IRC log started Tue Jun  6 00:00:01 2000
[msg(TUNES)] permlog 2000.0606
-:- SignOff hcf: #TUNES (Leaving)
-:- SignOff water: #TUNES (The Tao went that-a-way!)
-:- SignOff mark4: #TUNES (Ping timeout for mark4[purplecoder.com])
-:- mark4 [mark4@purplecoder.com] has joined #tunes
-:- NetSplit: varley.openprojects.net split from fastlane.openprojects.net [01:29am]
-:- BitchX+Deb1an: Press Ctrl-F to see who left Ctrl-E to change to [varley.openprojects.net]
-:- SignOff eihrul: #TUNES (Ping timeout for eihrul[usr5-ppp208.lvdi.net])
-:- eihrul [lee@usr5-ppp208.lvdi.net] has joined #tunes
-:- Netjoined: varley.openprojects.net fastlane.openprojects.net
-:- abi [abi@bespin.dhs.org] has joined #tunes
-:- water [water@tnt-10-53.tscnet.net] has joined #tunes
<water> re
* water/#tunes is looking at prograph for a bit
-:- SignOff abi: #TUNES (Ping timeout for abi[bespin.dhs.org])
-:- Closing Link: TUNES[bespin.dhs.org] by carter.openprojects.net (Ping timeout for TUNES[bespin.dhs.org])
-:- Connection closed from irc.us.openprojects.net: Success
-:- Use /Server to connect to a server
-:- Connecting to port 6667 of server irc.us.openprojects.net [refnum 0]
-:- PING  Register first.(from asimov.openprojects.net)
-:- BitchX+Deb1an: For more information about BitchX type /about
-:- Welcome to the Internet Relay Network TUNES (from asimov.openprojects.net)
-:- Your host is asimov.openprojects.net, running version u2.10.03.lagtime.desynch3.indent8.dead.whox2.zombie (from asimov.openprojects.net)
-:- This server was cobbled together Sat May 9 1998 at 12 25:56 EDT(from asimov.openprojects.net)
-:- asimov.openprojects.net u2.10.03.lagtime.desynch3.indent8.dead.whox2.zombie dioswkfcg biklmnopstv 
-:- [local users on irc(6)] 1%
-:- [global users on irc(444)] 47%
-:- [invisible users on irc(507)] 53%
-:- [ircops on irc(16)] 2%
-:- [total users on irc(951)]
-:- [unknown connections(0)]
-:- [total servers on irc(28)] (avg. 33 users per server)
-:- [total channels created(384)] (avg. 2 users per channel)
!asimov.openprojects.net Highest connection count: 10 (9 clients)
-:- Your default nick is already in use 
-:- Mode change [+iws] for user TUNES
-:- JOIN activated by "TUNES #tunes tunes@bespin.dhs.org "
-:- TUNES [tunes@bespin.dhs.org] has joined #tunes
-:- Topic for #TUNES: TUNES, Free Reflective Computing System: http://www.tunes.org/ || Slate Language: http://www.tunes.org/~water/slate-home.html || Tunes WebDB at http://diktuon.arrow.cx/
-:- topic set by hcf [Mon Jun  5 10:05:51 2000]
-:- [Users(#tunes:6)] 
[ TUNES     ] [ water     ] [ eihrul    ] [ mark4     ] [ coreyr    ] 
[ Fare      ] 
-:- Mode change [+f] for user TUNES
-:- Channel #tunes was created at Sun Feb 28 08:48:06 1999
-:- BitchX+Deb1an: Join to #tunes was synced in 12.994 secs!!
-:- Mode change [-ws] for user TUNES
-:- abi [nef@bespin.dhs.org] has joined #tunes
02:40am
-:- SignOff eihrul: #TUNES ([x]chat)
-:- SignOff water: #TUNES (The Tao went that-a-way!)
-:- Closing Link: TUNES[bespin.dhs.org] by asimov.openprojects.net (Ping timeout for TUNES[bespin.dhs.org])
-:- Connection closed from irc.us.openprojects.net: Success
-:- Connecting to port 6667 of server irc.us.openprojects.net [refnum 0]
-:- BitchX+Deb1an: For more information about BitchX type /about
-:- Welcome to the Internet Relay Network TUNES (from lackey.openprojects.net)
-:- Your host is lackey.openprojects.net, running version u2.10.05.18.(ipcheck4-5) (from lackey.openprojects.net)
-:- This server was cobbled together Wed Apr 28 1999 at 12 02:19 EDT(from lackey.openprojects.net)
-:- lackey.openprojects.net u2.10.05.18.(ipcheck4-5) dioswkfcg biklmnopstv 
-:- [local users on irc(55)] 6%
-:- [global users on irc(455)] 47%
-:- [invisible users on irc(509)] 53%
-:- [ircops on irc(17)] 2%
-:- [total users on irc(964)]
-:- [unknown connections(0)]
-:- [total servers on irc(28)] (avg. 34 users per server)
-:- [total channels created(373)] (avg. 2 users per channel)
!lackey.openprojects.net Highest connection count: 100 (99 clients)
!lackey.openprojects.net Welcome to Open Projects!  You are on 3 ca 1(2) ft 14(14) tr.
-:- Mode change [+f] for user TUNES
-:- Mode change [+iws] for user TUNES
-:- JOIN activated by "TUNES #tunes tunes@bespin.dhs.org "
-:- TUNES [tunes@bespin.dhs.org] has joined #tunes
-:- Topic for #TUNES: TUNES, Free Reflective Computing System: http://www.tunes.org/ || Slate Language: http://www.tunes.org/~water/slate-home.html || Tunes WebDB at http://diktuon.arrow.cx/
-:- topic set by hcf [Mon Jun  5 10:05:51 2000]
-:- [Users(#tunes:5)] 
[ TUNES     ] [ abi       ] [ mark4     ] [ coreyr    ] [ Fare      ] 
-:- Channel #tunes was created at Sun Feb 28 08:48:06 1999
-:- BitchX+Deb1an: Join to #tunes was synced in 7.167 secs!!
-:- Mode change [-ws] for user TUNES
-:- SignOff Fare: #TUNES (Ping timeout for Fare[esmeralda.enst.fr])
-:- lmaxson [lmaxson@adsl-63-194-24-207.dsl.lsan03.pacbell.net] has joined #Tunes
-:- lmaxson [lmaxson@adsl-63-194-24-207.dsl.lsan03.pacbell.net] has left #Tunes []
-:- morton [morton@AC9EEDDE.ipt.aol.com] has joined #tunes
<morton> where is everyone?
<morton> here's one for the record: 
<morton>    o      o
08:20am
<morton>        |
<morton> ---
<morton>    _____---
<morton> heh
-:- eihrul [lee@usr5-ppp183.lvdi.net] has joined #tunes
-:- morton is now known as AOLman
-:- AOLman [morton@AC9EEDDE.ipt.aol.com] has left #tunes []
08:30am
-:- smoke [smoke@15dyn102.delft.casema.net] has joined #tunes
-:- hcf [nef@207-172-225-222.s222.tnt1.pld.me.dialup.rcn.com] has joined #tunes
-:- Fare [rideaufr@esmeralda.enst.fr] has joined #Tunes
<coreyr> re fare.
10:00am
<Fare> err, coreyr
<coreyr> err?
