IRC log started Tue Jun 22 00:00:00 1999
-:- SignOff _QZ: #TUNES (BRiX [http://www.qzx.com/brix] :: sleep)
[msg(TUNES)] permlog 1999.0622
-:- Limbosuave [email@1Cust46.tnt3.covina.ca.da.uu.net] has left #Tunes []
-:- Limbosuave [email@1Cust46.tnt3.covina.ca.da.uu.net] has joined #Tunes
-:- lar1 [lar1@1Cust193.tnt3.sfo3.da.uu.net] has joined #tunes
-:- SignOff lar1: #TUNES (ircII2.8.2-EPIC3.004 --- Bloatware at its finest.)
-:- Limbosuave [email@1Cust46.tnt3.covina.ca.da.uu.net] has left #Tunes []
-:- water [water@ppp-tnt-111.tscnet.net] has joined #tunes
-:- NetSplit: koontz.openprojects.net split from clarke.openprojects.net [03:17am]
-:- BitchX+Deb1an: Press Ctrl-F to see who left Ctrl-E to change to [koontz.openprojects.net]
-:- Netjoined: koontz.openprojects.net clarke.openprojects.net
-:- water [water@ppp-tnt-111.tscnet.net] has joined #tunes
-:- s_r [s_rr@phila-dialup118.nni.com] has joined #tunes
-:- SignOff water: #TUNES (The Tao has left the building!)
-:- smoke [smoke@7dyn99.delft.casema.net] has joined #tunes
-:- Fare [rideau@quatramaran.ens.fr] has joined #Tunes
-:- SignOff Fare: #TUNES (Ping timeout for Fare[quatramaran.ens.fr])
-:- Fare [rideau@quatramaran.ens.fr] has joined #Tunes
-:- SignOff Fare: #TUNES (Read error to Fare[quatramaran.ens.fr]: Connection reset by peer)
-:- Fare [rideau@quatramaran.ens.fr] has joined #Tunes
-:- SignOff Fare: #TUNES (Ping timeout for Fare[quatramaran.ens.fr])
-:- Fare [rideau@quatramaran.ens.fr] has joined #Tunes
<Fare> Grrr!
05:00am
<smoke> :)
-:- SignOff smoke: #TUNES (hier we go)
05:10am
-:- SignOff Fare: #TUNES (Connection reset by pear)
-:- Fare [rideau@quatramaran.ens.fr] has joined #Tunes
-:- hcf [nef@me-portland-us345.javanet.com] has joined #tunes
-:- water [water@ppp-tnt-152.tscnet.net] has joined #tunes
-:- water [water@ppp-tnt-152.tscnet.net] has left #tunes []
-:- SignOff s_r: #TUNES (Ping timeout for s_r[phila-dialup118.nni.com])
-:- SignOff Fare: #TUNES (Ping timeout for Fare[quatramaran.ens.fr])
-:- Fare [rideau@quatramaran.ens.fr] has joined #Tunes
-:- tmf [s720@pissemaur.ii.uib.no] has joined #tunes
<tmf> hi
<Fare> lo
<tmf> I just had a representation of TOOL to a prof. at my uni :|
<hcf> Fare: got some urls for u to check out
<hcf> http://www.cs.chalmers.se/~andrew/cbneed-amb-abstract.html, http://cm.bell-labs.com/who/wadler/topics/monads.html, http://www.di.unito.it/~lambda/, http://www-sop.inria.fr/meije/personnel/Davide.Sangiorgi/AB/AB.html
<hcf> ftp.cis.upenn.edu/pub/papers/miller/mscs92.dvi.Z, ftp.cis.upenn.edu/pub/papers/miller/jlc91.dvi.Z, http://www.cs.vu.nl/pub/papers/theory/ABSTRACTS
<tmf> He recognized the fact that it worked and was elegant, but didn't quite see the point of it all.
<tmf> "what's the point of making things simple, when so much effort has been put into making them complicated all ready ?
<tmf> But I'm possibly getting 5 points ( half a semester ) for it if I manage to textually explain what I've done :) gee
<tmf> ( I guess that summs up my days as a student )
08:40am
<Fare> hcf: thanks
<Fare> tmf: who "He"?
<tmf> A lecturer in formal languages ( among other things )
<Fare> tmf: I got the impression that TOOL was essentially a string-rewriting system.
<Fare> is it that?
<tmf> yes, TML is a rewriting thingie, and TOOL is just an extention of that ( terms like 'object' defined )
08:50am
<Fare> why have string-rewriting instead of graph-rewriting?
<Fare> graph-rewrting is more general, more efficient to implement, conceptually simpler, etc
<Fare> and it's well-known already, and everyone understands it
<tmf> I never explisitly wanted to use string rewriting, it just what people like to call it
<Fare> but it *is* string-rewriting, isn't it?
<tmf> it is, but it's possible to have strings produce ( in their rewriting ) rules for further rewriting of things.
<tmf> I guess it could be used as a graph rewriting language as well
<Fare> sure, so it's a higher-order, maybe even reflective, string-rewrite system.
<Fare> but still, it's a string-rewrite system, which is unpractical.
<tmf> Definatly reflective. But TOOL is more an object oritented language, then an rewriting system. In it's usage anyway
<Fare> sure.
<tmf> Why are string rewriting unpractical?
<Fare> because it's unadapted to either human conceptual or machine implementational efficiencies.
<tmf> ok?
<tmf> what does that mean. What problems would that impose on the language?
09:00am
<Fare> difficulty to express even simple data structures, that you have to wrap into strings
<tmf> ok, like multidimentional lists and so?
<Fare> even simple lists
<tmf> TML is a very list friendluy language :)
<Fare> simple list operations suddenly have cost proportional to the length of the list!
<tmf> yeah, I know that, and that's a problem. I cannot just jump to an element, but have to walk all the way.
<Fare> that's due to implementing structures as strings on a memory-cell based architecture
<Fare> believe me: move to graph-rewriting instead
09:10am
<tmf> how does graph-rewriting look?
<Fare> you rewrite graphs instead of strings
<Fare> it's still reflective
<Fare> see Meseguer's Maude system (w/o backtracking), or Kirchner's ELAN (w/ backtracking)
<Fare> (both of these are pattern-matching based rewrite systems, not unification-based rewrite-systems)
<Fare> On top of Maude, Meseguer has reimplemented the CafeOBJ object system
<tmf> how would a rewriting rule look in such a system?
<Fare> much the same as in TML;
<tmf> like "swap all M's with 'FOO' " ?
<Fare> choose a particular system and see its syntax
<tmf> ok
<Fare> tmf: yeah, but "M" and "FOO" will represent nodes in the graph, rather than letters in a string.
<Fare> (or substrings)
<Fare> at the source level, of course, ASCII strings are used, for compatibility with existing technology
09:20am
<tmf> can string rewriting be seen as a special case of graph rewriting where the graph has only on path ( like A-->B-->C-->D ) ?
<Fare> yup
<Fare> a string is a particular case of graph
<Fare> can be considered as a graph in at least two ways!
<Fare> s/!/:/
<tmf> how does one decide which path to take?
<Fare> I mean, in many ways.
<Fare> Actually, it's a list of characters, so can be constructed as nil and cons nodes, or as atom and append nodes
<Fare> (in the latter case, modulo associativity)
<Fare> and associativity is costly to decide
<tmf> modulo associativity?
<tmf> or even -- assosiativity ?
<Fare> associativity
<abi> associativity is costly to decide
<tmf> sure, but what is it?
<Fare> that is, (("A"."B")."C") eq ("A".("B"."C"))
<tmf> ok!
