quotient, same thing as pattern matching :)
<water> tril: well, pattern matching can be done without an explicit algorithm
<water> tril: computers don't have to match patterns to add numbers
<water> sr: yes
<water> sr: oh
<s_r> tril how's work on Zelda?
<Tril> s_r, that's done for now
10:40am
<Tril> s_r: i've given my friend a monopoly on the information so he can make the first editor
<Tril> then i'll release the info
<water> bleh. curl sucks
<Fare> Tril: monopoly!!!!
* Fare/#Tunes says "down with monopolies!"
* water/#tunes thinks Fare bases language preference on how short it takes for a programmer to develop cps by programming in that language.
<water> s/cps/carpal tunnel syndrome
<Tril> water: that's what i thought you  meant by functions in tersm of predicates, i just wanted to check.  I'm familiar with that circularity, it's important to my system's reflection
<water> oops
<water> i should mention at this point that i'm really sick of undifferentiated use of the word "reflection"
<Fare> water: indeed! I prefer programming languages that allow you to be more productive, hence to risk cts if you work too long :)
<Tril> Fare: temporary monopoly. Plus he promised to release source
<water> "allow"
<Fare> Tril: then you have to find another term
<Tril> water ok i'll agree with that
<water> there's a loaded word
<Fare> like, "primeur" in french
<water> hehe
<Tril> Fare: There's another important reason, and that's that I'm not sure the information is correct, until someone tries to program an editor with it.
<Fare> french is a better language than english!
<Tril> abi: forget french
<abi> Tril: I forgot french
<water> Fare: i've made a language better than any other now spoken in my spare time in high school
<s_r> water with your own lettering system?
<s_r> water teach us this language
<Tril> abi french is a type of salad dressing
<water> sr: maybe later
<water> sr: the lettering system is fairly common
<Fare> s_r: also with its own communication medium.
<Fare> s_r: it is not a spoken or written language, but a though language
<Tril> abi french is a type of salad dressing, which is better than any other salad dressing, why doesn't everybody use french dressing?
<abi> bugger all, i dunno, tril
<s_r> water why not go the extra step and create a computer language?
<Fare> s_r: you must use telepathy to communicate with it
<Tril> abi french is a type of salad dressing, \which \is better than any other salad dressing, why doesn't everybody use french dressing?
<abi> i haven't a clue, tril
<water> this is pointless
* Tril/#TUNES gives up
<s_r> abi: French is a type of salad dressing that everyone should use.
<abi> ...but french is a type of salad dressing...
-:- Fare is now known as up
<up> Tril: what do you give me?
-:- up is now known as Fare
<s_r> abi: what is french
<abi> well, french is a type of salad dressing, which is better than any other salad dressing, why doesn't everybody use french dressing
* Fare/#Tunes is shall I tell Maciej to contact you?
<Tril> Fare: read http://www.tunes.org/~dem/tunes/spec/spec.html
<Fare> is: shall I tell Maciej to contact you?
<Tril> Fare: is is not in this channel, but still on irc
<s_r> fare: what kind of music do you prefer?
<Fare> s_r: good music ;->
10:50am
<Fare> s_r: I usually _listen_ to classical (actually, 19th century romantic) music
<s_r> i prefer TUNES
<s_r> :P
<Fare> s_r: by I try to appreciate a wide range of music, when I come to hear it
-:- bonobo [bonobo@ppp22.ijs.si] has joined #Tunes
<water> hi bonobo
<bonobo> hellou!
<bonobo> are you a robot?
<water> no
<water> abi is, though
<bonobo> of course
<water> what brings you here?
11:00am
<bonobo> tunes...
-:- water has changed the topic on channel #Tunes to: A Free Reflective Computing System (http://www.tunes.org/)
-:- water has changed the topic on channel #Tunes to: A Free Reflective Computing System at http://www.tunes.org/
<water> that kind of tunes?
<bonobo> music actually. i was just reading the paper about meta programmin on the TUNES site.
<water> ok
11:10am
<s_r> water what kind of music do you favor?
<water> hmm
<water> ambient electronica, acid, trance, some rock, new age, ...
<water> classical, acid jazz, jazz fusion
-:- smkl [sami@MCCCXVII.rdyn.saunalahti.fi] has joined #tunes
<bonobo> everything I like
<water> cool
<bonobo> right now I'm listening to Beethoven
<s_r> acid jazz?
<bonobo> the 5th
-:- Kaufmann [Kaufmann@200.224.105.201] has joined #Tunes
<Kaufmann> Hell-o!
<Tril> hi
<Kaufmann> In case anyone's interested, I'm building an archive of lore on the subject of code obfuscation.
<water> sr: what about it?
<s_r> obfuscate this!
<water> kauf: you have the story of Mel?
<s_r> err kaufmann what do you have now?
<Kaufmann> water: yep
<Kaufmann> s_r: very little. Just some stuff I've gathered from the Jargon File, plus a few notes I wrote myself.
<hcf> which oses / distros of oses use sendmail as default MTA?
11:20am
<s_r> kaufmann: hold a code obfuscation contest?
<Kaufmann> s_r: no, there's already IOCCC and IOPCC... I'm going to use material from both, of course, but not hold one myself
<Kaufmann> s_r: besides, the last time I tried to hold a contest, I made the mistake of offering a prize, and everyone got greedy. It sucked.
<water> kauf: got the intercal stuff?
<s_r> kaufmann: what do you think of code obfuscation in OS internals? is there any?
<Kaufmann> water: from ESR's page?
<Kaufmann> s_r: I wouldn't know, but prolly yes.
<water> kauf: dunno
<Kaufmann> water: it's one of the few INTERCAL pages around.
<Fare> UNIX is a huge contest a building an obfuscated OS!
<Kaufmann> LOL Fare
<Fare> (and NO, windows is NOT an OS)
<water> hehe. and Linux is a huge contest to improve on the obfuscation
<water> with more obfuscation
11:30am
-:- SignOff s_r: #TUNES (Read error to s_r[phila-dialup567.nni.com]: EOF from client)
<Kaufmann> If anyone has more stuff (especially stuff on Perl, Lisp or any other languages) send it to me (by email, at rnedal@olimpo.com.br)
11:40am
<water> Fare: care to explain implement1 to Tril?
<Tril> Fare: water says that Retro can't provide the safety and soundness required by Implement1, for your mini-scheme like language. I didn't know you were going to require that implement1 stuff until later
<Tril> water: i was under the impression it was going to be a requirement of the final system, but obviously not able to be provided during bootstrapping
<water> hmm
12:00pm
<Tril> well this conversation kind of died because Fare is not here
<water> indeed
<Tril> Fare: reply in email or some time on IRC when you are around
<Tril> I shoujld log off
<water> bye
* Tril/#TUNES is away: (afk) [BX-MsgLog Off]
<water> hmm. lots of unhelpful people here
<water> bye all
-:- water [water@ppp-tnt-193.tscnet.net] has left #tunes []
<Kaufmann> I should be going too. Peace
-:- SignOff Kaufmann: #TUNES ("Nationalism is an infantile disease. It is the measles of mankind." - Albert Einstein. For more information regarding the good and evil of nationalism, contact Fare at OpenProjectsNet #Tunes. (Heh.))