10:10am
-:- water [water@tnt-9-36.tscnet.net] has joined #tunes
<water> yo
10:30am
-:- SignOff eihrul: #TUNES (Ping timeout for eihrul[usr5-ppp183.lvdi.net])
-:- eihrul [lee@usr5-ppp142.lvdi.net] has joined #tunes
<water> so anyway
<water> prograph's ui/ide has a lot of things that would make morphic more slate-like
<water> namely, data-flow and the fact that it also uses a slot-"like" object model
<water> at some point today, i'm going to focus on the bottom-up programming that i should do in slate to demonstrate a lot of things
<eihrul> good :)
<water> i've been considering all the issues in the back of my mind for quite a while
<water> so i'm probably the only one who can really do it :/
10:40am
<water> is there anything you're getting from looking at pliant?
<eihrul> still looking
<water> wow /. readers totally missed the point of the "systems research is dead" paper
-:- SignOff coreyr: #TUNES (Ping timeout for coreyr[net255ip95.parklink.com])
-:- coreyr [coreyr@net255ip95.parklink.com] has joined #tunes
<eihrul> that's what they're good for :)
<water> yes, i suppose if they did get it, they wouldn't be the /. community, would they?
10:50am
<smoke> you're generalizing. there are some remarkably intelligent slashdot readers, presumeably you yourself :-)
<water> heh i never post
<eihrul> sure, but if you consider '/. readers' as an individual
<eihrul> the signals are really dampened :)
<water> i read /. like i would read poll results
11:00am
-:- SignOff eihrul: #TUNES (Ping timeout for eihrul[usr5-ppp142.lvdi.net])
-:- eihrul [lee@usr5-ppp142.lvdi.net] has joined #tunes
-:- NetSplit: fastlane.openprojects.net split from lackey.openprojects.net [11:17am]
-:- BitchX+Deb1an: Press Ctrl-F to see who left Ctrl-E to change to [fastlane.openprojects.net]
-:- Netjoined: fastlane.openprojects.net lackey.openprojects.net
-:- mark4 [mark4@purplecoder.com] has joined #tunes
-:- abi [nef@bespin.dhs.org] has joined #tunes
-:- smoke [smoke@15dyn102.delft.casema.net] has joined #tunes
-:- hcf [nef@207-172-225-222.s222.tnt1.pld.me.dialup.rcn.com] has joined #tunes
-:- Fare [rideaufr@esmeralda.enst.fr] has joined #tunes
-:- water [water@tnt-9-36.tscnet.net] has joined #tunes
-:- coreyr [coreyr@net255ip95.parklink.com] has joined #tunes
-:- FineWitMe [asdasd@node-64-248-41-202.dslspeed.zyan.com] has joined #tunes
<water> boo
-:- FineWitMe [asdasd@node-64-248-41-202.dslspeed.zyan.com] has left #tunes []
11:30am
<water> hm
<water> prograph also handles multi-methods
<eihrul> cool
<water> yeah, especially as it is data-flow oriented
<eihrul> abi: prograph?
<abi> prograph is a visual data-flow language with oo extensions at http://www.pictorius.com/prograph.html
<eihrul> hmm, isn't there a manual about it here that isn't 6 MB in size? :)
<water> not really... lots of pictures, you see ;)
<water> vpl's tend to be hard to explain with just words and source code ;)
<eihrul> well, how does it handle multi-methods?
<water> you just create a data-flow function outside of a class
<water> it's not really the greatest way to do it, as polymorphism has to be implemented all in the same data-flow diagram
11:40am
* eihrul/#tunes has been musing over multi-methods some.
<eihrul> very powerful but they kind of contradict the encapsulation model provided in a Self-like object model
<water> yes
<water> speaking of which
<eihrul> also if you base a language solely on multi-methods
<water> i'm considering the possibilities for actually *bypassing* the strong modularity of using oo for slate's namespace structure
<eihrul> object fields become a separative primitive
<water> why is that?
<eihrul> actually, well, maybe not... but they don't fit as well in the framework :)
<water> heh yeah that's what i was thinking
<water> here's what i was referring to about strong modularity....
<water> we can intercept selector-choice one at a time, as we've already discussed
<water> but if we want to protect code that starts from the prompt and reaches in N levels deep, we still have to intercept the calls selector-by-selector
-:- smklsmkl [sami@ppp52.dial-in.verkkotieto.fi] has joined #tunes
11:50am
<eihrul> hmm, oh yeah, i remember now why fields were trouble-some :)
<eihrul> there was a regression
<eihrul> if multi-methods themselves are used to access fields of an object
<eihrul> and your object model has a less primitive notion of "type" (i.e. type could be a field of an object)
<eihrul> then the type field must be access to check if a specific sub-method of the multi-method is to be applied
<eihrul> which itself requires another call to "type", and so on
<water> hm
<eihrul> so "type" would have to be applicable to all objects
<water> that kinda makes the self model a bit pointless
<eihrul> yes :)
<eihrul> that was why it was somewhat troubling
<eihrul> and why fields would have to be treated distinctly from multi-methods
<water> somehow i think this is the wrong approach
<eihrul> though perhaps that mirrors the same problem with meta-object lookup
<water> why would the multi-method need that type field?
<eihrul> a predicate that determines whether a method of the generic-function (multi-method) is applicable to an object
<water> that doesn't have to be a type slot, does it?
<eihrul> nope
<eihrul> just at some point you need some predicate that is primitive
<water> you could do the same thing by putting all your objects of a type in a namespace :)
<eihrul> just like in Slate you need a Top
<water> and the multi-method would only be able to work in selected namespaces
12:00pm
<eihrul> note, i'm not considering multi-methods as applied to Slate
<water> ah
<water> well in that case ;)
<eihrul> just multi-methods as a pervasive language concept and the consequences :)
* water/#tunes returns to gleaning info from prograph
<water> ok
<eihrul> okay, which file contains the manuals :)
<water> heh
<water> dude this language ain't that great
<water> i'm looking at it from a ui standpoint
<eihrul> okay, i'll just move onto montana then 
<eihrul> (pun not intended)
<water> abi: montana?
<abi> i haven't a clue, water
<eihrul> was an open compiler from ibm
-:- SignOff smoke: #TUNES (humz)
<water> oh
<water> hm.... prograph has absolutely no namespace concept
-:- smoke [smoke@15dyn102.delft.casema.net] has joined #tunes
-:- SignOff smoke: #TUNES (Read error to smoke[15dyn102.delft.casema.net]: EOF from client)
12:10pm
<water> this tutorial and manual are definitely geared to end users
<hcf> put this in ur queue for later, lots of PL etc poop, which u'v probly seen most of already, but worth a look, http://www.cs.mun.ca/~ulf/
<water> k
12:20pm
<hcf> hmm, has a section on writing PL definitions, /pld/write.html
<water> yep
-:- smoke [smoke@15dyn102.delft.casema.net] has joined #tunes
<water> bah, he doesn't list slate yet ;)
<hcf> email him :)
<water> lol
<water> no i'd like to see when he does it on his own
12:30pm
<water> it's funny, but the ui for prograph is not really very well OO
-:- SignOff water: #TUNES (Ping timeout for water[tnt-9-36.tscnet.net])
-:- water [water@tnt-9-36.tscnet.net] has joined #tunes
12:40pm
<water> ok you get polymorphism by using case-switch type features
<water> and you can separate out the different implementations of a multi-method that way
<eihrul> yeah, it basically inverts how dispatching is done
-:- SignOff water: #TUNES (Read error to water[tnt-9-36.tscnet.net]: Connection reset by peer)
-:- water [water@tnt-9-36.tscnet.net] has joined #tunes
<water> well, yeah i know how multi-methods work in general, i meant i discovered how to do it in prograph
<water> bah
12:50pm
<eihrul> oh, ok
<water> prograph has too many assumptions built in to the ui
01:00pm
<water> bbiaf
-:- SignOff hcf: #TUNES (Leaving)
-:- hcf [nef@207-172-225-68.s68.tnt1.pld.me.dialup.rcn.com] has joined #tunes
01:40pm
<smoke> is only a highlevel and a lowlevel language enough for a complete OS ? .. in unix i use about five or so languages for various tasks..