<Fare> abi: no, associativity of an internal binary operator @ is the property that forall a,b,c,  a@(b@c) == (a@b)@c
<Fare> abi: associativity of an internal binary operator @?
<abi> fare: i don't know
<Fare> abi: no, associativity is the property of an internal binary operator @ such that forall a,b,c,  a@(b@c) == (a@b)@c
<abi> okay, Fare.
09:30am
<tmf> what does it mean to rewrite something _modulo_ som axiom?
<Fare> it means that you consider that terms are equal if associativity says so, and that during matching, you may replace a term to be rewritten by any equal term
<Fare> another way to look at it is to add for every axiomatic equality two rewrite rules (one both way) so that any term may be rewritten in any "equal" term
<Fare> and that makes matching costly
<tmf> In TML you parse the text and replace each identifier with its according macro definition. Something can send or not send a message ( I guess the message is what the term is rewritten to ).
<tmf> The costly part is to find the matching definition for each identifier in the string.
<Fare> in graph rewriting, you parse once, and then you manipulate compiled graphs
09:40am
<tmf> I see how this is neccasary, but it may cause problems.
<Fare> what kind of problems?
<tmf> If I have a graph that evaluates into another graph, then it's fine. But I need thing to be able to evaluate into effects as well, and then I'm stuck with the possibility that I may need to recompile at evey step - which would make picking an element in a list a proportional to the length of the list. I guess.
<tmf> If '[ a <b> ]' is the form of a rewriting rule, then I can compile it prior to evaluation. But I need to be able to make rule of the form '[ a <[ b <c> ]> ]' and '[ a b ]' as well.
<tmf> When you know TML, you see that trying to compile TML code is pretty pointless.
-:- lar1 [lar1@2Cust81.tnt20.sfo3.da.uu.net] has joined #tunes
<lar1> Hey
09:50am
-:- jdl [jiml@ultra1.inconnect.com] has joined #tunes
<jdl> Hello, all
<lar1> hey jdl
* jdl/#tunes is just eavesdropping on the interesting TML convo
<lar1> Hmm, not much going on in here
<lar1> Whats TML?
<abi> TML is a rewriting thingie, and TOOL is just an extention of that ( terms like 'object' defined )
<lar1> like perl then?
<tmf> it burns down to TML, not beeing explisitly a rewriting thingie afterall
* lar1/#tunes is reading the irc log
<tmf> rewriting may be just a usefull thing to do - sometimes
<lar1> What would TML be useful... can't perl do sring manipulation and more?
10:00am
<jdl> tmf: Yes, what is TML's purpose?
<tmf> TML is supposed to be very simple, and thus reveal essential or significant properties. Wheter or not "simple" is a point at all is another question.
<tmf> It's used to implement TOOL - an oo language.
<jdl> tmf: So it's purely experimental?
<tmf> It has prooved very useful when designing TOOL. I meant it as something more than just and experiment.
<jdl> tmf: Have you published any docs?
<lar1> What can TOOL do that other languages can't?
<tmf> jdl, no
<tmf> lar1, it's fully reflective - reflection is not explisitly implemented, it is a bieffect - from good design :)
<lar1> Ok
<tmf> lar1, besides that, it's comparable to any OO language.
<jdl> tmf: How is it reflective?
<tmf> jdl, everything is objects, and what you can do to objects ( like modifying them ) can be done to everything.
<jdl> tmf: So classes & methods are objects?
<tmf> The reflective TOOL framework consists of three objects: 'object', 'references' and 'methods'
<jdl> tmf: Okay, so if I created a new method, what would I use?  Source code?
<tmf> jdl, it's a prototype based model, there is no distinction between objects and classes.
<jdl> olk
<jdl> oi
<jdl> argh
<jdl> o k
<tmf> you would create a new derivative of the method object, and specify its parameter and result ( method is a parametric "class" )
10:10am
<jdl> tmf: parameter and result type, right?  What about the method body?
<tmf> the implementation of the method is in 'action', a reference to the effect of the method. References are pretty general - not just identifiers or paths.
<jdl> tmf: Yes, but can you create new effects?  If so, how?
<tmf> jdl, effects are just things that alter the current context.
<jdl> tmf: Okay, I'm getting confused.  In a traditional OO language like Java, a method contains source code.  My question is, when you create a new method, do you provide source code or something else to determine the method's behaviour?
<jdl> (when you create a new method *in TML*)
<tmf> you provide source code, and put it in 'action'.
<jdl> tmf: ok.  Is TOOL interpreted or compiled?
<tmf> ( TML has only "terms" or macros )
<jdl> tmf: right, I meant TOOL, not TML
<tmf> it's interpreted with the TML interpreter. TOOL is basically TML with some useful macros predefined.
<jdl> Interesting... what does the syntax look like (for TOOL)?
<tmf> I'm rigging a little example now.
10:20am
<jdl> ok
<tmf> --------------------------------------------------
<tmf> ( method ( new: (:<some-method>:) !)
<tmf> 
<tmf>   ( ( specify: (:<result>:) !)
<tmf>     ( is: (:<some-type>:) !)
<tmf>   .
<tmf>   ( action (:<
<tmf>     METHOD SOURCE HERE  
<tmf>   >:) .)
<tmf> .)
<tmf> --------------------------------------------------
<tmf> ( just read the words, ignore the parenteses )
<hcf> proglang syntax as emoticons ;)
<jdl> :)
<tmf> emoticons?
<lar1> Its like  is x = to result?  Then x = some new value?  Reminds me of BASIC
<jdl> smilies everywhere :)
<lar1> Note the fact there are no frowns :)
<tmf> I guess it's kind of readalbe then :)
<lar1> Its plenty readable
<jdl> tmf: And this is all transformed to TML's core constructs by the TML interpreter?
<tmf> actually it is TML core constructs. Seen differently, most of the code above is translated into TOOL code on evaluation:)
-:- SignOff hcf: #TUNES (Ping timeout for hcf[me-portland-us345.javanet.com])
<jdl> tmf: Oh, okay.  Moving from TOOL to TML, what sort of core constructs do you provide?
<tmf> macro defintions -  [ m  def ]
<tmf> contexts - ( ... )
<tmf> and
<tmf> messages - <...> , which are sent to the surounding context.
<jdl> tmf: Any computation or IO constructs?
10:30am
<tmf> no, for example IO is done transparently, where macro defines are regarded as write and invokes are reads.
<jdl> I guess what I'm wondering is how TOOL does anything, since it's interpreted by the TML interpreter.
<tmf> It's rather nice, because you can have objects living on permanenty on disk.
<lar1> Who wants to get down to basics with me... _exactly_ how does a no-kernel os work?  It makes no sense to me
<tmf> eventually, TOOL objects will be able to translate themselves ( instead of just to the basic TML terms ) to machine code.
<jdl> (For example, how would "Hello World" be displayed on the screen?)
<jdl> tmf: Ok.  So at this time, TOOL can't compute or do IO, but it will once it can be transformed into some machine's code?
<jdl> (I'm not criticizing, just trying to understand)
<tmf> techincally, you need to hack things for doing this into the TML interpreter, but once you have the option to handle and produce binary objects, these "hacks" can be written in TOOL, and used transparently.
<jdl> tmf: I think I understand.. what sort of first-order types have you defined?
<tmf> ? like integers and strings and so?
<jdl> yes
<tmf> non :)
<jdl> :)  So those are reflectively provided by the programmer by using TML?
10:40am
<tmf> They're not required yet - TOOL is not implemented using for-loops and if-statements.
