IRC log started Tue Aug 24 00:00:01 1999
[msg(TUNES)] permlog 1999.0824
<ElGato> um
<ElGato> fare...are you awake?
01:30am
<ElGato> wb
[@TUNES     ] 
<ElGato> ok im leaving
02:00am
<Crimson_> hi, boys
<fare> lo, girls
<Crimson_> im on a faster, better and more stable line than ever. =)
05:20am
<Crimson_> I want to run BeOS on my computer. Were can I get it?
<Crimson_> C ya later.
05:50am
<ElGato> hello
<abi> que tal, ElGato
<fare> ole
<ElGato> hi
<ElGato> pop-11 isn't all that great
<ElGato> it just reminds me too much of c
07:20am
<ElGato> hey smkl
<smkl> hello ElGato
<ElGato> how's finland?
<smkl> fine ...
<ElGato> :D
<ElGato> i always ask people from finland that
<ElGato> every time we talk
<fare> EG: pop-11 is interesting in its cross-language capabilities
<fare> in its interactive stack architecture, too
<ElGato> it has the stack...
<ElGato> just not a very strong language
<fare> ?
<fare> strong?
<ElGato> er
<ElGato> as i said it reminds me of c :D
07:30am
<smkl> pop-11 ?
<abi> rumour has it pop-11 is interesting in its cross-language capabilities
<smkl> pop11 ?
<ElGato> um
<ElGato> it's the language that has a stack and stuff but is too much like c :P
<smkl> poplog?
<abi> i heard poplog was now freely available at ftp://ftp.cs.bham.ac.uk/pub/dist/poplog/freepoplog.html
<sr> elgato: is LISP still your favorite lang?
<smkl> stack sounds like a low level mechanism
07:40am
<ElGato> oh
<ElGato> sorry
<ElGato> i like scheme and forth
<ElGato> both equally
<ElGato> stack is low level for forth
<ElGato> your spoiled
<ElGato> or mabye now
<ElGato> damnit
<ElGato> mabye not
<ElGato> mabye im just dumb and talking too much
<sr> heh Scheme Underground
<sr> what is poplog?
<abi> rumour has it poplog is now freely available at ftp://ftp.cs.bham.ac.uk/pub/dist/poplog/freepoplog.html
<fare> stack is high-level in poplog or in HP RPL
07:50am
<ElGato> what's the difference of a stack being high or low level
<fare> GC, bounds checking, etc.
<ElGato> ah
<ElGato> someone mentioned a stack based implementation of c (very vaguely
<ElGato> heard of it?
<smkl> would asm be high level if it had GC and bounds checking?
08:00am
<fare> sure
<fare> well, it'd be higher-level than it is.
<fare> but it'd also be mostly a waste of silicon
<fare> also maybe not
<fare> (see LISP machines)
<ElGato> fare: stack based c
<ElGato> nm
<ElGato> this person took it back
<smkl> you can type check asm to made it mostly checked
<smkl> s/made/make
<fare> smkl: isn't it precisely what the Fox PCC thingy is about?
<sr> fare did you take a crack at crack.linuxppc.org? =)
<smkl> fare: not really sure .. there is also talc
08:10am
<Zhivago> hello
<abi> hi, Zhivago
<ElGato> hi 
<abi> what's up, ElGato
<ElGato> fare: zhivago is into the same stuff you are
<ElGato> im misleading him with my lack of understanding
<Zhivago> actually, I'm mostly after a version of retro which compiles :)
<ElGato> gee i wish fare would pay attention to the channel
<ElGato> just idle for a while
<Zhivago> hmm, fare and I had some miscommunications a year or two ago on a mail-list
<ElGato> uh oh
<sr> Dr. Zhivago
<Zhivago> Dr. sr
<sr> zhivago are you into OS development?
<Zhivago> sr: yes
<Zhivago> sr: and I freely admit that I want a working retro dist, so that I can rip off its bootcode.
<Zhivago> sr: but I'm entirely happy to pass on various bits back
<ElGato> im pretty sure boot.asm is error free
<Zhivago> el: that's ok
<ElGato> but lazy paul had to commit his shitty serial.asm
<Zhivago> el: I'm not into low level x86 to start with, so what I really need is something which mostly works
<sr> elgato: what's bad about serial.asm?
<ElGato> ah yes
<Zhivago> el: and from that position fixing is relatively easy
08:20am
<ElGato> sr: hundreds of errors
<Zhivago> and the retro module stuff seems to be broken too
<ElGato> well guys i gotta go
<Zhivago> at the moemnt I'm looking at LittleOS as well, but at this point I'd prefer to avoid grub, and just make simple images
<Zhivago> move onto grub as an option at some point in the future
<ElGato> btw im writing a language too
<sr> zhivago: isn't LittleOS at a stand-still?
<Zhivago> sr: yes, but I'm not developing it, I'm just nicking boot-code
<ElGato> i just happen to be somewhat unfamilier with some tunes concepts
<Zhivago> sr: I have a compiler, we produce native x86 code with it, and I have an elf producer
<Zhivago> sr: I'd like to add an x86-boot-image target
<ElGato> ok bye
<Zhivago> sr: and I don't know what I'm doing wrt real/pmode stuff and booting
<sr> cool
<Zhivago> sr: so I need a test platform to play around with
<sr> can someone here recommend a computer case
<sr> Zhivago: bochs?
<abi> somebody said bochs was a PC emulator at http://www.bochs.com that runs on win32, un*x, beos, mac, os/2 and amiga
<Zhivago> sr: retro looks ideal so far except for the non-compiling bit
<Zhivago> sr: I'm using bochs to run it in
<sr> did you look at OpenBLT?
<smkl> compiler for what kind of language?
<Zhivago> sr: but I when I say test platform, I mean, something which boots up and lets me play with stuff, etc.
<Zhivago> sr: hmm, I remember seeing OpenBLT mentioned, but I haven't looked.
<Zhivago> sr: I'll have a look now
<sr> www.openblt.org
<sr> rated #1 in "Little startup OS magazine, issue Summer 1999"
<smkl> then there is ChaOS
<sr> smkl?
<abi> i heard smkl was a spam bot who got stranded in #tunes
<Zhivago> I saw chaos, I didn't really like it
<sr> zhivago: BRiX and clementine, when they are released
<Zhivago> sr: I'm really just after some p-mode boot code that I can prepend to my binary image
<Zhivago> sr: which is what retro seems to do nicely.
<sr> zhivago: when do you expect your OS to hit the big time?
<Zhivago> sr: never :)
<Zhivago> sr: but I expect it to be useful in probably 12 months
<Zhivago> sr: and hopefully go open within 6
<sr> so many OS projects being started, so few will become useful
<Zhivago> sr: this is true, but then most of them are just rehashes
08:30am
<Zhivago> sr: or too ambitious in difficult ways
<sr> Zhivago: i hope your OS is more than a mere rehash
<Zhivago> sr: well, the focus here is non-conservative security, and reflection
<Zhivago> sr: so one address space, everything running at ring0/administrator mode/etc
<sr> everything running at ring0? 
<sr> do you have a trusted compiler or something?
<Zhivago> sr: persistent journalling virtual memory takes care of most of the rest of the issues, but limits us to 4 gig of space under a 32 bit system
<Zhivago> sr: yes, trusted compiler
* fare/#Tunes is back
<Zhivago> hello fare
<Zhivago> fare: do you know if there is a compiling version of retro about somewhere?
<sr> retro, retro, retro... that's all i ever hear
<Zhivago> sr: well, its the only product of tunes program-wise that I've seen so far :)
<Zhivago> sr: so its kind of important :)
<fare> Zhivago: just use cvs to extract a version that's a little back in time
<Zhivago> sr: besides which I like the design from what I can see without being able to compile it
<sr> well there is Tunes-0.81-PR1
<fare> Zhivago: or submit a bug fix
<Zhivago> fare: that is an excellent point, I keep forgetting why we use cvs.
<fare> Zhivago: why limit us to 4 gig space? that's only for *flat* *32-bit* pointers.
<Zhivago> fare: well, its not a bug fix, its more that that makefiles are broken.
<fare> the address space needs be neither flat no 32-bit
<Zhivago> fare: becasue I don't want to swizzle, and 64 bit machines will become ubiquitous very soon
<fare> no, the makefiles seem good; the kernel.asm lacks the SCAN_FAT primitive
<Zhivago> fare: oh, I worked around that, the problem was in the modules.
<fare> in the modules?
<fare> did you commit changes?
<Zhivago> fare: yeah, they didn't want to compile at all. no, I only checked it out
<Zhivago> fare: let me try now, incase it has become fixed, or there was some problem at my end
<fare> Zhivago: do a cvs diff, and send it to me and/or tcn
<Zhivago> fare: the machien its on is elsewhere, but I'm not that fussed with the latest version
<Zhivago> fare: the obvious suggestion of taking a previous version is what I need to do :)
08:40am
<sr> Zhivago: do you expect to own a 64-bit machine soon?
<Zhivago> sr: I'd hope so, but initially 4 gig is enough, if necessary, the memory system should be rewritable to adapt to other conditions, although I don't find the idea attrative
<sr> zhivago: how is it possible to have an address space larger than 4gb with a 32-bit machine?
<Zhivago> sr: well, x86 manages it
<Zhivago> sr: they use a segmented architecture
<Zhivago> sr: but I'm not keen on that really.
<Zhivago> sr: but you could always swizzle pointers on mapping
<Zhivago> sr: which is portable, but less efficient, this is probably the solution I'd use if I was going to extend it.
<Zhivago> sr: I like flat memory
<sr> swizzle?
<Zhivago> sr: well, you have 64 bit references which get mangled on loading into 32 bit pointers
<Zhivago> sr: then you unmangle them on unloading
<fare> well, removing the scan-fat and init-fat things from words.inc were enough to make it work, to me...
<Zhivago> fare: did that also make the modules?
<fare> sr: by not having a flat pointer space -- besides, that's how linux does it
<fare> sr: except you must *manually* manage the heterogeneity
<fare> make the modules?
<Zhivago> fare: linux addresses more than 4 gig of memory?
<Zhivago> fare: the .rfd (?) files.
<fare> I did a make floppy
<fare> and it compiles and boots
<Zhivago> fare: ok, I'll grab nasm for this machine, and try it out, either I did something stupid, or its been fixed recently.
08:50am
<fare> I just committed the change
<fare> since I've tested that it compiles and boots
<fare> Zhivago: certainly, linux can address terabytes or memory, split into files
<Zhivago> fare: a linux process can't.
<fare> yes it can
<Zhivago> fare: unless this is something I'm unaware of
<Zhivago> fare: how does the mapping occur?
<sr> how can you not have a flat pointer space?
<fare> no, it's just something you don't see with the right angle
<Zhivago> fare: I can't mmap() to any place above 3 gig
<Zhivago> fare: so its not addressing it in any traditional unix fashion
<fare> sr: by doing explicit read() and write(), for instance
<fare> stop thinking "flat address space".
<Zhivago> fare: yes, I undersatnd swizzling
<fare> a computational "pointer" needn't a machine-level address mask
<fare> swizzling is another method
<sr> Zhivago: is that like the cigar Swizzlers :)
<Zhivago> fare: and I understand what you're saying
<Zhivago> fare: well, you need to swizzle if you're going to preserve any illusion of homogenity over addressing
<fare> swizzling is a way of doing it implicitly at the low-level through interrupts.
<Zhivago> fare: which is what addressing kind of implies, otherwise its just processing of data
<fare> most people do it in an even simpler way, by translating object access into variable queries.
<Zhivago> fare: well, the swizzling I'm referring to happens as part of the read process, and hooks into a demand loader for foreign objects
09:00am
<fare> stop thinking with fixed limits between what the "hardware" and what the "software" do. From the user's point of view, it's just the whole "system" that does it.
<Zhivago> fare: yes, I understand all of this
<fare> all hardware tricks are to be considered; all software tricks are to be considered.
<fare> Zhivago: good.
<Zhivago> fare: but there is a reason we have hardware, so we can have reasonable perforamnce
<fare> not forcibly
<Zhivago> fare: which is why I fully intend to use it :)
<fare> RISC has shown that less hardware can lead to more performance
<Zhivago> fare: RISC systems still tend to use MMUs for some reason :)
<fare> the MMU aspect is independent from the RISC aspect.
<Zhivago> quite :)
<Zhivago> the MMU is the major hardware optimisation that I'm looking at using though
<fare> and sometimes, you wish there was better support for manual faulting, manual cache management, etc.
<Zhivago> and it makes a huge difference in terms of read/write barriers especially
<sr> fare: do you think the Merced will upset the world of processors?
<Zhivago> sr: I suspect the psx2 will have a greater impact, although indirectly.
<sr> zhivago: are you a fan of console gaming systems?
<fare> sr: not quite
<fare> sr: if I want a 64-bit chip, why not already take Alpha, or Sparc64, or HP PA?
<sr> Assembly HOWTO, by François-René Rideau <fare@tunes.org>. Information on programming in x86 assembly. Updated 22 June 1999. 
<fare> sr: 0.5a has been updated 1 August 1999
<sr> must make you proud to see your name in the HOWTO list
<fare> sr: no more; the HOWTO is now maintain by Konstantin Boldyshev
<fare> or the lightning project
<fare> s/or/of/
<Zhivago> sr: not particualrly, but I see the era of the pc coming to a close
<Zhivago> sr: I don't think pcs can compete 
<fare> abi: lightning is at http://lightning.voshod.com/
<sr> because the Alphas, sparc64s and HPs are hard to obtain, not widely supported, and undocumented
<sr> zhivago: Dreamcast :)
<Zhivago> sr: merced wil be aimed at the high end
<sr> did you see the dreamcast?