02:00pm
<eihrul> most of them are implemented in C
<water> sure, if the languages are powerful enough
<eihrul> on unix
<smoke> eihrul: ah, so you mean that for tunes people could implement multiple languages in the HLL ?
<eihrul> iow, C is the ubiquitous unix VM
<smoke> water: some languages will always be somewhat more powerful in a certain respect, no?
<water> maybe
<water> sorry, i have to go
<smoke> bye
<water> yeah maybe later
-:- SignOff water: #TUNES (The Tao went that-a-way!)
02:10pm
-:- SignOff hcf: #TUNES (Leaving)
-:- SignOff smoke: #TUNES (z)
-:- SignOff eihrul: #TUNES (Ping timeout for eihrul[usr5-ppp142.lvdi.net])
-:- eihrul [lee@usr5-ppp104.lvdi.net] has joined #tunes
-:- water [water@tnt-10-176.tscnet.net] has joined #tunes
<water> re all
04:10pm
-:- billh [billh@cx739861-a.dt1.sdca.home.com] has joined #tunes
<billh> Hello
<water> hey
<billh> Where is there a good CL library reference ?
<billh> Hello water. ;-)
<water> clhs?
<billh> http://www.xanalys.com/software_tools/reference/HyperSpec/
<billh> ?
<water> yeah
04:20pm
<water> abi: cltl
<abi> cltl is Common Lisp the Language, 2nd Edition online at http://www.cs.cmu.edu/Web/Groups/AI/html/cltl/cltl2.html
* billh/#tunes was watching a Linux bashing conversation earlier.
<water> cool
<billh> Yeah, I'm FreeBSD user now. I can't deal with Linux's VM performance. ;-)
<billh> Water is CL more complete than Scheme ?
<water> it has more features in it, yes
<billh> I'm learing Scheme now and it's a pretty elegant and simple LISP dialect.
<water> it is
<billh> Yeah, I'm trying to find out what specifically it's got over Scheme.
<water> you can do nearly everything that CL can do from within scheme
<billh> and whether it's worth learning.
<water> namespaces, for one
<billh> Really ?
<billh> what about OOP conventions ?
<water> yes, but you have to implement or add it yourself
<billh> Does it do that ?
<water> yes iirc
<billh> Yeah, I couldn't find it when I was reading over the Scheme doc.s
* water/#tunes opens DrScheme
-:- mark4 is now known as I440r
<billh> water I find CL to be too complex and I'm wonder if I can do just about everything in CL using Scheme
04:30pm
<water> you'll have to grab some scheme code from the net to enhance it to something comparable with CL
<billh> How much enhancement ?
<billh> Is it extensive ?
<water> i'm not an expert on this, unfortunately
<water> i'm looking to find out though
<water> what features are you most interested in?
<billh> Difficult to say.
<water> CL does just about everything
<billh> I'm looking for a new language that's neat to learn that'll be usefull for building this JIT core I'm writting.
<water> hm
<billh> Yeah, CL has too much clutter for me though. I'm not sure if I want to deal with that shit.
<water> if you like lisp's macros, scheme does have define-syntax
<billh> Also I'm possibly looking for a technology like Java's RMI.
<water> sure
<billh> yeah, the syntax doesn't matter to me as much as the clarity of the libraries, etc...
<water> rmi is a must have in some form or another
<billh> A buddy of mine has been pushing Kawa which is a Scheme interpreter writting in Java.
<water> i've heard of it
<billh> It looks pretty interesting.
<water> there's also scheme-48 and rscheme
<billh> I could also do this project in Java, but I'm hoping to avoid extensive syntax conversions, etc...
<water> hm
<billh> Yeah, I've been using scheme-48.
<billh> It's nice.
<water> i'd be more concerned with compiling to JVM than translating to Java
<water> usually if you want an object system in scheme you just write your own lightweight one
<billh> Using what facilities 
<billh> ?
<water> macros, as well as implementational things like hash-tables iirc
* water/#tunes looks
<smklsmkl> association lists
<water> yeah that's it
<smklsmkl> heh
<water> doesn't the PLT site offer an object system or two?
04:40pm
<billh> URL ?
* billh/#tunes has also been looking at Guile.
<water> http://www.cs.rice.edu/CS/PLT/
<water> yeah Swindle
<billh> Is DrScheme a common Scheme implementation ?
<water> Swindle is a CLOS rolled over scheme
<water> i think so
<water> it's definitely a good one
<billh> So I can use it over Scheme-48 ?
<water> it conforms to r5rs, so i guess so
<water> heh of course they have an xml package
<billh> DrScheme look like a winner.
<billh> XML is a good thing.
<water> hm a cgi package as well
<water> i forgot how much good code PLT had
04:50pm
<billh> water is DrScheme fast ?
<water> fast enough.... it's a good interactive compiler
<water> i haven't really pushed it, though
<billh> How's the speed of Schem48 in comparision ?
<water> i couldn't say, myself. i've heard scheme-48 is a little slow
<water> check at http://www.schemers.org
<water> hm the site must be down
* billh/#tunes is building DRScheme now.
05:00pm
<water> apparently MzScheme is faster than Dr
<billh> what out 48 ?
<water> probably simply due to less overhead
<billh> what about 48 then ?
<water> dunno... still looking
<billh> Ok.
<billh> The site won't load.
<water> yeah i know
<water> hm i found 4 more scheme object systems
05:10pm
<water> hm tinyscheme runs in less than 64k
<water> i can't seem to find any good benchmarkings
-:- SignOff smklsmkl: #TUNES (back tomorrow)
<water> this might apply: http://www.webcom.com/nazgul/change.html
<water> that's about all i can find
05:20pm
-:- SignOff I440r: #TUNES ([x]chat)
-:- lmaxson [lmaxson@adsl-63-194-24-207.dsl.lsan03.pacbell.net] has joined #tunes
<water> hi
<lmaxson> hi
<lmaxson> I should be able to converse somewhat better than that.
<water> heh
<water> so, what brings you here?
<lmaxson> I just ran across the website and thought I might check out the irc channel.
<water> ok
<water> we're currently working on a site re-build
<coreyr> here comes the infamous "so do you grok tunes" question
<coreyr> :)
<water> as well as working on slate
<lmaxson> Actually I normally frequent #voice and #scoug on the Webbnet.
<water> which is not the easiest language i can imagine designing
<lmaxson> What piqued my interest was the number of languages involved.
<water> but it's unified and it works for tunes hll, and that's what counts
<lmaxson> I've designed a language as well.
<water> really? what language is that?
