00:00:00 --- log: started forth/01.06.12
00:46:31 --- nick: adugone -> adu
01:37:45 <adu> hi
01:37:58 --- part: adu left #forth
01:37:58 --- join: adu (andrew@adsl-63-201-89-121.dsl.snfc21.pacbell.net) joined #forth
03:09:22 * aaronl is away: Yawn
03:41:33 <adu> hi
03:41:41 <adu> bye
03:41:45 --- quit: adu (IRCStep)
06:39:55 --- quit: aaronl (farmer.openprojects.net benford.openprojects.net)
07:23:23 --- join: grurp (grurp@pc19987.batc.tec.ut.us) joined #forth
08:05:10 --- join: cleverdra (jfondren@1Cust89.tnt4.florence.sc.da.uu.net) joined #forth
08:11:32 --- quit: grurp (Bye all)
08:24:02 --- join: cleverser (jfondren@1Cust251.tnt2.florence.sc.da.uu.net) joined #forth
08:25:15 --- quit: cleverdra (Ping timeout for cleverdra[1Cust89.tnt4.florence.sc.da.uu.net])
08:25:21 --- nick: cleverser -> cleverdra
08:38:28 <Trey> Good morning, cleverdra.
08:41:21 <cleverdra> g'morning, Trey.
08:42:39 <Trey> I went to add PCI bus support to Enth, only to find that Sean left out the 16 and 32 bit I/O primatives.
08:44:03 <cleverdra> Yes, he doesn't have a general assembly-level access to the machine built in yet, which is why I haven't tried to get an IDE disk or NIC working with Enth.  SASM should help that, which is actually in his version of Enth but not yet released.
08:44:13 <cleverdra> This is why some people have daily snapshots, ne? =)
08:44:17 <Trey> :-)
08:44:43 <Trey> He offered me a version of SASM, and I accepted but I haven't received it yet.
08:46:07 <cleverdra> Cool.
08:47:12 <Trey> Perhap he'll just roll out what he's got.  It seems to take him about 24hrs to respond to Emails, so I won't know until later today.
08:50:31 <Trey> cleverdra: have you signed up for his Enth mailing list?
08:51:00 <cleverdra> Yes, though it doesn't get much traffic and I don't check it much.
08:51:08 <cleverdra> I don't check it frequently.
08:51:23 <Trey> You have to "check it"?
08:51:43 <Trey> (you funnel it to a separate in-box, I suppose?)
08:53:00 <cleverdra> My personal computer isn't on the 'net, so I have to actually make an effort to check my email, Trey.
08:53:48 <Trey> Ah.
08:53:54 * Trey feels spoiled.
08:55:35 <cleverdra> Yes, I used to feel spoiled like that.
08:55:57 <Trey> I wanted to use the Enth mailing list to come to some kind of concensus about an appropriate word list for PCI access.
08:56:38 <cleverdra> And one day, I shall feel so spoiled again!  You know, I've four computers right now, and it's amazing how much you miss simply by not being connected to a global information passing network upon which all your peers are connected and from which your lifeblood comes.
08:56:47 <cleverdra> So why don't you, Trey?
08:57:15 <Trey> I will.  I didn't think of it until just a few minutes ago, though :-)
08:57:25 <cleverdra> Cool, OK.
08:58:45 <Trey> Have you come up against the need for a locking API yet?
09:00:08 <Trey> We'll need one Real Soon Now if multiple folks start writing drivers.
09:02:00 <cleverdra> No I haven't, Trey -- but understand that what little OS and hardware knowledge I've gained, I've mostly gained through Forth and Enth specifically.  Why would you need a locking API?
09:03:25 * Trey comes from an SMP background, and thinks of this sort of thing on a regular basis.
09:03:58 <cleverdra> Well, cool.
09:03:59 <Trey> Right now, Enth uses cooperative scheduling.
09:04:18 <cleverdra> Right.  futhin and I have had discussions on that.
09:04:21 <Trey> (tasks switch only when the programmer says)
09:05:49 <Trey> So, in theory, we don't _really_ need locking.  But it can simplify one's life if you have it.