12:10pm
-:- SignOff bonobo: #TUNES (Ping timeout for bonobo[ppp22.ijs.si])
-:- Oddity [d96-abo@brown01.nada.kth.se] has joined #tunes
-:- Kenn [cpe2@1Cust51.tnt7.tampa.fl.da.uu.net] has joined #Tunes
<smkl> hello Kenn and Oddity
<Kenn> hello
01:10pm
<Kenn> is anyone discussing anything?
<Oddity> Hi!
* Oddity/#tunes just posted the member application form...
<Kenn> does anyone know anything about Apertos?
<smkl> i remember the name ...
01:20pm
<Kenn> from what I've read recently it is a reflective, OO computing system
<Kenn> it's based on a meta-object protocol
<Kenn> they've done work on large distributed systems with mobile hosts
<Kenn> real-time scheduling
<Kenn> location-transparent inter-object communication
<Kenn> naming and addressing of objects without unique identifiers
<Kenn> persistence
<Kenn> authenticated distributed communication
<Kenn> OO device driver programming
<Kenn> a reflective, OO programming language using dynamic compilation
<Kenn> migration
<Kenn> are there truly eight others aware of what I've just said?
<Kenn> hello?
<Kenn> I shall leave you so that you may discuss weightier matters
-:- SignOff Kenn: #TUNES (Leaving)
01:30pm
<smkl> hmm Kenn was busy ...
-:- tcn [tcn@cci-209150250094.clarityconnect.net] has joined #tunes
<tcn> howdy
<smkl> hello tcn
01:40pm
-:- SignOff hcf: #TUNES (Leaving)
-:- hcf [nef@me-portland-us1002.javanet.com] has joined #tunes
* Fare/#Tunes is back
<tcn> hey fare
02:10pm
-:- SignOff tcn: #TUNES (tcn has no reason)
-:- Darkness_ [user9641@uia.gst.pm179.cyberg8t.com] has joined #Tunes
-:- Darkness_ [user9641@uia.gst.pm179.cyberg8t.com] has left #Tunes []
<Fare> lo tcn
<Fare> what's up?
<Fare> will you be continuing retro?
* Fare/#Tunes is trying to make implement1 readable
02:20pm
<hcf> how about ps2text'able too
<Fare> hcf: read the .tex!
<hcf> at least .html
<Fare> hcf: and fix an ascii backend to xypic
<Fare> hcf: why don't you get a .ps viewer?
<hcf> i do now
<hcf> i gave in
<Fare> apt-get install gv
* hcf/#tunes feels defeated
<Fare> hcf: feel free to convert the article!
* hcf/#tunes DOESNOT like .ps
<Fare> hcf: how do you propose I produce the figures?
<Fare> hcf: it's free software -- hack it!
<hcf> i propose: dont produce figures
<hcf> use ascii
<Fare> hcf: it's free software -- hack it!
<hcf> Fare: feel free to do the right thing
-:- SignOff binEng: #TUNES ( <k!14>)
02:30pm
<Fare> hcf: what is the right thing?
<hcf> make a ascii version, thus readable by everyone
<hcf> html would do even
<Fare> how do I put pictures/formulas in ASCII?
<hcf> "... yearning for the bad old days, before the Web, when you had very little chance of reading a document written on another computer, another word processor, or another network."--Tim Berners-Lee
<Fare> hcf: if you wanna stay in the linux console, I have a program to automatically spawn a new X with the right client, so as to view dvi/jpg/ps
<Fare> ASCII is lame. So is HTML.
02:40pm
<Fare> moreover, postscript standardized and viewable with free software.
<Fare> and again, you can less implement1.tex
<hcf> this is mute
* hcf/#tunes : away. food
-:- Kaufmann [Kaufmann@200.224.105.163] has joined #Tunes
<Kaufmann> Shalom!
02:50pm
-:- SignOff Kaufmann: #TUNES (Ping timeout for Kaufmann[200.224.105.163])
-:- Kaufmann [Kaufmann@200.224.105.237] has joined #Tunes
<Kaufmann> ...
03:00pm
-:- binEng [Anders@195.84.234.31] has joined #tunes
* hcf/#tunes is back
-:- SignOff Kaufmann: #TUNES (Ping timeout for Kaufmann[200.224.105.237])
* Tril/#TUNES is back from the dead. Gone 3 hrs 26 min 32 secs
<Fare> Tril!
<Tril> Fare are you here? good maybe you want to reply to what I said earlier?
<Tril> Or maybe comment on page 1 of my new spec?
<Fare> yup
-:- stef2000 [user9452@adsl-77-224-236.atl.bellsouth.net] has joined #Tunes
<Fare> wait a minute, which thing you said?
<Tril> Fare: water says that Retro can't provide the safety and soundness required by Implement1, for your mini-scheme like language. I didn't know you were going to require that implement1 stuff until later
-:- stef2000 [user9452@adsl-77-224-236.atl.bellsouth.net] has left #Tunes []
<Tril> He refers to this at the "Retro-Implement1 Problem"!!
<Tril> s/at/as/
<Fare> I see no problem
<Fare> first, what does he mean by safety & soundness required by implement1???
<Fare> implement1 is about implementing a given system with another
<Fare> retro is not trying to do that
<Fare> we don't yet have an abstract system to implement
<Fare> and then, since retro is open, we can adapt it to add whichever property we want
<Tril> well, his point was that Retro isn't any better than any other existing LLL
<Fare> no better, no worse
<Tril> but I dont know what he meant, so maybe you should ask him when you see him
<Fare> one advantage, tho: the author is enthusiastic about cooperating
<Tril> if you want, you can read my short paper at http://www.tunes.org/~dem/tunes/spec/spec.html
<Fare> ok
<Tril> it's short because I didn't write the rest yet
03:40pm
<Tril> Thanks to your FAQ I learned Linuxdoc DTD, I like it
03:50pm
<Fare> hehe, the FAQ wasn't 100% useless, then
<Fare> why should a function be an injection?????
<Tril> justasec
<Fare> second elapsed!
<Tril> ok you are right I'll change that
-:- MaCa [maca@ps4.S26.fortalnet.com.br] has joined #Tunes
<Fare> I feel the whole section is overkill
<Fare> either people know the concepts, or not
<Fare> just give a bibliography for the concepts
<Fare> also, I can see two distinct meanings for f:X->Y
<Tril> ok i changed it to:
<Tril> A function is a total mapping from one set (domain) to another (codomain).
<Tril> This is the standard mathematical definition of a function.