<lmaxson> Specification Language/I (SL/I)
<water> declarative, then?
<lmaxson> Yes and no.  Supports both processing types, assertion and assignment.
<water> hm ok
<lmaxson> Maybe you didn't mean that by declarative.
<water> it doesn't matter
06:00pm
<lmaxson> It's a superset of PL/I with the addition of logic programming (predicate logic) and all the APL2 primitive operators.
<water> is this for commercial development purposes only, or more general purpose?
<water> ok
<lmaxson> Actually it's to replace all other languages.
<water> heh
<water> you won't replace lisp with *that* ;)
<lmaxson> It's part of a software methodology in which the only language used is a specification language.
<water> or smalltalk or haskell .... :)
<water> sure
<lmaxson> You can create LISP from it.
<water> have you ever looked at Maude?
<lmaxson> Or Smalltalk or haskell or Maude.
<water> how?
<water> any language can implement any other language via direct implementation
<water> the question is how simple does the original language make it to do so :)
<lmaxson> You cannnot implement PL/I with C.
<water> url?
<water> c is totally not in the class of languages i consider
<lmaxson> None.  Most of the description exists on CompuServe.
<water> heh
<water> not much of a PR campaign :)
<lmaxson> I understand.  I have been using PL/I since its inception in 1964 and APL in 1962.
<water> oh really
* water/#tunes is used to starry-eyed younger programmers with grandiose schemes
<lmaxson> In 18 days I will celebrate 44 years in the business.
<water> so why do you suppose your spec language is better than others?
<lmaxson> I don't claim that it's better.  It's just that it doesn't need any other language in its implementation.
<lmaxson> As a specification language it is capable of specifying itself.
<water> you really should look at maude, i think
<water> abi: maude?
<abi> maude is a reflective rewriting logic language at http://maude.csl.sri.com/ or http://www.csl.sri.com/~duran/
<lmaxson> I will take a look at it.
<water> thanks
<water> are there any questions i can answer for you?
<lmaxson> The author has far different goals than does SL/I.
<water> or maybe there is something you'd like to discuss
<water> oh ok
<lmaxson> Languages will do just fine.
<water> well, i don't find maude useful in itself
<water> but the language semantics are very useful when embedded properly into other languages
<coreyr> lmaxson: so whats your assessment of tunes?
<lmaxson> tunes offers much more than I have had a chance to assimilate on such short experience.
<water> hm
<lmaxson> I am part of the FreeOS project.
<water> abi: FreeOS?
<abi> FreeOS is closed-source or http://www.freeos.grid9.net or forgetable
<water> heh
<coreyr> heh
06:10pm
<water> well, at least we looked at it :)
<lmaxson> FreeOS is an open source project to replace OS/2, Linux, and Windows(all versions).
<water> does abi speak of the same project?
<water> (per the url?)
<lmaxson> I think not.
<water> ok
<water> how do you propose that freeos replace these oses? what's its design and philosophy?
<lmaxson> FreeOS is probably non-descriptive.
<water> hm
<coreyr> how can it be free and closed source?
<coreyr> thats silly
<lmaxson> Uses the micro kernel and offers the various OS personalities concurrently.
<lmaxson> It's not closed source.
<coreyr> lmaxson: yeah i heard you.
<water> so how is it distinguished from, say, AtheOS?
<lmaxson> I'm not in a position to offer you a comparison.
<coreyr> didnt your freeos have a previous name?
<lmaxson> It is currently in the design stage.
<coreyr> seems familiar
<water> ah
<lmaxson> coreyr, I am not aware of any other name.
<water> perhaps the dolphin effort might enlighten
<water> hm no dolphin people are around currently
<lmaxson> I have an interest in software development as a process.
<water> dolphin essentially is built from objects which are mediated by an ORB (which is also made up of objects)
<water> ok
<lmaxson> Unfortunately I am not big on OO technology.
<water> technology?
<lmaxson> The use of objects and messages.
<water> uh right
-:- hcf [nef@207-172-225-237.s237.tnt1.pld.me.dialup.rcn.com] has joined #tunes
* water/#tunes thinks the word "technology" just makes good ideas sound pathetic
<lmaxson> I see no advantage to it, particularly in software maintenance.
<water> have you ever used a real oo language that's uniform?
<water> like smalltalk or Lisp/CLOS as opposed to Java?
<lmaxson> My only experience is with Smalltalk.  I did study Eiffel.
<water> what's wrong with smalltalk?
<lmaxson> I did LISP before CLOS.
<water> i don't doubt you did lisp before clos
<lmaxson> The problem is that you have no hll with assembler's capability of programming itself entirely.
<lmaxson> With SL/I you have that.
<water> i don't think i understand what distinction you're making
06:20pm
<lmaxson> I am talking about the ability to use the hll exclusively including the writing of the compiler.
<water> um smalltalk *does* that
<lmaxson> smalltalk is based on defining objects and passing messages.
<water> cecil does as well.... without a VM
<water> so?
<lmaxson> Why not a language without those restrictions?
<water> restrictions?
<lmaxson> Why not a language in which all are possible.
<water> everything is possible even within oo
<lmaxson> Why not a language in which it does not make a difference.
<water> it's a different way of thinking about the problem, that's all
<lmaxson> No.  OO is impossible without a standard library.
<water> i disagree
<water> most languages *are* themselves a standard library
<lmaxson> OO is dependent upon a library, a object classification system.
<water> like the static type systems they're based upon are standard libraries
<water> you're correct, but they're less based on stdlibs than other languages
<lmaxson> Standard implies that non-standard is possible.
<water> s/less/*less*/
<lmaxson> Why not a language in which non-standard is impossible.
<water> oh that's your assumption?
<water> because non-standard is often the *solution*
<water> standards usually just cripple the programmer newbie's thinking
<lmaxson> Why not a language in which non-standard is automatically transformed into standard?
<water> heh
<water> as in Common Lisp?
<water> or slate? :)
<lmaxson> Lisp or Forth (to be closer to self-defining and extensible).
<water> that suggests my proposals are *not* such
<water> which i find rather annoying
<lmaxson> Not my intent.
<lmaxson> Forth relies on a runtime dictionary in which the operators can be dynamically modified in terms of effect.
<water> otoh slate is rather poorly documented as of now
<water> i think you don't see OO in a friendly perspective
<water> you probably accepted the class libraries as unchangeable and decided OO was poor based on that alone
<water> and if you think slate is OO, you're quite mistaken
<lmaxson> My objection lies in class libraries and their internal use of methods being able to create incompatibilities.
<lmaxson> I have no knowledge of slate.
06:30pm
<lmaxson> I don't like the ability to introduce incompatibilities, only compatibilities.
<water> if you don't use the class libraries, or if you elaborate on them, you don't have that problem
<lmaxson> Why not design a language such that incompatibilities are impossible.
<water> and there are always wrappers in oo to solve that
<coreyr> lmaxson: for instance?
<lmaxson> SL/I.
<coreyr> heh
<water> LOL
<water> i'm not even going to go there
<coreyr> can you put it on the web?
<lmaxson> I would be happy to add it to the list.
<lmaxson> I am very much in favor of logic programming.
<water> i like logic programming, but the logic that's used in it is often deplorable
<water> (e.g. horn clauses)
<lmaxson> Are you familiar with the Z-specification language.
<water> yes
<lmaxson> Consider something along that.