<fare> they *are* documented (at least the Alpha and the sparc64)
<Zhivago> sr: dreamcast doesn't have what it takes
<fare> I hate the sparc64, but the alpha seems quite nice.
<Zhivago> sr: which is an ability to replace the pc at a much lower price
<fare> Linux/Alpha works just great
<sr> i'm debating whether to stick with intel for OS dev or buy an Alpha for os dev
09:10am
<Zhivago> sr: 32 megs of ram, dvd, usb, possible firewire does though.
<fare> and the Alpha will probably work on those K7 mobos, which will make it more affordable...
<Zhivago> hopefully alphas will get cheaper
<fare> its being pin-compatible with K7 will help a lot, I wish.
<sr> but i can't buy an alpha as i can an intel chip (www.pricewatch.com)
<Zhivago> it would be good, x86 is an ugly machine
<Zhivago> sr: do you expect merced to be different?
<fare> inhappily, an alpha in a X86 motherboard might be ugly, too :)
<Zhivago> sr: they're aiming it at high end servers
<fare> Zhivago: Merced will 1) be ugly 2) still have ugly hw compatibility.
<Zhivago> fare: at least it shouldn't be register starved
<Zhivago> fare: and have such a fuggly instruction set
<fare> self quote: As long as software is not free, we'll have hardware compatibility, hence bad, expensive, hardware that has decades-long obsolete design
<sr> Status: Semi-Supported, Driver Name: 3c501 
<sr> This obsolete stone-age 8 bit card is really too brain-damaged to use. Avoid it like the plague. Do not purchase this card, even as a joke. It's performance is
<sr> horrible, and it breaks in many ways. 
<sr> hahaha
<sr> merced won't be very different in those aspects from the current intel chips
* fare/#Tunes is against hardware compatibility
<fare> I'm for software compatibility
<Zhivago> source compatibility should be sufficient
<fare> and declarative software too
<sr> i declare this land in the name of Tunes!
09:20am
<sr> smkl how's finland?
<smkl> how is lameness?
12:10pm
<sr> heh
[ TUNES     ] [ smkl      ] [ hcf       ] [ sr        ] [ AlonzoTG  ] 
[ eihrul    ] [ Plundis   ] [ fare      ] [ Crimson_  ] 
12:30pm
[ TUNES     ] [ smkl      ] [ hcf       ] [ sr        ] [ AlonzoTG  ] 
[ Crimson_  ] [ fare      ] [ eihrul    ] [ Plundis   ] 
[ TUNES     ] [ abi       ] [ smkl      ] [ hcf       ] [ sr        ] 
[ AlonzoTG  ] [ Plundis   ] [ eihrul    ] [ fare      ] [ Crimson_  ] 
<dtm> yay
<smkl> ltns dtm
<dtm> !!!
<dtm> it's good to be back!! :)
<dtm> how are you ladies and gentlemen
<smkl> just like before
12:50pm
<dtm> tunes has changed its title  :}
<dtm> what made it expedient?
<dtm> did Fare get a little pep up, like maybe a lobotomy?
<dtm> poor poor Fare    my girlfriend is a psychology PhD student so I'm refreshing myself on my sympathy
<fare> dtm!!!
01:10pm
<dtm> alfkajs;alfajad
<dtm> FAIR!
<fare> dsfskjh oitus nsdfsa ldf!
<fare> datum!
<dtm> :D
<dtm> fare:   have you moved to CA yet
* dtm/#tunes checks fare's /whois
<dtm> i guess not!?
<dtm> >:-/
<fare> As I told you, I need 1) to finish my PhD thesis, and 2) to find a job
* dtm/#tunes is a personal supporter of famous starving students
<dtm> yes i remember that quite well
<dtm> you know what
<dtm> when i explain TUNES to people, most of them dont get it
<dtm> and those who can, dont tend to care
<dtm> i think it's the "NES" part
<dtm> combined with their "short attention span"
<fare> what do you try to explain them?
<dtm> i tell them that current software is junk
<dtm> especially considering that I am a gnu/linux y2k "specialist" now for Cygnus
<dtm> i'm thinking that my software should be able to tell me when it has a date problem
<fare> the N was recon'ed into Nevertheless
<fare> s/recon/retcon/
<dtm> yes i see
<dtm> the former N is what got them
<fare> you work for cygnus? great!
<dtm> to them, the definition is relative
<dtm> i'm contracted by cygnus
<fare> oh, contracted.
<dtm> RMS has appointed me the coordinator of canonical GNU y2k info
<dtm> i dont want to work here; i want to work at Apple!
<fare> so I guess you closely follow anything Y2K related...
<fare> at Apple???
<fare> but they are evil proprietary computer makers!
<dtm> haha   i closely follow the fact that y2k is a mental/social problem, not a computing problem
<dtm> yes and french people never bathe!
<fare> the only good guy there was Woz, whom they've buried in secondary projects
* dtm/#tunes cajoles and ribs fare
<dtm> get with the 90's, man.  John Sculley was fired :}
<dtm> do you realize how much source code apple has liberated
<dtm> http://publicsource.apple.com
<dtm> i have the entire source code to all sub-Finder portions of MacOS X Server
<fare> not that much.
<dtm> everything below the top user interface.  i assume it includes the window system too.
<fare> and their license was far from agreable
<dtm> hmm:/
<dtm> well then i'll influence it :)
<fare> I hope so
<dtm> is the APSL still viewed as "free bug fixes"?
<fare> dunno; haven't followed it all.
<fare> Just fuck'em for inventing yet another license.
<dtm> then why was it dissatisfactory in your opinion
<dtm> ah
<fare> Seeing how their stuff is 90% BSD, they should have kept a BSD license!
<dtm> care to educate me?
<dtm> ok
<dtm> is it true that the BSD license is close to public domain, in that it does not prevent someone from taking away freedoms, like the GPL does?
<fare> or maybe make a political move and be GPL.
<fare> Be don't need license fragmentation
<fare> (see ~fare/articles/ll99/)
<dtm> ha ok
<dtm> you come prepared
<fare> s/Be/we/
<fare> prepared?
<dtm> how very scientific
<dtm> got documentation and everything.
<fare> :-~
01:20pm
<fare> new self quote: Karl Schmidt said that the limit between philosophy and politics is when you have to choose your friends. Choosing the GNU GPL as a software license is definitely a political act under this criterion.
<dtm> hah
<dtm> i'm not sure about that
<dtm> can you briefly explain why
<dtm> I think gpl allows for everyone to play along.
<fare> the GPL is kind of a mutual friendship agreement.
<fare> The kind that was done being trading cities in the middle-ages: no one of us will impose barriers at our borders
<dtm> explain the barriers in a software context?
<dtm> afaik it just prevents outright theft, or anything less than total freedom of a unique, free product.
<dtm> it even encompasses playing along with nonfree products
<dtm> correct?
<dtm> i was thinking that it didn't require conformity in any particular way.  anyone can still implement a gpl'd product anywhere they want as long as they fully implement it.
<fare> IP barriers. Did you read my ll99 paper already?
<dtm> negative
<dtm> it shall be done@!
<fare> maybe it'll help a bit
<dtm> and me and my girlfriend are wanting to travel to europe together Someday...  we'll stop by your place.
01:30pm
<fare> you'll be welcome!
<fare> be sure to warn in advance, so I may not be abroad myself :)
<dtm> http://www-th.phys.rug.nl/~schut/gnulist.html
<dtm> that document , and the document behind the "booby trap" link at the bottom, are the most useful/meaningful y2k documents i've ever found in my 2 weeks of professional research!
<dtm> when i come visit i'll bring you some American object-oriented multitthreaded garlic pickles
<dtm> from the refrigeration section -- never boiled
<dtm> and then we shall capture you and make you our chauffeur!
<dtm> do you have a heavy french accent?
<dtm> mmmmm pickles
<fare> did you see that recent Y2K thing on /. ?
<dtm> yes 
<fare> when I speak english to french people I do have a heavy french accent;
<dtm>  /. generally regards y2k as a hoax/joke/annoyance
01:40pm
<fare> when I speak english to english-speaking people, I may be thought a native english speaker for a few seconds.
<dtm> wow.
<fare> only a few seconds, tho.
<dtm> when I speak french to french- or english-speaking people, i'm usually thought of as silent.
<fare> for I fumble on very common words.
<dtm> :-/
<dtm> i see
<dtm> well you type it very well
<fare> my written english is better than my spoken english
<fare> sometimes, I have to ask the person I'm talking to to spell a word he uses, and I find out that it's a word I write very often, but could not pronounce or recognize properly!
<fare> is there a single word for "person I'm talking to" ?
<dtm> that's fascinating
<dtm> hmm... "audience"? :)
<dtm> listener?
<dtm> recipient?
<fare> dunno. "vis-a-vis" or "interlocuteur", in french.
<fare> a powerbook G3 series?
<fare> could you run Linux on it?
<dtm> indeed
<dtm> Linux powerbook.lightsaber.net 2.2.6-15apmac #1 Mon May 31 03:54:09 EDT 1999 ppc unknown
<dtm> i had MacOS X Server on it for a week, as well
<fare> rats. I couldn't get internal IDE to work with those precompiled kernels.
<dtm> also there's a kernel patch and little applicatino that lets you boot another virtual machine, so you can run linux or MacOS as a linux app
01:50pm
<dtm> on which model?  this is G3 *Series*
<fare> was told to install MkLinux instead, but currently, my mom has it.
<fare> yup, G3 Series.
<sr> dtm: what made you switch from macos X to linux?
<fare> where do you get the "right" kernel sources for ppc?
<dtm> fare from any kernel mirror
<dtm> almost none of the extras that are available from linuxppc-dev list are necessary
<dtm> they're optimizations.
<dtm> or something.
<dtm> not necessary for general utility use
<dtm> sr:   1) linux is totally free, and compatible (X11 is a big one)   2) macOS X has no serial or ppp
<dtm> powerbook is my #1 computer.
<sr> did you like MacOS X server, though?
<dtm> i take it everywhere
<dtm> sr:   yes i took it apart thoroughly and appreciated it very much.  it's essentially NeXTStep for ppc
<dtm> it was very  easy to use
<dtm> i thought the user interface abstraction layers were done well
<dtm> kind of a dual standard in some ways
<dtm> but NeXTStep is a bit funky -- has some weird ideas that may be technically superior, or that may just be their own standard, but that aren't very common
<dtm> like NetInfo.....yuck
<dtm> it's the kind of thing that either everybody, or nobody, should adopt
<dtm> imho
<dtm> and in the opinion of many MacOS X users inside apple
<dtm> including developers
<sr> how does it compare to BeOS?
<dtm> but it is definately a gnu platform
<sr> Jean Louise Gassee is from apple :P
<dtm> BeOS is a funky gnu platform, like os/2
<dtm> beos is "different"    less unixlike    less legacy   
<sr> so are you into OS dev/language dev?
<dtm> they could all unite better, and probably won't without full integration of open source ideology
<sr> i've heard that BeOS's saving quality is its speed
<dtm> sr:   i'm as into it as a non-developer possibly can be, and more so than most
<dtm> sr:   and user friendliness, and economy of proprietary software... like a good photo/paint package costs $20 for beOS
<dtm> it's technically very robust
<dtm> one of the best OO/relational filesystems
<sr> i'd like to steal its speed ;)
<dtm> it's like OS/2 times ten, and with much less legacy/compatibility with ignorant ugly crap
<sr> and put it to use in my OS
<sr> it's purified essence of OS/2
<dtm> sr:   what's the name of yours?
<dtm> sr:    yes i guess that's a good description of beos
<dtm> sr:    but that's a very simple statement that means a lot -- it describes what it does and doesn't have.  beos is unixlike compared to os/2
<sr> haven't decided on a name, i can't tell you much about it because i've a) not started it and b) I probably won't finish it :)
<fare> JL Gassee is a stupid arrogant french man!
<dtm> sr:   heheheh
<dtm> fare:    yes down with french men!
<fare> OS/2 sucks, so does MacOS/X.
<dtm> fare:    you suck!
<fare> no, I *plain* suck
* dtm/#tunes waves torches in the air
* dtm/#tunes accuses the heretic
* dtm/#tunes persecutes
<fare> dtm: so, didn't you experience problems with the G3 series? With model of series is it?
<sr> witchhunt
<sr> dtm: get one of those neat iMacs =)
* fare/#Tunes has that horrid one with bad LCD, and no secondary cache.
<fare> iMacs suck
<dtm> fare:   absolutely none.  installing linux on an Apple machine is the best linux installation i've ever seen.  the only OS installation, PERIOD, that is better, anywhere, is macos.
<fare> macos installation is HELL!
<dtm> fare:    no problems.  the only configuration i did of any kind was hard drive partitioning, which i dont even have to do any more.
* fare/#Tunes doesn't want to see that again
02:00pm
<dtm> fare:    it didnt' ask and i didn't tell -- everything just works.
<fare> wierd. What model is it? What linuxppc version? R4? R5?
<fare> what kernel, initially?
<dtm> fare:    except power management and display switching.  but it inherits any display from MacOS -- ntsc, svga, etc
<fare> the kernel in R5 won't properly recognize the internal IDE -- it will generate partition corruption after a few hundred megs of transfer
<dtm> it's the Wallstreet model, from March 1999, wiht linuxppc R5 (used to have R4)
<dtm> fare:   i've heard of that -- try the FAQ-o-Matic at http://linuxppc.org
<fare> did the R5 kernel work?
<dtm> "interactive faq"
<dtm> everything works.
<sr> abi: Excelsior is a computer game similar to Ultima VI, at www.11d.com
<dtm> media bay hotswapping, etc
<fare> why play a game "similar" to U6, when you can have the real thing?