<tmf> it's a no-data language :)
<jdl> ? I'm talking about doing something like: "public int cube(int x) { return x * x * x; }
<jdl> I.e., computations using variables, which presumably come in a variety of types
<tmf> Eventually, there will be a 'number' or 'int' type in TOOL.
<jdl> Okay, that's what I was asking... what sorts of basic data types you were planning on providing
<tmf> TOOL has implisit strong typing
<tmf> ah!
-:- lar1 [lar1@2Cust81.tnt20.sfo3.da.uu.net] has left #tunes []
<tmf> Well, since there will be no destincion between primitive or basic data types or those built by the user, I don't need to decide this.
<tmf> But I guess there will be numbers, and lists and trees, and all the data-structure stuff.
<jdl> tmf: So any data type can be (a)defined by the user  or (b)reflectively defined in the program  or (c)both?
<tmf> A
<tmf> B
<tmf> C v
<jdl> :)
<tmf> I'm not sure what you mean by reflectively defined?
<jdl> Meaning you can add new data types at runtime within your program
<tmf> That would essentially be possible.
<jdl> Cool  :)
<jdl> That's pretty much all I was wondering about... thanks for explaining
10:50am
<tmf> happy to!
-:- SignOff tmf: #TUNES (Leaving)
11:00am
* jdl/#tunes is away (beep me if you want me -- ^G)
* jdl/#tunes is back
-:- SignOff jdl: #TUNES (Leaving)
* Fare/#Tunes is back
-:- SlyP3 [mark@Ndialup153.slkc.uswest.net] has joined #Tunes
-:- SlyP3 [mark@Ndialup153.slkc.uswest.net] has left #Tunes []
-:- jdl [jiml@ultra1.inconnect.com] has joined #tunes
<jdl> Fare, you here?
-:- binEng [Anders@dialup47-1-51.swipnet.se] has joined #tunes
<jdl> Hey, bin
01:00pm
<binEng> hi!
<jdl> How goes the game?
<binEng> um well, crawling
<jdl> :)
<binEng> yah. Lots to attend to, in spite of the vacation
<binEng> I've been trying to read the Arrows paper, among other things
<jdl> As always... me too.  Slow going.
<binEng> Do you know what the latest version is?
<jdl> I printed out v8 a couple of days ago.  I'm not sure what that means since Brian recently said he hates version numbers  ;)
<binEng> :) I think I've got that version too. Just checking.
<binEng> FYI, we had a most beautiful sunset here just minutes ago. So you know what awaits you ;)
<jdl> Which time zone are you in?  It's early afternoon here!
-:- SignOff binEng: #TUNES (Ping timeout for binEng[dialup47-1-51.swipnet.se])
-:- SignOff jdl: #TUNES (Leaving)
01:10pm
-:- s_r [s_rr@phila-dialup579.nni.com] has joined #tunes
-:- binEng [Anders@dialup44-2-54.swipnet.se] has joined #tunes
<binEng> If you said anything, I missed it...
* binEng/#tunes looks around... where did Jim go??
<s_r> bin do you know anything about segments and paging?
<binEng> probably not enough to be able to teach you anything.
<s_r> does BSD use flat memory model (no segments just one big one for the entire 4gb address space)
<s_r> or multiple segments?
<binEng> I'll check with my book(s)
<s_r> i have found that in using GDB i see addresses that are 32-bit (not offset) which are references to something in virtual address space
<s_r> that are not just offsets into segments
<s_r> and i think they should be referencing offsets instead
<s_r> if they use segments
<binEng> sorry, no idea
<binEng> unless...
01:20pm
* binEng/#tunes is checking another book
<binEng> Sorry, I can't say... It says here it's based on Mach
<s_r> can you tell me what Mach uses?
-:- change_m1 [user6187@user-38lcd3n.dialup.mindspring.com] has joined #Tunes
-:- change_m1 [user6187@user-38lcd3n.dialup.mindspring.com] has left #Tunes []
<binEng> one sec
<binEng> uh
<binEng> "It lacks mechanisms for memory sharing other than read-only sharing of text segments." Means it uses segments, no?
01:30pm
<binEng> This seems to be a nice book. I should read it sometime
<s_r> hmmm it just doesn't make sense
<s_r> bin you just buy books and never read them?
<binEng> heh
<binEng> well, I intended to read it. And I will :)
<binEng> but there's much to do
<s_r> Unix Internals: the new frontiers
<s_r> 4.4BSD the design and implementation
<s_r> Linux Kernel Internals
<s_r> Unix Internals: A practical approach
<s_r> those are some books you should get
<binEng> I was looking in the first of them...
<binEng> otherwise, I'm not *that* much into UNIX. I'm designing my own OS, ya know :)
<s_r> i hate it when i don't understand something
<binEng> well, who can't agree on that.
01:40pm
<s_r> doesn't it annoy you when there's a computer concept that you just can't find enough information about?
<binEng> indeed. But there's always much info I already have, that I need to plow throu... so generally I survive :)
01:50pm
<Fare> brb
-:- SignOff Fare: #TUNES (Connection reset by pear)
-:- Fare [rideau@quatramaran.ens.fr] has joined #Tunes
-:- _QZ [brand@p0wer.qzx.com] has joined #tunes
<Fare> gakuk!
<binEng> gakuk
<binEng> to both of you :)
<s_r> fare do you know how segments and paging work in OSes
<s_r> like, do they use separate segments for the kernel address space and the user address space
02:20pm
<_QZ> no
<s_r> i have been using gdb on a BSD kernel
<s_r> and i see that most memory references are to
-:- Limbosuave [email@1Cust80.tnt1.covina.ca.da.uu.net] has joined #Tunes
<s_r> things like 0xc*******
-:- Limbosuave [email@1Cust80.tnt1.covina.ca.da.uu.net] has left #Tunes []
<s_r> it seems to me that it is referencing in a flat memory space
<s_r> because those CAN'T be offsets into a segment
<s_r> what is happening? i don't understand
<_QZ> i has mapped whatever u are accessing into the last gig of memory
<_QZ> it has
<s_r> is there a segment in the last gig of memory?
<_QZ> was it u i just went over this with last night or the night before?
<s_r> i don't remember talking about this
<_QZ> go read the tunes log for the past few nights and search for everything that i said
<s_r> ok but it doesn't say anything about how the kernel has its own segment
<s_r> does the kernel have its own segment
<s_r> onto the top of teh address space
<s_r> like 0xc0000000 and up
<s_r> or does it just use all 4gb as one segment
02:30pm
<Fare> s_r: yes
<s_r> yes to what
<Fare> s_r: the initial question
<Fare> s_r: answer: modern OSes mostly do NOT use paging
<s_r> if it has its own segment then shouldn't the addresses that came out of gdb when looking at /kernel have been offsets
<s_r> into that segment
<Fare> s_r: they use a flat address space w/ paging, and part of it is reserved to system addresses, part of it to user addresses
<s_r> instead, I saw linear addresses like 0xc000a01b
<Fare> so 0xC0000000 and upwards is the typically kernel, while 0x00000000 to 0xBFFFFFFF is the userland
<s_r> i know that
<s_r> but is there a separate segment for 0xc0000000 andup
<Fare> so wtf are you asking?
<Fare> nope
<s_r> so that every address that the kernel went through would be an offset into that segment
<Fare> all segments map same linear space, but user segments stop at 3GB instead of 4GB
<s_r> ok thanks fare
<Fare> (early linuxen had kernel segment begin a 0xC0000000 which was a problem)
<s_r> that's what i wanted to know
<Fare> VSTa has a 2GB/2GB separation
02:50pm
<s_r> so the kernel accessing the kernel segment
<s_r> would have to provide an offset into the segment?