<water> no thanks
<water> z is terrible
<lmaxson> In what sense?
<water> they should have consulted someone with a real backgroung in logics
<water> background, even
<lmaxson> I feel the same way about Prolog.
<lmaxson> My guide is Paul Voda, the author of Trilogy.
<water> have you looked at combinators or linear logic?
<lmaxson> Not specifically perhaps.
<lmaxson> As I said my interest is the software development process.
<water> what about functional programming?
<water> bah
<water> that's a buzzword
<lmaxson> From specification through analysis, design, construction, and testing.
<lmaxson> What's a buzzword?
<lmaxson> I have looked at functional programming.
<water> formalizing it like that doesn't exclude non-business related forms of programming
<water> "software development process"
<lmaxson> I'm not looking to exclude anything.
<water> i believe that you already *have*
<lmaxson> You have to have a means of accepting requirements and transforming them in some manner into a version of a product.
<lmaxson> That's a process, call it what you like.
<water> a "product" is by definition alienated from the people who use it
<water> software does not always get made by clearly defined groups of people as distinguished from the community
<lmaxson> When the solution set maps directly onto the problem set no such "alienation" occurs.
<water> in fact, Tunes is against that!
<water> any level of indirection in the software development process beyond the end-user should be minimized
06:40pm
<lmaxson> I would agree.
<water> but formal specifications are not accessible to the end user
<water> either in practice or in theory
<water> not even a declarative language will help that
<lmaxson> There has to be a means of mapping the user's informal language with that of a formal specification (language).
<water> of course, but i seriously doubt that any PL/I derivative can seriously be considered to do so
<lmaxson> The only translation that has to occur is that of the user's requirements into formal specifications.
-:- lar1 [larman@adsl-63-204-134-1.dsl.snfc21.pacbell.net] has joined #tunes
<lmaxson> It is not a derivative of PL/I.  It is a superset.
<water> my statement still holds
<lmaxson> It only holds if there is an exception possible.
<water> heh
<lmaxson> That's logic programming.
<water> lol
<water> have you looked at goedel, mercury, etc?
<lmaxson> It only holds if a "false" result occurs.
<lmaxson> I'm afraid not.
<water> they are logic programming languages as well
<lmaxson> I use PL/I.  So far I have not seen any other language that comes close.
<water> and i know that they are not sufficient, though higher-order and functional
-:- Kyle_L [kyle@cr168790-a.nmkt1.on.wave.home.com] has joined #Tunes
<water> LOL
<water> how can you possibly prefer PL/I?
<water> i suppose you only compare to "commercially-viable" languages
<lmaxson> I've not tried everything else.
<water> try haskell
<lmaxson> The first thing I look at is supported data types.
<water> what if the language doesn't express things that way?
<water> not every language speaks like Algol familiy languages
<lmaxson> If it uses the term integer or float, then I scratch it from further consideration, along with short or long.
<coreyr> They should!
<water> heh
<lmaxson> There's something about fixed bin (31) and fixed dec (7) that tells me that not all integers are created equal.
* water/#tunes never has to consider these things when looking at languages
<water> if that crap is in the syntax, it doesn't exist to me
<lmaxson> When you want to match the solution set to the problem set.
<water> i don't get it... smalltalk handles this perfectly well
<lmaxson> That crap, as you call it, is inherent in the computer architecture.
<water> no, it's inherent in *specific* computer architectures
<lmaxson> First off, tunes talks about a "high performance" language.  That let's Smalltalk out.
<water> the use of primitive datatypes should not be in a language definition
<water> lol
06:50pm
<lmaxson> Name me a specific machine architecture not based on procedural logic instructions.
<water> you can dynamically compile a language like smalltalk
<water> self, for instance, is 50% of the speed of compiled C
<water> and self is more abstract than smalltalk
<lmaxson> Compile C is 10% of assembler.
<water> heh
<water> i suppose you miss the old mainframes? :)
<water> and teletypes?
<lmaxson> Why give up anything that assembler has to offer?
<lmaxson> I love the old mainframe.
<lmaxson> I spent 36 years with IBM.
<water> i guess this conversation is pointless then
<lmaxson> Why is the conversation pointless?
-:- m0rt0n [axg@AC8A1F23.ipt.aol.com] has joined #TUNES
<water> because you don't see the point of abstraction from hardware
<water> or meta-programming
<lmaxson> I see the point of abstraction.
<lmaxson> I certainly see meta-programming.
<lmaxson> I see no reason not to have it all in a single language.
<m0rt0n> does anyone here know how tcn's forth-os works?
<water> then why do you insist on hardware-typing in language?
<water> morton: l440r does
<lmaxson> Because at some point the rubber must meet the road.
<water> morton: no one here is familiar with it. try #forth
<water> lmax: that can be done transparently to the code, though
<water> lmax: even transparently to the language
<m0rt0n> so does a standard implementation of forth set up a jump table of words or what?
<lmaxson> The code is a language.  The code that does the code is a language.
<water> morton: mark4 on #forth *is* l440r
<lmaxson> Yes, on Forth.
<water> so?
<lmaxson> So why have more than one language?
<water> the self language is independent of the primitives used to implement it
<lmaxson> Why make that distinction?
<water> you can place primitives anywhere in the code that they fit
<water> geez
<lmaxson> I'm not too sure we are talking about the same thing here.
<water> so that i can port code trivially, or optimize it dynamically
<water> so i can optimize code according to actual use, not the use planned by a committee
<lmaxson> Let's talk about porting code, trivially or otherwise.
<water> why?
<lmaxson> I use only a single developer for the entire process, not a committee.
<water> i have things to do
<water> it doesn't matter
<water> you're not fast enough to optimize the code when it only takes 10^5 clock cycles sometimes
<lmaxson> A single developer for an entire operating system or application system.
<water> and only another 10^5 clock cycles to decide to optimize it
<water> the machine can do it better than you can
<lmaxson> Let the machine do it.  That's my philosophy.
<water> and when you make static choices in code, it hinders the machine and yourself
<water> so why don't you like smalltalk?
<lmaxson> Let people do what the machine cannot.  Let the machine do what people need not.
<water> vacuous maxims, coming from your mouth
<lmaxson> It's not a matter of liking smalltalk.  I have nothing against smalltalk.
* coreyr/#tunes is away: watching paint dry
<lmaxson> If you use it as a guide, you end up with tools that allow you to develop software 50 times faster and 200 times less costly at a minimum.
<water> i don't disagree
<water> i'm saying you don't have any concept of how they should be applied
<lmaxson> You haven't done that with smalltalk or any oo product.
<water> LOL
07:00pm
<water> </flamewars>
<lmaxson> The point is how do you design such a tool.
<water> go spread your gospel to people who haven't seen better languages than yours
<lmaxson> It's not a language issue.  It's a process one.
<m0rt0n> whoa
<lmaxson> It is the software development process.
<lmaxson> I don't have anything to spread, gospel or otherwise.
<lmaxson> I don't want to interfere with anyone's choices.  I want to enable them.
<m0rt0n> how did this start?
<water> you sound worse than a 17-year-old
<lmaxson> Not bad for a 68-year-old.
<water> i've seen ignorant coders who only know C++ and Java claim the same thing
<m0rt0n> whoz the 68 year old?
<lmaxson> lmaxson.