<sr> heh
<sr> diversity
<dtm> sr:   i saw an iMac with linuxppc and Applixware, at linuxworld expo
<sr> improvement
<sr> imacs just *look* cool
<fare> the faqomatic sucks!
<fare> but under the hood, they are lame
<sr> although i wouldn't really mind if someone gave me a new Powermac G3 with MacOS X server
<fare> dtm: how much RAM do you have? Do you also have a secondary-cache-deprived model?
<dtm> they're working on USB 1.6
<dtm> or is it 1.16
<dtm> i think it's 1.16.
<sr> fare: do you know PowerPC assembly?
<fare> sr: essentially, no.
<fare> dtm: what question in the faq-o-matic?
<fare> that's the nth time I look through it, w/o answer. I know someone else with same trouble, too
<fare> are you using a precompiled kernel? which?
<fare> if not, can you send me yours?
<dtm> fare:   i have 512K l2 cache and 96MB ram
<fare> maybe lack of L2 cache explains it
<fare> I have 96MB RAM, too
<dtm> fare:   i got a special deal...buy a floppy drive and get free 64mb ram.  other than that it's totally stock.
<dtm> i'm using R5's kernel
<dtm> Linux powerbook.lightsaber.net 2.2.6-15apmac #1 Mon May 31 03:54:09 EDT 1999 ppc unknown
* dtm/#tunes has a ppc book, including assembly
<fare> I got a similar deal, except that it was the cheaper model w/o sec cache
<dtm> fare:    how can you be so harshly critical on something like iMac  :)
<fare> and with lame  LCD
02:10pm
<dtm> btw mine was $2500, or $3000 plus tax and floppy
<fare> your uname -a doesn't help me much. Can you point me to the right vmlinux.gz (and associated modules files) ?
<dtm> fare:    you know that i run around calling things crap and generally speaking in generous amounts of hyperbolic verbosity -- but why do you just say iMac sucks
<dtm> fare:    it's potently powerful, compares favorably or superiorly to everything else on the market, created its own class, looks good, is built on ppc, runs linux, and is affordable to many :)
<fare> superiorly to what? A half cheaper PC is more powerful!
<dtm> fare:    plus is easily administerable, is a good thin client for MacOS X Server's netboot feature (thanks to OpenFirmware)
<dtm> fare:    now you're eliminating quality
<sr> fare: do you own an Alpha?
<fare> dtm: a same price PC is both more powerful and better quality
<fare> sr: not yet
<dtm> fare:    there is no pc that is more "powerful" than iMac.  
<dtm> iMac does exactly what it is intended to do, about as succinctly and affordably as possible.
<dtm> those ways in which peeseas are cheaper, they do not do as much.
<dtm> not to mention they can't do anything as well as a non-peesea because of their crippled architecture.
<dtm> not to mention that they are inherently less accessible and harder to maintain.
<ElGato> abi: where is gc?
<abi> bugger all, i dunno, elgato
<ElGato> abi: gc?
<abi> elgato: bugger all, i dunno
<ElGato> grrr
<ElGato> abi: garbage collection?
<abi> i heard garbage collection was meant for that
<dtm> in many simultaneous dimensions, apple's machines make irrelevant/obsolete the illusion that peeseas are worth anything.
<ElGato> oh geez
<dtm> and apple is merely an example.  find something that does not!
<fare> i find desktop PCs quite nice to maintain -- I upgrade them bit by bit, and five years later, they are still there. A five year old Mac is a useless piece of junk.
<dtm> i can get you an SGI Indigo^2 starting at $400 complete (minus monitor) with a warranty, so how do you like that?
<hcf> abi: gclist?
<abi> gclist is at http://lists.tunes.org/cgi-bin/wilma/gclist
<fare> that doesn't help me install linux on my mom's pbg3series :(
<dtm> fare:    well you're completely wrong and above all irrational :)
<sr> actually, using a five year old mac as a log machine is feasible :)
<ElGato> thankya kindly hcf :)
<hcf> ElGato: did want that?
<hcf> s/did/did u/
<ElGato> yes me did want that ;)
<dtm> fare:    piecemeal upgrades are not easy maintainability.   obsoleting the need for upgrade, even in common consumer perception, is.  and that's what apple does.
<fare> sr: then you can use a 15-year old PC as a logon terminal!
<ElGato> i would like some general info on gc though
<dtm> mac users do not think that old macs are junk.
<sr> dtm: but indigo is closed architecture and the source to IRIX is not widely available
<fare> EG: read the GC FAQ
<dtm> sr:    but it runs linux.
<ElGato> k
<fare> Review/Languages.html#GC
* dtm/#tunes spanks fare for his insolence
<sr> hehe
<ElGato> dtm!
<fare> dtm: I know a mac user who buys old Macs, and when I come to his place, it's a hell!
<dtm> fare does not think in terms of non-geek or total cost of ownership in terms of hardware
<sr> dtm, spare the rod, spoil the fare
<ElGato> that was uncalled for
<dtm> fare:   well too bad
<dtm> ElGato:   ????
<dtm> I called for it.
<fare> for the same price, he could have bought a second-hand PC on which to run a linux server for his Mac LAN...
<dtm> hence there it was!
<sr> fare: old computer museum
<ElGato> why the only place i know where that is accepted is #c and #nes2k on efnet
<sr> ElGato: you're forgetting #asm
<ElGato> oh yes
<ElGato> #asm too
<dtm> fare:   PC's are *crap*.  ok?  they're not built to have an identity, maintainability, ease of use, sustainability inside their *OWN* architecture, quality, good total cost of ownership, etc
<sr> if only... alphas were as easily obtainable as PCs...
<fare> what's #c, #nes2k and #asm doing in this conversation? did I miss something?
<dtm> fare:   they're meant for you to buy them, like Ginsu knives minus the quality.
<ElGato> n/m
<fare> dtm: PCs are a world of *competition*, ok?
<dtm> fare:    if i want something that's cheap, I'd eliminate TUNES
<fare> this means that the quality/price ratio is *superior*
<dtm> fare:   so is EVERYTHING ELSE
<dtm> fare:    your idea of what's proprietary and closed, is obsolete, irrational, and unprovable :)
<fare> no, Mac is a world of proprietary monopoly
<dtm> fare:    you're a nut, buddy :)
<ElGato> someday im gonna collect some 6502's  and build a computer from scratch
02:20pm
<sr> i want to buy an R12000 and build an SGI Octane from scratch
<sr> custom SGI lowrider
<fare> EG: once I dreamt about a bunch of 6502 with virtual memory coprocessor...
* ElGato/#tunes *drewl*
<dtm> fare:   i think you need better cache coherency there buddy because the world has evolved since apple had a primary identity being proprietary   that ended with this decade.
<sr> dtm: ever heard of IBM supporting open source? NO!
<dtm> fare:    Nintendo did pretty good with memory management hardware in 6502 (nes) and Z80 (gameboy)
<fare> dtm: they made a timid attempt at allowing clones, and rapidly backed off
<dtm> sr:    what?
<ElGato> for some reason im not authorized to go on 15 different efnet servers
<fare> dtm:  so any hint on getting Linux to run on that PBG3series?
<ElGato> im writing a funny little forth compiler for the nes (for writing emulated games) :D
<dtm> fare:    yes but I CAN MAKE A MACINTOSH CLONE     your generalizations are just as silly and ignorant as when I pretended to say that all French people smell bad.
<dtm> fare:   i have no idea what's wrong with your pb  :)
<sr> ElGato: did you ever consider writing games?
<fare> (needless to say, both the faq-o-matic and the mailing list proved useless)
<dtm> fare:    you should tell all this to the LinuxPPC people   i'm sure they'd want to know that you've decided that all their accomplishments are impossible.
<fare> sr: I did, and have a few drafts (in french)
<ElGato> the only games i write will be for nes emulation
<fare> from 1991
<ElGato> and that's that
<fare> impossible?
<dtm> ElGato:   do you have any games youve written?
<ElGato> 8 bits of pure enjoyment
<fare> what's impossible?
<ElGato> no...im just writing the compiler
<ElGato> just as a pet project to get some experience in compiler writting
<ElGato> before i write the big one (hopefully)
<sr> ElGato: how can i find out how to write games/programs for the NES (emulator)?
<ElGato> um
<dtm> fare:    yes     a proprietary anticompetative platform has miraculously produced some of the biggest open source developments ever, inside and outside the corporate world, including BOOTHS of powermac and powerpc clones (which run MacOS) at Linuxworld Expo
<ElGato> just learn how to rom hack
* fare/#Tunes wants a game named refhack, or RHACK, or something like that, in which you impersonate a guy who's trying to crack a computer...
<ElGato> i don't know yet but i know a bunch of nes hackers
<ElGato> :D
<dtm> fare:    i saw a very small embedded 4-cpu ppc board running MacOS last March at linuxworld expo
<dtm> fare:    custom built, commercially supported
<sr> you remind me of those authors who always talk of writing "The Great American Novel"
* ElGato/#tunes wants a forth machine
<dtm> sr:   who are you talking about :)
<sr> ElGato
<dtm> aha
<dtm> :)
<ElGato> :D
<ElGato> i love you all :D
<fare> dtm: so they accepted clones again? Or maybe those people found a way accross the ROM incompatibility?
* dtm/#tunes forgives fare for his silliness and bigotry because he's so fucking smart
<ElGato> sr: before i get into anything big i need some practice
<ElGato> bigotry!!!
<fare> maybe OpenFirmware and Darwin means that at last people can make mac clones?
<ElGato> ok everyone has a little bit of bigotry in them but fares no bigot!
<dtm> fare:     you suffer more than just mild delusion and self-perpetuated misinformation, my friend.  there is no "incompatibility".  and Apple doesn't have the *ability* to stop cloning!
<ElGato> brb
<dtm> fare:    can you just try applying logic to the situation?  how could apple possibly stop anyone from making a mac clone?
<fare> dtm: it *did* stop the clones, by refusing to license its ROMs anymore.
<sr> fare: if you find out, could you tell me where to buy Alpha processors?
<fare> dtm: and its MacOS used to not run unless you had the Apple ROMs
<fare> sr: dunno. ask google
<dtm> fare:    well      you dont need MacOS just to use those machines anyway
<sr> MacOS is evil software
<dtm> fare:     and you can get a MacOS ROM on your own.
<fare> dtm: if you copy a ROM, that's illegal. In practice, the clones were killed by the ROM non-licensing
<dtm> fare:   it makes little difference in practicality, whether apple allows an oem to distribute macos by default
<dtm> fare:   peopel can still do it.
<fare> dtm: did you read alan cox' article "I don't care if space aliens ate my mouse"?
<fare> see it on opencontent.org
<dtm> fare:   i dont remember   was it about m68k mac linux
<fare> dtm: but that means no one can buy a clone but someone who knows how to copy a ROM...
<tmf> [ hi ]
<fare> tmf!
02:30pm
<HickServ> sr: lengua shall own you
<dtm> fare:    but i've had enough of this silliness about proprietary architecture when you can get absolutely every piece of documentation straight from Apple's web site, for free, on how to write your own os for their hardware
<fare> dtm: that reduces the clone market to about zilch
<fare> dtm: proprietary means that you don't have right to copy
<HickServ> sr: you know now my ideas for the language are slowly becoming for like yours
<dtm> fare:    Not having any money is the only thing that can prevent someone from buying a Mac clone.  
<fare> it does not (necessarily) mean undocumented
<fare> as for documentation, read that alan cox article...
<dtm> fare:    then dont use proprietary things.  their hardware is not proprietary.
<dtm> fare  their hardware is *strictly* commodity pieces
<dtm> fare:     even ATI is doing linux development now
<HickServ> brb
<HickServ> er bbl
<dtm> fare:    besides, even with no video driver I have full use of my powerbook's video
<dtm> fare:    with any given piece of apple hardware i have a MUCH higher likelihood of compatibility, reliability, and ease of use under Linux, than is possible by throwing a dart at a random group of peeseas.
<dtm> not to mention open standards
<ElGato> looks like no one wants to quake right now :(
<dtm> it's built on Open Firmware, powerpc, fully documented parts and assembly, etc
<sr> quake1, 2, or 3?
<ElGato> quake 1
<fare> dtm: don't take a *random* PC. Take one made of pieces known to work with linux. Go to a linux PC vendor.
<fare> you'll get a machine more powerful, more reliable, more scaleable, and cheaper, than a Mac.
<fare> that's competition
<ElGato> fare: i was thinking about code factoring
<fare> EG: yes?
<dtm> fare:   you're nuts :)
<ElGato> fare: i was thinking about factoring my code in my lexer so there are inlines that act as operations over regular expressions
<ElGato> would that be a good idea?
<fare> dtm: oh, and I *do* have a video driver for my PC laptop, which is much better, faster, more reliable, hassle-free, etc, than my mom's PBG3series.
<fare> anyway, let's forget about Apple.
<tmf> i like the brightly colored transparent look, though!
<fare> dtm: you didn't answer my question about the kernel image you use...
<fare> tmf: I don't. Computers should be discreet, not flashy
<ElGato> fare: what do you think about my idea?
<fare> EG: I'm not sure what it means exactly
<fare> EG: you wanna partially evaluate the lexer and the parser?
02:40pm
<ElGato> you know like union, star, contecation
<fare> EG: that's a good idea, but maybe doing it with a metaprogram rather than by hand...
<ElGato> like lex?
<fare> for instance.
<fare> or your own metaprogram, if you don't wanna use lex.
<fare> or ocamllex
<fare> ocamllex is a good idea...
<ElGato> like make my own lex builder?
