00:00:00 --- log: started forth/01.06.25
00:19:22 --- quit: clog (Ping timeout)
00:19:22 --- log: stopped forth/01.06.25
00:19:30 --- log: started forth/01.06.25
00:19:30 --- join: clog (nef@bespin.org) joined #forth
00:19:30 --- topic: 'http://isforth.sourceforge.net -- http://cvs.sourceforge.net/cgi-bin/cvsweb.cgi/isforth/?cvsroot=isforth'
00:19:30 --- topic: set by ChanServ on [Sat Jun 23 01:03:40 2001]
00:19:31 --- names: list (clog adu @aaronl @rob_ert_sleep_ing @I440r @dired @Trey @klooie @nate37)
02:04:06 <adu> bye
02:04:07 --- quit: adu (IRCStep)
03:20:00 --- join: aaronl_ (aaronl@vitelus.com) joined #forth
03:20:14 --- quit: aaronl (Read error to aaronl[vitelus.com]: Connection reset by peer)
05:11:28 --- join: Fare (fare@ppp49-net1-idf2-bas1.isdnet.net) joined #forth
05:23:57 --- quit: klooie (Ping timeout for klooie[213-84-79-23.adsl.xs4all.nl])
05:24:39 --- join: klooie (kloo@213-84-79-23.adsl.xs4all.nl) joined #forth
06:06:25 * aaronl_ is away: LIGHT!! I'M MELTING!!
06:54:31 --- part: Fare left #forth
07:14:29 --- quit: klooie (Ping timeout for klooie[213-84-79-23.adsl.xs4all.nl])
07:14:47 --- join: klooie (kloo@213-84-79-23.adsl.xs4all.nl) joined #forth
07:18:35 --- join: grurp (grurp@pc19954.batc.tec.ut.us) joined #forth
07:40:12 --- join: scipient (scipient@ci156462-a.ashvil1.nc.home.com) joined #forth
08:05:06 --- quit: klooie (Ping timeout for klooie[213-84-79-23.adsl.xs4all.nl])
08:05:08 --- join: klooie (kloo@213-84-79-23.adsl.xs4all.nl) joined #forth
08:18:15 --- quit: dired (Ping timeout for dired[vitelus.com])
08:19:13 --- quit: aaronl_ (Ping timeout for aaronl_[vitelus.com])
08:49:11 --- quit: klooie (Ping timeout for klooie[213-84-79-23.adsl.xs4all.nl])
08:49:11 --- quit: rob_ert_sleep_ing (Ping timeout for rob_ert_sleep_ing[h3n1fls33o898.telia.com])
08:57:03 --- join: klooie (kloo@213-84-79-23.adsl.xs4all.nl) joined #forth
09:01:49 --- quit: grurp (Bye all)
10:10:47 --- quit: klooie (farmer.openprojects.net forward.openprojects.net)
10:10:55 --- join: klooie (kloo@213-84-79-23.adsl.xs4all.nl) joined #forth
10:36:05 --- join: grurp (grurp@pc19954.batc.tec.ut.us) joined #forth
10:44:51 --- join: rob_ert_sleep_ing (robost86@h3n1fls33o898.telia.com) joined #forth
11:00:41 --- join: edrx (edrx@200.240.18.89) joined #forth
11:01:12 --- mode: I440r set mode: +o clog
11:01:15 --- mode: I440r set mode: +o edrx
11:01:19 --- mode: I440r set mode: +o grurp
11:01:26 --- mode: I440r set mode: +o klooie
11:01:32 --- mode: I440r set mode: +o rob_ert_sleep_ing
11:01:36 --- mode: I440r set mode: +o scipient
11:12:13 --- join: MrReach (mrreach@209.181.43.184) joined #forth
11:12:21 <MrReach> hello, I440r
11:12:39 <MrReach> yes, I know why you cannot web from your firewall
11:12:47 <MrReach> do you still need to know?
11:13:31 <I440r> yes
11:13:33 <I440r> hi :)
11:13:41 <I440r> i asked in #debian and in #linpeople
11:13:44 <I440r> nobody seemed to know
11:13:49 <MrReach> ok, it has to do with the way IPMasq interacts with IP chains
11:14:11 <I440r> k
11:14:19 <MrReach> if your clear all the input and output rules, you should be able to web out
11:14:37 <I440r> input polacy is accept - chain is clear
11:14:44 <I440r> output polacy is accept - chain is clear
11:14:49 <I440r> the only RULE i have is
11:14:51 <MrReach> the masq forwarding rule bypasses the input and output chains.
11:15:07 <I440r> ipchains -A forward -i eth1 -j MASQ
11:15:18 <MrReach> did you chack manualy or via the config utility?
11:15:37 <I440r> i cant get anything to work at all if i try use any config util
11:15:40 <I440r> i have a little script
11:15:44 <I440r> ifconfig eht0 down
11:15:47 <I440r> ifconfitg eht0 up
11:15:50 <MrReach> ok, on a redhat system?
11:15:53 <I440r> route add default gw gatewayip
11:16:02 <I440r> ipchains -F
11:16:04 <I440r> ipchains -X
11:16:15 <I440r> ipchains -A forward -i eth1 -j MASQ
11:16:17 <I440r> debian
11:16:22 <MrReach> oh, ok
11:16:26 <I440r> thats my firewall script
11:16:31 <I440r> i run it by hand
11:16:36 <I440r> ive tried using mason - 
11:16:41 <MrReach> check that "ipchains -L input" returns nothing
11:16:44 <I440r> \if i implement ANY sort of firewall 
11:16:51 <I440r> ill paste that here
11:17:01 --- part: edrx left #forth
11:17:12 <I440r> root@gateway:/etc# ipchains -L
11:17:13 <I440r> Chain input (policy ACCEPT):
11:17:13 <I440r> Chain forward (policy ACCEPT):
11:17:13 <I440r> target     prot opt     source                destination           ports
11:17:13 <I440r> MASQ       all  ------  anywhere             anywhere              n/a
11:17:14 <I440r> Chain output (policy ACCEPT):
11:17:21 <MrReach> good forth site, btw, although a bit disorganized
11:17:46 <MrReach> hmmm ... can you ping your router?
11:17:53 <I440r> what router
11:18:04 <MrReach> what do you connect to the net with?
11:18:15 <I440r> dsl modem 
11:18:17 --- join: speuler (icafe@carambola.icafe.spacenet.de) joined #forth
11:18:21 <I440r> bongo!!!!
11:18:24 --- mode: I440r set mode: +o MrReach
11:18:25 <MrReach> heya, bongo!!!
11:18:27 --- mode: I440r set mode: +o speuler
11:18:30 <speuler> good day 2u
11:18:48 <speuler> just thought i'll check in while sitting in the internet cafe :)
11:19:00 <speuler> I440r:  tnx
11:19:02 <MrReach> still not at "home?
11:19:06 <I440r> :)
11:19:13 <speuler> just left again, two weeks ago
11:19:19 <MrReach> I440r: what do you connect to the net with?
11:19:23 <speuler> living on a camping site
11:19:26 <I440r> dsl modem
11:19:40 <I440r> sounds like fun bongo :)
11:19:43 <MrReach> heh, are you there for a contract? vacationing?
11:20:01 <speuler> MrReach:  contract work
11:20:03 <MrReach> can you ping the DSL card?
11:20:14 <speuler> MrReach: but using the van to live in
11:20:18 <I440r> i could if i knew its ip
11:20:20 <I440r> which i dont
11:20:24 <speuler> great fun now it's summer
11:20:31 <MrReach> oh! is the firewall daemon running?
11:20:36 <I440r> or - i think i prolly could
11:20:46 <MrReach> ifconfig should tell you
11:20:47 <I440r> i can ping my servers gateway at my isp
11:21:03 <I440r> no - that would just tell me the ip of my gateway at my isp
11:21:13 <MrReach> that's bizarre ... can you telnet to any other site?
11:21:37 <I440r> yes im talking to you from xchat FROM the gateway box
11:21:43 <I440r> irc is basically a telnet connection
11:21:53 <MrReach> yep ... as is the web
11:22:15 <MrReach> but you can't get to ebay?
11:22:35 <I440r> i cant get to www.etrade.com or www.vote.com
11:22:41 <I440r> im sure ther are others i cant get to
11:22:50 <I440r> i can get to most sites from here
11:22:52 <MrReach> how about the DNS? does each machine find their own IPs from the net?
11:22:53 <I440r> just not those 2
11:23:16 <I440r> the machines on the internal network use the gateway as their primary name server
11:23:24 <MrReach> you might have old IP addrs on the unix machine
11:23:26 <I440r> the gateway box uses my isp
11:23:36 <I440r> i can check that
11:23:37 <speuler> I440r:  how's your / your dad's german ?
11:23:41 <MrReach> ok, then they would all fail if it were an ip prob
11:24:08 <I440r> ill dig from server and from inside see if i get the same
11:24:16 <MrReach> can you ping etrade or vote?
11:24:17 <speuler> I440r: (plenty of docs written in german :-(
11:24:54 <I440r> i dont speak german nor does my dad - bu i bet we could learn :)
11:25:04 <MrReach> www.vote.com is pinging for me
11:25:14 <MrReach> www.etrade.com is not pining for me
11:25:16 <I440r> i can ping it from inside the net but not from the server
11:25:49 <I440r> aha!~!!!
11:26:00 <MrReach> ???
11:26:04 <I440r> when i dig from the server i dont get the same results as from INSIDE the server!
11:26:42 <MrReach> how long since a server reboot?
11:26:59 <I440r> speuler: are u thinking we might be needed if the language barrier didnt prevent ?
11:27:03 <I440r> yesterday :P
11:27:05 <I440r> by accident
11:27:19 <I440r> i said REBOOT from my laptop intending to reboot my laptop
11:27:27 <I440r> forgot i was sshing to the server hehehe
11:27:38 <I440r> up 21:45
11:27:44 <MrReach> well ... you might want to bring down the network altogether and bring it back up ... or reboot if you don't know how to do that
11:27:53 <MrReach> OOOPS!
11:27:55 <I440r> ifconfig eth1 down
11:28:00 <I440r> will bring the network down
11:28:09 <speuler> I440r:  my current impression is the project is going to go on longer than the initial 3 months
11:28:22 <I440r> speuler: thats always the case
11:28:25 <speuler> I440r: and it sounds big
11:28:32 <I440r> they bring you on for 2 weeks ur there for 2 years if they need u :)
11:28:45 <MrReach> no ... on redhat thats /etc/rc.d/init.d/network stop
11:28:46 <speuler> I440r: far too much for the few people doing the initial evaluation work
11:28:47 <I440r> i need a job :)
11:29:24 <I440r> i have /etc/init.d/networking
11:29:38 <I440r> speuler how much of a problem would the language barrier be
11:29:40 <MrReach> reinitializing the entire network params is quite different from bringing one of the cards down
11:29:50 <speuler> I440r:  may become years to work on
11:29:57 <I440r> cool :)
11:30:22 <speuler> I440r:  the first book i got was 800 pages, all written in german. 
11:30:51 <I440r> if they hired me on i would learn german :)
11:30:59 <speuler> I440r: the technology was pioneered by german and french companies, competing with each other.
11:31:16 <I440r> like sony with the cd
11:31:19 <speuler> I440r: so there's plenty of german in the docs and manuals
11:31:44 <I440r> mrreach i can do /etc/init.d/networking restart ?
11:31:54 <MrReach> is the firewall daemon running?
11:32:07 <I440r> what do you mean by daemon
11:32:08 <MrReach> should be able to ... will see you in a bit    @:^>
11:32:24 <I440r> am i still here ?
11:32:26 <I440r> :)
11:32:32 <I440r> root@gateway:/etc/init.d# /etc/init.d/networking restart
11:32:32 <I440r> Reconfiguring network interfaces: done.
11:32:44 <MrReach>   687 ?        S      1:40 /usr/lib/linuxconf/lib/firewalld -d -c /etc/heimdall/
11:32:47 <I440r> still cant ping www.etrade.com
11:33:00 <I440r> no i have no firewall
11:33:09 <MrReach> etrade doesn't ping ... as I said
11:33:23 <I440r> oh
11:33:31 <MrReach> on my system, shutting off that deamon totally screws ALL the networking
11:33:39 <I440r> lynx etrade
11:33:41 <MrReach> www.vote.com pings, though
11:33:46 <MrReach> as does www.yahoo.com
11:33:47 <I440r> making http connection to www.etrade.com -----
11:33:55 --- quit: speuler (changing servers)
11:33:55 <I440r> but i never get there
11:34:35 <MrReach> does lynx attempt the connection?
11:34:43 <I440r> it attemtps
11:34:43 <MrReach> I recall that lynx likes the full URL
11:34:46 <I440r> its attemtping
11:34:50 <MrReach> ok
11:34:55 <I440r> i can say lynx google
11:34:57 <I440r> and get there
11:35:02 <I440r> i say lynx etrade and i dont
11:35:29 <MrReach> ok
11:35:55 <MrReach> that's REALLY odd ... I'm runningout of ideas
11:36:05 --- join: speuler (icafe@carambola.icafe.spacenet.de) joined #forth
11:36:07 <I440r> do u have bongo in your icq ?