<water> your words about "enabling" people mean nothing without some kind of framework that we can discuss formally
<Kyle_L> lmaxson: you have a language?  What is the URL?  Thanks.
<water> compare/contrast works better than broad generalizations
<m0rt0n> maxson: what did you accomplish in 68 years of life before u became interested in computers?
<lmaxson> :-)
<water> fellows, she's more than experienced
<lmaxson> She is a he.
<water> sorry
<water> i assumed wrongly
<lmaxson> No reason.
<m0rt0n> grace hopper wannabe
<m0rt0n> heh
<lmaxson> Happen all the time.
<lmaxson> Not Grace Hopper.
<lmaxson> No such desire.
<lmaxson> I just want to make software affordable to everyone.
<water> um it already is
-:- SignOff m0rt0n: #TUNES (Leaving)
<water> making actual executables free of charge already has a large following
<lmaxson> Affordable is different from free.
<water> and free software has the FSF
<lmaxson> Affordable says that you have whatever resources necessary.
<water> this is so vauge that i'm leaving just so i don't have to hear it
<water> s/vauge/vague/
-:- water [water@tnt-10-176.tscnet.net] has left #tunes []
<lmaxson> Gets to have rough waters at times.
-:- mark4 [mark4@purplecoder.com] has joined #tunes
-:- mark4 is now known as I440r
<lmaxson> The Forth person.
<lmaxson> I feel like the third man out.
07:10pm
<lmaxson> I just came across tunes today.  A complete newbie.
-:- lmaxson [lmaxson@adsl-63-194-24-207.dsl.lsan03.pacbell.net] has left #tunes []
-:- water [water@tnt-10-176.tscnet.net] has joined #tunes
<water> so anyway
<water> damn it, the appearance of newbies is too erratic here
* water/#tunes now must relax before getting into the slate development again
07:20pm
-:- SignOff Kyle_L: #TUNES (Leaving)
<water> he spent a decent amount of time online before he quit and after he left
<water> i hope he had a good conversation with someone, then
<eihrul> that was odd... 
<water> what was?
<water> the flame war? :)
<eihrul> the above N hundred lines of discussion :)
<water> i can't believe he was seriously thinking about using a superset of PL/I
<eihrul> maybe its just a generation gap...
07:30pm
<water> that and a complete inability to grasp what smalltalk and related languages tried to achieve
<water> pl/i has to be the biggest joke of a language because it's trivially equivalent to every other procedural language out there
<water> not that that makes translation any easier
<water> damn it, i'm just pissed off now
* water/#tunes does some reading
<hcf> abi: freeos is also a os/2 clone at http://www.quasarbbs.com/daniel/freeosidx.html
<abi> okay, hcf.
07:40pm
-:- SignOff hcf: #TUNES (Leaving)
-:- SignOff water: #TUNES (The Tao went that-a-way!)
-:- water [water@tnt-10-203.tscnet.net] has joined #tunes
<eihrul> re
<water> care to interact on something?
<water> or not?
<water> damn it, the wind is gone from my sails
09:00pm
<water> eihrul: 'result' and '/'
<eihrul> okay
<water> they bother me
<eihrul> yeah, those two are 'vague'
<eihrul> guess its time to buy an engine...
<water> what?
<coreyr> isnt result a nessacery part?
<water> it performs a necessary function
<water> so does '/'
<coreyr> how else can it be done?
<water> they are not in themselves necessary
<water> i have thrown about a few ideas
<water> not satisfactory ones, but nonetheless...
<coreyr> are they over my head?
<water> the implicitness of 'result' only becomes ambiguous when you explicitly call it
<water> no, i just want to avoid them unless there is a logical necessity for alternatives
<coreyr> so these are the culprits for no spec?
<water> yes
<coreyr> well take them out back and shoot them
<water> there are certain to be objects for which 'result' will mean something rather vacuous
<water> if only it were so simple
<coreyr> well try approaching it from a different angle
<water> objects, for instance, are the data for slate as well as the programs
<water> in many cases you want the object in itself
<coreyr> other times?
09:10pm
<water> e.g. all literals need this approach
<water> when the object is not the value desired
<coreyr> hmm
<coreyr> is result a subconcept of something else?
<water> a literal returns the namespace it changes i think
* coreyr/#tunes is not very knowledgeble
<eihrul> that reminds me of one particular ambiguity :)
<water> it's a slot
<water> eih: which reminds you?
<eihrul> well, not so much an ambiguity but a quirk...
<coreyr> but its a essentially a return value?
<water> hold on
<water> a slate expression progresses through selectors
<water> like following a path in the namespace graph
<water> but to get this kind of thing to perform ordinary programming tasks requires certain care at special points
<water> which puzzles me
<water> what's odd is that in the graph model, you're not sending messages at all
<water> but it's ridiculously elegant how it captures almost all of the language ideas
<coreyr> is it just a facet of something underlying?
<water> which include oo and fp up to the basic idea of state or literal types, respectively
<water> i hope so
<coreyr> like a universal field theory
<water> maybe
<coreyr> i dont understand messages
<coreyr> they are just objects right?
<water> yes
<coreyr> if every message is an object, does communicate every happen?
<coreyr> communication
<coreyr> and ever
<water> hold on
<water> one thing i should remind us about is that ( and ) push and pop these selector accesses
<water> sorry, i had to say that explicitly
<coreyr> how does another object talk to another object
<water> yes communication happens, but we're formalizing at a substrate level where the entire system is one big graph where parts modify other parts
<coreyr> that sentence doesnt mean anything to me
<water> hm
09:20pm
<water> objects are immutable in slate... you can't change them.... assigning variables just makes new clones of the object
<coreyr> does an object make another object with the message it once as the value and then move it to the other object?
<coreyr> s/once/wants
<water> sort of
<water> there's the difference between the selector (a name) and the object invoked
<water> maybe 'result' should be explicit
<water> eihrul: opinion?
<coreyr> if its not universal, should it be explicit?
<coreyr> shouldnt
<water> it is universal, as we have it now
<coreyr> but its not universally used in the same way
<coreyr> right?
<eihrul> water: thinking...
<water> result is just another object
<water> have you heard the ":" and "result" interaction blurb for variable-like objects?
<coreyr> no
<coreyr> be simplistic and slow, im having concept overload
<eihrul> water: well, having an explicit result does solve one problem :)
<water> in a variable, ":" inherits the namespace containing the values desired
<water> it sets the "result" slot in the namespace that the ":" resides in to inherit from that object
<water> for example....
<coreyr> cont.
<water> myObject x : 4
<water> means that myObject x result *is* 4
<water> eihrul: which problem is that?
<eihrul> water: well, in the default scheme of things, (object message) can have two effects: invoke some particular snippet of code, or return a normal value
<eihrul> but what happens if you just want the literal value of the 'result' slot?
<eihrul> i.e. solely the latter effect, instead of both the former and the latter
<water> restate, please
* water/#tunes is thinking of how to put together the spec
<eihrul> if message's result is a method object
<eihrul> it is invoked
<eihrul> if message's result is not a method object, it is merely returned
<water> hm
<water> i don't think that that happens
09:30pm
<water> but
<eihrul> in slate.lisp it does :)
<water> heh
<water> but...
<water> it does make things more uniform if result always has to be called
<water> i suppose we can let result be part of userspace
<coreyr> are you looking to achieve perfection in a spec from the get go?