<sr> lex, lex, lex
<fare> ocamllex, ocamllex, ocamllex
<dtm> fare:   i answered your question several times.  it's the only one that comes with the R5 installer
<ElGato> myownlexer, myownlexer, myownlexer
<fare> the vmlinux.verystable.gz?
<dtm> dtm+tmf=dtmf
<fare> there comes several images.
<ElGato> :D
<ElGato> dtm: you into fones?
<fare> and none works with the internal IDE :(
<dtm> ElGato:   not like that :)
<ElGato> uh...ok :D
<ElGato> i would still like to write my own lex maker
<ElGato> that be cool
<fare> sure
<fare> but C is the wrong language for that, I hope you realize it!
<tmf> elgato, a new lexical analyzer or a reg-exp lexer clone?
* ElGato/#tunes HUG fare
<tmf> C is the wrong language .
<fare> HUG?
<ElGato> yes
<ElGato> yes
<ElGato> hug
<ElGato> fine
<ElGato> i shall write it in scheme and such\
<ElGato> if i decide to write it
<tmf> elgato, what particular additional feature did you want?
<ElGato> probably nothing...although lex doesn't have any features for regular expressions
02:50pm
<ElGato> you just name the tokens
<tmf> ? you define regular expressions and call then stuff
<ElGato> call then stuff?
<tmf> give them names
<ElGato> keywords?
<abi> keywords are the only type of filenames in brix
<fare> lex definitely has regular expressions.
<ElGato> it does?
<tmf> isn't that good?
<fare> what's an efnet server?
<smkl> irc.concentric.net
<ElGato> irc.primenet.com
<ElGato> i guess lex is good enough
<ElGato> does lex produce nfa's or dfa's?
<fare> no, use ocamllex instead
<fare> it directly produces dfa-based finite-state machines
<tmf> elgato, yes it eventually gives you an dfa
<ElGato> good
<sr> sexp?
<abi> sexp is, like, Symbolic EXPression
<fare> sexp is also just SEX renamed no to scare people
<abi> okay, fare.
<ElGato> what's so great about ocaml anyway?  
<fare> EG: what's so great about C anyway?
<ElGato> i asked first
<ElGato> :D
<fare> but we're LIFO
<fare> :)
<ElGato> LIFO?
<fare> LIFO is Last In, First Out
<ElGato> damn those stacks!
<ElGato> fine
<fare> LIFO is also the way stack data structures work
<abi> okay, fare.
<ElGato> ok the only reason i want c is because anyone can compile it easily but oh well
<fare> well, anyone can easily compile ocaml
<fare> (well, I heard that the ocaml rpm would screw SuSE)
<ElGato> is there ocaml on most linux and other unix distributions?
<fare> when it isn't, you can compile it out of the box
<tmf> [ gottogo ]
<ElGato> i'd rather it be compiled inside
<fare> something like tar zxvf ocaml.tgz ; cd ocaml ; ./configure ; make world ; make install
<ElGato> hmmm
<ElGato> why should i pico ocaml over scheme?
<ElGato> pick
<fare> maybe
<liar> fare: u know anything about java?
<fare> I'd say yes, unless you want reflection
<fare> liar: enough to not want to use it unless forced
<ElGato> i learned java and then quickly forgot
03:00pm
<ElGato> fare: i know you probably get this a lot but when will there be any coding done on the HLL?
<fare> EG: when you do it
<liar> do u know if there is a java2 compiler for linux yet?
<ElGato> im not really part of the hll subproject
<ElGato> although i sure have learned a lot lately
<liar> java2 is suppose to be really really good
<liar> so im gonna learn it
<fare> liar: anything is supposed really really good, if you listen to marketeers
<liar> this was from news articles
<ElGato> fare: is there a schemelex?
<fare> no. But there are many similar tools.
<fare> One is in tom lord's page
<fare> or should be
03:10pm
<ElGato> but not a lex builder?
<fare> lex builder?
<smkl> lexical analyzer generator
<ElGato> pardon my incompetence
<fare> something like lex?
<fare> I'm sure there are several written in scheme
<fare> none standard
<ElGato> not in scheme...for scheme
<fare> lex stuff that dump scheme code?
<fare> yes, there are several
<ElGato> :D
<ElGato> i got the itch to write my own 
03:20pm
<fare> maybe one in slib, one in bigloo, one in tom lord's thing, one around here, one around there
<ElGato> yes 
<ElGato> well it would be interesting
<sr> hey water
<ElGato> hi water
<ElGato> D:
<sr> hi water, bye water
03:30pm
<fare> bywater?
<sr> highwater
<ElGato> Vous êtes une bonne dame et je serais honoré pour mettre mon pénis dans votre bout. 
<ElGato> :D
<sr> hahahahha
<sr> well
<ElGato> sr: do you still have roger wilco
<ElGato> ?
<sr> yes
<ElGato> 209.68.229.171
<ElGato> connect
<sr> ok
<fare> sr: which roger wilco?
<fare> dans votre quoi?
<ElGato> sr: can you hear me?
<sr> yes
<sr> it's a walkie talkie app, fare
<ElGato> fare: im getting noughty with bable fish
<fare> something's wrong -- ladies don't have tips
<ElGato> naughty
<ElGato> er
<ElGato> You are a fine lady and I would be honored to put my penis in your butt.
* fare/#Tunes liked space quest IV (or was that III ?)
<sr> Le prix a eu une liaison amoureuse avec son ordinateur.
<fare> prix?
<sr> "fare"
<ElGato> !!!
<water> hey all
<fare> high water again!
03:40pm
<fare> put your boots on, guys!
<ElGato> Si vous n'êtes pas bon je mettrai mon pénis dans votre bout aussi. Vous améliorez non laissé le sope pour glisser. 
<sr> *** water has quit IRC (Connection reset by queer)
<sr> ???
<sr> L'eau est une expression du tao. 
<ElGato> water: enlighten me as to what taoism is about
<water> lag
<water> sorry
<fare> grrr! lynx+squid sux!
<water> taoism _is about effortless action
<ElGato> Avant que l'urinate de I sur vous aiment des urinates de pisses de hcf sur les personnes âgées
<sr> Les hommes courageux meurent une fois, alors que les lâches meurent beaucoup
<sr>                             de fois avant leur dépassement. 
<water> like how water flows downhill without having to consider where to go
<fare> EG talks about urinating, and you talk about water flowing down...
<sr> Before the urinate of I on you like urinates of pisses of hcf on the people old <--- babel fish translation
<ElGato> your bable fish sucks!
<water> what's news, all?
<ElGato> Urinate I sur vous poissons bable
<sr> Le Fare trahit-il son héritage avec l'élégance parisienne?
<fare> ???
<ElGato> !!!\
<ElGato> Pouvez vous vous brûler dans l'enfer Américain modifié. 
<water> lots o lag
>>> smkl [sami@CMLXXXIII.rdyn.saunalahti.fi] requested PING 935532262 946944 from #tunes
<fare> ?
>>> sr [depag@phila-dialup498.nni.com] requested PING 935535008 from #tunes
<ElGato> sr: did you hear that?!
<sr> [water PING reply]: 49 secs
<sr> ya
<water> brb
<ElGato> sr: did you hear ball and chain?
<sr> no
<ElGato> how about that?
<ElGato> are your speakers even on?
03:50pm
<sr> whoa i was afk for a while
<sr> say it again
<sr> i heard "hello"
<sr> "yes"
<sr> whoa
<sr> Roger Wilco kicked me off
<ElGato> 15 minutes
<ElGato> log on again
<ElGato> :D
<sr> hehe
<ElGato> :)))
<ElGato> i suck at singing
<sr> what the hell is this?
<ElGato> cool music
<sr> what is it?
<ElGato> doll- foo fighters
<sr> just get a realaudio server
<hcf> how bout u guys join #rogerwilco
<`shift> Hello :):)
<sr> hello :):)
<ElGato> :D
<`shift> so.. whats happening here then :):)
<ElGato> sr: how do i set up a real audio server?
<`shift> anyone here use 'Sonique' ?
<smkl> what is it?
<`shift> a media player for Windows
<`shift> no?
<`shift> www.sonique.com <-- 'tis better than winamp :):)
<smkl> not many of us uses windows
<`shift> BEOS version coming soon
<`shift> after that the LINUX version
04:00pm
<`shift> what you lot into here.. i new to this server
<`shift> :):)
<sr> OS development
<sr> Language development
<sr> elgato: www.real.com
04:10pm
<dtm> fare:   i read that article by Alan Cox, and he's very extremely irrationally biased.
<dtm> that article was viewed by me and by friends of mine who have NOTHING to do with apple at all (no particular interest of any kind) as irrational.
<dtm> technically ok but emotionally unreasonably biased.
<dtm> I actually read it in the bathroom while i was working at VA Research.
<fare> VA Research?
<liar> they make linux machines
<dtm> i'm reading that article again right now and am appalled at the *totally* immature, unprofessional, unreasonably emotionally biased attitude of the writer.
<dtm> it's something i might have written years ago and probably would have deleted right away once i read it five minutes later.
<fare> :)
<dtm> the most impressive thing is that every single last one of the irrational conclusions he comes to about Apple or Macintosh, can be replaced to describe a totally open system such as *gasp* Linux.
<dtm> it's a display of false bravado.
04:40pm
<dtm> the conclusions are derived from logical fallacy.  minute technical, social, or ethical mistakes lead to total misdirection, and the blatant underlying hostility doesn't help either  :)
<dtm> i showed it to my friends at VA just to show how embarrassing it was.
<dtm> and so i totally understand why you showed it to me, fare : )
<dtm> >:-}
<water> cool! a film adaptation of 'ender's game'
<water> i hope that they don't screw it up
04:50pm
<ElGato> HI FRIENDS
<fare> lo enemies
<ElGato> D:
<dtm> i'd like to also point out the fact that for a period of a few years in Apple's history, their entire reputation was fucked up.  since John Sculley joined, and until Jobs returned, their public interface ranged from disarray to complete disaster.
<dtm> and it's worthless to hold that against them still, even though a lot of things were permanently obfuscated
<ElGato> brb
<dtm> the reason why it was so hard to port anything to the Mac 68k was because of the proprietary nature of the CEO and board of directors at the time (john sculley)
<dtm> and the reason why it's not easy now, is because it was not easy then.
<dtm> they weren't dripping with detailed external documentation then.
<dtm> not even Motorola still has full docs for related chipsets such as 68881 and friends
<dtm> who complains about the Atari or Amiga?????!?!
<dtm> neither had the direct and broad impact on the human civilization that Apple's machines have had.
<dtm> which is another entirely different unfortunate issue.
<dtm> people cant FIND docs, in or out of Apple, that describe the DMA and memory controllers of the Mac IIfx and some Quadras.
<dtm> not even the apple engineers cared much as long as the software group booted their own OS on it.
<dtm> but those docs that do exist are completely public now, and have been for years.
<dtm> i was on the mac 68k port team.
<dtm> i bought a IIsi as my main machine just so i could suffer through that.
<dtm> i spent hours on the phone for them.
<dtm> and Alan was not there.
<dtm> Rob Pelkey and friends were.
<dtm> Alan was *not* on the developer mailing list.
<dtm> not during the time i was.
<ElGato> IM BACK FRIENDS
<dtm> GREETINGS MISTER CAT
<sr> is Alan Cox a Linux developer?
<dtm> WE WELCOME YOU BACK WITH RAISED TAILS
<dtm> sr:   yes he's second in command, to Linus.
<ElGato> meow
<dtm> i might add that Linus is just as irrational but by contrast is **HONEST** about it.
<ElGato> cause linus get cox from him :D
<dtm> whuh?
<ElGato> hehehe...lame joke
<sr> looks like Cox wants a piece of the Apple pie
<dtm> i got chur pie right here :}
<dtm> that Fare had *better* read all that shit i just typed!!!
<dtm> or i'm gonna kick his buttocks one day.
<sr> he purports to hate them and irrationaly does so, yet he secretly idolizes them
<dtm> sr:   i think so.
<water> i agree also
<dtm> sr:    in a strange and quirky way, apple has a completely unique identity and culture.
<dtm> sr:    right or wrong, their stuff fits in and stands out.
<dtm> sr:    it changes the shape, pushes the edge, etc
<dtm> CHALLENGES "THE MAN"
<dtm> mows the lawn
<dtm> shaves the dogs
<dtm> waxes the chickens
<dtm> cooks the liver
<dtm> *ahem*
<ElGato> steve jobs would be a cool boss when isn't having mood swings
<sr> mmm liver
05:00pm
<dtm> ElGato:   yer damn right coz otherwise you'd wish you'd never met him.
<sr> wozniak is cool
<water> hopefully, my idea for starting a company will do better than they have
<dtm> The Woz.
<ElGato> water: hire me :D
<dtm> water   go right ahead -- few companies' successes and mistakes have been as widely and accurately documented as those from Apple.
* water/#tunes idolized those two when much younger.
<dtm> water:    apple is a real model to learn from, mostly on how to not do things
<dtm> water me too
<ElGato> water: i'll work for $2 and hour
<dtm> water:   i know now better than to idolize anyone.
<dtm> ElGato:   where do you live
<sr> water: what's the idea behind your company?
<ElGato> san diego, CA
<water> i remember drooling over a NeXT box
<dtm> ElGato:   i'm in Sunnyvale
<dtm> water:   heheh
<water> "technology for the mind"
<ElGato> hmmm....central valley?
<dtm> ElGato:   yessir
<ElGato> cool
<dtm> ElGato:   come up and visit..  i know you HAVE to be cool if you hang out in #tunes!
<dtm> ;)
<ElGato> i've been up to fresno a few times but that's it
<ElGato> dtm: how old are you?