11:36:10 <I440r> aha there he is :)
11:36:12 --- mode: I440r set mode: +o speuler
11:36:16 <speuler> bloody irc client
11:36:26 <I440r> hehe
11:36:32 <I440r> which client ?
11:36:34 <speuler> tried to opem another server/channel in anotherwindow
11:36:42 <speuler> some tcl thing
11:36:52 <MrReach> yes, I have bongo on my icq, but he's not logged in because he's at an internet cafe
11:37:03 <MrReach> here, I'll send you his contact info
11:37:07 <speuler> sorry, i missed you last words
11:37:21 <I440r> oh yea
11:37:23 <speuler> try email:  speuler@blik.wox.org
11:37:25 <I440r> thats rite :)
11:37:35 <speuler> as i can pop my mail from this mail client
11:37:45 <speuler> whenever i'm around
11:38:15 <I440r> cool
11:38:22 <I440r> you know my email of corse - :)
11:38:41 <speuler> i440@purplecoder.com ?
11:38:43 <MrReach> speuler: when you're on icq ... are you "bongo" or "speuler" now?
11:38:45 <speuler> i440r
11:39:02 <I440r> yea ?
11:39:06 <I440r> ./me here :)
11:39:08 <speuler> mrreach: speuler. somebody nicked my nick
11:39:10 <I440r> no
11:39:15 <I440r> mark4@purplecoder
11:39:25 <I440r> i shud proly alias that one tho :)
11:39:32 <MrReach> yes, bongo is a common name on the IRC networks
11:39:39 <speuler> but i won't be on icq for another few months
11:39:52 * MrReach nods.
11:40:24 <speuler> I'll return to NL for a week, in august, i guess
11:40:33 <speuler> empty my mail box
11:41:15 <I440r> speuler you have an account here - you could .forward all ur email here and ssh in and read it!
11:41:19 <speuler> very object oriented, this project, by the way ...
11:41:25 <I440r> ouch
11:41:25 --- quit: grurp (Bye all)
11:41:34 <I440r> now THAT would be a problem
11:41:39 <speuler> i440: i got an .forward on your machine to my mail account
11:41:41 <I440r> OOP is the biggest mistake ive ever seen
11:41:47 <I440r> hehehe
11:41:51 <MrReach> heh, we had an earthquake here this morning, btw
11:41:54 <I440r> infinite .forward loop :)
11:42:01 <I440r> small one i hope :)
11:42:05 <speuler> MrReach: sheeet. heavy ?
11:42:27 <MrReach> 3.4 rictor, I'm about 4 miles from the epicenter
11:42:45 <speuler> richter ?
11:42:45 <MrReach> knocked the house around pretty good
11:42:54 <MrReach> yes, richter
11:43:11 <speuler> that's considerably strong.
11:43:17 <MrReach> little one
11:43:18 <speuler> any damage ?
11:43:36 <MrReach> its my 5th earthquake, but slept through the other four
11:43:38 <I440r> if anything can take a 3 month contract out to 3 years OOP will
11:43:44 <MrReach> speuler: not at this house
11:43:56 <speuler> I440r:  i have a pop server on my machine (blik.wox.org)
11:44:02 <speuler> so i can get to that mail
11:44:02 <MrReach> some cracked patios around here, and one street has a dangerous crack
11:44:18 <I440r> cool :)
11:44:30 <I440r> was it scary ?
11:44:36 <speuler> I440r:  it is permanently switched on
11:44:36 <MrReach> lots of aftershocks, and a secondarly quake at 2.3
11:45:00 <I440r> cool :)
11:45:12 <MrReach> I'm going to wait for this eve to do the foundation and chimney inspections, though, because the ground is still shaking
11:45:38 <speuler> are the damages covered by any insurance ?
11:45:59 <I440r> STILL shaking ?
11:45:59 <MrReach> it depends ... each home is insured sperately
11:46:02 <MrReach> yes
11:46:08 <speuler> klooie: u're dutch ?
11:46:26 <MrReach> and most insurance company have the "acts of god" weasel clause
11:46:41 <MrReach> this house is not insured for earthquake
11:46:50 <I440r> ouch!
11:47:02 <I440r> actually tho i find insurance to be a BIG CON
11:47:06 <speuler> mrreach: hope some sticky tape would help
11:47:19 <MrReach> damn it! I moved out of Seattle, and into one of the most geologically stable places in the USA ... it was part of my reason for moving here
11:47:20 <I440r> instead of paying insurance companies put the same money into an account for just this occasion
11:47:42 <I440r> :(
11:47:46 <speuler> my monitor soon will turn off
11:47:54 <MrReach> I'm not overly concerned about it
11:48:02 <speuler> machine stays on but i'll be in the dark then
11:48:05 <I440r> bongo if u disappear we'll know why - :)
11:48:08 <I440r> good 2 c u tho :)
11:48:17 <speuler> actually, it happened right now.
11:48:24 <speuler> i'm typing blindly now
11:48:28 <I440r> i used to do all my irc from an inet cafe in england
11:48:31 <speuler> can't see u anymore.
11:48:36 <I440r> but the guy at eh cafe was a buddy of mine hehe
11:48:36 <MrReach> heh
11:48:47 <speuler> so, if you don't mind kicking me (or remove the +o)
11:48:57 <speuler> so that the next user won't hasslke you
11:48:57 --- mode: MrReach set mode: -o speuler
11:49:02 <speuler> good by for now
11:49:06 <I440r> hehehe lets all beat up on bongo - he wont know :)
11:49:08 <speuler> cu
11:49:10 <I440r> cu dood!
11:49:20 --- quit: speuler (Read error to speuler[carambola.icafe.spacenet.de]: EOF from client)
11:49:29 <MrReach> heh, too bad he doesn't remember <ALT-F4>
11:49:34 <I440r> hehe
11:49:57 <MrReach> that's too funny
11:50:13 <I440r> where are u located again ?
11:50:25 <MrReach> I'm in Spokane, WA right now
11:50:29 <I440r> rite
11:50:42 <MrReach> will be moved back to Seattle area this summer, if I have anything to say about it
11:50:46 <I440r> how far is that from indiana ?
11:50:55 <MrReach> erm ... quite a ways
11:51:15 <MrReach> I will be visiting New Orleans for a couple of months this winter
11:51:34 <MrReach> am thinking of moving there, but want to see how much I like it, first
11:52:53 <I440r> coo
11:52:56 <I440r> cool even
11:53:01 <I440r> well actually - not cool hehe
11:53:06 <I440r> very hot and humid in the summer
11:57:02 <MrReach> yep, we'll have to see how it goes
11:57:09 <I440r> :)
11:57:19 <MrReach> there's an island in puget sound I've been thinking of purchasing
11:57:23 <I440r> so - no more gueses as to why my gateway is being stoopid ?
11:57:32 <I440r> wow!
11:57:36 --- nick: rob_ert_sleep_ing -> rob_ert
11:57:38 <MrReach> nope, not right now
11:57:46 <MrReach> heya, rob_ert
11:57:48 <rob_ert> Hi
11:58:06 <MrReach> the island will be a 5 year project, though, just in the purchase
11:58:14 <MrReach> and another 5 to fix it up
11:59:12 <I440r> :)
11:59:16 <I440r> we got our mortgage i think
11:59:29 <MrReach> on a new purchase?
11:59:31 <I440r> but its gone from a 20k down payment to a 30k to 40k down now
11:59:33 <MrReach> you're married?
12:00:16 <I440r> nope
12:00:19 <I440r> i wish
12:00:26 <MrReach> I finally just purchased this house outright ... I got *SO* tired of futsing around with this particular bank
12:00:48 <MrReach> they are real lame-asses
12:01:08 <I440r> im not allowed to retire the mortgage early
12:01:12 <I440r> that pisses me off
12:01:27 <MrReach> yes, but how tight is the prepayment penalty?
12:01:38 <I440r> dont know
12:01:56 <MrReach> is this your first house?
12:02:23 <I440r> yup
12:02:24 <MrReach> do you intend to be buying/selling several houses?
12:02:33 <I440r> nope
12:02:43 <I440r> i intend to buy this house and keep it :)
12:02:48 <MrReach> ok ... then prepay is not such a big deal
12:03:21 <MrReach> I'm intending to be buy/selling 52 houses/year by this time in 2002
12:03:38 <MrReach> so it's important to me to be on the ball with mortgages
12:03:51 <I440r> ouch!
12:04:02 <I440r> i used to have a teacher that baught a car a week
12:04:12 <MrReach> in that circumstance, prepays could cost me $30,000 a year or more
12:04:51 <MrReach> I've got to find a good lawyer to form a partnership with, though
12:08:26 --- join: adu (andrew@adsl-63-201-90-87.dsl.snfc21.pacbell.net) joined #forth
12:09:45 <MrReach> greets, adu
12:09:46 <adu> hi
12:09:46 --- mode: I440r set mode: +o adu
12:09:46 <I440r> adu didnt i add u to chanserv ?
12:09:47 <adu> i love forth
12:09:51 <I440r> hehe
12:09:53 <MrReach> heh
12:10:06 <adu> chanserv? i'm not a channel!
12:10:07 <I440r> whats not to love :)
12:10:19 <I440r> erm so you auto op
12:10:24 <adu> I440r: the fact that many forthers reinvent the wheel by nature
12:10:27 --- nick: MrReach -> MrGone
12:10:33 <rob_ert> Bye...
12:10:37 <I440r> which i say is GOOD
12:10:41 <adu> =) i'm op!
12:10:48 * rob_ert too
12:10:55 <adu> hi robert
12:10:58 <I440r> i opped you :P
12:11:01 <rob_ert> Hi adu
12:11:11 <I440r> every wheel should be reinvented each time its needed
12:11:22 <rob_ert> Yup
12:11:23 <I440r> because each wheel would then exactly fit the car it was put on
12:11:34 <adu> interesting
12:11:42 <I440r> in c that wont work because it would take too much time
12:11:46 <rob_ert> hmmm... 4 very different wheels on a car
12:11:47 <adu> ya, and people tend to know how to use things they made themselves
12:11:57 <I440r> in forth reinventing the wheel every time does not slow you down very much
12:12:08 <adu> people who use library, are only as good as the documentation
12:12:16 <adu> usually
12:12:53 <I440r> i wanna get coding isforth - but i need tcn's work
12:13:25 <adu> i've been wanting to do this hybrid kernel
12:13:47 <adu> i've maped out the connections and language bindings and implementations a bit
12:14:01 <adu> but i need a starting point
12:15:09 <I440r> i was pipe dreaming last nite - eventually have isforth run seperate from linux - where isforth is the OS hehe
12:15:11 <adu> i want a kernel in Ada & Lisp with a forth interface, maybe in forth with Ada/Lisp compilation, or something like that, and a unified graphics arch, with not only FB access but vector contexts, based on Mach only not in C
12:15:16 <I440r> and have it work like linux
12:15:25 <I440r> call it f-ix or fx hehe
12:15:34 <I440r> forthix hehe
12:15:35 <adu> fix =)
12:15:55 <I440r> and when you say startx it runs a word taht COMPILES x and runs it hehe
12:16:00 <I440r> everything stor3ed as sources 
12:16:06 <I440r> when you run things you compile them :)
12:16:24 <adu> interesting
12:16:24 <I440r> like i said - pipe dream heeh
12:16:31 <adu> JIT to the max? =)
12:16:39 <adu> pipe as in bong?
12:17:04 <I440r> hehe exactly
12:17:55 <adu> =)
12:18:13 <adu> what do you think of my Ada/Lisp kernel based on Mach-like arch?
12:18:48 <adu> i'm not sure if it should have only a forth interface/shell, or be implemented in forth\
12:19:17 <I440r> i was thinking of the forth kernel being the only thing that stays compiled
12:19:20 <I440r> when you say startx
12:19:25 <I440r> forth sees no word called startx
12:19:26 <adu> i don't think about forth & lisp being HLL i think of them as MLL
12:19:33 <I440r> searches for a source file called that
12:19:35 <adu> C & Ada are HLL
12:19:37 <I440r> interprets the file
12:19:45 <I440r> which compiles all the x shit rite then
12:19:47 <adu> startx.fth?
12:19:53 <I440r> no - just startx
12:19:57 <adu> hmm
12:19:58 <I440r> everything would be a .fth
12:20:12 <I440r> all commands you find in unix would exist 
12:20:18 <adu> i mean a file called startx.fth would be called by typing startx?
12:20:19 <I440r> but everything would be a SOURCE file
12:20:33 <adu> ya, with forth you can do that
12:20:36 <I440r> the system would not have the ability to save executables
12:20:43 <I440r> it would FORCE open source hehehehe
12:20:48 <adu> but i want multiple runtime archs, including the compiled nature of C/Ada/Lisp
12:21:22 <adu> i also want to make an Ada95 compliant INTERPRETER
12:21:27 <adu> call it AdaScript =)
12:21:37 <I440r> ugh!