<water> corey: for certain things, yes
<water> so i take it we agree to make result explicit in all cases?
<water> and if so, are we still going to require every object to have a 'result'?
<eihrul> if result needs to be explicit, then i don't see the point in every object having it :)
<water> hm
<eihrul> or atleast, it could be an identifty operation in Root
<water> that bugs me... it's like making lambda a user construct of scheme
<eihrul> s/identifty/identity
<water> oh well, to hell with my feelings
<eihrul> this is the default behavior of slate
<water> ya
<eihrul> it shouldn't be the definitive behavior of slate :)
<water> i suppose
<water> the problem is if everyone starts using different names for their function-forcing mechanisms
<water> for no good reason
<water> what do you think?
<coreyr> right lots of code showing the right way
* water/#tunes is sitting alone in a room talking through a keyboard to people thousands of miles away about programming language specification in engrossing detail, and it bothers him
<coreyr> and put it in the tutorial that you will be crucified if you dont do it your way
<water> 2 miles away there's a party of friends at club mercury, but i can't get in because i need a member to get me in, and the only people who will won't go out
09:40pm
<water> sorry, i guess i'm not the most psychologically well put-together person
<coreyr> they wont meet you outside
<coreyr> (the ones there already)
<eihrul> well, i prefer sitting at my chair talking through a keyboard to people thousands of miles away
<coreyr> water: your starting to sound like lar
<water> :P
<water> anyway
<water> it just seems wierd that we'd have to enforce an arbitrary rule
<eihrul> well, there's lots of them in slate already :)
<water> gah maybe it's for the best
<water> huh?
<water> screw you, eih! :)
<water> just because you don't get why i picked something to be some way doesn't make it arbitrary
<water> :P
<water> :b
<water> :p
<eihrul> i didn't mean it that way
<water> :B
<water> what did you mean, then?
<coreyr> eihrul: you cant escape now
<eihrul> well, the underlying idea of your above statement without the self-implication :)
<eihrul> one person's arbitrary is another person's obvious
<water> yeah i guess
<coreyr> good save
<eihrul> and vice versa
<water> ok fine
* water/#tunes waves the magic slate spec scepter
<eihrul> but, it is not necessarily arbitrary
<water> voila! 'result' is userland fodder
<eihrul> unless you're choosing to make 'result' explicit for no particular reason at all :)
<eihrul> then it is arbitrary...
<water> well we already discussed the reasons
<coreyr> err feelings
<water> hey! now *there's* a good use for irc logs :)
-:- SignOff billh: #TUNES (Ping timeout for billh[cx739861-a.dt1.sdca.home.com])
<eihrul> okay, then, is 'result' decided?
<water> now ppl can find out exactly why we picked certain things to be in the spec even if we forget to mention it in the spec
<coreyr> water: on to /!
<water> yes
<water> exactly, what of "/"?
<coreyr> fix it
<water> :P
<water> gee thx
<coreyr> ok fix it please.
<water> eihrul: what are your thoughts as of now on "/"?
<water> i've stated before that we could simply omit ")"s from expressions to get the same effect
09:50pm
<water> but this is not something that comforts most programmers
<eihrul> yes
<water> they're used to balancing expressions with care
<eihrul> however, the absolute flavor of "/" is trickier
<water> which i think is just a matter of giving them the right tools for their code
<water> and "/" doesn't have a good OO model without getting into reflection
<water> although if we want to fiddle with ( and ), then we'll be doing the same thing
<water> so?
<eihrul> well, what about the ".." and "/" interaction here :)
* water/#tunes taps a silent asynchronous melody on eihrul's inner ear
<water> hm
<water> without ".." you don't get reversals of namespace accesses
<water> and you're stuck with a hierarchy
<water> since it only leaves you with )
<eihrul> "/" with ".." creates a new reference object that refers to the real object
<water> we've decided that result has to be explicit, so there's no longer any worry about invoking code implicitly
<water> you mean ".. /"?
<water> er
<water> "(..) /"?
<eihrul> well, i mean consider "/" with the existence of the ".." slot on objects :)
<water> uh why don't you just say what happens?
* water/#tunes isn't following eihrul's phrasings
<eihrul> "/" bridges namespace to access an object
<water> bridges?
<water> quit using metaphors
<eihrul> if we're in namspace "/Foo" and from within that we get access to an object "/Bar"
<eihrul> and object we got access to must have a ".." of "/Foo"
<water> uh
<eihrul> but "Bar"'s .. is "/"
<water> corey: is he making sense to you?
10:00pm
<coreyr> nope
<water> ok, i'm not the only one who found that a bit convoluted, and probably wrong
<coreyr> but what your saying doesnt make sense either. :)
<water> bah :)
<eihrul> okay...
<eihrul> let us put this another way:
<eihrul> if some code is being evaluated in /Foo (where /Foo serves as the activation record)
<water> what's with the /?
<eihrul> "/Foo" refers to selector Foo off of the root namespace
<coreyr> root?
<water> a namespace is just an object.... / doesn't have to be part of its names
<coreyr> hmm
<water> oh
<eihrul> now... this particular code needs to access and manipulate an object /Bar
<eihrul> Bar's immediate .. is the root namespace
<water> ok
<eihrul> to access and manipulate Bar in any way from within /Foo
<eihrul> there must be some (/Foo message) such that (/Foo message) => Bar
<water> yeah there is
<water> you say ".. Bar" from within Foo
<water> no / needed afaik
<eihrul> this is an example :)
<eihrul> its irrelevent
<water> so....?
<eihrul> the Bar in (/Foo message) => Bar must have an .. such that (Bar ..) => /Foo
<water> huh
<water> no, () takes care of that
<water> remember we're working in a sugared syntax
<water> ".. Bar" from within Foo is really "(.. (Bar))"
<water> at least, that's the assumption i've been working under
<water> and you?
<eihrul> i think you've taken a detour :)
<water> um no
<eihrul> anyone, i have lots of stuff i should be reading right now :)
<water> what detour are you suggesting i've taken?
10:10pm
<eihrul> s/anyone/anyway
<water> no you don't
<eihrul> unfortunately, yes, i do :P
<water> explain where "/" fits into your argument
<eihrul> well, perhaps it doesn't :)
<water> geez
<coreyr> heh
<water> perhaps it's irrelevant then
<coreyr> water: sometimes i wonder if two waters would solve your problems
<eihrul> okay, i took the detour, you went straight... and all things being relative, you seemed to take the detour :)
<water> heh
<water> the point of "/" is that ( and ) operate strictly under the assumption of a hierarchy
<water> which is great for expressions entered at the prompt
<water> wait
<water> hm
<water> (.. is distinguished from ..
<water> one of the original points of "/" was also that it prevented 'result' from getting invoked
<water> which is now unnecessary
<water> i need a life
<coreyr> its overrated
<water> there are so many damned intricacies to this language
<water> corey: yes but i miss having good friends
<coreyr> you said it yourself. control
<water> control? what about it?
<coreyr> nm i dont want to discuss this, you always win
<eihrul> well, slate is a lot like life, a lot of simple rules that combine to make complex and intricate concepts :)
<water> heh
<coreyr> you feign ignorance to let me walk into a trap. very sneaky.
<water> things like this are why slate will be much better as a background language for visual or macro-driven programming
<coreyr> mobius?