<dtm> ElGato:   do you have any friends/family in the bay area
<dtm> ElGato:  22 yrs
<ElGato> my aunt is in reno
<water> a company that would make "toys" and systems that would revolutionize their fields by applying the kinds of concepts i have in mind with my "arrow" idea
<ElGato> closest :D
<sr> water: OSes too? :P
<dtm> water:   what is 'arrow'?  that was the name of some software.
<dtm> ElGato:    where's reno?
<ElGato> dtm: i'll come up if you buy me beer
<dtm> ElGato:   negatory
<ElGato> nevada
<ElGato> close to CA
* water/#tunes envisions the day when he can market a six-dimensional puzzle to kids that they would love.
<dtm> ElGato:   no mind-altering or blatantly toxic substances
<water> abi: arrow?
<abi> arrow is a homo-iconic information representation system intended to support programming and human languages in the way that Tunes should. at http://www.tunes.org/papers/Arrow/, http://www.tunes.org/~water/
<dtm> water:   bad ass   i'll take a hundred bucks' worht!
<dtm> abi:   wow
<dtm> abi's really smart
<ElGato> oh...well why else would i want to visit with someone over 21
<ElGato> j/k ;)
<dtm> ElGato:   yo'ure a fool!  how old are you
<ElGato> 14
* dtm/#tunes pokes elgato in the head
<ElGato> :D
<water> i'm trying to get some really good academic papers written to detail the theory now
<ElGato> i love you too
<dtm> well then how would you get here?  sailing?
<ElGato> i won't
<dtm> ElGato:   have you contributed to tunes other than your lively conversation
<ElGato> well if you consider retro part of tunes then yes
<dtm> what would you want with beer?  to drink it?
<dtm> ElGato:   what is retro
<ElGato> a pet os of a tunes developer
<ElGato> forth based os
<ElGato> pretty text colors
<ElGato> ;)
<water> be back in an hour... have to pick my roommate up from work
<liar> ElGato: u havent helped out with retro
<ElGato> i tried
<ElGato> but paul messed it up
<ElGato> by putting so many errors in the code
<ElGato> i can't test my code
<liar> hmm so since i looked at the retro cvs i also helped out with it?
<sr> liar: have you helped the TUNES project in any way?
<ElGato> um i am helping
<liar> remind me to have pyro add my name to the credits
<ElGato> i just haven't tested my code
<ElGato> :D
<liar> sr: i guess my presence here is helping the tunes project too :)
<ElGato> liar: you are mean
<ElGato> :(
<sr> liar you're like me but not clueless =)
<hcf> liar: u sure help a lot by lying to newcomers about tunes
<sr> <_QZ> Tunes: the musical OS! the motto: Just sing it!
* dtm/#tunes snorts
<ElGato> <liar> tunes is a musical os
<ElGato> hehe
<dtm> oh i get it.
<dtm> musical.
<liar> hcf: so when i said tunes does not exist and will probably never exist, that was all just a lie? :)
<ElGato> liar should be kix0red
<dtm> we should keeeel heeem.
<dtm> liar:  blasphemer!
<liar> fare himself said he wants tunes to catalog his mp3's
<dtm> probability imho is an ancient and stupid lie.
<liar> that was even on the web page
<dtm> liar:   neat
<dtm> liar:   so get to work!
<ElGato> liar: keep your BRiX crap out of here
<dtm> far'es mp3's cant wait
<sr> fare has a 3 gig mp3 archive, and according to some, an equally large porn archive
<dtm> oh boy   liar == qz??!
<ElGato> yes
<dtm> HI THERE QZ
<ElGato> hehe
<dtm> i almost dinnet recognize ya!
<liar> dtm: :)
05:10pm
<ElGato> <Fare> or pr0nfix which is having an having it on both sides
<dtm> liar:   i forgot where it was that you live.  i'm trying to rebuild historical context here :}
<ElGato> <fare> pop-11 will give you a woody
<dtm> ok.
<sr> <liar> i live in corvo flores
<ElGato> that porn freak!
<dtm> where is corvo flores
<sr> but we know it's really salt lake city
<dtm> ElGato:   do not underestimate how many kinds of a freak, is fare.
<liar> off the coast of portugal
<ElGato> :(
<ElGato> im just joking
<ElGato> fahree is my daddy btw
<ElGato> :D
<sr> does anyone here know anything about AmigaOS?
<abi> sr: i don't know
<dtm> ElGato:    he's more of a freak that you can POSSIBLY IMAGINE
<dtm> sr:   i do
<ElGato> ebonicslly speaking
<dtm> sr:    I LIKE IT
<dtm> s/that/than
<sr> dtm: can you point me to some explanations/details
<sr> i want to find out about this OS and how it manages to be kernel-less
<sr> or semi kernel-less
<dtm> sr:    yahoo.com !!  http://www.amiga.com  :}
<sr> dtm: are amigas still alive?
<dtm> afaik it's mostly because it's so heavily componentized
<dtm> sr:   they'll never die
<liar> sr: yes, goto www.amiga.com
<dtm> amigaOS has such a strong component model that it can dynamically insert/remove kernel code from ROMs that are autodetected onboard a peripheral.
<dtm> saving ram.
<dtm> and so forth.
<fare> remove code from ROM? wow!
<dtm> I *just* might also add that Macs have been able to autodetect and boot ROM driver code on peripherals since just about FOREVER
<ElGato> fare: hi dad
<ElGato> :D
* dtm/#tunes punishes fare for his insolence
<sr> <fare> hi son. time for your spanking
<fare> how do you do that? It points a gun at the operator, and threatens to shoot if he doesn't plugin a different ROM?
<dtm> fare:   get back to writing your nonexistant but expedient OS!
<dtm> fare:    it can remove KERNEL CODE (i.e. code that is IN KERNEL SPACE) and which has been inserted FROM A ROM
<dtm> shall i use the term "allocate/unallocate"?
<liar> dtm: yes please do
<dtm> liar:   thank you; i will.
<fare> dtm: in what is that ROM code written? If it's x86 or m68k assembly, then too bad.
<sr> l' Fare est un ignorant et smelly un Français. 
<ElGato> !!!
<dtm> fare:   unknown sir
<FareSON> don't talk about my daddy that way!
<dtm> fare:   i would imagine that Amigans had a heavy assembly trend..
<dtm> sr:   that's what i was thinking.   dip him in the brine!!!
<sr> determinate
<dtm> fare:   and i would have to guess that an Amiga-based peripheral's driver was NOT written in x86 anything.
<dtm> sr:   yes thank you
<sr> datum
<dtm> sr:    digital terrain modeler
<sr> what kind of processor does the amiga use?
<dtm> sr:   68k
<sr> isn't that 32-bit CISC?
<dtm> sr:   AHA!!!  but it's not dependant upon an MMU, or 32-bit external addressing -- it runs on 68000.
<sr> wow that amiga spinning globe logo looks really cool
<fare> the amiga doesn't exist anymore
<dtm> sr:   it's 32bit internally, for sure.
* dtm/#tunes beats fare with the largest, hardest, heaviest, closest thing he can find
<ElGato> soy El Gato 
<fare> the 68000 is 16-bit internally
<ElGato> soy tu padre
<fare> but it has 32-bit registers;
<liar> fare: yes it does
<fare> and its 68008 variant has a 8-bit data bus
<fare> go figure!
<liar> fare: the new amiga will have linux as the core of its os too :)
<sr> abi: ElGato is also the love child of fare and liar
<abi> okay, sr.
<dtm> fare:    back to the OS design bench with you!  get to work!
<dtm> sr:   i dont doubt that
<ElGato> liar isn't my daddy
<dtm> abi, who is ElGato?
<ElGato> !!!
<abi> it has been said that ElGato is annoyed when you ask that or the love child of fare and liar
<ElGato> abi: forger ElGato
<ElGato> abi: forget ElGato
<abi> ElGato: I forgot elgato
<dtm> @RL@!K!L@
* liar/#tunes would kill himself if elgato was his son
05:20pm
<ElGato> ElGato is annoyed when you ask that
<ElGato> ElGato?
<abi> you are annoyed when you ask that
<ElGato> liar: you probably don't have the sperm to produce  me
<fare> Linux learnt computing on his 68008-equipped Sinclair XL, IIRC
<fare> s/ux/us/
<sr> hah
<dtm> hahhehahehehehkjlsjf
<fare> dtm: ok
<fare> dtm: where to begin from?
<sr> does the production of elgato require superhuman sperm?
<dtm> fare:   you start with repenting for being difficult!
<liar> sr: retarded sperm
<dtm> fare:   you're supposed to be incredibly bright, not inconvenient ;)
<ElGato> no
<dtm> fare:    you're an inspiration to all of us!
<ElGato> it take superhuman sperm
<ElGato> sr is right
* fare/#Tunes flushes the discussion
<fare> let's go back to tunes
<sr> /clear
<ElGato> and cat eggs
<ElGato> for vice versa
<liar> ElGato: but if u were made from retarded sperm u would think u were superhuman and u would be running around with yer superman cape on :)
<dtm> fare:   just as long as you read the text above that i worked hard to educate you on
<dtm> fare:  that was a benevolant act from me.
<ElGato> sr IIIIID -- - - (: liar
<ElGato> III(:
<liar> ElGato: do u think u could have been made from retarded sperm and yer just in denial?
<ElGato> IIIID   (:
<fare> I read the part of it that was still on my screen when I came back
<ElGato> III(:
<ElGato> IID
<dtm> fare:   then read the rest
<ElGato> actually i lied
<ElGato> you see a forth machine must be fertalized by cat sperm
<ElGato> strange process it is
<dtm> ElGato:   i hear ya
<liar> were u made from cat sperm or do u consume cat sperm?
<ElGato> no no no
<ElGato> made from cat sperm silly ;)
<liar> are u in denial?
<ElGato> that is a loaded question
<ElGato> one in which i will not answer
<liar> :)
<ElGato> liar: do your parents know your gay?
<liar> i will assume that yer in denial then
<liar> yes they do
<ElGato> glad to see your coming out and being proud of it :D
<fare> ok, 1) what language to start from 2) what kind of features to add first?
<ElGato> asm
<ElGato> write the damn interpreter in asm and bootstrap it into a compiler!!!
05:30pm
* fare/#Tunes hits EG
<fare> iepos!
<iepos> fare!
* ElGato/#tunes cry
* fare/#Tunes is talking about HLL, not LLL
<ElGato> :\
<ElGato> Perly!!!
<iepos> fare, are you still planning to implement your lambdaND system?
<ElGato> from #c?
<Perly> elgato, hi, what's this about
<Perly> yeah
<ElGato> :D
<ElGato> hey guys here's one of those freaks...er...guys from #c on efnet ;)
<sr> welcome to the Musical Lounge called #tunes
<ElGato> hahaha
<Perly> herm
<Perly> elgato, i mean, what's this channel about? :)
<ElGato> the tunes os/language
<Perly> oh
<smkl> Perly: go away. you are even more lame than HS
<Perly> smkl: uhm, why is that?
<Downix> hey-lo all
<fare> iepos: dunno. Why not?
<sr> wow this channel is getting larger every second
<ElGato> D:
<fare> iepos: actually, as I concluded in lambdaND, I need communication, too
<ElGato> lame?!?!
<ElGato> HS isn't lame
<ElGato> he's my friend ;)
<dtm> fare:   what kind of communication do you need?
* dtm/#tunes snuggles up to fare
* dtm/#tunes bats his eyes up at fare
<iepos> fare: yeah, lambdaND is not really a language i suppose... just a computing model...
<fare> dtm: inter-agent communication
<dtm> Hmmm.
<fare> iepos: no; only the pure subset thereof
<dtm> Oh.  In the software.
<Perly> smkl?
<abi> i heard smkl was a spam bot who got stranded in #tunes
<fare> dtm?
<iepos> fare: well, applicative languages seem like good languages to me...
<ElGato> perly: check out tunes.org and idle here for a while
<fare> iepos: sure. See CL, Scheme, SML, OCAML
<sr> ElGato: halt! catch fire!
* Downix/#tunes is an odd person, only codes in C
<dtm> fare:   you mean that you need a communication model in software?
<dtm> fare  is that the comm you were referring ot
<dtm> s/ot/to
<ElGato> !!!
<ElGato> hcf!!!
<fare> dtm: yes, I want concurrency from within the HLL
<hcf> ElGato?
<abi> ElGato is annoyed when you ask that or the love child of fare
<dtm> fare:   oh ok
* ElGato/#tunes HUG hcf
<fare> dtm: did you read my glossary entry on uK?
<hcf> ElGato: oic, nm
<smkl> >>> Perly [C@perly.xnet.com] requested unknown ctcp VERSIOIN  from smkl
<smkl> gee
<ElGato> now smkl be nice :)
<sr> <ElGato> hcf reminds of a 12 year old girl
05:40pm
<ElGato> i didn't say that
<iepos> fare: hmmm... scheme doesn't really seem applicative to me (although i'll admit, i've never used it, only looked at it a bit)... it has the common problem of confusing procedures that yield things with the things themselves, right? hmm... i guess that doe'
<ElGato> *ElGato imagines hcf as being a 12 year old asian girl
<fare> iepos: the confusion is only in your mind
<dtm> fare:   negative  :(
<fare> iepos: it's as applicative as ML (and anterior to it)
<fare> dtm: read on!
<dtm> fare:  i just might!
<fare> dtm: did you read the article/ll99/ thing?
* dtm/#tunes computes his y2k work
* Downix/#tunes laughs
<dtm> fare:   not yet!
<iepos> fare: all right... some day i've going to have to try out ML, i guess. 
<dtm> fare:   i'm goig to refresh myself on the entire site soon!
<fare> iepos: what applicative language do you know, if it's neither LISP, neither ML?