<Tril> end of section
<Fare> the first is that forall x . x\in X \implies f(x) \lhd y \implies y \in Y
<Fare> the first is that forall x,y . x\in X \implies f(x) \lhd y \implies y \in Y
<Tril> whwwat is \lhd
04:00pm
<Fare> in presence of non-determinism, I use A \lhd B to denote the fact that all values of A are values of B
<Fare> uh, I mean, the contrary
<Fare> err, wait
<Tril> no i mean what symbol is it
<Fare> <|
<Fare> the triangle pointing to the left
<Tril> oh one of those symbols that has no standard meaning?
<Fare> yup
<Fare> unless you find somewhere a standard symbol for such things...
<Tril> what's wrong with subset
<Fare> it's already taken
<Fare> A \lhd B is same as values(A) \subset values(B)
<Fare> but A is not same as values(A)
<Tril> did you mean to say "the second" the second time you said "the first"?
<Fare> actually, in a value-less formalism (purely reduction-based), you can still define \lhd but no more values()
<Fare> the second, well, err, I wrote it in one of my notebooks
<Fare> but not the latest
<Tril> hehe.. I lost one of my notebooks the other day. It's not important, I remember everything important anyway
<Tril> or did I? it might be this one right here ...
<Fare> the second is about the same, but requires f(x) to have values
<Tril> well, there is no non-determinism so your definition doesnt apply. right?
<Fare> that is, the first says "if x\in X and f(x) converges to y, then y\in Y"
<Fare> the second also says "if x\in X, then f(x) converges"
<Fare> if you consider only total functions, then both definitions are the same.
<Tril> are you using "fuinction" to mean non-deterministic partial function , again?
<Fare> (btw, total functions == mappings)
04:10pm
<Fare> Tril: yup
<Fare> Tril: oh, even in presence of determinism, you have these two definitions, if you don't have totality
<Fare> the question is "must functions always terminate/converge/yield a result/whatever)
<Fare> s/)/?"/
<Tril> ok. well in my spec I mean a total function, so forall x \in X  xRy \implies y \in Y  , ok?
-:- water [water@ppp-ip86-b21.tscnet.net] has joined #tunes
<Tril> hi water
<water> hey tril
<water> hi all
<water> 2 new ppl?
<Tril> Fare they must have exactly one result if they terminate, but they don't need to terminate
<Tril> water: Oddity just applied to be a member, I haven't been introduced to MaCa
<Fare> ok, so there's still the question of whether being of type X->Y means you terminate on X or not
<water> Tril: ok
<Fare> water: read the log about retro & implement1
<water> ok
<Tril> Fare: I don't require that either
<MaCa> Tril: Hello Mr. Tril.
<MaCa> hehe
<Tril> Fare: I think, though, if x is not in X, then the member predicate of X must terminate in a negative answer, and evaluation of f will be stopped with a type exception
<Tril> MaCa: Hi
<abi> salut, Tril
<Tril> Fare: maybe not, haven't thought about it
<MaCa> Uh... It may sound silly but, what is the channel all about besides this TUNES project?
<water> hehe
<Tril> In absence of good reason I just allow everything
<Tril> MaCa: the TUNES project is everything
<water> maca: whatever is worth talking about to us
* water/#tunes thinks the Arrow project is closer to everything than Tunes is.
<water> :)
-:- _QZ [brand@p0wer.qzx.com] has joined #tunes
-:- _QZ [brand@p0wer.qzx.com] has left #tunes []
-:- _QZ [brand@p0wer.qzx.com] has joined #tunes
<water> hey qz
<Tril> water: What is your definition of "think" or "closer" or "project" or "is" (tril pokes at water for complaining about the meaning of words earlier)
<water> Fare: concerning implement1, ok
<water> hehe
<_QZ> hey water
<Tril> or more importantly, "everything"
* water/#tunes plans on booting english just like Tunes will boot LLL.
<water> ;)
<Tril> what about LLL?
<Tril> I don't know anything about a LLL :)
<water> nothing, just that i'll dscard english just like you will LLL
<Tril> oh, ok
04:20pm
<water> i guess i assumed too much about what implement1 was trying to say.
<Tril> well, we'll program a LLL into HLL and allow it to be subject to lots of dynamic changes, is that what you mean about English?
<water> sort of, yes.
<Tril> I agree the english thing is not part of TUNES requirements, but I'll still work on it anyway
<water> it's funny, but making my HLL over english was much simpler than for Tunes.
<Tril> when was this?
<water> high-school
<Tril> oh, that was when you were 6th grade?
<water> hehe
<Tril> or whatever
<water> yeah
<Tril> that's the one that you won't tell anyone about
<Tril> or, is it that you can't describe it because of language
<Tril> that's opposiute of what you just said, never mind
<water> it won't be presentable until i am satisfied that i can explain arrow to anyone
<Tril> aobut implement1, what did you assume
<water> that the lll would have to correspond somehow with the hll in some high-level way
<water> key words being "high-level way"
<Tril> wel... kind of..  after bootstrap, when we say "throw away LLL" we mean gradually replace LLL with things in HLL, i.e. extend HLL which does not have "levels" so it has functionality previously reserved for "low level" systems.
<water> to me, that's merely a question of proving that the HLL can do what we want it to.
<Tril> it's a question of describing hardware in formal specifications in the system!
<water> not just that
<water> you could manually enter all the necessary compiler and os code
<Tril> even declaratively?
<water> but that would just show it to be turing-complete
<Fare> water: the lll will indeed be used to implement the hll
<Fare> water: but since we don't have a hll yet, 1) we needn't worry too much, and 2) we must think about generic infrastructure to implement stuff
<water> there has to be a better test. we'd have to be able to transform the hll to fit all of the possible needs
<Fare> water: also, retro is not a compiler, only an executive
04:30pm
<water> Fare: i understand, but i thought that low-level bugs might disrupt the system
<water> of course i agree about all the other evaluations of retro
<Tril> it's not worse than alternatives? you agree with Fare about that?
<water> yes, but it's definitely not better
-:- SignOff MaCa: #TUNES ("I may be drunk, but in the morning I will be sober, while you will still be stupid and ugly.")
<binEng> Tril: It was you who recommended putty to me?
<Tril> yeah
<water> the Squeak or Self or Lisp VM would probably make much better choices.
<binEng> Is it possible to save the settings? login scripts?
<Tril> enter a name for the session and hit save
<Tril> never looked at scripting, it would be on one of those tabs
<binEng> That doesn't seem to work... do I have to install putty in any way first?
<Tril> after saving it appears on the list
<Tril> no
<binEng> that's strange. And I can't write [ or ] either. I need to!
<Tril> you need a host name and a session name to save.
<Tril> you have a non-English keyboard probably
<Tril> non US
<binEng> still I need it to work
<Tril> i dont know anything about that, maybe putty has translation in one of the tabs. Do you need me to go to the other room and open putty to look?
<binEng> well... *I* don't know how to do it... it's up to you
<Fare> what "Lisp VM" ???
<binEng> What is 'character classes'?
<Tril> Fare: a mailing list
<Fare> bE: where from?
<Tril> binEng: hold on I'll go run it
<water> Fare: basic Lisp
<binEng> Fare: in putty :)
<Fare> Tril: I know it, but apart from that, there's no such thing
<Fare> water: why not COBOL Lisp???