<s_r> not just a linear address?
* s_r/#tunes is using VSTa so i want to be sure
<s_r> on anotehr computer
-:- AlonzoTG [Alonzo@209-122-253-149.s149.tnt8.lnh.md.dialup.rcn.com] has joined #tunes
<Fare> s_r: 1) why care? 2) if you do, why not just UTSL?
<s_r> utsl?
<Fare> utsl!
<abi> well, utsl is the acronym of "Use The Source, Luke!"
-:- Limbosuave [email@1Cust223.tnt2.covina.ca.da.uu.net] has joined #Tunes
-:- Limbosuave [email@1Cust223.tnt2.covina.ca.da.uu.net] has left #Tunes []
03:00pm
-:- tcn [tcn@cci-209150250082.clarityconnect.net] has joined #tunes
<tcn>  'sup fellas?
<tcn> hmm.. exciting
<binEng> hi tcn
<tcn> life
<tcn> !
<tcn> well, I was just stopping in
<tcn> gotta
<tcn> gotta go
* binEng/#tunes waves bye
-:- SignOff tcn: #TUNES (39n~~!1~11~~~~~~~~~~~~~~~.`.~.`~~~>~~~~~~~)
<Fare> tee-see-en!
<Fare> oops. gone.
03:10pm
-:- SignOff AlonzoTG: #TUNES (Have Nice Day :))
<s_r> movl	__ucodesel,%eax
<s_r> movl    __ucodesel, %eax
<s_r> in AT&T syntax, does that move the contents of __ucodesel into EAX or the address of that into EAX?
<Fare> s_r: 1) read the assembly HOWTO 2) try it out
<Fare> 3) see the output by as -a 4) compare the binary code produced to your spec sheet
<s_r> don't you have a quick answer for me :P
<Fare> 5) see what kind of code gcc produces
<Fare> these were quick meta-answers
<Fare> I wrote the HOWTO not to have to answer such questions anymore.
<s_r> where's the assembly HOWTO?
<Fare> http://metalab.unc.edu/LDP/HOWTO/Assembly-HOWTO.html
<Fare> latestest is in .sgml at http://tunes.org/~fare/files/asm/
<s_r> WOOHOO
<s_r> Fare's assembly HOWTO _actually_ helped
<s_r> heh who would have thought!
* Fare/#Tunes hits s_r with a clue stick
-:- mode/#tunes [+o Fare] by ChanServ
<s_r> heeh
-:- s_r was kicked off #tunes by Fare (LART!)
-:- s_r [s_rr@phila-dialup579.nni.com] has joined #tunes
<Fare> Ah, feels better!
<s_r> Fare you don't like people making fun of your HOWTO
<Fare> Fun is ok. It's just that my sense of humor is biased.
<binEng> I'm off.
-:- SignOff binEng: #TUNES (whee!)
03:40pm
-:- Fare was kicked off #tunes by Fare (You susceptible git!)
-:- Fare [rideau@quatramaran.ens.fr] has joined #Tunes
-:- mode/#tunes [+o Fare] by ChanServ
-:- AlonzoTG [Alonzo@216-164-230-144.s144.tnt7.lnh.md.dialup.rcn.com] has joined #tunes
-:- water [water@ppp-tnt-213.tscnet.net] has joined #tunes
<water> 'lo
<Fare> hi
<water> what's news?
<AlonzoTG> om
<Fare> mu
* AlonzoTG/#tunes needs to learn how to pogrom linuks.. =\
<water> "needs"?
<AlonzoTG> yes, I am going to write my prototype under linux
<AlonzoTG> basically ask linux if my OS is possible...
<water> i forget which prototype is yours?
<water> ask linux?
<water> you mean linus torvalds?
<AlonzoTG> I am going to Ask LINUX
<water> oh
<water> n/m
<AlonzoTG> wheather my OS is possible..
<AlonzoTG> I don't want to bug linus...
<water> you think they'll have the answer?
<AlonzoTG> Linux is a very powerful os.
<AlonzoTG> if my prototype can't run in it then it can't run at all. =\
<water> not true
<water> running ~= emulating
04:40pm
<water> you can often emulate without running
<AlonzoTG> either or.
<AlonzoTG> Basically if my prototype that emulates the core features of my OS can't run under linux then it can't run..
<water> well, that's an easy question
<water> you can always emulate hardware
<water> and then debug your os on that hardware
-:- _water [water@ppp-ip107-b42.tscnet.net] has joined #tunes
<_water> sorry i was cut off
-:- SignOff water: #TUNES (Ping timeout for water[ppp-tnt-213.tscnet.net])
<Fare> what does DTD stand for, already?
-:- _water is now known as water
<water> hell if i know
<Fare> vera told me
<water> what is it?
-:- billyboof [hatefull@nrwc-sh5-port54.snet.net] has joined #tunes
<billyboof> hello
04:50pm
<water> hi bill
<billyboof> hows it going water
<water> i'm ok
<billyboof> cool... did you see my os spec?
<water> trying to make my entire arrow paper obsolete
<water> yeah
<billyboof> was it .txt or .htm?
<water> umm. html i think
<Fare> vera is M-x info g vera
<water> fare: huh?
<billyboof> Fare: huh?
<Fare> abi: no, vera is C-h i g v e r a RET
<abi> okay, Fare.
<water> oh
<water> vera?
<abi> vera is, like, C-h i g v e r a RET
<Fare> you may need to apt-get install vera, first
<water> fare: english please
<abi> english is no more the right language
<water> abi: no one knows that more than i
<Fare> abi: VERA is also Virtual Entity of Relevant Acronyms
<abi> okay, Fare.
<Fare> vera?
<abi> vera is C-h i g v e r a RET or Virtual Entity of Relevant Acronyms
<water> ahh
<water> so what did vera tell you about DTD?
<Fare> Document Type Definition
<water> ok
<billyboof> Fare: did you see the os spec i wrote?
<Fare> depends; which was it?
<billyboof> http://pages.cthome.net/consultant/os_spec.htm <updated a few times>
<water> oh yeah. now i remember
<water> how is this going to beat M$?
<water> open-source?
<billyboof> not sure yet... probably
<water> you should concentrate on cloning BeOS, then
<water> and using open-source to improve development
* billyboof/#tunes has never seen BeOS
<water> oh my god!
<water> you're designing a multi-threading SMP os, and you haven't looked at the best design out there?
<Fare> bb: I read it yesterday, and wasn't impressed
<water> ok, maybe just close to the best
05:00pm
<billyboof> what didn't you like about it?  <other than the things that i forgot to put in yesterday that are in now>
<AlonzoTG> b00f!!!!
* AlonzoTG/#tunes wants to reed speck
<Fare> bb: looked like yet another list-of-random-features.
<AlonzoTG> Warez speck?
<billyboof> you mean like not enough about implementations?
<Fare> no, not enough about a programming model
<water> i agree
<AlonzoTG> Show Me The Spec!
<Fare> it's a random list of features, come out of nowhere.
<AlonzoTG> Show Me The Spec!
<AlonzoTG> Show Me The Spec!
<AlonzoTG> Show Me The Spec!
<water> you're os should provide a compiler-writer a clear path for writing programs
<Fare> not that it be bad
<water> http://pages.cthome.net/consultant/os_spec.htm
<AlonzoTG> thanks!
<Fare> it's even got interesting ideas; but it comes from nowhere and goes nowhere.