<sr> dtm is a mystical entity who somehow knows fare and entered this channel in mysterious circumstances
<fare> dtm: btw, I have no driver's license!
<ElGato> fare: you only need to hold onto a programming language for a short time til you create the HLL
<iepos> fare: oh, i don't know an applicative programming language... but it would seem like an applicative language would be good.
<ElGato> just settle on scheme or ml
<ElGato> a while ago i was flipping through /users on bespin and i discovered dtm
<ElGato> he didn't stumble in here
<dtm> sr:   i've been into TUNES since about 1994 or 1995
<sr> hehe but my explanation sounds better
<dtm> sr:   over a period of several years since then i'd been writing to fare every 6 or 12 months to ask him what i could do to help and only got extremely latent and nearly totaly useless examples
<dtm> sr:   yes it does.
<dtm> i like it
<fare> iepos: oh  :)
* dtm/#tunes stumbles over elGato
<ElGato> meow!
<dtm> where the hell is Dem?
<abi> i think Dem is logged in more than once; writing to ttya0
<dtm> wtf?!
<Downix> hmm
<iepos> fare: anyhow... i was thinking... one good feature of a language is that it fixes an interpretation to symbols, without regard to context. this seems like a must for a language suitable for logic.
<dtm> how does abi know so much
* ElGato/#tunes points a gun to fare's head
<ElGato> it's scheme or ml fare
<dtm> be careful with that thing elG
<ElGato> take your pick
<iepos> fare: i mean automated logic...
<Downix> hehe
<hcf> dtm: she learns from channel chatter, 'x is y' things
<sr> it?
<Downix> so, how is Tunes going?
<iepos> fare: this requirement would rule out languages with pronouns or words like "now"...
<hcf> abi: forget dem
<abi> hcf: I forgot dem
<ElGato> red tape rivals recycling bibles
<Downix> ElGato:  I say recycle them into something useful, like porn magazines.  8)
<ElGato> :D
<iepos> fare: although it would still be fine if the language provided a "now" procedure that resulted in the time that it was executed.
<fare> iepos: you want the language to have pure semantics
<ElGato> like forth :D
<fare> iepos: unhappily, this isn't always possible
<sr> fare: from where do you expect to buy an alpha?
<fare> intentionality is part of life, it's part of computing
<iepos> fare: definitely, if it's going to be suitable for automated logic... syntactic sugar is fine though of course, for users.
<fare> sr: I had addresses, long ago. At worst, at Compaq.
<ElGato> synacttic sugar bothers me
<iepos> sr, alphas? i have a multia i got off ebay. still haven't got it to boot linux though :-)
<ElGato> minimal syntax maximal semantics
<iepos> sr, 
<sr> fare: i really want to buy one, and if you can tell me where i'd really appreciate it
<iepos> multias sell for around $100 on ebay these days.
<fare> In France, acticiel.com has Alphas
<Downix> I'd rather buy a MIPS box myself
<ElGato> syntax makes the programmer feel good
<Downix> but I'm biased
05:50pm
<ElGato> semantics makes the programmer express himself
<iepos> (a multia is a 166Mhz aplha with a network card/pcmcia slots, keyboard, mouse, ..., typically with 64MB of ram, hard drive not recommended, external if you do)
<dtm> one of the head maintainers of Stampede Linux wants to sell his multia
<dtm> gabe@stampede.org
<Downix> iepos:  I don't like Alpha's terribly much I'm afraid.
<sr> i want to buy an alpha chip OEM
<ElGato> i like 6502's
<ElGato> yeah baby
<Downix> ElGato:  Hey, the 6502's were nice, but I think the 68000's toast them
<ElGato> i haven't read up on 68000's
<iepos> fare: you said "pure semantics" was not always possible. how so?
<ElGato> dudex0rs
<ElGato> forth
<liar> can anyone tell me why netscape asks if i want to save any html file that is try to view from my drive?
<liar> s/is/i/
<sr> http://www.digital.com/alphaoem/developers.htm
<Downix> ElGato:  Do it, they're fantastic project CPU's.  2-stage pipeline, ASM that makes sence, and a bus that's easy as pi to develop
<fare> http://www.microway.com/ has alpha machines
<sr> i want the chip from the factory :P
<fare> the 6803 beats the 6502
<sr> alpha OEM lists motherboard products
<ElGato> what do you mean by project cpu's?
<sr> but they don't list prices
<Downix> fare:  Agreed. but the 68030 beats the 6803
<fare> Downix: :)
<Downix> sr:  I wouldn't Alpha OEM, it's EXPENSIVE!
<water> back
<iepos> greetings, water!
<ElGato> agua :)!
<sr> downix: how so? isn't getting them from the factory cheaper?
<dtm> liar:   it's because of your mime types
<Downix> sr:  Not by much.  You can get equivelent performance from SPARC and MIPS CPU's for much lower cost
<iepos> but getting it used is cheaper still.
<dtm> liar:   those are set from the users' perspectives, in ~/.mailcap and ~/.mime.types or in some system wide file, and you might have some that are overriding the built-in static binary types as well
<fare> iepos: see the way Haskell does I/O
<dtm> liar:    go check in your preferences, under 'navigator'
<Downix> iepos:  Not by much.  Unlike SPARC and MIPS series, noone continues the older, cheaper Alpha series
<fare> iepos: I/O and communication are the antithesis to pure computation
<dtm> what do you ladies think of WD65C816
<sr> Downix: how much would you sell me an Alpha for?
<fare> iepos: they embody intentionality as opposed to extensionality, whereas pure computation is extensionality
<Downix> sr:  a million dollars
<iepos> fare: the way haskell does i/o is the way i'd like to do i/o, maybe. a fuller model of time would be nice though, so one could do multiple things at once with arbitrary synchronization constraints.
<sr> so you wouldn't sell one?
* fare/#Tunes liked the ARM design. Too bad no one pushes it.
<Downix> fare:  Agreed, ARM is very well done.  
<fare> iepos: to do concurrent computation like that is impossible
<Downix> sr:  I would, for a million dollars
<iepos> fare: impossible? 
<fare> I exaggerate;
<fare> but you'd end up with something very ugly
06:00pm
<fare> in the end, you'd pass around whole environments and do clumsily what impure languages do easily
<Downix> fare:  the problem is that ARM entered into a very compeditive field w/o a huge advantage.  It's one level, StrongARM is firmly controlled by Intel, not good prospects for ARM in the end.
<iepos> fare: hmmm... the key to me seems to be to separate the logic/system (or reduction system) from the i/o... i don't suggest passing around environments... (...)
<fare> Downix: yeah. Too bad.
* fare/#Tunes would like the F-CPU to be ARM-compatible!
<Downix> fare:  I wouldn't.  WOuld be too limiting
<fare> plus that would be a *great* market for it!
<fare> why limiting?
<iepos> fare: the system should could keep a general knowledge base which it does logic on, and also a set of pending instructions.
<liar> dtm: well it was saying 'type application/x-httpd-php3' but i removed that from the helpers list and even restarted netscape but it still asks to save as that type
<Downix> fare:  A great market, but if it was an "ARM clone" it would be shunned by the other ARM users.  They are proud of their chip, and don't want people copying them.
<iepos> fare: forget 'should'
<fare> it's a doable first target, and compatibility with that well-spread design would attract attention and mindshare...
<dtm> liar:    that wasn't the right type if you're just talking about plain html.
* dtm/#tunes has a 20MHz StrongARM in his Newton MP100
<fare> it would force the other ARM users to open more, and create exactly what we want!
<Downix> fare:  Not true.  I would note that other chip manufacturers have tried this before
<liar> dtm: it is plain html but the netscape save as dialog says that type
<fare> 20MHz StrongARM? Ain't that a contradiction in terms?
<dtm> liar:    right
<Downix> fare:  Let me take keen note of the Z80, which has never been a popular mainstream CPU
<fare> Z80 was *very* popular!
<fare> and still is
<iepos> fare: the system i am thinking of will probably not do most of its work through reduction, but rather through i/o ("procedures"). but these procedures will be described in an applicative language...
* liar/#tunes has a 100mhz dragonball in his pilot
<Downix> fare:  Not it's not popular, tell me a time when Z80's were selling 800,000 units a year?
<ElGato> wb agua
<Downix> fare:  I can name other CPU's that do
<fare> iepos: we need to model concurrent computations, such as naturally happen when lots of people connect asynchronously to the system!
<sr> i found an Alpha 533MHz system for $600 on ebay
<fare> Downix: what do you suggest, then? that the F-CPU be an incompatible chip with an unknown design that cannot be used in any way by any one?
<dtm> fare:   that sounds like a good idea.
<sr> fare: will the F-CPU ever exist? :P
<dtm> I like it.
<dtm> Let's take votes!
<iepos> fare: hmm... i wasn't clear. i'm thinking of tossing the lambda-reduction-as-computation approach  generally (it may be of some interest on a low level)...
<Downix> fare:  sounds smart to me.  because that way people will learn the F-CPU, and not ARM, MIPS, or some other architecture.  F-CPU was to be better than these other ones.  If we copy one too closely, we won't manage to be better
<fare> iepos: what would you retain, then?
<iepos> fare: i'd retain an applicative language and the haskellish way of modelling procedures...
<fare> Downix: maybe start from one of those proven research designs?
<iepos> fare: well, haskell is not really applicative ... but ...
<Downix> fare:  We have.
<fare> Downix: those proven with formal methods...
<Downix> fare:  Our design borrows elements from Power and MIPS architecture
<dtm> i dont think it's all that winning of an idea to want to imitate other architectures except in concept.
<fare> iepos: Haskell is the antithesis to applicative programming
<dtm> i'm leaning toward downix's standpoit here
<Downix> fare:  And avoids pitfalls both fell into
<fare> Downix: what about the bus?
<dtm> Downix:   what are some pitfalls of ppc and mips?
<Downix> fare:  The bus is still up in the air.
<fare> Downix: was the design chosen already?
<dtm> i hope the bus should be independant.  not something overly simple or legacy like pci
<Downix> dtm:  Power started off too heavy in the overhead.  Hard to reduce the cost to a point that's good for embedded w/o sacrificing performance.
<fare> can it be extended with type tags on register and/or memory, as in LISP machines?
<Downix> fare:  Semi.  We have only one proposed, and noone is complaining
<dtm> rather, not *dependant* on something like pci.
<Downix> dtm:  Agreed
<fare> will it be compatible with existing chipsets? or will a free chipset have to be designed, too?
<dtm> See, guys?  I'm smart, just like Downix.
<Downix> fare:  Yup.  and we're working on making it compatable with existing chipsets.
<dtm> Downix likes non-bus-dependancy too!
06:10pm
<dtm> is there any sane way to abstract all these things away, so that you could implement a tunes architecture however you want?  such as in ibm's new open ppc design, or on an intel or an apple?
<dtm> or a c64?
<iepos> fare: i like the way that haskell separates lambda-abstraction from procedure-calls (unlike C and ...)... that's really all i meant to say about haskell i guess :-)
<fare> how long the pipeline? to keep it full, will you use software management of stalls? speculation? multithreading?
<Downix> dtm:  always possible.  Hell, someone ported BSD to a Sega Dreamcast
<ElGato> downix: what do you mean by a project cpu?
* dtm/#tunes speculates across multiple threads, about fare's query
<dtm> Downix:   indeed   but they might have been seemingly highly irrational to start with :)
<Downix> fare: the pipeline length is to be between 3 and 5 stages.  Any more and latency can kill the system.  
<dtm> Downix:   i'm just wondering if it's possible to make it easy
<Downix> ElGato:  I'm part of the F-CPU project.
<ElGato> ...
<fare> Downix: have MISC proposals be considered?
<dtm> ElGato:   yeah, so dont mess with Downix!
<Downix> dtm:  I don't see why not.
<ElGato> could you point me out to a url or something
<ElGato> >
<Downix> fare:  Yes and rejected.  MISC has never had performance.
<ElGato> ?
<Downix> ElGato:  http://f-cpu.tux.org
* dtm/#tunes points elgato to a url
<dtm> http://slashdot.org
<ElGato> danku schon
<dtm> http://www.y2k.gov
<fare> iepos: sure. It's nice to be able to isolate pure parts of programs. But I think it's wrong-headed to not support impure programming directly
<iepos> fare: i've been working on my kludgy interpretive system that i promised awhile ago. it's not really interesting at this point though...
<fare> iepos: usual self-quote: Ob(Pure Programming Languages): implicit state _is_ modularity. If we want to be really radical, and eliminate implicit state, then we should eliminate it at the meta-level, too, and remove named variables, leaving us only combinat
<ElGato> downix: so you're a l33t hardware hax0r? ;)
<fare> ors.
<Downix> fare:  And to get MISC to perform would require far more hardware than would be profitable.
<iepos> fare: not sure what you mean by "impure" programming. you mean procedures with side-effects? in the system i am thinking of, there will be almost no procedures _without_ side-effects, since if you didn't have a side-effect, you probably didn't need a proce
<iepos> fare: ... procdure at all.
<fare> Downix: never had performance? It looks quite performant, to me.
<Downix> ElGato:  semi, kinda-sorta.  I am mostly a researcher, but I can apply some of it
<Downix> fare:  Performance comes at a cost, always.  
<ElGato> what are you actually making?
<fare> Downix: remember that you can pack many instructions in a memory cycle.
<fare> Downix: even in a static memory cycle!
<iepos> fare: and ... the key is that one should be able to refer to things directly, without needing a procedure.
<fare> iepos: uh?
<fare> iepos: not sure what you mean, there.
<iepos> fare: hmm.... well... (....)
<fare> impure == has side effects, communication with other agents, etc
<ElGato> wait wait
<ElGato> are you actually making a processor or something?