<water> yeah mobius is part of that
<eihrul> well, is slate a language or is it a virtual machine? :)
10:20pm
<water> hell if i know
<water> what's Tunes HLL?
<water> i doubt you'd get a better answer about that than about slate
<coreyr> whats a virtual machine?
<water> heh
<water> you know what the JVM is, right?
<coreyr> yeah
<water> similar idea, though more abstract
<coreyr> well isnt a vm an implementation of a lang?
<water> the lambda calculus or the combinator calculus are both virtual machines
<water> no
<water> not in the sense that the JVM implements Java or the Smalltalk VM implements Smalltalk
<water> the VM is it's own language that acts as a substrate for other languages
<coreyr> like a spec language? :)
<water> uh
<coreyr> im kidding
<water> good
<coreyr> so the lambda calculus is a substrate for other languages?
<water> alright, "/" is gone
<water> (until further notice)
<water> in a sense, yes
<water> you can build functional programming languages (real ones) from it
<water> and you can build other lambda calculi from it
<coreyr> is there a vm that can be a substrate for all other langs?
<water> yeah, it's called a Universal Turing Machine ;)
<coreyr> so why dont we use the universal turing machine as a lang?
<water> you see you make this really big mechanical device that reads and writes little numbers onto an infinitely long tape..... :)
<eihrul> not enough paper tape to satiate a whole world of turing machines? :)
<water> picking the right substrate is very crucial to maintaining simplicity
<coreyr> so a vm defines a lang?
<eihrul> actually, if you just loop the paper tape :)
10:30pm
<eihrul> you have a nice model of today's computers...
<water> no, looping the tape doesn't work
<water> oh
<water> i thought you weren't going to qualify that
<eihrul> a finite amount of recycled tape with lots of obfuscated instructions...
<water> anyway
<water> yes, a vm defines a language of sorts
<coreyr> a vm defines what a lang can be used for?
<water> none of this is a science yet.... it's all about metamathematics
<water> a language can be used for anything, though
<eihrul> a vm is a language intended to be used as the target for other languages, imo
<coreyr> a vm defines how a lang can be used?
<water> the nature of the language determines how you have to go about using it, though
<water> damn it, this is rambling
<eihrul> well, we got "result" and "/"
<water> get a book on logic and goedel and lambda calc
<eihrul> we managed to interject some ".."
<coreyr> heh
<water> yes we did
<eihrul> what's next?
<water> hm
<water> corey: get a cheap dover book
<water> hm
<coreyr> there are many publishers named dover right?
<water> mutation i suppose
<water> huh?
<water> no, there's just one... and they have no website
<coreyr> ok
<water> they re-publish books
<coreyr> "dover publications" ?
<water> yep
<water> mutation is definitely the next thing to consider
<water> after that, the mop
<coreyr> a little case d with the word "DOVER" in the circle of the d?
<water> yep
<water> did you find a website?
<coreyr> http://207.122.105.90/
<coreyr> but its not much
10:40pm
<coreyr> looks like they are playing around with something
<eihrul> water: actually, mind if i go to my other readings? you can beep me if you really need something discussed...
<water> holy cow they finally put something online
<water> eih: sure
<water> lol there's nothing there
<coreyr> the site isnt live obviously
<water> at least they leave their address and email
<coreyr> theres like discussion of how the site should be org'd.
<water> yes i saw
<water> it's hysterical... i assumed they made enough money to do what i've had done in my spre time :P
<water> although my favorite music recording companies are also not online :)
<coreyr> probably gonna be doverpublications.com when they're thru
<water> it already is actually
* water/#tunes checked that url
<coreyr> heh
<coreyr> i just spent 3 minutes using whois 
<water> what for?
<coreyr> i didnt think they were using a domain yet
<water> oh
<water> say... how are those diktuon mods going? :)
<coreyr> hmm
<coreyr> i have to think how best to convert names to be usable for urls
<water> k
<coreyr> im sure you dont care the details
<coreyr> any specific title i should be looking for?
<water> it depends on what aspects of the subject you want to know about
<water> strictly cs stuff is online, actually
<water> just go to RI and search for "lambda calculus"
<coreyr> but i want to know the rest of logic also
<water> hm
10:50pm
<coreyr> i suppose ill learn most thru indoct. at university
<water> "set theory and logic" by robert r. stoll
<water> it's the best dover book on logic i have
<coreyr> 12 bucks, cant beat that
<water> heh
<water> that's expensive for their books
<coreyr> hmm
<coreyr> ucf library has it
* coreyr/#tunes is cheap.
<coreyr> heh
<coreyr> but it has the original 63 version
11:00pm
<coreyr> dover's is 78
<coreyr> think there is a difference?
<water> hey if my local library had this stuff, i'd never buy dover
<water> otoh i spend lots of time on a ship
<water> just corrections in the 79 version
<coreyr> that guy wrote some other stuff
* water/#tunes looks
<coreyr> but dover only has that one
<coreyr> and they're al in.
<coreyr> cool.
<water> dover doesn't publish everything under the sun
<water> where are you looking?
<coreyr> library.ucf.edu
<coreyr> :)
<coreyr> you can search the whole florida library system
<water> damn it, i'm 30 miles from a good college lib
* coreyr/#tunes is across the street
<coreyr> how fast do you read?
<water> bah
<water> quickly enough
<coreyr> hmm
<water> i usually read crap that has to be deeply grokked or you pay dearly in the next paragraph
<coreyr> i can keep them for like months
<coreyr> maybe i can ship some to you
<coreyr> if you cant find something local
<water> like the one book i have that starts off in the first paragraph with tensors and then leaves for alien worlds
<water> most people freak out when i show them that book
<coreyr> name?
<water> "curvature and homology" icuc
<water> another dover book.... totally unreadible to 99.9% of the population
<coreyr> by Greub
<coreyr> not checked out either
<water> uh no, goldberg
<coreyr> oh wait
<coreyr> i read to quick
<coreyr> Connections, curvature, and cohomology
<coreyr> is Greub
<water> ah
<water> that book is a little easier on the head
<water> c&h took me 7 months to finally grok
<coreyr> hmm
<coreyr> ucf doesnt have it
<coreyr> but just about every other uni does
<coreyr> i could get it thru interlibrary loan
<coreyr> Riemann surfaces
<water> :)
* coreyr/#tunes only knows a encyclopedia entries on this kinda stuff
<coreyr> student of gauss
<water> n-dimensional curvature rocks
<coreyr> influence of einstein
<coreyr> non euclidean
<water> yep
<coreyr> thats all i know
<water> heh
11:10pm
<coreyr> parallel lines may meet
<coreyr> stuff like that
<water> bolyai-lobachevsky geometry
<water> (sp?)
<coreyr> seen those names
<coreyr> dont know what they are
<water> ther're people who invented the geometry :)
<water> i learned that when i was 15 in geometry class
<water> we spent two weeks on proofs
<water> just for that geometry alone
<coreyr> ive never really had a math class
-:- _ruiner_ [DIY@ppp146.wi.centurytel.net] has joined #tunes
<water> uh hi ruiner
-:- _ruiner_ [DIY@ppp146.wi.centurytel.net] has left #tunes []
11:20pm
* coreyr/#tunes is away: to die, to sleep, to sleep, to perchance dream
* coreyr/#tunes is away: sleep
[msg(TUNES)] newlog 2000.0607
IRC log ended Wed Jun  7 00:00:00 2000