<Fare> there are implementation details that shouldn't be user-visible
<billyboof> yeah.... i get what you're saying (i think) ... i guess this is written more for people i'm going to be working with than anyone else
<Fare> and it is not obvious how the result would ultimately differ from a souped-down UNIX. Or perhaps from VSTa or such.
<Fare> if it's a personal draft, then ok. But it's not a OS spec.
<billyboof> yeah... i guess that's more of what it is... just to make it so everything i think of isn't just in my mind--so i don't forget and so people i'm working with can know
<AlonzoTG> =\
<AlonzoTG> I like the effort
<AlonzoTG> But it seems to be too rooted in Eunichs...
* billyboof/#tunes doesn't even use linux
<AlonzoTG> And it doesn't answer the most important question...
<AlonzoTG> HOW????
05:10pm
<AlonzoTG> In my spec I list functions, Tell a little about how each one does what it does, why it is there, what data structures it uses, which files it goes in....
<AlonzoTG> Yours seems to lack that.
<billyboof> yeah... you're right... i started writing this yesterday... haven't gotten there yet.. i guess i was looking for people opinions on the ideas....
<water> you should look around at the existing os's features
<water> i mean, really closely
<water> get the linux source code, or *bsd
<AlonzoTG> I am very glad to see that you even tried to write a spec. =)
<water> yes, that is a very good way to make your ideas clear. :)
05:20pm
<Fare> bb: yeah, that's a first attempt. Don't get discouraged. Rather, learn a lot, and a few weeks/months from now, begin again, and compare!
<Fare> bb: don't forget: if you need a new OS model, it's to change the way people program. So the question is: what programming model do you want?
<billyboof> so would what i was saying about file io be an example of how i almost defined that?  like how i want file io to be mem mapped as much as possible, and as little as possible procedural?
<Fare> that'd be a *little* progress, but not much.
<Fare> Unix has mmap(). How do you propose things be better?
05:40pm
<billyboof> i was just decided on it for ease of use, and to make it so that disk caching was done with virtual memory, instead of having seperate buffers set aside.
-:- NetSplit: king.openprojects.net split from sterling.openprojects.net [05:41pm]
-:- BitchX+Deb1an: Press Ctrl-F to see who left Ctrl-E to change to [king.openprojects.net]
<Fare> again, how does your design improve upon UNIX mmap()? 
-:- Netjoined: king.openprojects.net sterling.openprojects.net
-:- _QZ [brand@p0wer.qzx.com] has joined #tunes
<billyboof> i don't know exactly how mmap() works, so i can't really say.. i just know that to me it seems better than procedural file reads/writes
<Fare> questions that arise with mmap() are or consistency management: consistency between core memory, disk caches, disk, swap, multiple processes, multiple CPUs, etc, etc.
<Fare> what is the consistency to be? Who manages it?
<Fare> what are detected events?
-:- AlonzoTG [Alonzo@216-164-230-144.s144.tnt7.lnh.md.dialup.rcn.com] has joined #tunes
<Fare> what operations are atomic?
<AlonzoTG> B00F!!!!
<Fare> what operations involve locks?
<Fare> these are the important questions
<Fare> depending on the answers, you may get widely different systems
05:50pm
<AlonzoTG> hehe
<AlonzoTG> y0
<AlonzoTG> I am off to go play this dumb old game...
<AlonzoTG> only 4kilobytes!
<AlonzoTG> =P
-:- SignOff AlonzoTG: #TUNES (Have Nice Day :))
06:30pm
-:- _water [water@ppp-tnt-139.tscnet.net] has joined #tunes
-:- SignOff water: #TUNES (Ping timeout for water[ppp-ip107-b42.tscnet.net])
-:- _water is now known as water
<s_r> water have you looked at BeOS?
<water> yeah. defnitely
<water> why?
<s_r> i think BeOS is cool
<s_r> except for being bullishly closed source
<water> ok. i agree
* s_r/#tunes listens to BeOS_theme_song.mp3
<water> it should be open, too, in my opinion
<water> there is one?
<s_r> and they call themselves "hackers"
<s_r> shameful.
<s_r> water what's your favorite OS for its design
<water> hmm
<s_r> have you looked at VSTa
<water> nokernel, i guess
<water> not closely
<water> i'd like language=os
<s_r> water likes "Arrow OS" ;)
<water> with no vm
<s_r> but vm is neccessary
<water> no!
<water> only for languages that suck
<water> :)
<s_r> how can an OS get by without virtual memory?
<water> lots of ram :)
<water> hehe
07:00pm
<water> or you have high-level programming with a persistent store
<s_r> how?
<water> well, usually the vm handles that
<water> take a look at the review to-do list
<water> it has pointers to "univ of st andrews" persistence group
<s_r> what do you mean by persistant store, exactly?
<water> sort of the opposite of virtual memory
<s_r> hmm i want to find out what you mean
<water> the ram acts as a cache for objects on the hard drive
<s_r> once you get many programs, they are bound to exhaust physical ram
<water> well, it's not for programs
<water> per se
<water> it's for objects
<water> or functions or whatever
<s_r> i like the idea of a persistant machine state
<s_r> where the OS saves all the information being transacted at one state before rebooting or shutting down
<s_r> and then restores them exactly and continues after being booted the next time
<water> besides, part of the vm code handles the caching, so that extraneous results are shifted to the hard drive
<water> sr: os/2 has that
<s_r> is os/2 somewhat like BeOS
<water> well, it has the same overall user-interface
<water> the workings are quite different
<s_r> BeOS is multithreaded
<water> yes, os/2 completely lacks thread support
<s_r> water do you think multithreading is good?
<water> os/2 also doesn't have posix interface
<water> multi-threading is much better than multi-processing
<water> i mean, multiple processes
<water> but not as good imho as single-process model with threads and co-routines
07:10pm
<s_r> water have you considered developing your own OS?
-:- AlonzoTG [Alonzo@209-122-199-171.s425.tnt4.lnh.md.dialup.rcn.com] has joined #tunes
<water> yes, in fact i had one patched together in college
<AlonzoTG> B00F!!!!
<water> it was a 32-bit pmode os without a filesystem yet, based on my own hacking
<water> i cracked the euphoria DOS interpreter
<water> and merged it with DOS
<s_r> .dcc send water BeOS_theme_song.mp3
<water> it had list-processing and was fairly well optimized
<water> it could even spawn new vm's
<s_r> how does the concept of a vm work?
-:- _water [water@ppp-tnt-20.tscnet.net] has joined #tunes
<AlonzoTG> ekselent! =)
<_water> it also had a gui based on sexp-expressions
-:- SignOff billyboof: #TUNES (hatefull@antisocial.com)
<_water> did i miss anything?
-:- lar1 [lar1@1Cust170.tnt20.sfo3.da.uu.net] has joined #tunes
<lar1> hey
<abi> hoy, lar1
<_water> hi lar
-:- SignOff water: #TUNES (Ping timeout for water[ppp-tnt-139.tscnet.net])
<lar1> _QZ:  whats the liner address for the middle of page 0?
<_water> s_r: what do ya think of my old design?
<s_r> lar1 having problems with paging
07:20pm
<s_r> water i haven't seen it
<lar1> s_r: coding an os, and am getting up to speed on paging
<_water> well, i just described it
<_water> it's pretty simple
<_water> just a pmode dos vm
<_water> with list-processing and a gui
<s_r> lar1 YAOS?
<s_r> yaos?
<abi> yaos is truly the acronym to describe it
<lar1> s_r:  huh?  Its called raven currently but it desprately needs a name change
<_water> yaos is also "yet another operating system"
<abi> okay, _water.
-:- _water is now known as water
<AlonzoTG> =)
<lar1> Hehe alaonzo
<s_r> lar1 have you got any work done on it?