12:21:39 <I440r> heh
12:21:52 <adu> then you could force that to be open source =)
12:22:08 <I440r> there would be no executables
12:22:25 <I440r> every script would end in a command to forget everything that the script compiled
12:22:37 <I440r> so the kernel would have to be able to compile to different contexts
12:22:48 <adu> like JIT
12:22:56 <adu> only save to memory instead of to a file
12:23:03 <adu> makes sense
12:23:10 <I440r> you might run a script to start applications x y and z and want the script to end and be purged but not x y and z
12:23:21 <I440r> no executables
12:23:30 <I440r> you write a program and you save its source
12:23:48 <adu> ya i know
12:23:49 <I440r> when you execute it forth loads it into some execution context and compiles it there
12:24:23 <I440r> when script finishes it forth frees up that execution context and all its memory
12:24:42 <I440r> and if it started any other applications they all die too
12:24:47 <I440r> unless specified otherwise
12:24:54 <adu> you'd still need an executable bootloader
12:25:02 <I440r> a complete unix written in forth where EVERYTHING is source
12:25:10 <I440r> that would be forth 
12:25:18 <adu> no
12:25:20 <I440r> rebuild of kernel would be a metacompile
12:25:28 <adu> it hasto be executable
12:25:53 <adu> at least the bootloader hasto be
12:25:59 --- quit: adu (farmer.openprojects.net zelazny.openprojects.net)
12:26:19 --- join: adu (andrew@adsl-63-201-90-87.dsl.snfc21.pacbell.net) joined #forth
12:26:35 --- mode: ChanServ set mode: -o nate37
12:26:35 --- mode: ChanServ set mode: -o Trey
12:26:35 --- mode: ChanServ set mode: -o I440r
12:26:35 --- mode: ChanServ set mode: -o clog
12:26:35 --- mode: ChanServ set mode: -o scipient
12:26:35 --- mode: ChanServ set mode: -o klooie
12:26:35 --- mode: ChanServ set mode: -o rob_ert
12:26:35 --- mode: ChanServ set mode: -o MrGone
12:27:00 <adu> whatever the bootloader loads could be anything like if the boot loader was a forth interpreter it could load fth source
12:27:34 <adu> but you can't point BIOS to a forth file and expect it to understand
12:29:10 <adu> damnit i wish people wouldn't be so stubborne
12:29:30 <adu> i want to work on a common os, that will please everyone, and it won't if only one person is working on it
12:29:40 <adu> computers aren't like the real world
12:29:57 <adu> you don't haveto provide only one thing
12:30:06 <I440r> THAT would be the ONLY executable to actually get written to disk
12:30:06 <I440r> yes the forth kernel would be
12:30:06 <I440r> but nothing else
12:30:10 <adu>  you can please everyone
12:31:02 <adu> uber
12:31:07 <adu> yey
12:31:07 <adu> i'm crazy
12:31:12 <adu> i just went insane
12:31:23 <adu> hey did i tell you about a new lang i made with 3 keywords?
12:33:53 <adu> SDL is easy, but isn't as centralized
12:33:59 <rob_ert> Yup
12:34:15 <rob_ert> no?
12:34:15 <rob_ert> sounds interesting
12:34:15 <rob_ert> no, i mean
12:34:23 <adu> oops
12:35:34 <rob_ert> hehe
12:35:34 <rob_ert> what keywords?
12:35:44 <adu> all, none, thru
12:35:55 <adu> they're state and scope modifiers in one
12:36:38 <adu> instead of volatile/static/const & public/protected/private its all/thru/none respectively
12:37:00 <adu> and all the keywords of other langs are expressed implicitly by syntax and structure
12:37:22 <adu> like if/case/for/do/while/class/struct/function/method
12:37:43 <adu> cool huh?
12:37:50 <I440r> ugh method = oop
12:38:01 <adu> ugh?
12:38:09 <I440r> ugh
12:38:11 <I440r> nuff sed
12:38:12 <I440r> :P
12:38:43 <adu> heres an example, d=statements c=condition v=variable e=error type:
12:38:44 <adu> TRY CATCH	{d.}, (e):{d.}, {d.}.
12:38:44 <adu> IF ELIF ELSE	<c>{d.}, <c>{d.}, {d.}.
12:38:45 <adu> SWITCH CASE	(v):{l;d.}
12:38:45 <adu> FOR IN	(d.d.)<c>{d.}
12:38:45 <adu> DO WHILE	:{d.}<c>
12:38:46 <adu> WHILE	<c>:{d.}
12:39:10 <rob_ert> have you made a compiler for that lang?
12:39:17 <adu> no
12:39:26 <rob_ert> interpreter?
12:39:31 <adu> i've started with a BNF/Yacc description tho
12:40:30 <adu> i was thinking of doing a version in forth with special instances of ':{' and }< and ): operators and things like that
12:40:55 <rob_ert> hehe
12:41:13 <rob_ert> IF 2 @_@ 5 THEN ;-)
12:41:49 <adu> whats @_@?
12:41:54 <rob_ert> Hmmm
12:41:59 <rob_ert> Strange smiley ;)
12:42:05 <adu> o hehe
12:42:28 <adu> and here is how i got rid of the return keywords:
12:42:29 <adu> break;	<-.
12:42:29 <adu> continue; 	<--.
12:42:29 <adu> return #;	<- #.
12:42:29 <adu> throw T;	<-: (T) .
12:42:30 <adu> catch T;	(T){ ... }
12:44:36 <adu> anywayz, i'll probably never get it working until i get some other langs working on this machine
12:44:44 <adu> i want to get clisp and bigforth working
12:45:04 <adu> maybe even port MINOS to MacOSX
12:45:13 <rob_ert> MINOS?
12:45:15 <rob_ert> What's that?
12:45:20 <adu> bigFORTH's GUI library
12:45:47 <rob_ert> How do you do function calls in forth?
12:45:54 <adu> function calls?
12:45:58 <adu> uh you don't
12:46:09 <adu> every word is a function call
12:46:21 <adu> if is a function call
12:46:25 <adu> then is a function call
12:46:32 <adu> + is a function call
12:46:39 <rob_ert> heh, ok
12:46:43 <rob_ert> but how do you call a procedure
12:46:45 <rob_ert> ?
12:47:00 <adu> by typing its name
12:47:23 <rob_ert> ok
12:47:36 <adu> theoretically if oyu wanted to call a C function or procedure, you'd haveto build a wrapper, and push values on the stack then call its name
12:47:39 <rob_ert> the procedure, does it start with a label?
12:47:54 <adu> no
12:48:09 <adu> everything in forth is separated by a space
12:48:19 --- quit: rob_ert (Read error to rob_ert[h3n1fls33o898.telia.com]: Connection reset by peer)
12:48:20 <adu> and everything between the spaces is a word, or a mini-program
12:48:27 --- join: rob_ert (robost86@h3n1fls33o898.telia.com) joined #forth
12:48:41 <rob_ert> Damn, I was disconnected
12:48:49 <adu> so if you wanted to call printf("Hi %s", name); for example
12:49:00 <adu> forth[12:47] : and everything between the spaces is a word, or a mini-program
12:49:26 <adu> so you'd haveto have a wrapper that knew to call printf when the word printf was called
12:50:07 <adu> then you'd do:  " Hi %s" name printf
12:50:20 <rob_ert> ok
12:50:25 <adu> the " is a word that reads until it finds another " and pushes it onto the stack
12:50:25 <rob_ert> like in asm
12:50:42 <adu> notice that the " is surrounded by spaces
12:51:19 <rob_ert> yes
12:51:26 <adu> now name is a variable that pushes its contents to the stack when executed, and printf would be an assembled routine with a C function call that takes the values on the stack, and calls the C function with them
12:52:10 <rob_ert> When you type a vars name in forth ,do you push its address?
12:52:15 <adu> i wish there was a SWIG port to forth
12:52:28 <rob_ert> SWIG = ?
12:52:29 <adu> no its value
12:52:43 <adu> SWIG = Standard Wrapper Interface Generator
12:52:55 <adu> it converts C/C++ APIs to perl/python/tcl
12:52:59 <rob_ert> Ok
12:53:00 <adu> and ruby i think
12:53:48 <adu> rob_ert: if your just learning about langs, i suggest Ada, Eiffel, Lisp, and Forth
12:54:11 <rob_ert> Hmmm.... I want to learn them if they're useful
12:54:25 <adu> C is what every body says is useful
12:54:28 <rob_ert> Forth seems to be a small language, and I really like that
12:54:33 <adu> just bcuz every1 uses it
12:54:41 <adu> but as history has proven time and again
12:54:50 <adu> the majority is often wrong
12:55:09 <rob_ert> So C is bad you mean?
12:55:35 <adu> C isn't bad, but its unsafe, and that means its dangerous, not bad
12:56:00 <adu> but its the only widely known portable lowlevel language
12:56:21 <rob_ert> Yeah
12:56:22 <adu> C's pointers are one of its downfalls
12:56:41 <rob_ert> Hmm
12:56:49 <rob_ert> are they? why?
12:56:50 <adu> thats one reason i like Ada, it can do everything C can without C's mistakes
12:57:50 <adu> rob_ert: bcuz its the programmer's responsibility to make sure that pointer resolution is resolved correctly, and to ensure that the memory at that address is allocated corectly, Ada also comes with Garbage collection
12:58:06 <adu> C doesn't
12:58:43 <rob_ert> "pointer resolution is resolved correctly" means?
12:58:57 <adu> Ada uses the in/out specifiers so passing by reference can still take place
12:59:17 <adu> a pointer is an address, and to resolve it you have to load the value at that address
12:59:43 <adu> and you can have multiple pointers, like the char **argv you often see
12:59:57 <adu> that hasto be resolved 2 times, in order to get the value
13:00:03 <rob_ert> ok
13:00:09 * rob_ert is not very good at english
13:00:13 <rob_ert> me english bad yes
13:00:17 <rob_ert> ;)
13:00:24 <adu> then you haveto use pointer aritmetic to get the rest in that array
13:00:31 <adu> its also seen char *argv[]
13:00:32 <adu> same thing
13:00:53 <rob_ert> I fond C pointers logical
13:00:59 <rob_ert> I find C pointers logical
13:01:01 <adu> in C an array is a pointer to memory that contains pointers to a list of objects
13:01:07 <adu> they are
13:01:11 <adu> but they're dangerous
13:01:14 <rob_ert> hehe
13:02:00 <adu> it does it like that partly to save efficiency, but ada can do the same without loosing that efficiency
13:02:36 <adu> other langs like perl/python/java aren't as efficient
13:02:40 <adu> they're just easy to use
13:02:48 <rob_ert> yes
13:02:59 <adu> and one thing i hate is that languages aren't used in their place
13:03:15 <adu> every lang has a purpose, and most often their used for purposes they aren't designed for
13:03:37 <rob_ert> BASIC is for learning and small math programs
13:03:47 <rob_ert> C is for a lot of stuff
13:03:55 <rob_ert> Assembler if for Demos and OS's
13:04:00 <rob_ert> ;)
13:04:18 <adu> the best language for applications is CommonLisp
13:04:27 <rob_ert> Hmm
13:04:31 <adu> the best language for quick webapps is Python
13:04:34 <adu> imho
13:04:36 <rob_ert> ¨never tried (or heard about)  that
13:04:47 <adu> the thing is is that many langs are designed for multiple purposes
13:05:01 <klooie> common lisp is the most impressive language i've ever seen, still learning it.
13:05:13 <adu> its a big language
13:05:28 <adu> scheme is good for learning lisp-like langs, but its not right for big apps
13:06:11 <adu> forth is good for OSs too, and debuging and testing
13:06:22 <adu> good for real-time environments
13:06:29 <rob_ert> any OS's written in Forth?
13:06:42 <adu> heh, most forth environments ARE OSs
13:06:47 <adu> the early ones at least
13:06:51 <rob_ert> Hmm
13:06:58 <rob_ert> then BASIC is OS language too
13:07:16 <adu> now they're degraded by making them emulated programs that can only interpret, not control the underlying hardware
13:07:37 <adu> igtg
13:07:40 <rob_ert> ok
13:07:43 <rob_ert> see you
13:07:45 <adu> i want to see a movie
13:07:54 <adu> rob_ert: good luck, languages are awsome!
13:08:01 <adu> just keep in mind:
13:08:26 <adu> languages aren't like girlfriends, you're allowed to know more than one!
13:08:33 <rob_ert> Hehe
13:08:35 --- quit: adu (IRCStep)
13:33:08 --- nick: rob_ert -> rob_ert_aw_ay
14:06:07 --- nick: MrGone -> MrReach
14:08:51 --- join: cleverdra (jfondren@1Cust57.tnt4.florence.sc.da.uu.net) joined #forth
14:08:51 --- mode: ChanServ set mode: +o cleverdra
14:09:00 --- mode: cleverdra set mode: +oooo clog I440r klooie MrReach
14:09:01 --- mode: cleverdra set mode: +oooo nate37 rob_ert_aw_ay scipient Trey
14:10:35 --- topic: set to ': 2pick 2>r dup 2r> rot ;' by cleverdra
14:12:11 <cleverdra> Hello, #Forth.
14:18:28 <MrReach> heya, cleverdra
14:18:48 <cleverdra> how're you doing, MrReach?
14:19:10 <MrReach> pretty good, had an earthquake this morn
14:19:48 <cleverdra> oh?  Where are you?