<ElGato> i'd imagine that would be expensive ;)
<Downix> fare:  You can do that w/ other architectures.
<Downix> ElGato:  it is, which is why all of it is theoretical for now.
<iepos> fare: in many programming languages, to refer to, say, a list of prime numbers, you would write a procedure that generates them, but in a good system, you would not need a procedure.
<ElGato> lengua == pure
<Downix> fare:  MISC has no serious advantages.  On the same die size, you can get better performance from both CISC and RISC
<fare> Downix: or you could put several MISC processors in parallel
<Downix> fare:  Again, nothing that cannot be done cheaper with other CPU's
<fare> Downix: have measurements been done to back this assertion?
<Downix> fare:  Measurements were done in 78, when MISC was first tried out.
<ElGato> downix: do you work with a silicon compiler or something?
06:20pm
<sr> Downix: is the Z80 a good processor?
<fare> Downix: at that time, the memory/CPU speed discrepancy was not what it is today
<iepos> fare: oh, your self-quote i missed... actually it does seem interesting to eliminate variables at the meta-level, although remember the language i'm looking for is for computers... less-pure extensions for people might be interesting still.
<Downix> sr:  VEry good for it's time.
<sr> i would like to get a good project CPU
<sr> i want to make a gameboy-like device
<Downix> fare:  No, but the CPU's today perform even better.  Perhaps MISC today could work, but it's a dead architecture for the most part for a reason.  It's EXPENSIVE!
<Downix> ElGato:  Semi, I have a start-up company
<Downix> sr:  I'd go 68000, much better documentation
<ElGato> i bet i could get a 6502 for $2
<sr> where can i buy some 68000s?
<fare> D: what makes it so expensive? I'm interested...
<ElGato> how much are 68000's?
<iepos> fare: does that make sense to you, what i said before? that one should be able to refer to the set of prime numbers using a set abstraction of some other ordinary construct, without needing a procedure... ?
<Downix> sr:  Any electronics catelog
<Downix> ElGato:  $7 is that
<ElGato> for a 68000?
<Downix> fare:  The size of the routing units.
<Downix> ElGato:  Right.
<ElGato> !!!
<ElGato> i want one
<Downix> fare:  plus other things in the ALU
<fare> iepos: I'm not sure what's wrong about using procedures.
<Downix> ElGato:  They're 8Mhz CPU's, first consumer model w/ a full pipeline tho
<iepos> fare: procedures are still necessary, if you want the computer to actually output/input something of course. and still one could describe procedures using ordinary language constructs. the act of describing a procedure would not be the same as executing it
<sr> are there any sites online that sell 68000s?
<ElGato> i gotta learn electronics though
<Downix> sr:  Motorola made them
<fare> D: routing?
<fare> D: the ALU might be very simple
<iepos> fare: procedures could be useful in other cases as well, besides i/o, also. but itvery unnatural to 
<iepos> e to have to use a procedure to refer to the set of prime numbers or such
<ElGato> sr: i'll be steve jobs and you be steve wazniak
<ElGato> sr: we'll have a startup ;)
<fare> D: or when you say "procedure", do you miply something about side effects?
<sr> downix: would it be possible to make a small terminal device like a gameboy and a unix prompt with a limited budget?
<iepos> sorry, keep bumping enter key. :-)
<Downix> fare:  I don't really know how a MISC CPU works, but as my college professor pointed out, "the cores of MISC CPU's are always much larger due to the increased transistors to handle the tasks a MISC device needs"
<fare> s/D:/iepos:/
<Downix> sr:  Easily.
<sr> i want something like a graphing calculator with more capabilities
<Downix> fare:  I'm not positive, as I said.
<sr> hahaha iepos
<Downix> sr:  68000 may be a good choice for you then.  BSD runs on it
<fare> Downix: has this professor studied the ShBoom or the F21?
<ElGato> now i can be a l33t hardware hax0r like downix :D
<fare> Downix: why are there increased transistors?
<water> sr: get a pda with squeak loaded :)
<iepos> fare: when i say "procedure", i mean an instruction about something to do... i like to compare procedures in my system to commands in ordinary languages like english rather than in programming language
<Downix> fare:  I don't know, I'd have to see how a MISC CPU works first.
<ElGato> i want a forth machine
<fare> because of barrel shift registers all the way in the implicit register stacks?
<water> laggg.....
<ElGato> are there any cheap forth cpu's out there?
<Downix> fare:  I've studied the F21, found it a curiosity.
<sr> where can i get information on making these devices? where are there stores online that sell the necessary equipment?
<dtm> sr:   i used my Newton as a graphing calculator :)
<Downix> sr:  JDR microdevices does, and sells books with electronics projects on it
<Downix> fare:  I also found it a relic from the era of VAX, which was a similar concept I'd note
<sr> jdr, any URL for that?
<iepos> fare: "go home", "eat cheese", "spit out a bunch of characters on the screen"... those would be procedures in the sense that i use...
<fare> iepos: whether a procedure is expressed with lambdas or jsut combinators doesn't change the overall semantics of the system.
<iepos> fare: oh, a lambda extension probably wouldn't be too bad, although i don't see the need.
<iepos> fare: but a procedure would need to be expressed with more than just comibnators or lambdas.. some other constructs would be necessary also, of course. 
<fare> iepos: I still don't see what you object exactly to what.
06:30pm
<fare> Downix: the VAX was not quite a MISC. A prototypical CISC, maybe.
<Downix> fare:  And I'd note that you don't see VAX's around much anymore, do you?
<iepos> fare: i'm not exactly sure either.. i don't like the confusion in many systems that prevents one from talking about procedures without actually having the system execute them. that is the biggest thing, i suppose.
<Downix> fare:  no, VAX was a similar idea to MISC, just more rigid.  A CPU designed around programming language optimization, not for performance.
<dtm> can someone define 'MISC'?
<dtm> expand the acronym please
<iepos> fare: haskell doesn't have that problem... but C does. and i think scheme does too. not sure.
<Downix> Modular Instruction System Command
<dtm> HA!  we *finally* scared off ElGato!
<dtm> perhaps once and for all!
<dtm> Downix:   is there any simple way to quantify the difference it has over cisc and risc
<water> he'll be back
<dtm> Downix:   i'm assuming that is the context in which is it derived
<dtm> did i assume wrongly
<dtm> water:   doh!()@#%
* dtm/#tunes wonders what you guys will say after dtm is gone
<fare> D: MISC is not designed for programming language optimization
<Downix> dtm:  Ok, with MISC, despite all other claims, the end result is trying to save memory on the system by allowing the CPU to run code more efficiently
<Downix> fare:  The original concept behind MISC was actually just for that.
<fare> D: it's designed for minimal and fast hardware, and pushing complexity into software.
<Downix> fare:  That's another way to phrase it, but in the end, the same thing
<Downix> focusing on the software over the hardware
<dtm> fare:   that sounds like a good concept -- does it work
<dtm> fare:   consistantly
<Downix> dtm:  It does, but the problem comes when performance on a MISC or similar ideas such as VAX does not come near what a RISC system can do w/ the same memory latency and cache
<dtm> Downix:   so misc tends to have a lower performance ceiling?
<dtm> is it a ceiling or is it a curve or is it just something hard to define
<Downix> dtm:  it's hard to define, but there is a degredation in performance over RISC.  But it has clear advantages over standard RISC designs.
<Downix> most notibly in the area of memory useage
<dtm> Downix:   are the advantages in that it's more practical, or easy to design, or cheaper, or what
<Downix> MISC is idea for memory-intensive apps like 3D work
<dtm> oh really
<Downix> dtm:  However, it is expensive to produce in any MISC model made to date
<dtm> Downix:    but if it's ideal for intensive apps, then how does its performance crunch come into play?
<sr> www.jdr.com
<dtm> Downix:   what?  you're saying that MISC is expensive to design so far?
<Downix> dtm:  Because during the 98% of all CPU time other than memory peak periods, MISC has bad latency
<Downix> dtm:  that's one problem.  MISC designs usually end up with lower performance as a result, because noone can afford to produce the high-performance potential of the chips.
06:40pm
<dtm> Downix:   oh    why cant they afford it
<Downix> since F-CPU was for affordable performance, MISC was rejected.  Along with such ideas as TTL, 5-level cache, etc
<dtm> Downix:    it's interesting that i've never heard of any of this. 
<fare> why was TTL rejected?
<dtm> Downix:    what about RISC for F-CPU
<Downix> dtm:  Because they could get buy more RISC chips than a single MISC chip w/ the same performance
<fare> ahem. TTL opposes to gate-level logic, doesn't it?
<Downix> fare:  Again, cost.  TTL may have the key in speed, but it's expensive to synth
<Downix> dtm:  F-CPU is RISC
<fare> why expensive to synth? That depends on the software.
<fare> I know a guy who programmed a LISP system to do chip design from TTL to block design.
<fare> it could go even down to mask design
<fare> and you'd program it all in LISP, with the right level of abstraction at every step of your design.
<Downix> fare:  We are looking at TTL for later on, but for the initial design, we're sticking to slower, cheaper tech.
<fare> so much for rejecting MISC on alledged slowness
<Downix> also there's the other problem we had:  Noone knows how the heck a MISC design actually works
<Downix> we are all CISC and RISC people
<Downix> one word of advice I got a long time ago:  Do what you know!
<dtm> hehehhe
<dtm> Downix:   but properly abstract it!
<dtm> Downix:   so as to not obfuscate any given implementation!
<dtm> Downix:   otherwise that would kind of reduce the purpose of Tunes eh?
<Downix> dtm:  Right.  We've grouped all of our knowledge into this.  From the PPC folks, to the 68k guys to us MIPS-a-holics and even the ARM gent's had their turn
<dtm> i feel like trying out Quake 3 test tonight.  i wonder if there's a new MacOS client
<Downix> dtm:  Right
<dtm> Downix:  in what group is this that you all collaborated?  i dont remember if i've heard of f-cpu but it's starting to ring a bell
<dtm> from maybe years ago.
<dtm> the 'freedom architecture'?
<dtm> was that the one that's supposed to cost something like $20 
<dtm> perhaps i'm confabulating.
<dtm> on the value.
<fare> Downix: seing how you can do a 21-bit MISC CPU in 10000 transistors, I'd guess a 32-bit CPU would be less than 20000 transistors
<Downix> dtm:  kinda, but the name is right.
<dtm> fare:   i can do one in five.
<Downix> fare:  Nope, to jump from 21-bit to 32-bit would go to 28000
* dtm/#tunes etches a drawing of five transistors into fare's forehead
<Downix> fare:  But we're doing 64-bit, square that
<dtm> there!  now you're moving faster.
<fare> Downix: why square?
* dtm/#tunes antagonizes fare relentlessly and meaninglessly
<fare> Most of the logic will scale linearly!
<Downix> fare:  Not so I've found.  In the end, squaring is usually safer to deal with, and 90% of the time you're damned close
<Downix> fare:  Remember, double the bits, double the transistors to control those bits, double the security for the system, double everything.
<Downix> you end up squaring the whole problem
<fare> I disagree. a badass adder or router would be like quadratic in bit size, not exponential!
<fare> double what security?
<Downix> fare:  Yes, but I always go higher for safety's sake
<fare> Downix: no, you don't end up squaring the whole system
06:50pm
<Downix> fare:  your pipeline stall
<Downix> fare:  But you don't end up doubling it either.  It's in the middle
<fare> Downix: oh, yeah; you may have pipeline problems; but if the CPU is so much faster than even the static RAM, that'll be a noise in the memory access...
<Downix> fare:  here's an example, it is diffcult to make a 32-bit CPU for under 100,000 transistors
<Downix> difficult, not impossible
<fare> ARM did in in 70000 transistors, didn't they?
<Downix> fare:  they did, yes.  And the 68000 did it in 68000 transistors
<fare> (and it's not even MISC)
<Downix> fare:  No, it's not.
<fare> well, the 68000 is a microcoded 16-bit core.
<Downix> fare:  No, it's a microcoded 32-bit core w/ 16-bit bus
<Downix> and there was a 32-bit bus design version of it as well
<fare> depends on what you call the size of the core
<Downix> fare:  In the 68000 I consider everything inside of the pins the core
<fare> the ALU was 16-bit, wasn't it?
<Downix> fare:  nope, 32-bit.  The 68000 was the 32-bit version of the 16-bit 6800
<fare> 6800 was 8-bit!!!
<fare> aren't you confusing with 68020???
<dtm> fare, 68000 has 32bit internal addressing 
<fare> the 68000 had 16-bit ALU, so that 16-bit add would take 3 cycles, and 32-bit adds would take 7 cycles
<fare> (or so)
<Downix> fare:  I got the 68000 system control spec right here, it states "the 68000 is a full 32-bit CPU with the last 16-pins of the bus deactivated, in order for it to fit onto a 100-pin chip layout"
<dtm> perhaps you guys are disagreeing on parameters for a definition relating to 32bits
<fare> 68000 used 32-bit registers for addressing (with only 24 bits of effective address or so)
<Downix> fare:  Not quite, the 68000 was pipelined, could do 320bit apps in a single cycle while featching the next instruction
<dtm> but 32bit os's run on 68000
<Downix> quite a few 32-bit OS's I'd note
<fare> Again, you must be confusing with 68020 or 68030
<Downix> BSD, UNIX, Amiga
* dtm/#tunes would note
<liar> fare: u are wrong man, give up
<Downix> fare:  I am sorry, you are very wrong
<fare> liar: am I?
<dtm> liar:   let's kill him
<liar> fare: yes u are
<dtm> fare:    dont worry; he's just a liar
<fare> ok, I will have to find back some 68000 documentation...
<Downix> fare:  I have a ton of it, want some?