09:06:11 <cleverdra> What do you mean by locking?
09:06:18 <Trey> Also, if he switches to preemptive scheduling, we will have problems.
09:06:50 <Trey> You don't want two different threads trying to read/write the PCI bus configuration registers at the same time.
09:07:02 <Trey> (for example)
09:07:24 <cleverdra> Ah, right.  OS's with filesystems deal with this in the FS.
09:07:40 <Trey> They deal with it in a number of places.
09:07:52 <cleverdra> Yes, I'm just specifying one that I am strictly aware of.
09:07:56 <Trey> Right.
09:08:42 <cleverdra> In a preemptivly schedules system, something like 'bload bload' would be bad if the editor's FDC stuff works in its own thread, ne?
09:09:16 <cleverdra> ...without some kind of locking being done.
09:09:34 <Trey> If you mean 'bload' in two different threads, both trying to read from the FDC, yes.
09:09:46 <Trey> The FDC needs to serialize access to the device.
09:10:06 <Trey> Rather, something needs to serialize access to the device.
09:10:41 <cleverdra> It'd be simple for FDC to do that with a FIFO stack of sorts.  Are you thinking of a general solution?
09:12:00 <Trey> Probably something more general.  It all depends on where Enth ends up going.
09:13:33 <cleverdra> Yeah.  login passwords and memory access controls and runtime security checks.  Blah.
09:13:41 <Trey> We probably ought to just work on a PCI wordset first, and see where that leads us.
09:14:44 <Trey> I don't know if we need all of that stuff for Enth.  But a flexible pre-emptive multi-tasker can simplify a programmer's life in some cases.
09:14:45 <cleverdra> That sounds like a plan.
09:15:14 <Trey> I need to bail for an hour or so.  Later, all.
09:15:20 * Trey is away: errands, etc. [11:15]
09:15:21 <cleverdra> bye Trey.
11:15:49 <Trey> This forum sure does get a lot of traffic.
11:26:46 <cleverdra> The forum off of pringle.sphosting.com?  Most of that traffic is mine, with replies
11:40:55 <Trey> I meant this one.  Nary a word between when I left and returned.
11:53:05 <cleverdra> Oh.  No it doesn't.  Somebody mentioned it on clf and that's what brought be here, but others don't seem to be in the coming.
12:03:28 --- join: aaronl (aaronl@vitelus.com) joined #forth
12:08:17 --- quit: aaronl (zelazny.openprojects.net farmer.openprojects.net)
12:08:20 --- topic: set to 'http://isforth.sourceforge.net -- http://cvs.sourceforge.net/cgi-bin/cvsweb.cgi/isforth/?cvsroot=isforth' by ChanServ
12:08:28 --- mode: ChanServ set mode: -o Trey
12:08:28 --- mode: ChanServ set mode: -o lar1
12:08:28 --- mode: ChanServ set mode: -o clog
12:08:34 --- join: aaronl (aaronl@vitelus.com) joined #forth
12:09:07 --- quit: aaronl (simak.openprojects.net farmer.openprojects.net)
12:10:46 --- join: aaronl (aaronl@vitelus.com) joined #forth
12:20:44 --- quit: cleverdra (Ping timeout for cleverdra[1Cust251.tnt2.florence.sc.da.uu.net])
12:21:43 --- join: cleverdra (jfondren@1Cust243.tnt4.florence.sc.da.uu.net) joined #forth
12:40:54 * aaronl is away: breakfast
12:56:38 --- join: grurp (grurp@pc19976.batc.tec.ut.us) joined #forth
13:09:15 --- quit: cleverdra (Ping timeout for cleverdra[1Cust243.tnt4.florence.sc.da.uu.net])
13:12:43 --- join: cleverdra (jfondren@1Cust149.tnt3.florence.sc.da.uu.net) joined #forth
13:29:07 --- quit: grurp (Bye all)
13:30:18 * aaronl is back (gone 00:49:24)
13:56:26 --- join: dired (infobot@adsl-207-214-111-232.dsl.snfc21.pacbell.net) joined #forth
13:56:53 --- quit: dired (regrouping; bbiab)
14:17:09 --- join: dired (aaronl@vitelus.com) joined #forth
14:17:21 --- quit: dired (Read error to dired[vitelus.com]: EOF from client)
14:18:23 --- join: dired (aaronl@vitelus.com) joined #forth
14:19:14 --- quit: dired (regrouping; bbiab)
14:19:35 --- join: dired (aaronl@vitelus.com) joined #forth
14:19:39 --- quit: dired (regrouping; bbiab)
14:20:09 --- join: dired (aaronl@vitelus.com) joined #forth
14:20:12 <aaronl> dired: emacs?