14:20:03 <MrReach> Spokane, Wa
14:20:31 <MrReach> one of the most geologically stable places in the USA ... it was one of the reasons I moved here.
14:21:08 <cleverdra> Huhn.
14:21:19 <MrReach> so go figure
14:23:00 <cleverdra> I've just finished the FIGGY BAR logs, and some of the bestofGEnieu stuff, and am now working my way through the conferences.  Interesting stuff.
14:23:24 <cleverdra> I had a thunderstorm or something this morning =)  Not an earthquake... any damage?
14:23:41 <MrReach> yes, amazing the sophistication that some of the early forth writers displayed
14:23:54 <MrReach> no, no damage to speak of
14:24:28 <MrReach> lot of stuff shaken off the walls, some cracked patios ... and one street is unusable because of a big crack in it
14:24:45 <cleverdra> I don't think these are "early".  A few years back, about the time of the ANS Forth effort.
14:24:54 <MrReach> oh, ok
14:24:58 <cleverdra> Oh, a big crack.
14:25:18 <cleverdra> Just like in the movies.  I'm sure the children along that street (if there are any?) are enjoying themselves.
14:25:59 <MrReach> pretty much business as usual here, except that the aftershocks are rather unnerving
14:26:15 <cleverdra> how frequent/strong are they?
14:26:49 <MrReach> there was the initial quake of 3.4, about an hour later, another quake of 3.2
14:27:02 <MrReach> for a long time, tremors every 30-40 minutes
14:27:25 <MrReach> just had one about 10 mins ago, but it was 2 hours after the one before it
14:30:04 <cleverdra> Yeah, that must be disturbing.
14:32:41 <MrReach> heh, the animals HATE it (my dogs and cats) ... they're all in the office with me, huddled around my feet.
14:33:33 <cleverdra> =)  How're the other humans doing?
14:35:30 <MrReach> don't know ... haven't left the house today
14:35:58 <MrReach> big frenzy on the news, but that's what the news does
14:36:04 <cleverdra> Yeah.
14:36:54 <cleverdra> Any Forth-related news?
14:37:08 --- nick: rob_ert_aw_ay -> rob_ert
14:37:23 * cleverdra is still wondering about that CR/LF thing of Hohensee's.
14:37:26 <cleverdra> hello rob_ert.
14:37:37 <MrReach> haha! not much here
14:37:41 <rob_ert> Hi
14:39:35 <cleverdra> I thought you'd decided to stick with C or something, rob_ert =)
14:55:18 <cleverdra> http://www.jwdt.com/~paysan/dragon/ - hee =)
14:58:56 <MrReach> bye, take care
14:59:02 --- quit: MrReach ()
15:02:04 --- join: aaronl (aaronl@vitelus.com) joined #forth
15:13:06 --- mode: cleverdra set mode: +o aaronl
15:13:12 <cleverdra> hello aaron
15:19:51 <aaronl> hi
15:40:29 <I440r> does an amd k6-2+ support intel rng ?
15:40:45 <I440r> or does that mean "INTEL ONLY" :)
15:40:51 <rob_ert> rng = ?
15:40:57 <I440r> random number generator 
15:41:05 <I440r> build into the processor or something
15:41:13 <rob_ert> ah
15:41:21 <rob_ert> wtf
15:41:25 <rob_ert> if you hear me
15:41:28 <rob_ert> i have to reboot
15:41:44 <cleverdra> I don't even have a pIII for that thing, and I don't use RNGs enough in cases where a hardware one would matter.
15:41:52 <cleverdra> you have to reboot?
15:41:56 <cleverdra> well, bye.
15:42:06 <rob_ert> maybe
15:42:07 <rob_ert> wait
15:42:19 * cleverdra favorite RNG is a teeny little thing implemented in Forth.
15:42:48 <cleverdra> I use it on Enth/Flux.  It's written in Color Code, and uses the RTC to add a little randomness.  I dunno how much.
15:43:15 <cleverdra> I've only used it so far for a display hack (random colors) and a Shakespearean Insult Generator, anyway.
15:43:43 <rob_ert> hehe, no need for rebooting
15:44:15 <I440r> :)
15:44:42 <I440r> its a kernel compile option - cat/proc/cpuinfo shows my processor doesnt support it so no poing in compiling it in :)
15:45:02 <rob_ert> oh
15:45:17 <rob_ert> don't build in CPU specific functions for a CPU you don't have
15:45:34 <cleverdra> Rule #342 of Assembly Programming.
15:45:50 <rob_ert> lol
15:46:14 <rob_ert> maybe #2 or something
15:46:20 <rob_ert> (what is #1?)
15:47:43 <cleverdra> #1 is the same as in any other language -- The program must work!
15:47:54 <rob_ert> hmmm.... why?
15:48:20 <rob_ert> My programs never work
15:48:32 <rob_ert> but I use some piece of code from it in the next prog
15:48:45 <cleverdra> If you ever compile them, they've worked *that* far =)
15:48:56 <cleverdra> but you're right.  I dunno what #1 is.
16:10:39 <cleverdra> www.figTaiwan.org
16:10:44 <cleverdra> int'resting
16:19:04 --- join: adu (andrew@63.201.90.87) joined #forth
16:19:11 --- mode: cleverdra set mode: +o adu
16:19:14 <cleverdra> hello adu.
16:36:04 --- nick: rob_ert -> rob_ert_sleep_ing
16:37:01 <rob_ert_sleep_ing> night
17:01:53 --- nick: rob_ert_sleep_ing -> rob_ert
17:06:44 * cleverdra returns.
17:06:51 <cleverdra> good night sleep, robert? =)
17:06:55 * rob_ert salutes him
17:06:56 <rob_ert> hehehe
17:06:58 <rob_ert> no
17:07:00 <rob_ert> short sleep
17:08:33 --- nick: rob_ert -> rob_ert_may_be_sleep_ing
17:11:58 * aaronl is away: I'm busy
17:12:31 --- quit: rob_ert_may_be_sleep_ing ()
17:33:00 --- join: futhin (thin@h24-67-113-99.cg.shawcable.net) joined #forth
17:33:15 <futhin> hi all
17:33:42 <cleverdra> hello futhin.
17:34:01 <futhin> somebody tell me if forth is an interpreted lang or a compiled lang or what?
17:34:13 <nate37> heh
17:34:14 <cleverdra> it's both.
17:34:22 <cleverdra> If you wait an hour or so, I'll back and explain it fully.
17:34:23 --- quit: cleverdra (Leaving)
17:34:35 <futhin> heh
17:38:30 <I440r> hi fu!!
17:38:46 <I440r> its both
17:38:50 <I440r> interpreted and compiled
17:38:54 <futhin> hi i4!! :)
17:39:01 <futhin> how so?
17:39:02 <I440r> it seems complex but its realy very simple
17:39:26 <I440r> when you type FOO on the keyboard forth will either interpret it or compile it depending on the current state
17:39:34 <I440r> the default state is interpret
17:39:38 <I440r> so when i type 
17:39:44 --- quit: adu (Read error to adu[63.201.90.87]: Connection reset by peer)
17:39:50 <I440r> : foo  some more words here ;
17:39:58 <I440r> forth starts interpreting that line
17:40:08 <I440r> the word : (pronounced colon) does a number of things
17:40:15 <I440r> it switches us into compile mode
17:40:41 <I440r> then creates a header in the dictionary for the new word (with the name field of the header set to 'FOO;
17:40:46 <I440r> then it exits
17:40:55 <I440r> so when forth sees 'some more words here'
17:41:08 <I440r> it compiles each of them into the definition for FOO
17:41:19 <futhin> hm
17:41:20 <I440r> the word ; is special - its called an immediate word
17:41:28 <I440r> even tho we are in COMPILE mode
17:41:33 <I440r> forth INTERPRETS it
17:41:38 <I440r> it executes its code
17:41:41 <I440r> ; does this
17:41:48 <futhin> yeah ok, gotcha
17:41:57 <futhin> i know the immediate stuff.. like IF THEN
17:42:07 <futhin> <--- read starting forth a few days ago
17:42:08 <I440r> it compiles an UNNEST onto the end of the just compiled : definition
17:42:20 <I440r> it then switches us out of compile mode back into interpret mode
17:42:26 <I440r> futhin thats a very good book
17:42:31 <I440r> get thinking forth too if you can
17:43:10 <futhin> yeah :)
17:43:14 <nate37> UNNEST
17:43:18 <nate37> hmm mine is just EXIT
17:43:18 <nate37> heh
17:43:19 <I440r> so as u can see - forth is interpreted AND compile
17:43:25 <I440r> exit is another name for unnest
17:43:31 <futhin> gotcha
17:43:32 <nate37> yea..
17:43:42 <nate37> EXIT is in Starting Forth
17:43:43 <I440r> dependintg on context i use each
17:43:50 <nate37> futhin: did you finish the book?
17:43:54 <I440r> i consider it to be an unnest if its at the end of a : definition
17:44:03 <I440r> but an exit if its in the middle of a : def hehehe
17:44:25 <nate37> heh
17:44:45 * nate37 making forth for ti-85
17:45:11 <futhin> yeah i finished the book..
17:45:18 <nate37> hmm
17:45:27 <I440r> futhin
17:45:29 <I440r> read it again :)
17:45:31 <I440r> and again
17:45:32 <nate37> Moving Forth isn't bad also
17:45:32 <nate37> yea 
17:45:32 <I440r> and again :P
17:45:33 <futhin> yeah
17:45:38 <nate37> and thinking forth
17:45:43 <nate37> Moving Forth <-- online
17:45:52 <futhin> there was some parts that i skimmed
17:45:55 <nate37> Thinking Forth <-- another book by Leo Brodie
17:46:00 <futhin> because i have to code the stuff first
17:46:06 <nate37> futhin: "Under the Hood"? ;)
17:46:12 <I440r> leo brodie now says he disagrees with most of what he said in thinking forth
17:46:19 <nate37> heh oh well
17:46:20 <I440r> he has gone all OOP on us
17:46:43 <nate37> OOP == changing dictionary :)
17:46:51 <I440r> no
17:47:02 <nate37> heh
17:47:04 <I440r> thats not oop
17:47:08 <I440r> oop = =evil
17:47:15 <I440r> vocabularies = good
17:47:16 <nate37> i like OOP a bit
17:47:30 <I440r> except in ans where they dont even call them vocabularies
17:47:36 <I440r> they invented a new arty farty name for them
17:47:42 <nate37> what is that name?
17:47:48 <I440r> i forget
17:47:58 <I440r> i try to forget as much as i can about ans forth
17:48:05 <nate37> i've been stealling words i should implement into ti-forth from ans.. otherwise i havn't read it much
17:48:10 <nate37> heh
17:49:09 <I440r> i read part of it - am diametrically opposed to what i read so much i didnt bother reading the rest
17:49:24 <I440r> ans is as bad as oop
17:49:25 <nate37> what part did you read?
17:49:33 <I440r> 1.e fulll of BS
17:49:42 <nate37> what is better then?
17:50:07 <I440r> ok .. lets change the subject slightly for a second or 2
17:50:18 <nate37> why? boss comming around or something?
17:50:19 <nate37> heh
17:50:22 <I440r> i ahve always said taht anything you can do WITH recursion you can do better without
17:50:30 <I440r> that statement applies equally to oop
17:50:35 <nate37> hmm
17:50:56 <nate37> recursion == iteration? or any type (including tail-end)
17:51:16 <I440r> OOPs biggest failing is that it makes it so very easy to turn a 3 week project into a 3 year project that will never work and will be scrapped eventually
17:51:45 <I440r> OOP was invented by "career academics" who never fscking coded a line of REAL code in their lives
17:53:40 <nate37> are you aiming at Smalltalk-80,Self,C++ (woo way different from first two heh)?
17:55:13 <nate37> I440r?
17:55:15 <I440r> any oop
17:55:23 <nate37> python?
17:55:36 <I440r> ive yet to see a single instance of oop thats not totally fscked in the head
17:55:47 <nate37> example
17:55:51 <I440r> PARTS of oop are ok - most of it is bullshit
17:56:00 <I440r> something BAD from oop
17:56:14 <I440r> information hiding
17:56:18 <nate37> altho i've been coding in z80 asm the last 2 weeks so any higher language (cept forth) atm is painfull ;)
17:56:25 <I440r> classes
17:56:28 <I440r> inheratance
17:56:32 <nate37> python untyped
17:56:40 <nate37> python doesn't have information hiding
17:56:44 <I440r> all do nothing to make the code easier to read OR write
17:56:48 <nate37> and hardly any inheratency
17:56:49 <I440r> but do a hell of alot to make it fat
17:57:39 <I440r> every time you create an instance of a class you DUPLICATE the structures associated with that class
17:57:47 <I440r> duh!
17:58:07 <I440r> why not jsut have a single function and pass it different parameters depending on what you are trying to do
17:58:10 <I440r> SHARE code
17:58:13 <I440r> SHARE information
17:58:23 <I440r> make you code small and fast
17:58:30 <I440r> oop is less than useless
17:59:16 <nate37> a function be an object?