<water> hehe
<water> no comment
<hcf> BASIC Lisp
<Fare> there is no such thing as basic lisp
<water> _core_ Lisp
<Fare> no such thing
<water> i beg to differ
<Fare> oh, you mean LISP 1.5 ?
<Fare> the 1963 thing?
<water> hmm
<Fare> or LISP 1.0, the 1958 thing?
04:40pm
<smkl> did 1.5 have lexical closures
<smkl> ?
<Fare> nope
<water> i hope not
<water> good
<Fare> Scheme did it around 1974.
<Fare> CommonLISP adopted it around 1984
<water> i was thinking of the basic grammar without hardware
<Fare> between the two, some LISPs had an ugly behavior whereby the interpreter had dynamic variables only, and the compiler lexical variables only -- ouch!
<water> i mean, hardware types
<Fare> LAP?
<Fare> LISP Assembly Programming?
<Tril> Fare: VM is probably the same as an interpreter
<water> oh yes
<Fare> VM is EMACS' View Mail mode
<Fare> abi: VM?
<abi> VM is the thing defined by the current interpretation of a particular arrow, so that multiple (or infinite numbers of ) VMs may exist simultaneously.
<Fare> what's that thing???
<abi> fare: wish i knew
<water> abi forget vm
<abi> water: I forgot vm
<water> that's part of an explanation i was giving
<water> within my system, it's quite true
<Fare> lemme define a system by the fact that the sentence "abi is gaklosmontic" be true.
<Fare> well, in dis system, abi is gaklosmontic, indeed!
<Fare> now, just what does "gaklosmontic" mean?
<water> well, within the arrow system, the only type available is "arrow"
<Fare> not very useful
<water> and the only identifier available is ""
<water> very useful!!
<water> the entire interpretation is left to what is made implicit
<water> constructively
<Fare> ?
<Fare> could you speak english?
04:50pm
<Fare> (or better, french)
<Tril> or better, release the new Arrow paper :)
<water> sigh...
<water> tertium quid sounds best about now.
<water> (i.e. what Tril said)
<Fare> well, be sure to write it in english
<Tril> lol
<water> Fare: there's no point in my trying to please you.
<Fare> well, in France, too, we have "philosophers" who use a language that superficially looks like french, but actually is a dialect only understandable by each other "philosopher"
<Fare> water: there's a point in trying to be understood.
<water> i understand
<water> fare: yes, but you aren't looking beyond Lisp-variants, so what's the point in showing you what's out there?
<Fare> I wish I did
<Fare> water: I know ML, Haskell, Perl, zsh
<Fare> prolog
<water> so? you'd implement them all in Lisp if you had the time.
<Fare> mercury
<Fare> yup, Lisp allows that
<water> no kidding
<Fare> the contrary isn't trye
<Tril> writing lisp in FOO?
<water> not exactly
<Fare> there's already been prolog-in-lisp
<water> i know
<water> i've seen it
<Fare> Screamer. Works quite well. And _portable_ macros.
<water> here we go again
<Fare> try to write an efficient prolog _embedded_ into C
<Fare> in a portable way
<water> Fare: you don't need to convince me of anything about Lisp.
<Fare> so what, then?
<Fare> are you reproaching me of making the right choice?
<Fare> "hey, why not choose FORTRAN as your HLL?"
<water> i'm reproaching you for thinking that you have the right choice for any reason
<Fare> "uh, I mean ENIAC assembly!"
<smkl> is screamer efficient?
<Fare> smkl: sure
<water> Lisp is NOT a valid HLL candidate
<Tril> agree
<Fare> water: why not?
<smkl> what implementation strategy it uses?
<Fare> smkl: the user-specified strategy, really;
<water> because it's only designed to encapsulate certain kinds of ideas
<Tril> lisp is ok for prototypes, though
<Fare> smkl: it's a backtracking engine written using macros and a lisp code walker
<Fare> water: any formalism is "only designed to encapsulate certain kinds of ideas" -- namely the kinds of ideas the formalism is designed to encapsulate
<Tril> Fare: so what's an anti-formalism?
<Fare> what's an anti-what?
-:- kermit [kermit@md19.netserver1.tower.com.ar] has joined #tunes
05:00pm
<Tril> Fare I suggest that maybe we should exceed formalisms the way you conceive of them
<Tril> Hi Kermit.
<kermit> hi
<kermit> may i ask a question ?
-:- SignOff water: #TUNES (Ping timeout for water[ppp-ip86-b21.tscnet.net])
<Fare> kermit: may we not answer if we cannot?
<kermit> sure.
<Fare> sure, then.
<kermit> what is the tunes project all about ? an OS ?
<Fare> kermit: have you read the FAQ?
<kermit> nope ...
<kermit> should i do that ?
<kermit> (;
<Tril> yes always
<Fare> kermit: well, it'd be better if you come back after reading it
<Tril> you can never read too many FAQs
<Fare> this way, we may determine what is wrong about the FAQ
<Fare> if we tell you before you do, we cannot measure the FAQ's efficiency on you
<kermit> ok, hold on. im reading it right now.
<Tril> you know how reality turns into streams of numbers in the Matrix? Well, TUNES is trying to do that with the software in your computer
<Fare> Tril: wasn't it streams of numbers turning into reality?
<kermit> i think i got a pretty good idea. I didnt read it all, but i believe i do have an idea ...
<Fare> Tril: I saw the movie last night
<Fare> kermit: any more questions?
<kermit> yes ...
<kermit> do you need help ?
<Tril> Fare: right
<Fare> kermit: sure
<Fare> kermit: a lot!
-:- water [water@ppp-tnt-223.tscnet.net] has joined #tunes
<kermit> ok, let me tell you about me ... then if you like what you hear, you can "hire" me ...
<kermit> (:
<Tril> wb water
<water> b
<kermit> first of all, my first language isnt english, so i apologize for my poor use of it.
<smkl> your first language is spanish?
<kermit> thats true.
<Tril> Fare: if your formalisms encapsulate only certain kinds of ideas, they aren't good enough
05:10pm
<Tril> who knows Spanish. Water and Fare?
<Fare> hablo pocito el espanol
<Tril> feel free to create a #tunes-es
<kermit> i do speak spanish
<Fare> poquito?
<water> Fare: ever considered informal yet precise ideas?
<Fare> water: sure
<Fare> water: see the debate of extension vs intension
>>> hcf [nef@me-portland-us1002.javanet.com] requested PING 931649653 from TUNES
<Fare> s/debate/duality/
<kermit> so, you said you need help ... in which areas ?
<water> fare: where?
<Tril> that's exactly what we're talking about
<Fare> kermit: all -- design, implementation, documentation, administration, ...
<kermit> ok, count me in
<smkl> kermit: do you know anything about AI?
<water> but implementation is straight out until we decide on design
<Fare> Tril: by definition, the formalism can only tackle the extensional part; at most, it can provide "hooks" for external intention
<kermit> Artificiall Intelligence ?