<AlonzoTG> Lar: Tell me about yer OS =)
<lar1> s_r: Some very baic stuff... nothing to brag about
<s_r> lar1 having problems developing it?
<lar1> s_r:  Well, I tryed to jump in without recearch... I know how to code just fine, but I just recenlty learned x86 pmode and now paging
<lar1> AlonzoTG:  Umm, Its going to be a simple os for the x86... I don't have too much of a goal but it needs to be fast and rock solid
<_QZ> any1 know howto delete items from the win98 taskbar from the filesystem?
<s_r> qz how's work on BriX?
<AlonzoTG> Okay. 
* lar1/#tunes just got the new blink 182 cd
<AlonzoTG> Will it support multiprocessor configurations?
<water> where are they in the filesystem?
<_QZ> like desktop icons are stored in /windows/desktop
<water> i thought they were in the registry
<_QZ> damn
<lar1> AlonzoTG:  As soon as I get a multiprosser machine to devlop on ;)
<water> actually, i KNOW they're in the registry
<AlonzoTG> 'doze 78 CRAWLS on my father's machine...
<water> qz: got norton utils?
<_QZ> i need to remove them from another machine via file sharing :)
<AlonzoTG> It was okay for a whilest but now it broke a leg and crawls...
<water> qz: n/m then
<lar1> _QZ:  whats the liner address for the middle of page 0?
<_QZ> water: is there anyway to reboot the machine thru file sharing? :)
<s_r> jmp F000:FFF0
<water> qz: i don't support that kind of thing any more
<s_r> but you have to do that with files
<_QZ> i have full r/w access to the hdd but a program is running that im trying to delete so i cant remove it
<s_r> hehe
<s_r> maybe you could represent the memory location with a file
<water> qz: there are ways around that, but they're convoluted
<water> qz: n/m
<s_r> QZ how's work on BriX?
07:30pm
<_QZ> hmm know of any nukes that work on win98
-:- SignOff water: #TUNES (Ping timeout for water[ppp-tnt-20.tscnet.net])
<_QZ> that could reboot the machine?
<lar1> _QZ:  I know there is a way to do what you want... I just don't rember how to do it
07:40pm
-:- SignOff AlonzoTG: #TUNES (Have Nice Day :))
<lar1> Anyone able to answer simple paging question?
<_QZ> lar1: goto #asm on efnet
<lar1> _QZ:  Ok... thats were hcf told me to come here... BBL
-:- SignOff lar1: #TUNES (changing servers)
07:50pm
-:- s_r [s_rr@phila-dialup579.nni.com] has left #tunes []
-:- water [water@ppp-tnt-1.tscnet.net] has joined #tunes
-:- iepos [iepos@d6.t1-8.tecinfo.com] has joined #TUNES
<water> hi
<abi> what's up, water
<iepos> hi
<iepos> hmmmmmmmmm......... i think i finally understand that i don't understand the paradoxes... 8(
<water> what's on your mind, iepos?
<water> :)
<water> what paradoxes are those?
<iepos> i thought my logic was free of them... i thought throwing out excluded middle, double negation, and non-contraction and stuff would get rid of them...
<iepos> consider this statement: "if this statement is true, then the moon is made of green cheese".
<iepos> now, assume for a moment that it is true...
-:- SignOff _QZ: #TUNES (Ping timeout for _QZ[p0wer.qzx.com])
<water> it's reflective, though
<iepos> if so, then the moon is made of green cheese... 
<iepos> at least if you believe in the deductive theorum or whatever they call it
<water> but it's a statement that refers to itself
<iepos> we'll call that weird statement Z
<iepos> and the "moon is made of green cheese" X
<iepos> so we beleive Z -> X we think
<water> sigh...
<iepos> but Z -> X is basically the same as just Z if you substitute the "this statement" for the whole thing...
<iepos> so we can prove X... anything...
<water> yes... and?
<iepos> i thought that reasoning was okay...
<iepos> self-referencing statements make odd things happen maybe...
<water> what do you mean?
<water> oh
<water> hehe
<water> reflection is rough on your brain
<water> :)
<iepos> but if you have combinators and implication, self-referencing statements can't be avoided i don't think...
<iepos> 8(
<water> hehe
<water> but you can restrict the logic from making contradictions
<iepos> sure... just get rid of all your inference rules and your axioms.
<iepos> :-)
<water> not all of them
<water> just some
<water> logicians in research do it all the time
<iepos> yeah..... have you heard of intuitionnistic logic? i don't understand... all i always hear is that it is "constructive"...
<iepos> do you know what it really is? if you do, i'd like to know...
<water> bleh, i've heard too much
<iepos> oh, you don't like it?
<water> it's a stack-based system of assumptions
<water> i hate stacks
<water> it's a personal thing
<water> anyway...
<water> every assumption pops a new predicate onto the stack
<water> you can push things from the stack, too
<water> but then you loose the results of that reasoning unless you reflect
08:20pm
<iepos> then when you pop it off, you get "A -> B" where A was your assumption you pushed and B was what you concluded, right?
<water> well, only if that's what you had beforehand
<iepos> well... what kind of logic do you like then?
<water> intuitionist logic is turing complete, but it's too simple
<water> umm
<water> sort of a reflective relativised arrow logic
<water> well, that's not my favorite, but it works
<water> it's also mine
<iepos> hmm...... so what are some axioms you like? or does that apply ... ?
<water> just stuff about arrow worlds
<water> combinations, reversals, identities
<iepos> you don't like anything about implication or quantification ?
-:- _QZ [brand@p0wer.qzx.com] has joined #tunes
<water> yes, but i don't like picking just one way to do them
<iepos> hmmmmmmmm.....
<water> there's sequent calcli
<water> a close second
<water> sequents are of course implications
<iepos> is the sequent calculi a logic?
<water> pretty much
<water> but calculi is a plural
<water> "there are many sequent calculi"
<iepos> oh
<water> it's the logic of making implications
<water> logics
<iepos> hmmmm....
<water> i don't play favorites as much as try to transcend
<iepos> well.....
<water> i have lots of papers about new ideas for quantification
<water> hey! wtf is up with the channel policy!
<water> ?
<water> FARE!
<water> TRIL!
<iepos> i'm still puzzled by that odd paradox .... i guess maybe the deductive theorum doesn't always work...
<iepos> channel policy?
* water/#tunes appeals to the mighty opers
<water> the topic is closed and empty
08:30pm
<water> the paradox indicates an incomplete logic
<water> in fact, it proves that the standard logic is mathematically incomplete
<iepos> well... this means something is wrong with the logic... because anything can be proved...
<water> right
<water> which is why logicians make new logics all the time
<water> they explore the landscape for better possibilities
<iepos> well... i'd be happy to find a pretty nice general one now that doesn't have inconsistencies ...
<water> why just one?
<iepos> well one is better than zero
<iepos> two might be even better
<water> you're results will be stuck in that single logic
<water> your
<iepos> well then after i look at one 
<iepos> i'll look at another
<water> well, i'll just address that with the Arrow system
<water> you just stick with your logics for now.
<iepos> heh heh...... aren't arrows just another way of representing information, though?
<iepos> do they really make logic any easier?
<water> what do you mean, "another" way?
<water> they're not the subject for logic!
<water> they're not logical!
<water> geez
<iepos> so what is the purpose in using them, then? for communication?
<water> another person who has no f***ing idea
<water> communication is always in a language
<water> arrows are not a language
<water> the arrow system has no implicit part
-:- billyboof [hatefull@nrwc-sh7-port15.snet.net] has joined #tunes
<water> except when viewed by a language
<water> hi bill
<iepos> so... what do you plan on using arrows for, then? if not for logic or communication... computation?