17:59:35 <I440r> no
17:59:41 <I440r> who needs objects
18:00:05 <I440r> an object is just a structure that defines the properties of a function
18:00:10 <I440r> who gets to see what parts of it
18:00:18 <I440r> what data structures the code needs
18:00:21 <I440r> who creates them 
18:00:26 <I440r> who destroys them
18:00:27 <I440r> argh
18:00:28 --- join: adu (andrew@63.201.92.85) joined #forth
18:00:28 <nate37> python basically has objects as dictionarys.. untyped.. when you call objcet methods it actually calls the method with the object as arguments....
18:00:30 <nate37> altho slow
18:00:39 <I440r> lets just tie our code into one giant gordian knot why dont we
18:00:48 <nate37> heh
18:01:01 <nate37> hmm
18:01:13 <nate37> but providing an interface, for some things, can be useful
18:01:15 <adu> sure
18:01:22 <I440r> i had a friend of mine who wrote a puzzle game
18:01:35 <I440r> you removed coloured blobs from the board and the ones above it fell down
18:01:50 <I440r> you could only remove blobes that had 4 in a row
18:02:05 <I440r> when the ones above fell down you ahd to find anotehr 4 in a row to remove
18:02:06 <adu> cool
18:02:11 <I440r> to win you had to clear the board
18:02:27 <I440r> he had a recursive function to find 4 adjacent blobs
18:02:35 <I440r> they didnt need to be in a row..
18:02:51 <I440r> jsut any 4 of the same colour next to each otehr horizontally or vertically
18:02:57 <I440r> maybe diagonally also
18:02:59 <I440r> i cant remember
18:03:06 <nate37> ok...
18:03:12 <I440r> his recursive function was realy realy slow hehe
18:03:16 <nate37> heh
18:03:24 <I440r> so i sent him a copy of my regeon flood fill function 
18:03:27 <I440r> non recursive
18:03:31 <nate37> depending on the size of the board and how its stored.. find appropriate method
18:03:41 <nate37> math, recursion or data structure
18:03:47 <I440r> recursion is never appropriate
18:03:49 <nate37> eh
18:03:51 <nate37> heh
18:03:54 <I440r> even with tree traversal
18:03:59 <I440r> its jsut fucked up
18:04:02 <I440r> what is recursion
18:04:04 <I440r> well
18:04:10 <I440r> its a method of looping
18:04:17 <I440r> actually - its 2 loops :)
18:04:31 <I440r> the loop where you call self till a limit is reached
18:04:38 <nate37> it is applying a set of instructions repeadely to data
18:04:39 <I440r> followed by a loop of the same number of returns 
18:04:46 <futhin> you know the quicksort method? i've been meaning to implement it in assembly for fun and practice..  haven't gotten around to it, but the version that i know of is recursive..
18:04:57 <I440r> why flood your stack with loop branc vectors ?
18:05:02 <I440r> why do it - its stupid
18:05:11 <nate37> ah not iteration?
18:05:13 <I440r> quick sort
18:05:16 <I440r> anything but
18:05:25 <I440r> go to www.purplecoder.com
18:05:32 <I440r> click on the link to my code
18:05:36 <I440r> and then to my sorting code
18:05:39 <futhin> there's a quicksort interative version i think
18:05:42 <I440r> read whta i wrote
18:05:48 <I440r> its written by gnu
18:05:59 <I440r> its the gnu itterative quicksort
18:06:08 <I440r> and - its even worse than the recursive version
18:06:13 <I440r> its sphagetti code
18:07:15 <nate37> how old are you?
18:07:36 <I440r> me ?
18:07:38 <I440r> 37
18:07:55 <nate37> 15
18:07:56 <nate37> heh
18:08:03 <I440r> then you should be nate15 :)
18:08:08 <futhin> i440r: i only see heap, shuttle, shell, and count sort
18:08:17 <I440r> i mention quick sort
18:08:25 <nate37> I440r: 3 = march 7 = 7th.. cheezy but..
18:08:31 <nate37> at least it won't change with mood :)
18:08:41 <I440r> the count sort is probably the fastest sort method ever invented
18:08:55 <I440r> i couldnt thing of a faster method - 
18:09:04 <I440r> it doesnt compare any items in the list to sort with any other
18:09:08 <nate37> brb
18:09:16 <futhin> it's fastest for large files or small files? :)
18:09:25 <I440r> but its an out of plae sort
18:09:27 <I440r> for any
18:09:35 <I440r> you need 2 buffers
18:09:38 <I440r> one for unsorted data
18:09:43 <I440r> one for sorted
18:12:33 <futhin> hm
18:13:11 <adu> ii want to make my own kernel
18:14:07 <I440r> me 2
18:14:18 <I440r> a 100% forth based unix clone :)
18:14:27 <I440r> now THAT would rock :)
18:15:37 <adu> I440r: lets start
18:15:37 <futhin> heh :)
18:15:43 <adu> =)
18:16:20 <I440r> hehe
18:17:46 <adu> : ipc-init lock-task ... unlock-task ;
18:17:50 <nate37> back
18:17:51 <adu> hehe
18:17:53 <futhin> making it posix compatible isn't necessarily a good idea..
18:18:05 <adu> dang this is taking alot of time
18:18:11 <futhin> why not just make a good os instead :P
18:18:17 <adu> i'm deleting the XFree source cvs
18:18:30 <adu> 'xc'
18:19:04 <adu> as i recall, mach by itself isn't posix compliant
18:20:55 <adu> yet its still cool =)
18:21:14 <I440r> win 2k is posix compliant
18:21:48 <adu> i want to make a mach ernel in forth
18:22:04 <futhin> win2k is posix compliant? ugh
18:27:29 <I440r> :)
18:27:48 <I440r> ugh is MY word :P
18:28:09 <adu> no way!
18:28:13 <I440r> :)
18:28:19 <I440r> is to too!!!!
18:28:29 <I440r> ugh and bleh (tm)
18:29:18 --- topic: set to ': 2pick 2>r dup 2r> rot ; - do drop >in - turn on, boot up, log int, drop out' by I440r
18:29:34 --- mode: I440r set mode: +o adu
18:29:36 --- mode: I440r set mode: +o futhin
18:35:24 <I440r> hrm that definition for 2pick gives me an idea
18:35:38 <I440r> have a word called grap and another called drop
18:35:49 <I440r> 2>r dup grap 2r> drop
18:35:59 <I440r> grab even
18:36:33 <I440r> could use di as a temp store for grab as long as we dont use any word taht might use it
18:36:48 <I440r> maybe we could have grab-bx grab-cx grab-dx etc
18:36:58 <I440r> and stay where to hold onto the item :)
18:39:31 * aaronl is back (gone 01:27:33)
18:41:14 <nate37> back
19:23:35 --- join: edrx (edrx@copacabana-ttyS22.inx.com.br) joined #forth
19:34:07 --- join: cleverdra (jfondren@1Cust245.tnt6.florence.sc.da.uu.net) joined #forth
19:34:07 --- mode: ChanServ set mode: +o cleverdra
19:34:12 --- mode: cleverdra set mode: +o edrx
19:34:39 <cleverdra> hey futhin.  You still wondering if Forth is interpreted or compiled?  Has somebody explained threading to you?
19:36:06 <futhin> well my understanding of it is that the code in the words are compiled into the dictionary and when it encounters an immediate word within the code it executes that (interpret)
19:37:02 <futhin> that's what i gleaned from what the other people said
19:37:10 <cleverdra> Forths can be more complex than that, but yeah.  Forth is both an interpreter and a compiler.
19:37:32 <cleverdra> Do you know about the inner and outer interpreters?
19:39:07 <futhin> um.. sorta? :)
19:39:17 <futhin> <-- read starting forth a few days ago
19:39:28 <futhin> but it didn't talk about inner and outer interpreters..
19:39:31 <cleverdra> OK.  Here's a brief rundown:
19:39:33 <futhin> i read underview for forth
19:39:45 <futhin> and it said an text interpreter and an address interpreter
19:39:51 <cleverdra> The outer interpreter is what deals with Forth source code.
19:40:01 <cleverdra> yeah.  text->outer, address->inner.
19:41:28 <cleverdra> The outer interpreter reads from the input stream, or from blocks, and does some intelligence on what it sees there.  The outer interpreter can do typical Forth, or Color Forth, or all kinds of wierdness.  The result is that it compiles words into the dictionary and executes words from the dictionary (by looking up the text it sees in the dictionary...)
19:42:22 <cleverdra> The inner interpreter is what actually deals with words, what executes compiled Forth code in the dictionary.  The best way to describe the inner interpreter, I think, is to give an example of its operation:
19:44:27 <cleverdra> Letsay I've defined three words... : under+ >r + r> ; : bump 1 under+ ; : 2count 0 0 begin 2dup . . bump over 10 = until 2drop ;
19:44:34 <cleverdra> OK, that's kinda wierd code.  I just made it up =)
19:45:39 <cleverdra> But that's all in the dictionary.  There's an NFA (name field address) somewhere that holds the counted string "2count", followed by some information and a list of addresses that correspond to the words in 2COUNT, and some extra stuff for the literal numbers.
19:46:04 <cleverdra> (numbers are compiled as double cells in the dictionary, a DOLIT or like word followed by the number...)
19:46:56 <cleverdra> If you type 2COUNT in now, the outer interpreter will find it in the dictionary, find the CFA (code field address) of it, and then the inner interpreter will step through the CFA executing Forth code.
19:48:11 <cleverdra> It'll find DOLIT, push 0 on the stack, find DOLIT, push another 0 on the stack, execute 2DUP (begin doesn't compile to anything -- it's an immediate word), execute . and ., execute bump, over, put 10 on the stack, execute =, and then conditionally jump back to 2DUP, or execute 2drop and then NEXT.
19:48:45 <cleverdra> You don't have to understand all this right now =)  Just get the idea.
19:49:04 <cleverdra> Those words, words like 2DUp and BUMP and OVER, are just numbers.  Cell-sized numbers.
19:50:12 <cleverdra> The inner-interpreter goes to the address in memory correspondant to these numbers and then proceeds to execute all the code there.  Eventually it'll get to actual machine code, and run that =)  Every word in Forth eventually gets down to this machine code, but it may go very deeply through other words and jump around and play with the stack and do all kinds of things.
19:51:17 <cleverdra> The inner interpreter DOES NOT see "2dup" or "bump" or any of that.  It only sees numbers.  It only moves through memory running the machine code deep down in the dictionary.
19:51:48 <cleverdra> This is called "threading", because the inner interpreter acts kinda like a needle going deep into words and moving around and coming out... see?
19:53:53 <cleverdra> Letsay you've a word ``: a b ;'' and some code ``a''.  The outer interpreter reads ``a'', parses the word A out of it, looks that up in the dictionary, and then the inner interpreter finds itself at 'B'.  It doesn't see the string "B", it only sees a number there.  So it puts the address it's at right now, the address in memory where 'B' is inside the word 'A', on the return stack, and then jumps to the address in memory that 'B' in t
19:54:07 <cleverdra> That's kinda twisted to say.  Do you get what I mean, futhin?
19:55:42 <cleverdra> After it does who-knows-what from 'B', when it finally comes to its last NEXT, it'll get the number you've put on the return stack and jump to that part of memory.  Now you're back inside the word 'A' in the dictionary, and the next word is NEXT, so it jumps to that part of memory... and so on.
19:56:35 <cleverdra> the word ';' compiles NEXT into the dictionary before it changes the state of Forth back to interpreting.  This NEXT is implicit, but it's there.
19:57:19 <cleverdra> (well, sometimes it ain't, and the inner interpreter continues along.  In some systems you get from this what are called "cascading definitions")
19:58:15 <cleverdra> I'm done, futhin.
20:02:39 <futhin> ooh
20:02:43 <futhin> i was gone, talking to a friend :)
20:02:50 <cleverdra> Oh :/
20:03:30 * aaronl is away: I'm busy
20:05:34 <cleverdra> Well, do you get what I'm saying, futhin? =)
20:05:38 <futhin> doesn't it get to be kinda wasteful when you've got all these layers and you gotta go thru them all before you actually execute code?
20:06:03 <futhin> i understand pretty much everything you said :)
20:06:13 <futhin> i get it
20:06:16 <cleverdra> Not really.  You get tremendous flexibility and power from this, futhin.  Some forths don't have those layers, and optimize it all into machine code.
20:06:38 <futhin> hmm? some forths optimize it all onto one layer of machine code? :)
20:06:45 <cleverdra> Yeah, you can de-layer it, so to speak, and just have your application words with machine code and a super-small dictionary.
20:06:50 <futhin> that's cool
20:07:53 <futhin> why "super-small dictionary" ?
20:07:54 <cleverdra> It's not normally useful.  Forth is, can be, extremely fast even doing all this "wasteful" jumping around.  You can write small definitions, and should(!), and you can factor things intelligently, and all it costs is a jmp and putting something on the return stack.  C, for instance, does all kinds of messy things that make function-calling strenuous and inefficient.
20:08:18 <cleverdra> futhin - because you'll only have your application words, you won't need DUP and stuff because it'll be inlined.
20:08:23 <futhin> jmps take lots of cycles though :(
20:08:32 <cleverdra> Still faster than a call.