14:20:12 <dired> it has been said that emacs is /ee'maks/ [from Editing MACroS] n. The ne plus ultra of hacker editors, a program editor with an entire LISP system inside it. It was originally written by Richard Stallman in TECO under {ITS} at the MIT AI lab, but the most widely used versions now run under UNIX. It includes facilities to run compilation subprocesses and send and receive mail; many hackers spend up to 80% of their {tube time} inside it. Some versions running under 
14:25:11 --- quit: dired (regrouping; bbiab)
14:25:16 --- join: dired (aaronl@vitelus.com) joined #forth
14:25:26 * cleverdra looks up.
14:25:29 <cleverdra> That's interesting.
14:25:36 <cleverdra> dired: forth?
14:25:36 <dired> no idea, cleverdra
14:25:50 <cleverdra> dired, forth is at www.forth.org
14:25:50 <dired> OK, cleverdra.
14:26:00 <cleverdra> dired, forth is also at www.ultratechnology.com
14:26:01 <dired> okay, cleverdra.
14:26:08 <cleverdra> dired, forth is also at www.forth.org.uk
14:26:09 <dired> okay, cleverdra.
14:26:16 <cleverdra> dired, status?
14:26:16 <dired> Since Tue Jun 12 14:25:15 2001, there have been 3 modifications and 1 questions.  I have been awake for 1 minute, 1 second this session, and currently reference 5091 factoids. Addressing is in optional mode.
14:26:29 <cleverdra> 5091?  From what?
14:26:30 --- join: futhin (thin@h24-67-113-99.cg.shawcable.net) joined #forth
14:26:37 <cleverdra> hello futhin
14:26:37 <aaronl> cleverdra: jargon, mostly
14:26:50 <aaronl> forth?
14:26:50 <dired> forth is probably at http://www.forth.org or at http://www.ultratechnology.com or at http://www.forth.org.uk
14:26:53 <cleverdra> dired, cleverdra is cleverjulian@hotmail.com -- but he doesn't check it much
14:27:04 <cleverdra> aaron - cool.
14:27:11 <futhin> hello all
14:27:13 <aaronl> dired: FORTH?
14:27:14 <dired> FORTH is at http://www.forth.org or at http://www.ultratechnology.com or at http://www.forth.org.uk
14:27:15 <aaronl> hmm
14:27:22 <aaronl> dired: cobol?
14:27:22 <dired> cobol is /koh'bol/ [COmmon Business-Oriented Language] n. (Synonymous with evil.) A weak, verbose, and flabby language used by {card walloper}s to do boring mindless things on dinosaur mainframes. Hackers believe all COBOL programmers are suits or {code grinder}s, and no self-respecting hacker will ever admit to having learned the language. Its very name is seldom uttered without ritual expressions of disgust or horror. See also {fear and loathing}, {software rot}
14:27:23 <cleverdra> dired, cleverdra =~ s/much/frequently/
14:27:23 <dired> I didn't have anything matching 'cleverdra', cleverdra
14:27:30 <cleverdra> erm.
14:27:35 <cleverdra> Hah, that's cool.
14:27:49 <futhin> you know what i find weird..  forth is supposed to be extremely portable, but all these losers keep writing forth in C?!?! 
14:27:59 <aaronl> lol
14:28:11 <cleverdra> lots of assembly Forth going on too, futhin.