<water> in my terms, i generalize a language to an "ontology"
<water> they're a formalism for cybernetics
<iepos> so, what do you plan to do with them?  ( :-) )
-:- lar1 [lar1@1Cust170.tnt20.sfo3.da.uu.net] has joined #tunes
<lar1> Hio
<iepos> greetings lar1
<billyboof> hey water... whats up
<iepos> hmm... you say arrows are for modeling concepts... but why would you model concepts if you are not going to do logic on them or communicate them... ?
-:- SignOff water: #TUNES (Ping timeout for water[ppp-tnt-1.tscnet.net])
08:40pm
-:- water [water@ppp-tnt-149.tscnet.net] has joined #tunes
<_QZ> 3com sucks
<water> whew back
<water> qz: why?
<_QZ> they charge $60 for the program to upgrade my courier from 33.6 to v90
<water> oh
<lar1> That really sucks
<water> indeed
<_QZ> this modem was $400 when i had my buddy swipe it off the usr assembly line for me
<lar1> One hell of a modem
<_QZ> and a new v90 modem is just barely over $60 and they think i should pay to upgrade my $400 modem
<_QZ> lar1: it was one hell of a modem in 1996
<_QZ> but 3 years later the 31200 connection sucks
<lar1> 1996 $400 for a 33.6?  Seems pricy
<_QZ> its a courier V.everything and sold as a 28.8. i had to upgrade it to 33.6
<lar1> Ohh
-:- SignOff iepos: #TUNES (Leaving)
<_QZ> at the time it was the only modem that could connect to any other modem
<lar1> And probly one of the only flash modems
<_QZ> and it was used by all the ISPs for dialup connections
<_QZ> ya
<lar1> Why does this channel always fall deadly silent after 5 minutes?
<_QZ> cuz it sucks
<lar1> Hmmm, what does dir:10, page:10, offset:12 mean?
<lar1> when describing a linear address
-:- _water [water@ppp-tnt-88.tscnet.net] has joined #tunes
08:50pm
<lar1> My head is about to explode
<_QZ> the high 10bits are the entry in the page directory
<_water> lar1: because you always discuss dry stuff like code and hardware!!!
<_QZ> the middle 10 bits are the entry in the page table pointed to by the page dir entry
<_QZ> the bottom 12 bits are the offset into the page
-:- SignOff water: #TUNES (Ping timeout for water[ppp-tnt-149.tscnet.net])
<_water> blech
-:- _water is now known as water
<lar1> Thank you, thats all I needed
<lar1> water:  So what should I talk about?
<water> dunno. use your imagination
<water> what belongs on the tunes server for all posterity?
<lar1> Hehe, all this reflective stuff goes way over my head
<water> then learn or get out!
<lar1> Well, I'll learn when someone teaches me
<lar1> I'm only 14 sooo
<water> then you _are_ lazy
<water> pick up some books
<water> i taught myself programming for 15 years and still going
<water> :P
<lar1> I don't understand all the books...
<water> so pick a simple one
<lar1> Hey!  I taught myslef plenty
<water> buy an apple ][
<water> use hypercard on the mac
<lar1> I should get reflection for you and me
<water> huh?
<lar1> hypercard... oh my god... nightmare
<lar1> There should be a book, "Reflection for you and me"
<water> whatever
<_QZ> hahaha
<water> bwahahahaha
<water> the general public can't deal with reflection
<lar1> I was just joking
<water> yeah sure, whatever
<billyboof> water: how old are you?
<water> nice try
<water> 21
<lar1> I am serious... I was joking! hehe
<lar1> I tought myself x86 asm, thats gotta be good for something
<water> i bet you weren't joking about hypercard
<billyboof> so you're saying you started coding when you were 6?
<lar1> No hyper card I was dead seriout about
<water> billy: yes
<_QZ> man 3coms page is out of date. it says u will receive a free upgrade when v90 comes out
<lar1> I started coding around 6 or 7
<water> billy: apple basic, forth, asm
<lar1> Water: asm at 6?
<water> yeah
<lar1> Water:  for what chip???
<water> without a friggin manual, either.
* billyboof/#tunes started at 9 or 10..
<water> 65c02
<water> motorola
* billyboof/#tunes feels like a retard baby.. a bit slow
<water> the only apple ][ chip
<water> well, cpu anyway
<lar1> Without a manual? Howd that happen??
<water> i learned how to break program execution
<water> and trace the assembly
* lar1/#tunes feels that BASIC was the main language water was in when he was 6
<water> by poking bytes into the program using basic
<water> an older friend showed me how to write assembly
<water> apple ][ asm is extremely simple
09:00pm
<lar1> See, there was no one around me that knew asm when I was 6
<water> oh
<lar1> My dad, who was the computer god of the house, didn't even know basic
<water> hehe
<billyboof> my parents can just barely use windows
<lar1> I remember me telling mom what to type into the BASIC interperter.. I couln't type at the time ;)
<water> lol
<water> cool
<lar1> It was a lot of fun... I was always annoyed at the limits of BASIC, though
<water> of course
<water> my best programs were 3d wireframe renderers
<lar1> I wanted to make one of those menu programs and didn't know the shell command... you should have seen the crap I came up with to try to do that! heh
<water> and some games
<lar1> Yea
<lar1> I am just picking up 3d now
<lar1> I did some simple games though
<water> i spent most of my free time in 97 making graphics engines
<water> they were much better than quake's or unreal's
<lar1> So you were like 18
<water> 19, mostly
<lar1> Where did you learn about 3d?  Just books?
<water> well, i figured out the basics myself on the apple (about age 8)
<water> i just saw stuff in movies and tried to figure it out
<water> it took a while
<lar1> Umm, you just figured out how to texture map and rotate?
<water> well, no textures
<lar1> Oh wait
<lar1> thats what I was going to say
<water> but lots of rotation and eventually perspectives
<lar1> But 3d in like 4 colors is pretty simple
<water> hehe yeah
<water> i actually have some 3d games for my apple ][ emulator
<lar1> That you wrote?
<water> no
<water> i found them on the web
<lar1> oh
<lar1> When the apple ][e was out I was too young to understand it... I would have probly just slobbered all over it
<water> too bad
<water> it was a real hacker's computer
<water> this guy i know was on the nsa's 100 most wanted list
<lar1> tight
<water> his favorite computer to hack with was an apple ][e
09:10pm
<water> he's in the navy now to avoid prosecution
<water> a real shame
<lar1> What kind of stuff did he do to get on the nsa's list?
<lar1> With an apple ][ I don't think we are taling breaking encryption or anything ;)
<water> he hasn't told me everything, but it seems he's good at playing with bannk accounts
<lar1> Ah so
<water> ya know, low level simple stuff
<water> like network hubs
<lar1> Yea
<lar1> 300 baud modems with accustic couplers
* lar1/#tunes is sorry he missed the dawning of the computer era... about 10 years too late
<water> as long as the encoding is simple or published, you can manipulate the system at will
<lar1> Thats true
<water> his favorite game was "core wars"
<lar1> Core wars!!!!
<water> ever heard of it?
<lar1> Yea!
<water> no, not the one you play with the mouse
<lar1> I know
<water> well, what is it? you know?
<lar1> Yes, it is a game whos purpse to make yer opponent execute an invalid instructions
<lar1> like dat
<water> sort of
<lar1> Where they used to play with the real ferite core memory
<water> his version of the game was a bit different
<lar1> Not like pmars?