20:08:41 <cleverdra> lots of cycles?
20:09:00 <futhin> yeah, a jmp is like 50 cycles??
20:09:03 <futhin> cpu
20:09:06 <cleverdra> With cascading definitions you can get more speed from this, but really the only way to get faster is to write pure assembly and be really good at it.
20:09:53 <futhin> does C use calls? why'd you mention calls? :)
20:09:55 <cleverdra> There *are* optimizing Forth compilers.  Some of them are very good at it.  You can get Forth hardware that makes things unbelievably(!) fast.  Forth is normally quite fast and efficient besides.  See?
20:10:48 <I440r> i dont like optimizations at the processor level
20:10:51 <I440r> they obfuscate the code
20:10:54 <cleverdra> futhin - some Forths use calls.  Nevermind.  C does stack-frames and local variables and implicit stack mussing and all kinds of things that can be very wasteful.  If you follow with common optimizing philosophy, you say that this is nothing and that machines are fast and that, anyway, only 10% of your code is the bottleneck.  This is all just as true in Forth.
20:11:09 <I440r> i dont like any compiler to do anything other than a 1:1 source <--> object
20:11:11 <cleverdra> i440r - uh.. ever seen the machine code for the F21?
20:11:22 <futhin> i440r: why?
20:11:26 <cleverdra> i440r - oh, I see what you mean.
20:11:29 <I440r> impossible to debug
20:11:36 <futhin> ah
20:11:45 <futhin> why would you debug it at the assembler level? :)
20:11:48 <I440r> wanna debug c ? turn optimizations off
20:11:59 <I440r> no guarantee that the code will work the same when you turn them back on tho --
20:12:05 <I440r> sounds like an arguement NOT to use it
20:12:11 <I440r> optimize your algorighms
20:12:15 <I440r> NOT your code
20:12:29 <futhin> why would you debug it at the assembler level??
20:12:42 <cleverdra> The biggest benefit from threading is the incredible flexibility and power you gain, and how small your programs can become through good factoring.  Other reasons... people more familiar with Forth-building could probably tell you more.
20:12:44 <I440r> actually tho - i DO like to mess with optimizing my asm but not to extreemz
20:12:54 <I440r> yes - sometimes debugging the asm is the only way to fix c code
20:13:03 <I440r> sometimes the compiler does TREALY REALY stupid shit
20:13:28 <cleverdra> futhin - C optimizers sometimes make incredible changes to your source.  If you want to debug your program, it's usually a better idea to turn these optimizations off so you know what you're looking at.
20:13:48 <cleverdra> and that's just C level stuff.  I440r seems to know about deeper problems.
20:13:55 <I440r> but when you turn them back ono again dont expect the code to work the same :)
20:14:25 <I440r> experience tells me that if you let the compiler decide how your code is going to be then your code probably wont work very well
20:14:31 <futhin> why would you debug forth code at the assembler level?  (that's the whole point of not using an optimizing forth isn't it?!)
20:14:42 <I440r> futh i wouldnt
20:14:46 <I440r> i wouldnt need to
20:15:05 <futhin> what's wrong with an optimizing forth then?
20:15:09 <I440r> because i would knnow that the source i was looking at in the debugger would EXACTLY match the object code produced
20:15:18 <futhin> uh
20:15:20 <I440r> because it would destroy that
20:15:29 <nate37> I440r: ah ur count sort is similar to something in "Programming Pearls"
20:15:32 <futhin> why would you look at the source in the debugger?
20:15:40 <I440r> if i had to use an optimizing forth compiler i would need to debug it as the assembler level
20:15:46 <I440r> i would not trust it at the forth level
20:15:53 <I440r> imho it would not BE forth
20:16:20 <nate37> it would be forth the language, the forth the environment
20:16:24 <I440r> a forth debugger shows the name field of each token as you single step
20:16:34 <I440r> but NOT forth object
20:16:49 <I440r> a forth compile is supposed to compile FORTH object code
20:16:54 <I440r> if it doesnt - its not forth
20:17:05 <futhin> hm
20:17:28 <I440r> it might be good - but not forth
20:18:43 <futhin> hmm..  but in cases where you want to compile an executable that you are going to put on the internet so people can download it (a game perhaps)  why not use an optimizing forth - it's unlikely that bugs will appear suddenly when you compile the executable with an optimizing compiler..
20:18:58 <cleverdra> I'm just getting back to Enth here.  It's a subroutine-threaded Forth, with a Color Code (Flux) system built on it, a metacompiler and SASM.  I'm going to see how well I can apply my new assembly knowledge to this 486 machine =)
20:19:26 <I440r> :)
20:19:42 <cleverdra> futhin - you typically do not compile an executable... though admitedly you can.  Metacompilers can be used to create tiny binaries of various format with just your application code if you want.
20:21:00 <I440r> turnkey
20:21:08 <I440r> or saveexe/savesystem
20:21:16 <futhin> uh? no compiling the software into an executable??
20:21:18 <cleverdra> Yeah.
20:21:31 <futhin> yeah to what i said?
20:22:06 <cleverdra> futhin - no, read what I said.  I said that typically you don't do this... at least I don't.  You *can* do that though.  It's hardly difficult -- see what i440r said.  I was 'yeah'ing to him.
20:22:32 <adu> really
20:22:37 <adu> interesting
20:22:47 <cleverdra> You can remove the compiler, the assembler, the dictionary headers, and all that from your resultant binary.
20:22:57 <I440r> thers no such thing as a "forth program"
20:23:03 <cleverdra> I think Pygmy is used pretty effectively to do that.
20:23:20 <I440r> what ppl call forth programs are realy just forth - extended a little to do the particular job at hand
20:24:32 <cleverdra> Any Forth with a metacompiler can do that.  Actually Pygmy would be a harder medium unless you're producing dos COM files, because it comes with a simplistic clone compiler.
20:24:32 --- quit: cleverdra (Read error to cleverdra[1Cust245.tnt6.florence.sc.da.uu.net]: Connection reset by peer)
20:24:42 --- join: cleverdra (jfondren@1Cust245.tnt6.florence.sc.da.uu.net) joined #forth
20:25:05 <futhin> why is it there's so many forth's implemented in c?  i don't get it :(
20:25:06 <cleverdra> Any Forth with a metacompiler can do that.  Actually Pygmy would be a harder medium unless you're producing dos COM files, because it comes with a simplistic clone compiler.
20:25:09 <cleverdra> With a metacompiler, futhin, you can compile a binary executable for a totally alien system.  I could go into Pygmy today, write some assembly and set things up for my Pentium Linux computer, and then compile a Pygmy for Linux.
20:25:47 <futhin> like forth is "the most portable" lang, yet it keeps getting implemented in C.. that's fucked up heh
20:25:58 <cleverdra> futhin - there are lots of them implemented in assembly, and some that're only implemented in Forth and use assembly in their dictionary and a metacompiler to reproduce.
20:26:42 <futhin> i sort of know what a metacompiler is, but in that context, a metacompiler is another forth?? or what
20:27:00 <cleverdra> Ancient history, futhin :-/  People wanted to use Forth on their unix machines.  They wanted to sneak Forth into their C-based jobs.  They came from a C-flooded academic world and wanted to write what they thought were portable implementations of Forth.
20:27:08 <adu> so a metacompiler is different than a cross-compiler?
20:27:16 <futhin> yes
20:27:18 <cleverdra> um, let me get a link for you adu.
20:27:20 <futhin> i know that much :P
20:27:25 <I440r> no
20:27:27 <I440r> well
20:27:29 <I440r> yes and no
20:27:34 <I440r> they both serv the same function
20:27:36 <adu> I440r: ?
20:27:43 <cleverdra> adu - metacompilers can be clone-compilers, or cross-compilers, but they're both of these and more complex.
20:27:45 <I440r> they are both target compilers
20:27:53 <I440r> they just compile for different targets
20:28:23 <I440r> normally when forth compiles it taggs the new stuff onto the end of itself
20:28:32 <adu> right
20:28:48 <I440r> a TARGET compiler compiles a whole new image off in space somewhere - when finished it saves this out
20:28:49 <cleverdra> with a metacompiler you can, inside Forth, produce another Forth and write it as a binary for your system, or another system, or an imaginary system :-/  You just set the format up and write the assembly for the target system.  Sometimes you don't even do that.
20:28:58 <I440r> with a meta compiler the target is a new copy of itself
20:29:18 <I440r> wieht a cross compiler the target is an image of the binary you want executed on the target :)
20:29:22 <I440r> same difference heh
20:29:22 <adu> one thing i always feared about a huge forth system was if i wanted to remove something that i installed before the most recent installation
20:29:50 <cleverdra> Yeah, a metacompiler is like... your Forth is creating a copy of in memory.  It's reproducing.  Sometimes it's just splitting, sometimes it's producing something different, sometimes it's producing something *totally* different but with part of its dictionary, see?
20:30:11 <adu> ok
20:30:21 <adu> but not neccesarily a different arch?
20:30:28 <cleverdra> No.
20:30:36 <cleverdra> Not necessarily a different OS.
20:30:41 <adu> like can you metacompile x86 to ppc?
20:31:17 <I440r> yes
20:31:22 <I440r> well
20:31:26 <I440r> not x86 assembler
20:31:55 <cleverdra> You can from an x86-based Forth metacompile a Forth for PPC, yes.
20:32:00 <I440r> you could take forth source and have the same compiler compile it for x86, 8051, 68k or anthing
20:32:14 <adu> huh cool
20:32:16 <I440r> same compilers - just different meta compilers
20:32:23 <I440r> as long as it was pure forth
20:32:24 <cleverdra> I say 'Forth', but it's more flexible than that... it's really incredibly powerful.
20:32:34 <I440r> if it had any asm it would be more difficult
20:32:37 <adu> metacompilers are?
20:32:43 <cleverdra> metacompilers are.
20:32:50 <adu> o ok
20:32:50 <I440r> forth is
20:33:05 <I440r> meta compilation is just a small part of what makes forth so powerful :)
20:33:11 <adu> no i'm talking about x86 bytecode ->ppc or visa versa
20:33:55 <adu> you can't metacompile other languages?
20:34:18 <cleverdra> probably machine code, adu.  You're writing machine code for PPC and building your Forth around that, with your higher-level Forth words based on the new target-arch's machine-coded words.
20:34:33 <cleverdra> s/probably //
20:34:40 <cleverdra> metacompile other languages?
20:34:57 <I440r> ok. lets do the unthinkabke and code isforth in 100% PURE assembler
20:35:05 <I440r> we end up with a compiler that will not port
20:35:16 <I440r> you could write an equiv compiler on ppc tho
20:35:28 <I440r> so now we have two compilers
20:35:37 <I440r> identical except running on different machines
20:35:48 <I440r> so you write your super duper forth program
20:35:52 <I440r> no asm in your code
20:36:02 <cleverdra> If isforth is 100% pure assembler, your metacompiler would have to implement all the words in isforth (or those that you want) in the target arch :/  Your high-level Forth code is fine.
20:36:05 <I440r> it will compile and execute identically on BOTH machines without modification
20:36:17 <cleverdra> yes, what i440r said.
20:36:29 <I440r> isforth wont be 100% pure assembler
20:36:35 <adu> that sounds like a simple cross-compiler
20:36:37 <cleverdra> adu - in other languages metacompilation at the level of Forth is nontrivial.
20:36:47 <I440r> but i believe that ALOT of the inner core of a forth kernel should be pure assembler
20:36:58 <I440r> so - isforth kernel WONT be portable
20:37:11 <I440r> once the x86 kernel is working i want a ppc version
20:37:15 <I440r> i want an alpha version
20:37:20 <I440r> etc etc etc
20:37:25 <adu> I440r: we should provide C routines for other archs still tho
20:37:32 <I440r> the individual complers do not need to be portable
20:37:40 <cleverdra> adu - ... C routines ... ?
20:37:45 <I440r> each will utilise the specific processors they were written for
20:37:54 <I440r> USERS source code will port
20:37:58 <I440r> thats important
20:38:03 <cleverdra> Yes, that's cool.
20:38:08 <adu> then should it be ANS forth?
20:38:11 <I440r> no
20:38:23 <I440r> but it should follow a standard
20:38:30 <adu> our own?
20:38:37 <cleverdra> it doesn't have to be, adu.  Where do you get that?  i440r is saying that user isforth code will be portable between implementations of isforth, each of which will be optimized for their arch.
20:38:42 <I440r> each should contain the same words and they should execute identically on each
20:39:02 <I440r> thats the idea anyway
20:39:03 <cleverdra> polyForth does this.  As does LMI, as do most other multi-arch forths.
20:39:12 <adu> so we need an execution standard
20:39:14 <cleverdra> polyForth code is internally portable, you might say.
20:39:30 <adu> pForth is too
20:39:34 <cleverdra> adu - an execution standard?  Does that mean that the Forths act the same?  Yes, that's what i440r is saying re isforth.
20:39:48 <I440r> yes - poly is not one forth - like coke isnt one coke hehe (quote from pepsi guy -- coke are it) hehe
20:39:58 <adu> do you want to code than std'ize? or std'ize then code?