<Tril> don't forget more help in public relations, marketing, research, grunt work and morale support
<Fare> smkl: don't you scare the newbies!
<Fare> Tril: and sex?
<water> Fare: that's pr ;)
<kermit> not really.
<Tril> Fare: not in this channel. go create another if you must
<smkl> well, i don't too
<kermit> i have a really basic idea though.
<kermit> i mean REALLY BASIC ...
<Tril> smkl: HickServ wants to help with AI. I havent seen any ideas from him though
-:- SignOff hcf: #TUNES (Ping timeout for hcf[me-portland-us1002.javanet.com])
<kermit> im actually an OS guy.
<kermit> if theres such a thing.
<Fare> kermit: what is your technical background?
<kermit> currently studying comp. science
<Fare> and scientific background?
<kermit> undergraduate
<kermit> i expect to get my degree next year.
<kermit> scientific ? theres no such thing as "science" down here.
<kermit> all the research ive done ive done by myuself ...
<kermit> myself.
<kermit> hmm ... hello ?
<Fare> ok
<Fare> have you specifically implemented/hacked any program?
<Fare> what languages do you know?
<Fare> which part of Tunes would you like to hack?
<kermit> im currenttly "trying" to write an OS (just for fun)
<kermit> languages: C, Java (some time ago), Scheme (not much), ML
<kermit> im also wrting a compiler right now.
<kermit> writing.
-:- hcf [nef@me-portland-us1002.javanet.com] has joined #tunes
<water> wb hcf
<Tril> water: touching up arrow paper?
05:20pm
<kermit> which part ... hmmm ... its hard to decide ... i dont know, i guess that i would hack anything you tell me too (if i can handle it of course)
<smkl> compiler for what language?
<water> tril: the replacement papers, yes
<kermit> Tiger
<Tril> kermit that may be a problem, we dont give orders, usually people have to find something to do themself.
<Fare> Tiger?
<kermit> its a kind of functional lang.
* water/#tunes really wants to have the power of LaTeX, but doesn't have the time to learn.
<kermit> Tril: ok, then when i read more about the project i guess ill find the are that interests me ...
<Fare> kermit: well, one thing you could do is write the core of a scheme compiler
<kermit> Fare: sounds good to me.
<Tril> kermit: or if you want to give orders you could try doing that, too
<kermit> Tril: i dont tyhink id like that ... im not a "suit"" yet ... (:
<Fare> kermit: ok. If you want to write a lexer, follow the spec given by the CommonLISP HyperSpec
<kermit> sure
<Fare> kermit: alternatively, assume read is present already
<Fare> (and write the lexer later)
<water> :P
<kermit> uh ?
<Tril> Fare: Is CLHS downloadable in its entirety? I can't find a link to a big file on their site. Also, is CLHS going to be taken down since Harlequin is going out of business?
<water> Tril: got a site grabber?
* Fare/#Tunes wants to be cultivated but doesn't want to spend time reading
* Fare/#Tunes wants to be muscled, but doesn't want to spend time doing sport
<Tril> I tried wget, (not very hard) it didnt seem to get enough files
<Tril> is there a good link to start spidering from
<water> yes, the clhs is very large
<Fare> Tril: I remember I got the CLHS from Harlequin.
<water> dunno
<Fare> I don't remember how
<Fare> I think it was a .tar.gz
<Fare> at worst, you can use w3mir
<Tril> FTP, maybe? or some other site besides harequin?
<Fare> I think it was at Harlequin's
<Tril> actually, I think there's a deb
<Fare> oh
<kermit> i have a question ...
<kermit> a lexer for scheme ?????
<water> what about it?
<kermit> you dont need a lexer ...
<water> hehe
* water/#tunes thinks kermit sees the futility in Fare's plan
<Fare> kermit: if you want a standalone compiler, you need a lexer
<smkl> kermit: well, you would need to tokenize it to words, ., ' and parenthesis
<Fare> kermit: i.e. an implementation of read
<Fare> kermit: that's not the most important
<kermit> correct me if im wrong ...
<kermit> but, after youve tokenized, and parsed, and youve checked the datatypes, all you got is a kind of "lisp" sintax ... right ???
<Tril> found it
<Fare> kermit: you may begin with a compiler that takes SEXP with variables, lambda, set! and call/cc in input, and output symbolic "assembly" for a virtual machine that has a stack and a heap
<Fare> kermit: right
<kermit> so, why would i want a lexer for scheme ??
<water> hehe
<Fare> kermit: the lexer is the thing that tokenizes!
<kermit> i know ...
<kermit> but, did you read my question ?
<Fare> kermit: your question begins with *after*
<Fare> kermit: there is a before, too
<kermit> yes ...
<kermit> so, wen you write lisp (scheme) code, its like youve already "tokenized" and parsed and stuff.
<kermit> right ?
<smkl> when you write it, it's just one big ascii file
05:30pm
<Fare> kermit: parsing isn't done magically. Someone must write the program.
<kermit> look, this is what i mean ...
<smkl> parsing and lexing lisp is just extremely easy
<water> and pointless
<Fare> easy != magic
<kermit> (:
* Tril/#tunes is definitely confused
<Tril> water: care to explain what is going on here?
<_QZ> stril: no i think kermit is the confused one :)
<water> hehe
<kermit> it happens to be that lisp syntax is the same as the intermediate code syntax ...
<water> there are two groups here
<Fare> again, if you don't want to write the lexer, don't
<kermit> im i right ? (please correct me if im wrong)
<water> neither understands the other
<Fare> the compiler is more interesting, perhaps
<Fare> and we need it, too
<Fare> or write a pattern matcher
<water> "need" <- bwa ha ha!
* Tril/#TUNES is away: (food) [BX-MsgLog Off]
<Fare> kermit: don't be confused between _syntax_ and _structure_
<kermit> youre right ... i see your point nowe.
<kermit> now.
<water> bye Tril
<Tril> oh, so you want to parse into s-exp even if the language isn't lisp. i get it
<kermit> yes, i see it now.
<Tril> because s-exp is a general purpose data structure
<Tril> ok. bye
<binEng> bye
<kermit> hey, i should be going now, but ill be back ... if thats ok w/you ...
<Fare> s-exp is a _syntax_ for recursive lists, that are a general purpose data structure
<kermit> Fare: yes.
<Fare> sure
<water> code doesn't have to be a recursive list
<water> a sexp could represent a persistent data value
<water> obtained by computation
<kermit> btw, id like to code some of the OS core rather than the compiler if thats ok.
<Fare> kermit: sure
<water> ok, no one's listening i see
<kermit> im not a languages guy as youve already saw ... (:
<Fare> water: lists represent code, that can evaluate to something else than lists
<Fare> kermit: what do you prefer?
<kermit> what do you need ?
<water> fare: lists prohibit value re-use if there are no identifiers or pattern-matching
<Fare> kermit: say, integrating retro with oskit drivers?