<water> well, there was this hackers' club, and they all had apple ]['s. their core was a hard drive that was mailed in a loop
<lar1> mailed?  Oy vey
<lar1> That could get expensive
<water> within 48 hours from the time of receipt, the player had to fill a certain amount of the drive with a specific code, his code.
<water> if he couldn't, he gained a point, and re-formatted the drive
<water> and blanked it
<water> then it was passed on
<lar1> What was the objective of the game?
<water> that's about it
<water> to not get points, of course, and to make stories
<water> i think the game was over when the drive became unusable
<lar1> Hehe
<water> :)
09:20pm
<lar1> Not quite the modern core war...
<lar1> I assume we are taking 10 mB drives
<water> they were probably only 1mb
<lar1> wow
<water> maybe 1.5
<lar1> heh, my first computer, the 286 had a 40 mb drive
<water> wow
<lar1> That was like... you'll never ever fill it
<water> my family's apple ][ came with tape recorders
<lar1> woah!  No floppy?
<water> i had a2/fs1 turbo
<lar1> :)
<water> sublogic flight sim
<water> M$'s flight sim buyout
<lar1> Hmm
<lar1> I leaned to type on an apple ][e
<water> cool
<lar1> The school was too cheap to buy new machines... I din
<lar1> t mind
<lar1> But recently the swapped them all out for new PIIs
<water> hehe
<lar1> You know its interesting... when you think orignal family computer... youthink apple ][e... _not_ 8088(xt)... but now you think x86 not apple...
* lar1/#tunes has an orignal IBM XT
<water> PowerPC's rock.
<water> x86's are dinosaurs
<lar1> but everyone uses x6
<lar1> I would love to learn to code asm for G3's
<water> so? everyone is dumb, too.
<water> they're subject to the marketplace
<lar1> Well if apple wasn't so screwed in the head we'd all be on macs now
<water> lar1: it's much simpler than for x86
<water> yeah
<water> lar1: you can get the processor tech man's from motorola and ibm's websites
<water> lar1: just like from intel
<lar1> IBM?  What do they do with the macs?
<water> ibm _makes_ powerpc chips
<lar1> Really?  I thought that was motorola's place
<water> they also did the main research for the copper laced chips
<water> it's a joint project
<water> they also make the Power chip, which is the powerpc's parent
<lar1> cool
<lar1> The new G3s are sooooo cool
09:30pm
<water> the g4's will be even better
<water> by a long shot
<lar1> Ohhhhh g4... *drool* *drool*
<water> hehe
<lar1> I almost fell off the chair when I read the g3 specs
<lar1> I might consider getting a g3 and put linux on it
<lar1> The thing is x86 has market domonation...
<water> it's a shame you can't run beos on a g3 motherboard
<lar1> Yah
<lar1> That would be tight
<lar1> It should be ported
<water> ported to what?
<lar1> g3
<water> it can't
<lar1> Why not?
<water> apple won't give be the specs
<lar1> Ohhh
<lar1> Stupid apple... shooting them selves inthe foot again!
<water> sorry, i watch computer news very closely
<water> i mean, the important stuff
<lar1> I read /. religiously now... jsut found out about it 5 months agao or so
<water> have you studied its design?
<water> i mean, how it works?
<lar1> the g3?  no
<water> no, beos
<water> ok, g3 too
<lar1> Oh, nope...
<water> you should really pick up those processor manuals
<lar1> So IBM's site?
<water> ibm or motorola
<water> i assume you have the x86 ones from intel
<lar1> Problem is I don't have a ppc... just an o-l-d performa 476 (68040)
<water> so? it's to learn from
<lar1> I got the intel cd... but I want the pentium manual in paper... I've been meaning to go to the site and get em
<water> if you want to understand chips better, compare two designs
<lar1> But learning for me is hard without accually being able to apply what I just read
<water> ok
<lar1> I'll still read the books
<lar1> My friend just got rid of 2 ppc... damnit... should have picked up one
<water> it's much easier to write fast code for the ppc
<water> i learned by writing those graphics engines
<lar1> Is programming for macos easier then windows?
<water> definitely
<lar1> Hmm, I have a old version of think c
<water> win32 is very obtuse and complicated
<lar1> Hehe, no kidding
09:40pm
<water> on the other hand, the macos has some inconsistencies between versions
<lar1> Uh oh... inconsistencies... I guess between 7.5 and 8.0 is biggest?
<water> well, more like between 7.0 and 8.0
<water> but yeah
<lar1> What ever happened to rapsody?
<water> rhapsody -> macos X
<water> it already exists!
<lar1> Ohhh
<lar1> So it was introduced in mac os 8?
<water> hmm? not really
<water> macos 8 just had more interface improvements and ppc code
<lar1> Ahh
<water> also some applescript stuff
<water> 8.5 is almost totally ppc
<lar1> Well, its been real nice chatting with you... but mom needs the connection and I don't have ipmasq setup ;)
<water> ok
<water> bye
-:- water [water@ppp-tnt-88.tscnet.net] has left #tunes []
-:- SignOff lar1: #TUNES (ircII2.8.2-EPIC3.004 --- Bloatware at its finest.)
-:- jdl [jiml@ultra1.inconnect.com] has joined #tunes
09:50pm
-:- SignOff jdl: #TUNES (Leaving)
-:- SignOff billyboof: #TUNES (hatefull@antisocial.com)
-:- Alaric [alaric@automan.ivision.co.uk] has joined #Tunes
-:- Alaric [alaric@automan.ivision.co.uk] has left #Tunes [Bye]
-:- NetSplit: sterling.openprojects.net split from koontz.openprojects.net [11:17pm]
-:- BitchX+Deb1an: Press Ctrl-F to see who left Ctrl-E to change to [sterling.openprojects.net]
!koontz.openprojects.net!! Remote CONNECT sterling.openprojects.net 8005 from lilo
-:- Netjoined: sterling.openprojects.net koontz.openprojects.net
-:- Fare [rideau@quatramaran.ens.fr] has joined #tunes
-:- ServerMode/#tunes [+o Fare] by sterling.openprojects.net
-lilo(lilo@varley.openprojects.net)- [GlobalNotice] Hi all.  Sorry about that brief service interruption, problems with a server clock.
-lilo(lilo@varley.openprojects.net)- [GlobalNotice] There may be a couple more quick problems, please bear with us
-:- NetSplit: sterling.openprojects.net split from koontz.openprojects.net [11:18pm]
-:- BitchX+Deb1an: Press Ctrl-F to see who left Ctrl-E to change to [sterling.openprojects.net]
!koontz.openprojects.net!! Remote CONNECT sterling.openprojects.net 8005 from lilo
-:- Netjoined: sterling.openprojects.net koontz.openprojects.net
-:- Fare [rideau@quatramaran.ens.fr] has joined #tunes
-:- ServerMode/#tunes [+o Fare] by sterling.openprojects.net
-lilo(lilo@varley.openprojects.net)- [GlobalNotice] I think that solves the problem.  Thanks all.
-lilo(lilo@varley.openprojects.net)- [GlobalNotice] Hi all.  Just to recap, we had a slight server clock problem which caused a couple of splits.  It's fixed now.  Thanks again for your patience.
-lilo(lilo@varley.openprojects.net)- [GlobalNotice] Please welcome santaclara.ca.us.opirc.nu (lewis.opirc.nu), courtesy Christoph Lameter and Captech.com!
-:- water [water@ppp-tnt-231.tscnet.net] has joined #tunes
[msg(TUNES)] newlog 1999.0623
IRC log ended Wed Jun 23 00:00:00 1999