<liar> fare: u go do that
<fare> Downix: sure, from the software side, the 68000 was a 32-bit computer
<fare> but in its internal, it had a 16-bit core.
* Downix/#tunes digs through his box marked "Motorola tech manuals"
<dtm> and it can dance a mean jig
<Downix> fare:  Afraid not dude.
07:00pm
<Downix> fare:  Unless of course we're referring to a different 68000 than the one labelled MPU68000  from a small company called Motorola
<liar> Downix: small, hehe
<Downix> I even have a very old tech book from 1981 here
<Downix> man I love these chips
<fare> can you confirm from those books?
<Downix> fare:  yes I can.
<Downix> fare:  Call up motorola and ask for a manual for it, they'll ship you onw
<fare> do you have timings for 32-bit and 16-bit adds?
<Downix> lemme dig em out
<fare> can you confirm the "pipeline" assertion?
* Downix/#tunes needs to find the BIG book for that one
<Downix> fare:  By "pipeline"  it is simply the chips ability to featch an instruction while an instruction is being processed
<fare> IIRC, the 32-bit ALU ops would take 2-4 cycles more than 16-bit ALU ops.
<fare> Downix: oh, instruction prefetch, sure.
<Downix> not a true pipeline, but closer to one than existed prior, in the x86 series for example
<Downix> fare:  You're right about the timings BTW.  But allow me to refer to something from the MIPS R4000 manual
<fare> of course I'm right. It had a fucking 16-bit ALU!
<Downix> "even tho the MIPS has a 64-bit ALU core, processing 32-bit and 16-bit commands will be finished in far less cycles than 64-bit commands"
<fare> perhaps for the R4000.
<Downix> fare:  if you were 100% right, it should be a 6-stage difference, not a 4
<fare> why, not a 4?
<Downix> fare:  the tie-in of splitting up a command
<fare> it's addition, not multiply-accumulate
<fare> uh?
<Downix> fare:  that's a common ALU procedure when dealing with higher bit-depths than the ALU can process.
<fare> 4 is already 2 cycles of tie-in!
<dtm> well guys it's been fascinating and i've been following your conversation so far but i think it's time for me to go home.
<Downix> fare:  no, 4 would be 3 plus 3 plus 1, where's the last 1?
<Downix> fare:  I can't find a single manual here that says it's a 16-bit ALU, only 32-bit
<liar> is there a way to change the path variable used by the shell that X is being run in without exiting from X?
<dtm> i'lll talk with you ladies perhaps tomorrow
<Downix> I DID find one that stated that there was a crippled version of the 68000, called the 68005 w/ a 16-bit ALU however
<Downix> liar:  set PATH
<dtm> liar   probably no more likely than it is to do so with any other process's environment so get to work on TUNES so we can obsolete that problem
<liar> dtm: heh
<dtm> uh
<dtm> GUYS
<dtm> BYE!!
<Downix> liar:  It's all a government conspiracy to force you to buy microsoft and Intel!
* dtm/#tunes waves
<liar> BYE!
<dtm> BUY!!
<dtm> BIGH???!
<liar> BI!!
<Downix> bye=bye!
<dtm> ok
<liar> he is an addict
<Downix> yup
<dtm> who?
* Downix/#tunes rips out dtm's power cord
<liar> u
<dtm> addict?
<dtm> well
<dtm> i'm on a ups so there
07:10pm
* dtm/#tunes 's powerbook goes to battery
<dtm> good night ladies
* Downix/#tunes drains dtm's battery by starting his car off of it
<liar> heh
<Downix> trust me, getting thast car to start would drain any battery
<fare> well, actually, the 68000 had 2 16-bit ALUs for data
<fare> and one for addresses
<Downix> fare:  That could be what this one ment "32-bit split into 2 16-bit ALU's"
<liar> Downix: he will never give up, just kill him now
<Downix> liar:  hehe
<Downix> fare:  Technically, we'd both be right then, wouldn't we?
<liar> Downix: just say "yes fare, u are correct" or he will go on all night
<Downix> liar:  I'm more stubborn than he is.  
<Downix> liar:  Plus I have a secret weapon... Motorola R&D
* Downix/#tunes hasn't talked to Art for awhile
<Downix> I should see how the 080 is doing
<Downix> anyways
<Downix> this is VERY off the topic
<Downix> we were talking why the F-CPU is RISC
<Downix> and now I'm the only one talking
<Downix> hey liar
07:20pm
<fare> 080?
<fare> what a 080?
<liar> hey downix
<liar> 68080
<Downix> 68080
<fare> does that exist????
<fare> I left at the 060...
<Downix> same idea as the 68060 but w/ a starting bus speed of 90Mhz
<Downix> fare:  Semi, it's in prototype right now.  Not official, like with all other motorola products, until weeks before shipping
<Downix> unlike Intel which blabbers it all over
<fare> oh. Will it be 000 binary-compatible, or rather ColdFire like?
<liar> what is still using 68k chips?
<Downix> fare:  Yup.
<fare> yup what?
<Downix> liar:  Embedded apps use them quite often, so are systems like the Amiga
<fare> liar: PalmPilots
<Downix> fare:  000 binary.
<liar> fare: which ones?
<fare> hum. I'm not so sure it's good news.
<fare> liar: all
<liar> my pilot uses a dragonball 68328
<fare> liar: also, the linux-on-a-simm
<liar> its an embedded 68k chip
<fare> liar: yup, that's what we said
<Downix> fare:  the 68k gets great performance, and is very solid.
<liar> ok let me rephrase my question. what uses a friggin 680x0 chip
<Downix> liar:  Whirlpool
<liar> whirlpool is gonna use a 68080?
<Downix> liar:  They're using the 68060 now, why not upgrade?
<liar> in what?
<abi> well, in what is that ROM code written? If it's x86 or m68k assembly, then too bad.
<liar> abi: forget in what
<abi> liar: I forgot in what
<Downix> liar:  Washingmachines, dish washers, even microwaves
<Downix> abi:  No clue
<liar> why do they need a friggin 060?
<Downix> liar:  don't have a clue
* Downix/#tunes thinks that's overkill
<Downix> of course I've made hand calcs out of pentium II's
<liar> whatever
<Downix> BTW, never do that, the pentium II's still have a FPU bug
<Downix> of course it could have been my own doing
* Downix/#tunes shrugs
<Downix> ok, who here thinks people really need a washing machine w/ equal CPU power to a home computer?
07:30pm
<liar> well if it will get my clothes when they are dirty and wash them then ya it can have an 080
<Downix> lol
<liar> and it would also need to put them in the dryer and then hang them up when done
<Downix> with an 080 it had better be able to play MP3's while it rinses
<fare> Dx: PII's have FPU bug?
<Downix> now THAT would be cool, the washer's going, suddenly it pipes up "at the car wash!"
<Downix> fare:  It's slight, but during cos functions I'd never get a steady function
<Downix> fare:  get a spike I shouldn't have
<sr> downix: do you have any experience in electronics?
<sr> err electronics engineering, designing devices
<Downix> sr:  Semi, I'm a hack, not a pro.
<sr> i really want to learn how to make electronic devices
<fare> bug in cos???
<Downix> fare:  it's a mathematical formula
<fare> what kind of bug? precision bug?
<Downix> fare:  more like in a certain order of numbers the FPU would be 1-2 off
<fare> oh, like cos^2 + sin^2 != 1
<Downix> fare:  right
<Downix> I don't get it
<Downix> I trashed the thing
<fare> "order of numbers"?
<Downix> fare:  Don't ask me, someone else designed it, I just tried it out
<fare> oh
<Downix> built it, kinda.  
<sr> hmm Falco rules
<Downix> sr:  Well, I can recomend taking an electronics course.  I'm enrolled for one
<sr> fare: what kind of games do you want to develop?
07:40pm
<liar> oh give me a break
<liar> java is case sensitive
<liar> and i thought those parenthesis in list were stupid
<Downix> hmm
<liar> s/list/lisp/
<sr> liar: develop your own language
<fare> sr: a universal game
<liar> sr: i have
<sr> liar: give me the details
<fare> sr: an adventure game, where the hero must crack a computer
<sr> fare: so it's going to be text based? :)
<liar> it has no parens, no brackets, no semicolons, and no other stupid useless characters
<sr> like a virtual unix prompt
<fare> sr: no. For the computer has a graphic mode
<Downix> hehe, I semi-designed a quake TC where you were a valiant computer hacker braking into M$'s mainframe
<fare> sr: actually, one of the first trick to crack the (virtual) computer will be to get it into backward compatible text mode.
<sr> a mind game
<liar> tc?
<sr> liar: can i see a specification of your language?
<Downix> Total COnverstion
<Downix> where the game isn't even Quake anymore, just the same engine
<sr> Downix: just use the axe to chop at the computer :D
<Downix> sr:  hehe
<liar> Downix: never heard of it, i like the halflife quake engine
<Downix> liar:  I do too, which is why I was desining a game based on it
<fare> I wanted information to crack to be hidden in games included in the (virtual) computer
<fare> so that so as to solve the main game, you'd have to recursively solve subgames.
<sr> fare so will you start coding it?
<fare> it would be the "killer app" for TUNES :)
<fare> the virtual computer is obviously something that TUNES should be doing...
<eihrul> what will be?
<fare> eihrul: read the irc log...
<sr> liar: i would like to learn more about your language
<liar> ah my first hello world java app :)
<eihrul> fare: how so? excuse my naivety...
<Downix> cool
<Downix> I've yet to try Java
<liar> i will soon be making $100/hr writing java apps :)
<fare> eihrul: got an account on bespin?
<Downix> cool
<fare> else see ftp.tunes.org:/pub/tunes/irc/
<eihrul> fare: nope
<fare> or www.tunes.org/files/irc/
07:50pm
<liar> "If you couldn't care less about Java applications, you're already familiar
<liar>          with object-oriented concepts"
<liar> that is on sun's java web page
<eihrul> so true
<Downix> hmm
<Downix> hey, a java 3D modeller, sound good?
<liar> they have java 3d stuff to make that faster
<Downix> I know
<Downix> brb
<eihrul> liar: but then Mesa itself is just slow...
<eihrul> so that pretty-much evens it out
<liar> heh
<fare> in my game, there'll be 3D
<fare> s/3D/no 3D/
<Mr_Wrong> 3d is overused/overrated.
<eihrul> not to meant it just makes writing games hell
<Mr_Wrong> There are few situations where it's used where an isometric view would not contribute as much or more to game play
<eihrul> er mention
<liar> Mr_Wrong: so 3d is overused in quake?
<sr> no
<fare> yes. I prefer CastleWolfenstein
<fare> not the 3D, the II
<liar> fare: ha
<eihrul> liar: yes
<Mr_Wrong> Quake is a first person shooter.  It wouldn't be a first person shooter without 3d
08:00pm
<eihrul> there's 2.5D
<Mr_Wrong> Doom runs on the Psion 5
<Mr_Wrong> I haven't actually downloaded it yet though.  I have Nethack and Frotz and Python though
<liar> doom runs on my hp48gx
<Mr_Wrong> Cool
<sr> ascii quake runs on my 486
<Mr_Wrong> Not as nice of a screen though
<Mr_Wrong> Is it reasonably fast on the 48gx?
<Mr_Wrong> What processor is the 48gx, anyway?
<sr> Mr_Wrong: are you a fan of Apple?
<fare> it's a Saturn 4/64bit CPU
<eihrul> sr: that's cheating
<fare> very wierd
<Mr_Wrong> 64 bit?
<Mr_Wrong> sr:  Depends on what you mean by a fan
<fare> really a 4-bit ALU, but with 64-bit register
<fare> registers
<Mr_Wrong> Tjat
<Mr_Wrong> err that's a strange processor.
<Mr_Wrong> The Psion has a 32 bit processor with a 16 bit instruction set to keep code size down
<Mr_Wrong> The ARM710T
<liar> there is a hyperprogramming part to it tho
<liar> oops
<Mr_Wrong> Which is actually a macrocell on a single chip computer
<Mr_Wrong> the Cirrus Logic CL-7110 or something like that
<Mr_Wrong> 36 MHz
>>> sr [depag@phila-dialup498.nni.com] requested PING 935550556 from #tunes
08:10pm
<liar> why is it that windows cant determine what hardware profile to use?
08:30pm
<dtm> uhuhuuuhuhuh.
<Mr_Wrong> It might have something to do with the fact that Windows sucks.
<Mr_Wrong> Which is why I don't use it and neither should anyone else.
<dtm> Mr_Wrong:   that's right
<dtm> who's talking about windows and why?
<Mr_Wrong> [20:23] <liar> why is it that windows cant determine what hardware profile to use?
<dtm> ??????
<Mr_Wrong> I was quoting liar
<dtm> that's obviously an irrational question, i think.  unless he has an exemplary will, which he doesn't because he's QZ.
08:50pm
<dtm> so who's awake here
<hcf> zzz
<liar> zzz
* dtm/#tunes smacks yer ass with a towel
<dtm> make me my reflective software system, damn it!
09:00pm
<Downix> back
<dtm> hi downix
09:10pm
<Downix> what's up?
09:20pm
<ElGato> HI FRIENDS
<Downix> hi
<abi> niihau, Downix
<ElGato> hey 
<ElGato> what's going on?
<Downix> nada, you?
<ElGato> not much
<ElGato> me am programming
<ElGato> :D
09:30pm
<Downix> I am compiling
<ElGato> cooly
09:40pm
[ TUNES_    ] [ eihrul    ] [ Mr_Wrong  ] [ liar      ] [ Crimson_  ] 
[ fare      ] [ Plundis   ] [ abi       ] [ TUNES     ] 
[msg(TUNES)] newlog 1999.0825