<smkl> we need a clue
<kermit> retro ?
<abi> retro is at http://bespin.cx/~tcn/retro.html
<hcf> is this the entire CLHS: http://www.harlequin.com/education/books/HyperSpec/HyperSpec-4-0.tar.gz ?
<kermit> btw, do we have to write the kernel ???
<water> hcf: it should be
<Fare> kermit: that depends
<Fare> another hot project is building OTOP
<kermit> because i saw a pretty cool micro-kernel, im actually working w/it.
<Fare> kermit: do you prefer implementing a persistence layer hitting the page table directly?
<smkl> kermit: what is it called?
<Fare> kermit: maybe that's the kind of things you like
<kermit> fiasco. 
<kermit> do you know it ?
<Fare> yup
<Fare> it's the free L4 reimplementation
<kermit> yes
<kermit> im "kind of" working w/it
<hcf> abi: fiasco is at http://os.inf.tu-dresden.de/fiasco/
<kermit> finally someone who knows about fiasco ... (:
05:40pm
<Fare> if you prefer, you may adapt the OS kit to fiasco
<kermit> thats already done.
<Fare> oh
<Fare> well, then write that persistence layer atop fiasco
<kermit> actually fiasco uses a part of the OS Kit
<kermit> and theres also a version of the linux kernel that runs on top of fiasco.
<kermit> actually the linux OS
<smkl> what kind of performance it has?
<kermit> well, i ve got to go now.
<smkl> be back tomorrow
<kermit> smkl: theres a paper about that, i havent read it but it seems to be pretty good.
<kermit> btw, at what time do you people meet ??
<water> whenever we can
<Fare> smkl: some 3-5% slower than linux w/o uK
<kermit> oh. ok, ill be back tomorrow. it was nice meeting you, and i hope we can work together.
<Fare> smkl: but with the possibility of writing programs directly over the uK
<Fare> in parallel with linux programs
<kermit> bye now.
<water> bye kermit
-:- SignOff kermit: #TUNES (kermit)
-:- SignOff hcf: #TUNES (Leaving)
-:- hcf [nef@me-portland-us822.javanet.com] has joined #tunes
<water> wb
-:- SignOff binEng: #TUNES (Sleeping time <k!14>)
* water/#tunes is reading a very depressing slashdot article
<water> anarchy!!!
<water> centralized control sucks.
05:50pm
-:- SignOff smkl: #TUNES (sleeping)
* Tril/#TUNES is back from the dead. Gone 0 hrs 26 min 8 secs
<water> wb tril
<Tril> hi
<abi> que tal, Tril
06:00pm
<water> i just realized that if we managed to make Tunes that the government would, once it understood it, _have_ to have it.
<Tril> and make it proprietary?
<water> the US alone could wipe us out just to get it.
<water> yes.
<Tril> A proprietary tunes would not be able to compete against a free one.
<water> because it would violate "national security" to have a system that can recompile itself in any conceivale way.
<Tril> It's completely designed to be free, if you restrict that, it won't really work (be useful)
<water> for instance
<water> who needs competition?
<water> they'll just take everything
<Tril> water: It's free. they can have it already
<water> i don't want them to. but then that just encourages me to do it better
<water> i don't want to have to compete with paid people over hackers
<water> especially the linux addicts
<water> i don't consider them to be useful, although they would be allies
<Tril> i used to think like you do, but I realized that great software helps big corporations and governments.  The tech industry is driving US economy, and soon (if not already) free software is driving the tech industry.  If anything, tunes will allow corporations to make MORE money and allow govenrments to get more power.
<Tril> however, they will also get that power at the cost of being unabel to stop others from having it as well. That, in itself is inevitable even without tunes.
<water> hmm
<water> we'd better take care of what we make
<water> i've considered the capabilities of the agencies as best i can
<water> if they had tunes as well as the public, and meddled with us to foil development...
<water> well, actually...  they're probably not able to pull that kind of thing off.
<water> the government is well known to be incapable of grasping the concept of non-linear control
06:10pm
<water> ooh. here's an interesting article about Stormix, the new Linux distributor
<Tril> water: trust me!  "they" are looking for attacks by people like themselves, seeking profit.  They can't comprehend behavior that gives them profit at no cost.  We're changing the system not by braking it, but by growing it so fast that the "system" becomes unnecessary.
<water> ok
* Fare/#Tunes is back
<water> Storm Administration System
<water> it allows you to write programs as modules that access terminals or X interfaces equally via a Simple Interface Language
<water> sounds very low-level
<Tril> water: a fundamental limitation of a business is it is limited in how quickly it can grow (one company=micro). In the macro world if technology develops faster than the speed a business can expand at, it will find other more equitable outlets!
<water> hmm
<water> let's move on
<Tril> in other words some technolgoy seems unprofitable because it is TOO profitable
<water> saturation by public availability
<water> tril: any arrow stuff to discuss?
<Tril> anyway, no one can have a monopoly on ideas
<Tril> OK arrows? Um, oh, i thought you were goign to give a paper, like, today
<water> hehe
<water> sorry
<Tril> well, my fault for misreading that
<water> i have the reflective relativised arrow logic paper nearly ready, but Word 97 format sucks for math notation
<water> so the examples and problems are vaporware
<water> :(
06:20pm
<water> the paper references my arrow logic books and discusses the sexp-notation for arrows, but also always assumes a base frame of logical symbols.
<Tril> why don't you get a volunteer to make a paper in some other format with math examples?
<Tril> (like someone who knows tex like fare)
<water> i'm too proud to ask fare for anything
<Tril> ask the net in general
<Tril> like the tunes list
<Fare> water: <subliminal>use LaTeX</subliminal>
<Fare> err, I mean, \begin{subliminal}use LaTeX\end{subliminal}
<water> Fare: <subliminal> f*** off</subliminal>
<Tril> DocBook has math symbols
<water> is the typesetting good quality?
<Fare> <sin><plus><var name="a"><var name="b"></plus>
<Fare> </sin>
<water> uhh
<Tril> water: A lot of books use it. it must be
<water> where is it available?
<water> i hope it's free. i don't even have a credit card
<Tril> there's a lot of links to it on www.sgmltools.org
<water> ok
<Fare> Is there already free support for DocBook???
<Tril> DocBook is an SGML DTD.
<Tril> Fare: The LDP is asking all HOWTO / FAQ authors to switch to it
<Fare> I thought the SGMLTools project had been discontinued after Cees de Groot was hired away...
<water> it seems so from the home page
<Fare> Tril: no. They are _intending_ to switch to it
<Fare> Tril: but no one is developing the tools anymore
<Tril> Fare: he is in discussion with someone to become the new maintainer
<Fare> why don't I receive anything from the ldp anymore?
<Fare> Tril: would bespin be blocking some sites, including the one with the ldp list?
<water> hey! no linux-specific stuff here! :)
<water> wow. an open-source project that doesn't pander to linux.
<Tril> i'm only blocking 2 sites, see /etc/hosts.deny
<Fare> water: f*** on
<water> more power to them.