20:40:19 <I440r> i want the behaviour of the compiler standardised
20:40:35 <I440r> isforth uses ebx to cache top of stack
20:40:48 <I440r> ppc version need not cache top of stack if thers a better way
20:41:02 <I440r> but + - / . over etc etc  etc HAVE to function identically
20:41:10 <adu> what do pop&push do in x86?
20:41:24 <I440r> push eax  ; pushes register onto stack
20:41:33 <I440r>  pop eax ; pops item off stack into eax
20:41:43 <adu> so you can specify a reg, ok
20:41:53 <adu> ya that sounds like a good idea on ppc
20:41:57 <I440r> or a memory location
20:42:03 <I440r> push [12345678]
20:42:07 <adu> only theres no stack, concept, just mem, still doable tho
20:42:14 <I440r> push the contents of the memory at address 12345678
20:42:32 <adu> huh ya x86 is much more flexible than ppc
20:42:36 <I440r> push mem is in my opinion a true memory to memory move on x86 :)
20:42:41 <adu> ppc has so few things that do so much
20:42:41 <I440r> tho ppl dont think of it in that way
20:42:53 <adu> there are usually at least 3 args to each inst
20:43:29 <I440r> some with none
20:43:31 <I440r> some with 1
20:43:35 <I440r> most with 2
20:43:39 <I440r> some new ones with 3
20:43:58 <adu> the most on ppc is 5
20:44:46 <cleverdra> is PPC as CISC arch?
20:44:53 <adu> cmp crN, state, cmpOp, val1, val2
20:45:03 <adu> cleverdra: no RISC
20:45:15 <cleverdra> ah.  I guess that makes more sense actually, with the 5 arguments.
20:45:21 <adu> it has aprox 100, where x86 has like 350 not including mmx
20:45:49 <adu> and thers a comparison register, in 8 sections
20:46:15 <adu> then you can specify which condition register when you do conditional jump
20:47:43 <I440r> who besides apple uses the ppc ?
20:47:52 <adu> amiga, and ibm
20:47:59 <I440r> amiga doesnt exist
20:48:06 <adu> RS/6000 is a very popular high-end ppc arch
20:52:52 <cleverdra> bye.
20:52:54 --- quit: cleverdra (Leaving)
20:54:28 --- quit: edrx ([x]chat)
21:31:26 <nate37> hmm
21:31:54 <nate37> am i correct in saying that when an interupt occurs on x86 that it has to save via pushing and poping each time?
21:32:01 <futhin> yes
21:32:11 <futhin> i
21:32:13 <nate37> someone else...
21:32:18 <futhin> i'm pretty sure that's correct :)
21:32:24 <nate37> (wanted to get more oppionions ;))
21:32:38 <futhin> heh
21:33:50 <Trey> The "pushing" and "poping" include lots of details, but you have the basic idea right.
21:34:12 <futhin> do you have to do the push and pop yourself?
21:34:22 <nate37> Trey: but that takes a good number of cycles.. right?
21:34:31 <nate37> Trey: why not have two register sets like z80?
21:34:43 <nate37> so it only has to exx to change which set it uses
21:34:50 <Trey> Nested interrupts.
21:34:58 <nate37> ahhhhh
21:34:59 <nate37> :/
21:35:21 <futhin> i thought there was an instruction that automatically pushes everything and pops when it comes back..  like call or something
21:35:26 <futhin> and interrupt
21:35:35 <nate37> call pushes EIP 
21:35:37 <nate37> and stuff
21:35:41 <nate37> not sure what else
21:35:51 <Trey> call pushes EIP only.
21:36:02 <nate37> thought so.....
21:36:03 <Trey> PUSHA pushes all of the regs.
21:36:12 <nate37> but since i only know z80 asm well..
21:36:13 <nate37> hmm
21:36:16 <nate37> Trey: how many cycles?
21:36:24 <nate37> if you happen to know heh
21:36:25 <Trey> For which instruction?
21:36:29 <nate37> PUSHA
21:36:51 <futhin> the INT command pushes flags onto the stack
21:37:02 <Trey> (I don't know for current hardware.  I have a x386 book handy with cycle counts.)
21:37:20 <futhin> you can't do nested interupts in z80???
21:37:24 <nate37> i just need "the book of cycles"
21:37:25 <nate37> heh
21:37:41 <nate37> don't need details.. just cycles and maybe what instructions generally do
21:38:03 <nate37> futhin: not to my knowledge..
21:38:37 <futhin> nate: why not use INT?
21:38:52 <futhin> altho that's software interupts and i don't know about hardware interupts
21:38:56 <nate37> has 3 modes of interupts: 0 = device pushes what code to jump to onto bus 1 = go to a specific location (forget the location.. $33 or something)... 2 = interupt vector
21:38:58 <nate37> no INT
21:39:14 <nate37> no software interupts
21:40:01 <nate37> well i guess you could manually jump/call to the interupt vector if you wanted to emulate software interupts
21:40:43 <Trey> brb
21:40:46 --- quit: Trey (Read error to Trey[ns.TreySoft.com]: EOF from client)
21:40:55 <futhin> that means manually pushing and poping
21:41:21 <nate37> or exx
21:41:24 <futhin> how do you survive not nesting interrupts? :)
21:41:35 <nate37> umm?
21:41:44 <nate37> not many interupts on a ti-85? i dunno :P
21:41:54 <futhin> heh
21:42:31 --- join: Trey (bowser@ns.TreySoft.com) joined #forth
21:43:23 <nate37> wb
21:48:54 --- quit: futhin (bbl)
22:03:32 --- quit: Trey (Bowser[1.1-release]: server window terminating...)
22:13:12 --- join: futhin (thin@h24-67-113-99.cg.shawcable.net) joined #forth
22:13:54 <futhin> does forth keep the Top two elements of stack in registers?  or only TOS in a register?
22:14:40 <nate37> depends on implementation
22:14:53 <nate37> mine doesn't keep any
22:15:56 <futhin> hm
22:19:53 <futhin> don't you think you should at least have TOS in register?
22:22:32 --- join: Mike_L (mike@66-2-65-17-chi-03.cvx.algx.net) joined #forth
22:22:41 * Mike_L looks around
22:23:41 <futhin> hi
22:24:15 <adu> futhin: mike's making a bytecode that will make things really cool
22:24:19 <adu> any ideas
22:24:51 <futhin> why would it make things really cool?
22:24:57 <futhin> what's the purpose?
22:25:03 <Mike_L> ya stack machine
22:25:06 <futhin> what kind of ideas are you looking for? :)
22:25:31 <Mike_L> I want the VM to be able to work with data in registers and in memory
22:25:47 <Mike_L> and also immediate data for the instructions
22:25:52 --- mode: I440r set mode: +o futhin
22:25:54 --- mode: I440r set mode: +o Mike_L
22:26:01 <Mike_L> i am really going for a KISS method
22:26:12 <I440r> keeep it SINPLE stupid :)
22:26:15 <futhin> i440r: why not add me to the chanserv :P
22:26:18 <I440r> thats the #1 forth moto :P
22:26:29 <I440r> when u learn forth :P
22:26:30 <I440r> hehe
22:26:37 <Mike_L> so I'm thinking whether I should just make one instruction that takes immediate data and loads it in a stack slot
22:26:41 <I440r> or when i get tired of opping you :)
22:26:52 <I440r> (lit)
22:26:53 <Mike_L> two instructions that load/save data from stack to memory
22:26:54 <futhin> i don't understand bytecode?  where's the bytecode in forth?
22:26:56 <I440r> thats what (lit) does
22:27:16 <Mike_L> and then have all arithmetic instructions work only on the stack slots...
22:27:26 <Mike_L> futhin: I'm not doing this specifically for forth
22:27:42 <Mike_L> futhin: I'm trying to write an all purpose VM that will be an acceptible target for many programming languages
22:27:58 <futhin> what lang are you writing the VM in?
22:28:12 <futhin> please don't tell me something super lame like C :P
22:28:23 <futhin> heh
22:28:28 <Mike_L> I'm working with C right now because that is what I am most familiar with
22:28:37 <Mike_L> I am learning Scheme
22:28:49 <Mike_L> and I really didn't feel like writing a VM in python or perl =)
22:28:52 <futhin> why not implement it in forth?
22:28:56 * aaronl is back (gone 02:25:27)
22:29:20 <Mike_L> i don't have much experience with forth
22:29:22 <futhin> what does bytecode mean?  
22:29:27 <Mike_L> *sigh*
22:29:35 <futhin> i sort of know what bytecode is
22:29:39 <Mike_L> bytecode is the `machine code' of the Virtual Machine
22:29:40 <futhin> but not fully
22:29:50 <futhin> ok
22:30:11 <futhin> okay
22:30:17 <Mike_L> it is the sequence of instructions that the VM executes
22:30:31 <futhin> so you are implementing a bytecode for a VM, and it's going to be a nice "architecture" 
22:30:45 <futhin> you want to implement a stack for this virtual architecture?
22:30:53 <futhin> hmm
22:31:07 <Mike_L> i have compiler 50% ready... it just needs this bytecode spec ironed out so it can start emitting the bytecode
22:31:50 <Mike_L> yes there is a stack, immutable code segments (called definitions or defns), and a linear memory area
22:32:30 <futhin> you are making an all purpose VM so that everybody can code for it and it is "portable"..  but what's the point? how is it going to better than other ways?
22:32:55 <Mike_L> better than what?
22:33:28 <Mike_L> currently i see only Java and it is undesirable because of a number of characteristics
22:33:34 <futhin> like?
22:33:45 <I440r> forth
22:33:49 <I440r> and postscript
22:33:58 <I440r> im told both java and ps are forth based
22:34:08 <I440r> tho i dont believe it with java
22:34:10 <futhin> java isn't really forth based
22:34:11 <Mike_L> like the implementations are huge, the class libraries are bloated, it is non-free
22:34:25 <I440r> fat slow and evil
22:34:26 <I440r> hehe
22:34:36 <futhin> java has some "stack" stuff "borrowed" from forth.. but it's bs mostly i think
22:34:37 <Mike_L> i don't know a whole lot about the java vm, and I don't particularly want to
22:34:51 <Mike_L> I would research it though to make my own VM better
22:35:13 <futhin> i440r: can forth be considered a vm ?
22:35:27 <Mike_L> futhin: forth is a language I think
22:35:28 <futhin> like..  can i tell mike to go with a forth vm ? :)
22:35:43 <Mike_L> no no no =)
22:35:52 <Mike_L> i've read about forth and tried it a bit
22:36:11 <Mike_L> i can see people compiling C, Scheme, LISP, Forth, and python code to run on Sebae
22:36:14 <futhin> mike: its more than that.. it's like interactive, interpretive, compiled, extensible, and very portable  (more portable than c)
22:36:26 <Mike_L> but I can't see them compiling to Forth code to run in a Forth-only VM
22:36:34 <I440r> c isnt portable
22:36:47 <I440r> thats why all this "portable code" is full of
22:36:51 <I440r> #ifdefined x86
22:36:56 <Mike_L> futhin: i'm not making this VM for development.  I'm making it for real-life everyday use
22:36:57 <I440r> and #ifdefined powerpc
22:37:25 <Mike_L> I440r: yeah that sucks =(
22:37:25 <futhin> mike:  you can extend forth and EASILY create your own syntax, structures, etc.
22:37:40 <Mike_L> futhin: I have a lot of really good ideas that I think will work out very well
22:37:52 <Mike_L> futhin: but I don't want to limit myself to forth
22:38:01 <futhin> mike: they don't have to code in "Forth"  the forth vm can be extended so that the code looks a lot different.
22:38:38 <futhin> mike: what's the difference between  asm>c>vm bytecode>other languages  with  asm>forth>otherlanguages ??
22:38:45 <Mike_L> anyway if i decide to use the name `forth' then i'm going to run into trademark issues and people who have ever heard of forth will make all sorts of wrong assumptions about my VM
22:39:04 <futhin> forth isn't trademarked AFAIK
22:39:19 <futhin> and you don't have to mention it loudly :P
22:39:22 <Mike_L> ok
22:39:42 <Mike_L> I want to do this VM in a non-language specific way
22:39:42 <futhin> you can extend the forth "syntax" to look like your bytecode language
22:39:51 <futhin> i imagine
22:40:15 <Mike_L> anyway using forth would severely limit performance
22:40:17 <futhin> i440r: say something, i'm not a forth expert, please make sure i'm not bullshitting :)
22:40:42 <Mike_L> I need to work out this bytecode format
22:40:58 <futhin> not really..  forth is faster than c afaik
22:41:05 <I440r> erm ive no idea what byte code is
22:41:14 <Mike_L> this whole VM is designed with the different implementations in mind: interpreted, JIT compiled to native code, and hardware implementation
22:41:15 <I440r> it sounds like another method of semi compiling
22:41:24 <futhin> he defined it awhile ago .. bytecode is the "machine code" for the virtual machine
22:41:34 <I440r> so is forth
22:41:46 <I440r> in fact forth has been used as the micro code for many processors
22:41:55 <I440r> including the pci chipset
22:41:55 <Mike_L> bytecode : Virtual Machine :: machine code : real cpu and machine
22:42:10 <I440r> sounds like forth
22:42:27 <Mike_L> you mean it is used as the programming language to write BIOS code for motherboards
22:42:29 --- join: Fare (fare@ppp49-net1-idf2-bas1.isdnet.net) joined #forth
22:42:34 <I440r> fare!!!