<Fare> PARANOID may be the problem...
<Fare> hum. Does it affect SMTP???
<Fare> sendmail doesn't read /etc/hosts.deny, does it?
<Tril> no, hosts.* is for tcpwrappers (/usr/sbin/in.tcpd)
<Tril> tcpd is run for inetd stuff. sendmail has its own daemon
06:30pm
<Fare> ok, so I don't understand why I don't receive mail from the ldp list anymore :( :( :(
<water> well, i was wrong about them being linux-independent
<water> :(
<Tril> what to do when everyone uses tunes?
<water> i'd work on something else, silly
<Tril> Fare: maybe you filtered it to another folder?
<Tril> no , i mean when you can't find anyone who's tunes-independent
<water> oh
<Tril> besides, we're not supposed to care if other people use tunes
<water> i'll speak against tunes then
<Fare> Tril: nope. Nothing from them in procmail.log
<Tril> Fare: what is the list address?
<water> i like their advertised features list
<water> sweet and simple
<Fare> used to be ldp-l@linux.org.au
<Fare> whose feature list?
<water> sgmltools
<Fare> oh
<water> bleh. only for linux
<water> Tril: i can't find actual DocBooks links
<water> wait. found it
06:40pm
<Tril> abi: DocBook is at http://www.oasis-open.org/docbook/
<abi> ...but docbook is an SGML DTD....
<Tril> abi: DocBook is also at http://www.oasis-open.org/docbook/
<abi> okay, Tril.
<water> why the hell must typesetting be so difficult?
<Fare> water: if it's soooo easy, just write a typesetting program!!!
* water/#tunes thinks Fare can take his sarcasm and shove it.
<Tril> maybe Adobe Acrobat (full ver) is better?
<water> actually...
<water> it seems FrameMaker supports it
<water> and i have a pirated copy
* Fare/#Tunes denounces water to the SPA
* Tril/#tunes pities water's soul
* water/#tunes has no soul
<Fare> you sold it to the devil already!
<water> i never had it to sell
* water/#tunes envies people who have ever had souls.
<water> great. i need FrameMaker+SGML
<water> only $2000 US
06:50pm
<Tril> hmm, the dark secrets of the elite academic community begin to be faintly seen
<Tril> They make it really hard to write papers!
<Fare> How the hell can I deactivate these popups-on-enter and popups-on-leave from netscape?
<water> disable javascripts
<water> ?
<Tril> popups-on-enter only show the first time you run, or if a .netscape/lock exists, what's a popup-on-leave?
<Fare> isn't enough, apparently!
<Tril> when you run netscape, or go to a site?
<water> i'll just apply to my university's department for a copy. oh wait!
<Fare> it looks like  netscape runs javascript stuff even when disabled
<Fare> or maybe it's done with HTML extensions
<Fare> anyway, netscape 4.5 will do annoying things
<Fare> happily, that makes it die, often
<water> use Lynx
<water> better, emacs :)
<Fare> emacs isn't multiprogrammed
<water> so? :)
* Fare/#Tunes hits water with a large pan
<Tril> abi what is multiprogrammed
<abi> tril: wish i knew
<hcf> Fare: how bout mmm?
* Tril/#tunes logs off to write
* Tril/#TUNES is away: (afk) [BX-MsgLog Off]
* water/#tunes hits Fare in the head with the island of Gibraltar.
<Fare> mmm won't render slashdot
-:- SignOff water: #TUNES (The Tao has left the building!)
<Fare> neither will chimera
<hcf> oh dear, what a tragedy
07:00pm
<Fare> :)
<Fare> and closure isn't multiprogrammed, either (and has a bug in picture rendering)
<hcf> f*** gfx
<hcf> whats the url for closure?
<abi> closure is the Common LISP web browser, a counterpart to CL-HTTP the web server or at http://www.uni-karlsruhe.de/~unk6/closure/
<hcf> abi: good girl
<abi> :)
<Fare> also, I couldn't run closure but on proprietary ACL5
<Fare> (actually, I couldn't get CMUCL to run at all)
07:10pm
-:- water [water@ppp-tnt-223.tscnet.net] has joined #tunes
-:- s_r [sr@phila-dialup281.nni.com] has joined #tunes
<s_r> i have a question about data addressing and alignment.
<s_r> can anyone answer it?
<Fare> hum
<Fare> under what architecture?
07:30pm
<s_r> x86
<s_r> of course
<Fare> why, "of course"?
<Fare> and what is the question?
<abi> the question is "must functions always terminate/converge/yield a result/whatever)
<Fare> the answer is 42
<Fare> abi: forget the question
<abi> Fare: I forgot question
<s_r> functions must always terminate
<s_r> they musn't always yield a result though
<s_r> answer?
<abi> well, answer is 42
<s_r> ok, i have heard that memory can only be addressed in multiples of word size (32-bits)
<Fare> what memory are you talking about?
<s_r> if that's true, then why does nasm output 'db's  in just bytes
<s_r> and not word sizes
<s_r> like it doesn't align data
<s_r> is memory addressed in word sizes?
<Fare> the data bus on 32-bit processors is 32-bit
<s_r> so i can't access a single byte at a time?
<Fare> which means that the processor accesses memory by words of 32-bit
<Fare> now, 1) some processors (including the x86 family) have some additional logic to do multiple word access so as to access unaligned data, at a cost
<Fare> 2) some processors get a trap/fault when accessing unaligned addresses (like the 486+ if proper flags are set)
<Fare> 3) some processors cannot even express unaligned access, and only address whole words
<Fare> 4) even if the processor natively talks 32-bit, the memory-access unit might have a different memory bus size (16-bit on the 386SX, 64bit on the Pentium, god knows how much on a PIII)
<s_r> ok so can i access just one byte that isn't aligned on a pentium?
<s_r> if i wanted to, how could i do that?
<Fare> a byte is always byte-aligned, dude!
<s_r> if i sent out a mov ax, <unaligned memory> does the process access the whole word and just give me the byte
<s_r> or do i have to access the whole word in my code?
<Fare> the pentium takes care of alignment problems in hardware
<Fare> s_r: hey! RTFM
<Fare> why don't you read a x86 tutorial?
<water> hehe
<s_r> so some C compilers just output data that is word aligned?
<Fare> 1) the x86 will let you access unaligned data, albeit more slowly
07:40pm
<Fare> 2) compilers typically align all data, for speed
<s_r> ok thanks
<s_r> that's what i wanted to know
<s_r> now you should work on including all this in your ASM howto
<s_r> ;>
<Fare> 3) if you put the x86 in special mode, it will trap when accessing unaligned data
<Fare> s_r: fuck you!
<Fare> s_r: there's nothing linux-specific about that
<Fare> s_r: the asm-howto is no x86 tutorial
* Fare/#Tunes hits s_r with a Pentium User Guide
<water> hehe
* Fare/#Tunes hits s_r with a truckload of intel manuals
<s_r> heh i should read that manual
* Fare/#Tunes hits s_r with the last section of the asm howto
<s_r> what's the location of the ASM howto?
<Fare> abi: HOWTOs?
<abi> HOWTOs are at http://metalab.unc.edu/LDP/HOWTO/
<s_r> does the processor only access things at even addresses?
07:50pm
* Fare/#Tunes hits s_r with a cluestick
* s_r/#tunes hits Fare with a trout
* Fare/#Tunes is trout-proof
<hcf> enuf hitting, do something else painful
<Fare> ok
-:- mode/#Tunes [+o Fare] by ChanServ
-:- s_r was kicked off #tunes by Fare (s_r)
<Fare> ah!
-:- s_r [sr@phila-dialup281.nni.com] has joined #tunes
<s_r> heh
<water> hey hcf
<water> what's going on for you?
<hcf> water: not much, u?
>>> hcf [nef@me-portland-us822.javanet.com] requested PING 931659529 from TUNES
<water> a little depressed today over a slashdot article about privacy, as well as some frustrations with typesetting
<water> sorry i had to answer the phone
<water> forgot to say "brb"
<hcf> whats the M$ usually typesetting "solution"?
<hcf> s/\$/\$\'s/
<water> hehe Word 97
<water> my current pirated fav
<water> it's terrible
-:- s_rr [sr@phila-dialup007.nni.com] has joined #tunes
<hcf> theres no word (9[89]|2000) ?
<_QZ> Fare: pentium thru p3 are 64bit data bus
-:- SignOff s_r: #TUNES (Ping timeout for s_r[phila-dialup281.nni.com])
<water> 98 is for mac
-:- s_rr is now known as s_r
<water> i don't have word 2000 pirate yet
08:00pm
<water> and i refuse to pay
<water> maybe i should just get the right linux tool and force myself to learn Yet Another typesetting language
<water> :P
<Fare> QZ: K7 are 128-bit memory bus, aren't they?
<water> i really hate linux
<water> it wasn't until i had it installed for more than a week and had been using it for that long that i realized it
<Fare> water: if you learn LaTeX, you'll never have to learn another one afterwards, since LaTeX is free software!
<water> shut up Fare, i have EmTeX which supports LaTeX
<water> LaTeX is about as usable as C
<water> :P
-:- water was kicked off #tunes by Fare (bad idea)
-:- water [water@ppp-tnt-223.tscnet.net] has joined #tunes
<water> sorry
<Fare> that's ok.
-:- mode/#Tunes [-o Fare] by Fare
<Fare> C is free software too
<water> i guess i shouldn't bother someone as sensitive as you.
<water> so?
<Fare> if you want something better, pay for it!
<water> if someone offered you poison for free, you'd take it?
<Fare> i.e. pay the development of a free software alternative
<Fare> LyX? Curl?
<Fare> KWord? Abiword?
<Fare> Amaya?
<water> what?
<water> you speak Hindi or something?
<Fare> apt-get install lyx abiword amaya
<water> oh yeah, you speak Linuxese
<water> don't try to program me, Fare
<water> i really hate it when hackers use shell scripts to talk to people
<water> i'm not a f***ing machine
<Fare> I hate it when people are close-minded and don't want to LEARN
<water> learn ~= acting like a machine
<Fare> learn = copy-paste the script I made an effort to write for you
<Fare> which script will install three software packages for you to use
<water> forcing my thought-process into ascii text is not my idea of learning
<water> it sounds like de-evolution
<water> are they available for anything besides Linux?
<Fare> water: 1) I'm not your secretary 2) if you want to pay me as a teacher, say so. I've given courses in the past. But not for free.
08:10pm
<water> F*** you. you have nothing to teach me, machine
<water> you're linux words are worth nothing
<water> your
-:- hcf has changed the topic on channel #Tunes to: rage against the machine
<water> you go 'apt-get' all you want. it will gain you nothing
<water> go bit-shift all your friends
<water> re-direct their output to a console
<water> encapsulate them in a bunch of Athena widgets
<Fare> poor water
<s_r> water has Linux Rage
<Fare> pour water
<s_r> water do you like synth music?
<Fare> s_r: put some reggae on the tape player
<Fare> s/s_r/water/
* s_r/#tunes is listening to Beastie Boys' "Sure Shot"
* water/#tunes is babysitting now
<Fare> poor baby
<s_r> water are they driving you crazy?
08:20pm
-:- SignOff water: #TUNES (Ping timeout for water[ppp-tnt-223.tscnet.net])
-:- Kaufmann [Kaufmann@200.224.105.209] has joined #tunes
<Kaufmann> Shalom!
<hcf> hoy Kaufmann
09:00pm
<Kaufmann> hoy hcf
-:- SignOff s_r: #TUNES (Ping timeout for s_r[phila-dialup007.nni.com])
-:- NetSplit: varley.openprojects.net split from asimov.openprojects.net [09:08pm]
-:- BitchX+Deb1an: Press Ctrl-F to see who left Ctrl-E to change to [varley.openprojects.net]
-:- Netjoined: varley.openprojects.net asimov.openprojects.net
-:- Fare [rideau@quatramaran.ens.fr] has joined #Tunes
09:10pm
-:- NetSplit: varley.openprojects.net split from tolkien.openprojects.net [09:16pm]
-:- BitchX+Deb1an: Press Ctrl-F to see who left Ctrl-E to change to [varley.openprojects.net]
-:- Netjoined: varley.openprojects.net tolkien.openprojects.net
-:- Fare [rideau@quatramaran.ens.fr] has joined #Tunes
<Kaufmann> abi
<abi> yes, Kaufmann?
<Kaufmann> abi: are you having fun tonight?
<abi> wish i knew, kaufmann
-:- water [water@ppp-tnt-215.tscnet.net] has joined #tunes
<water> b
<water> sigh....
<water> good night all
-:- water [water@ppp-tnt-215.tscnet.net] has left #tunes []
09:30pm
<_QZ> Fare: u still here?
09:40pm
-:- william428 [william@dt130n87.tampabay.rr.com] has joined #Tunes
-:- william428 [william@dt130n87.tampabay.rr.com] has left #Tunes []
-:- SignOff Kaufmann: #TUNES ("Nationalism is an infantile disease. It is the measles of mankind." - Albert Einstein. For more information regarding the good and evil of nationalism, contact Fare at OpenProjectsNet #Tunes. (Heh.))
-:- noise [user6613@genius.hypestream.com] has joined #Tunes
-:- noise [user6613@genius.hypestream.com] has left #Tunes []
-:- NetSplit: varley.openprojects.net split from lackey.openprojects.net [11:50pm]
-:- BitchX+Deb1an: Press Ctrl-F to see who left Ctrl-E to change to [varley.openprojects.net]
-:- Netjoined: varley.openprojects.net lackey.openprojects.net
-:- Fare [rideau@quatramaran.ens.fr] has joined #Tunes
[msg(TUNES)] newlog 1999.0711
IRC log ended Sun Jul 11 00:00:00 1999