22:42:38 --- mode: I440r set mode: +o Fare
22:42:43 <I440r> have u seen tcn at all dood ?
22:43:06 <Fare> not recently
22:43:16 <I440r> grrrr he is awol :P
22:43:25 <futhin> mike: from what i know about forth, it should be a lot easier to implement this VM in forth than c.  forth has a number of advantages that i think will make it fairly easy to implement the VM in it.
22:43:26 <I440r> and i need his latest 
22:43:54 <Mike_L> futhin: that's great, but that's not the issue at hand
22:44:19 <futhin> what's the issue?
22:44:50 <I440r> the issue is that my fscking mouse wont work in x and console on my laptop - fscking synaptics touchpad - grrr
22:45:07 <Mike_L> the issue at hand is the bytecode spec itself... do I simplify it by making the arithmetic operations work only with data on the stack, or do I make 50 versions of each opcode to accomodate acting on data which is on the stack, immediate data, pointer on stack, immediate pointer
22:45:51 <Mike_L> also there is the question of the product of the operations
22:45:53 <futhin> the fortmer
22:46:00 <futhin> go with the former..
22:46:18 <futhin> you don't have to do static typing..   
22:46:18 <futhin> uh
22:46:31 <Mike_L> that means that to modify data in memory it requires 3 instructions
22:46:45 <Mike_L> load from memory into stack, do operation, store result to memory
22:46:56 <futhin> there's some forths that implemented static typing, but i'm not sure how they do it..  but there might be some ideas from how they implement it that'll help you keep your bytecode simple...
22:47:17 <futhin> does anybody in this chan know about static typing w/ forth??
22:47:18 <Mike_L> lol is forth all you ever talk about? =)
22:47:27 <futhin> mike: heheh, sorry :P
22:47:38 <futhin> mike: if you gonna talk about stack, i gonna talk about forth ;)
22:48:05 <Mike_L> this stack is not going to work like forth's at all
22:49:13 <Mike_L> that's another thing I'm debating... how to address the stack
22:49:23 <futhin> mike: imagine forth as a stack-based language at the most basic level, with nothing added on..  no syntax, no structures, no static typing, etc.   so if you talk about stack it's practically the same as talking about forth :)
22:49:27 <Mike_L> C programs will define a `frame' for each code block
22:49:48 <Mike_L> futhin: there are many kinds of stacks
22:49:57 <futhin> how so?  i don't think so
22:50:23 <Mike_L> um yes
22:50:37 <I440r> a stack is a stack
22:50:39 <futhin> the only stack I know of is a LIFO stack.. last in, first out
22:50:46 <Mike_L> forth uses the stack as a constantly flowing thing... each instruction adds or removes from the stack
22:50:47 <I440r> spme processors have multiple stack pointers
22:50:56 <I440r> sp and bp are both stack pointers in x86
22:51:05 <I440r> forth usuallyt has 2 stacks
22:51:08 <Mike_L> some languages use the stack in chunks... a chunk for each code block
22:51:10 <I440r> parameter stack and return stack
22:51:14 <I440r> my 8051 forth has 3
22:51:44 <Mike_L> i want to implement one stack
22:51:47 <futhin> C usn't a stack-based language,  rather, it uses a stack in memory as a frame of reference for compiling..  not all c compilers do that tho.. they can, and some do
22:51:53 <Mike_L> and it won't be a pure stack either... each stack element will behave like a register
22:52:35 <Mike_L> instructions will be able to take data from any element on the stack and store the result in any element
22:53:53 <futhin> uh.. it doesn't even sound like a stack
22:53:57 <futhin> it sounds like an array
22:54:07 <Mike_L> this style of stack could facilitate forth code
22:54:15 <Mike_L> it is an array and stack combined
22:54:22 <Mike_L> and registers
22:54:58 <Mike_L> every code segment will have a specified stack effect, that is the number of stack elements it consumes and the number it produces
22:55:10 <futhin> uh.. if you treat the stack like an array, you are gonna probably end up with all sorts of nasty code to work with the array
22:55:15 <Mike_L> this will allow for efficient translation of the code to the host cpu's native format
22:55:51 <futhin> that "specific stack effect" really sounds like forth code ;)
22:56:13 <Mike_L> it won't be like an array in that the bytecode will not be able to work with computed indexes into the stack
22:56:34 <Mike_L> all bytecode will have immediate data representing the stack element to work with... relative to the top of the stack
22:56:36 <adu> oooo still talking? good job!
22:56:52 <futhin> heh :)
22:57:00 <Mike_L> adu: futhin won't shut up about forth
22:57:09 <futhin> heheh
22:57:18 <Mike_L> adu: is he a bot? =)
22:57:18 <adu> =) that was my plan
22:57:19 <futhin> i discovered forth very recently :P
22:57:25 <futhin> but
22:57:28 <futhin> stack == forth
22:57:36 <Mike_L> futhin: ah that's why
22:57:38 <futhin> :)
22:57:40 <Mike_L> futhin: no
22:57:41 <adu> most people in this channel are estatic about the language, thats why there are so few
22:58:03 <Mike_L> I discovered python a few months ago and thought it was the best thing since sliced bread... till I found its flaws and shortcomings
22:58:12 <futhin> adu: naw.. that can't be true! :P
22:58:27 <Mike_L> adu: futhin also has this really bad concept of what a stack is
22:58:37 <adu> no i mean thats why there aren't that many ppl in the channel
22:58:44 <adu> lol
22:58:47 <Mike_L> inflexible concept I should say
22:59:05 <futhin> heheh
22:59:07 <adu> well in forth there are actually 3 stacks
22:59:10 <futhin> (to adu)
22:59:14 <futhin> um
22:59:15 <adu> one user stack
22:59:21 <adu> one return stack
22:59:27 <adu> and one floating-point stack
22:59:34 <futhin> um
22:59:34 <Mike_L> hmm
22:59:39 <Mike_L> sounds complex
22:59:50 <adu> the use stack is the most used
22:59:52 <futhin> that floating-point stack is only implemented on a few forths
23:00:11 <adu> futhin: really? does it use the main stack on others?
23:00:12 <futhin> 98% of forth implementations only have 2 stacks
23:00:22 <Mike_L> can't the user stack handle floating point numbers?
23:00:41 <adu> Mike_L: probly if it dosen't have a stack specifically for fp
23:01:21 <adu> the main stack also accepts structures and complex data types
23:01:53 <Mike_L> hmm
23:02:05 <Mike_L> ya i dunno about that
23:02:17 <Mike_L> i think for this VM I want to just use 32bit stack elements
23:02:21 <adu> a good place to start is forth.org
23:02:31 <futhin> you don't do floating point in most forths.. you convert the floating point to integers (when inputting) and work with that, then convert back to floating point (when you display it..)
23:02:33 <adu> its the FIG ws i think
23:02:45 <Mike_L> i'm not going to learn another programming language right now
23:03:07 <Mike_L> futhin: sounds pretty limited
23:03:15 <futhin> mike: no
23:03:19 <futhin> it's not
23:03:35 <futhin> you don't need to do the arithmetic in floating point because that'll only consume cycles
23:03:38 <futhin> but
23:04:17 <Mike_L> futhin: I can't imagine any time one would use floating point data types if one didn't need the precision that integers lack
23:04:37 <futhin> the newer implementations are including floating point stuff.. so it doesn't really matter.. you can implement floating point stuff with 2 stacks anyways,  but a floating-point stack probably helps simplify some of the implementations that do want floating point
23:05:30 <Mike_L> multiple stacks will always simplify the code written for them... and complexify the code used to implement them
23:06:09 <adu> Mike_L: not necesarily
23:06:26 <adu> it can be visaversa
23:06:32 <Mike_L> i want to keep this VM simple... even if that means added complexity for the compilers
23:06:52 <adu> if you wanted to implement a single stack ontop of a multiple stack arch, its very hard
23:06:55 <adu> i've tried
23:07:07 <adu> i wrote a small pythonVM interpreter in forth once
23:07:19 <futhin> mike: you work with the integers.. suppose you give a number like 1.2345  it'll just become 12345 on the stack and you work with that..  and keep track where the decimal place should be at yourself.  or you could put two numbers on the stack. 12345 and -4 and avoid keeping track of the decimal place yourself
23:08:16 <Mike_L> futhin: that can be hard with some arithmetic ops
23:08:28 <futhin> how so?
23:08:49 <Mike_L> multiply 12345 by 100
23:08:55 <adu> like 1.2345 + 198760000.3
23:09:17 <Mike_L> your product is 1,234,500  if you keep your -4 then you have 123.4500
23:09:24 <Mike_L> hmm which is right
23:09:41 <Mike_L> ya you have to put your nums on the same base too before you add/subtract
23:09:57 <adu> but fp have a wider range, with less acuracy
23:10:04 <futhin> um?  you could modify the -4 itself to times it by 100
23:10:23 <adu> its all algebra
23:11:45 <futhin> floating point math ultimately ends up using integers..  the older "traditional" forths didn't have floating-point support so basically  you just kept track of it yourself, which was fine because the forth coders for the traditional forths were usually concerned about speed
23:12:07 <adu> uh not really
23:12:33 <Fare> nowadays, forth isn't considered speedy anymore
23:12:37 <adu> but the fpu on intels are faster than int units
23:12:40 <Fare> if it ever was
23:12:47 <Mike_L> modern cpus have FPUs
23:12:53 <futhin> adu: that's partially true
23:12:57 <adu> Fare: why is that
23:13:26 <futhin> fare: uh... afaik forth is faster than c
23:13:37 <Mike_L> heh
23:13:45 <Mike_L> no
23:14:10 <Mike_L> that is your new-language-syndrome talking
23:14:40 <adu> Mike_L: in some ways it is
23:14:44 <Mike_L> C is a lower level language than forth
23:14:48 <futhin> cough cough
23:14:53 <futhin> forth is lowerlevel
23:15:08 <I440r> mike erm c is NOT even slightly low level
23:15:11 <Mike_L> oh I'm certain that there are some things that forth does much more efficiently than C
23:15:18 <adu> since forthers usually reinvent the whell often, there is much less abstraction, and as a result, fewer layers that slow things down
23:15:27 <I440r> forht is not just a step above asm and higher level than ada all in one
23:15:38 <I440r> Mike_L: yea - everything P
23:15:47 <futhin> heh
23:15:54 <Mike_L> ok I don't want to argue about languages
23:16:11 <futhin> we could explain it "thoroughly" :P
23:16:18 <adu> I440r: you beleive that?
23:16:20 <Mike_L> i'm not overly concerned with performance of the apps in this VM
23:16:52 <adu> Mike_L: forth isn't very high level, its very ll, but portable
23:17:01 <adu> Mike_L: do you know about the forth dictionary?
23:17:05 <futhin> it's portable because it's ll
23:17:16 <futhin> and very small/easy to implement 
23:17:55 <Mike_L> *sigh*
23:17:58 <I440r> adu wrong
23:18:01 --- part: Mike_L left #forth
23:18:05 <I440r> forth is both very high and very low
23:18:21 <adu> Mike_L: the 'dictionary' is the place where all the 'words' are placed, and it itself is kinda like a stack, when a word is defined, it gets added to the top of the dictionary, and each dictionary entry is usually a name, then a pointer to code, thats it
23:18:27 <Fare> f/li
23:19:08 <futhin> this is funny
23:19:45 <adu> and if you define a word thats already defined it just places it ontop like any other word, but words defined before it, still use the previous definition, bcuz it searches backwords in the dictionary
23:19:58 <futhin> i just ended up talking about forth the whole time to mike_l :P
23:25:10 --- join: dired (aaronl@vitelus.com) joined #forth
23:31:29 <futhin> fare
23:31:56 <futhin> how much do you know about forth?
23:32:30 <Fare> I FORTH ABOUT BIT A KNOW
23:32:35 <futhin> heheheh
23:32:40 <futhin> what does f/li mean?
23:32:51 <Fare> means that I can't type properly
23:33:01 <futhin> huh
23:33:16 <futhin> what was it supposed to mean? or is it some acronym?
23:33:25 <adu> lol
23:33:34 <adu> 15432
23:33:40 <Fare> Float division with low imprecision
23:33:58 <Fare> or just /li (go to channel #lisp) with a "f" being prepended
23:34:55 <adu> wtf?
23:34:55 <futhin> um, why did you type it? or was it a mis-chan?
23:35:06 <Fare> fnord
23:35:19 <adu> futhin: you asked
23:35:30 <futhin> huh? :)
23:35:46 <futhin> oh
23:47:11 --- quit: aaronl (The name's X. Bitch X.)
23:47:29 --- join: aaronl (aaronl@vitelus.com) joined #forth
23:55:20 --- quit: Fare (Connection reset by pear)
23:58:31 <nate37> hmm
23:58:38 <futhin> hmm
23:59:18 <nate37> zup?
23:59:59 --- log: ended forth/01.06.25