14:28:17 <aaronl> cleverdra: use s///
14:28:37 <cleverdra> brb.  Very cool of you to put an infobot in here, aaron =)
14:28:57 <aaronl> Thx :)
14:29:02 <futhin> cleverdra: yeah, but i'm not sure there are any assembly Forth for windows and linux??
14:31:32 <futhin> were older computers stack-based rather than register based?  or were they register based from the get-go??
14:31:38 --- quit: cleverdra (Ping timeout for cleverdra[1Cust149.tnt3.florence.sc.da.uu.net])
15:15:41 --- quit: dired (Read error to dired[vitelus.com]: EOF from client)
15:16:04 --- join: dired (aaronl@vitelus.com) joined #forth
15:45:32 --- quit: futhin (sleep)
17:19:34 * Trey is away: life, etc. [19:19]
18:21:32 --- join: Chaynez (chaynez@cc461258-b.srst1.fl.home.com) joined #forth
18:21:38 <Chaynez> anyone' seen Mark/I440r?
18:28:14 --- nick: aaronl -> aaronl_
19:27:54 --- join: I440r (mark4@purplecoder.com) joined #forth
19:31:35 <Chaynez> hey Mark
19:31:37 <Chaynez> u get my ICQ?
19:31:49 <I440r> nope let me go look :)
19:31:58 <Chaynez> s'ok
19:32:05 <Chaynez> just point chaynez.purplecoder.com to 24.250.80.13
19:32:24 <aaronl_> heh
19:32:30 <I440r> im takin my sis round her friends house man ill brb
19:32:33 <I440r> then ill do it :)
19:32:36 <aaronl_> Chaynez: do you use forth?
19:32:53 <Chaynez> naah
19:32:58 <Chaynez> I finally got a copy of Kylix though :)
19:32:59 <Chaynez> heh
19:33:01 <aaronl_> neither do i :/
19:33:03 <Chaynez> thanks Mark
19:33:16 <aaronl_> I want to learn forth but I440r has to finish isforth first :)
19:33:25 <Chaynez> Hey Mark.. you know where I can buy a cheap (like $20) PCI Dual Parallel Port card?
19:33:34 <Chaynez> I ordered 4 new cams, but my machine's only got one parallel port card :)
19:33:34 <Chaynez> heh
19:34:12 --- join: cleverdra (jfondren@1Cust205.tnt6.florence.sc.da.uu.net) joined #forth
19:35:00 <cleverdra> dired, status
19:35:01 <dired> Since Tue Jun 12 15:15:31 2001, there have been 37 modifications and 4 questions.  I have been awake for 4 hours, 19 minutes, 7 seconds this session, and currently reference 5141 factoids. Addressing is in optional mode.
19:35:14 <cleverdra> dired, seen futhin
19:35:14 <dired> futhin was last seen on #forth 5 hours, 3 minutes and 19 seconds ago, saying: were older computers stack-based rather than register based?  or were they register based from the get-go?? [Tue Jun 12 14:31:32 2001]
19:37:40 --- join: cleverser (jfondren@1Cust186.tnt4.florence.sc.da.uu.net) joined #forth
19:39:30 * cleverser pokes cleverdra.  Die.
19:39:33 --- quit: cleverdra (Ping timeout for cleverdra[1Cust205.tnt6.florence.sc.da.uu.net])
19:39:40 --- nick: cleverser -> cleverdra
19:41:32 <cleverdra> mmm dired, Enth is at pringle.sphosting.com
19:41:36 <cleverdra> dired, Enth is at pringle.sphosting.com
19:41:36 <dired> OK, cleverdra.
19:42:04 <cleverdra> dired, forths?
19:42:04 <dired> cleverdra: i haven't a clue
19:42:19 * cleverdra thinks of a good factoid.
19:44:48 <Trey> dired, time?
19:44:48 <dired> trey: no idea
19:44:55 <Trey> Hmm.
19:45:13 <Trey> dired, seen cleverdra?
19:45:14 <dired> cleverdra was last seen on #forth 2 minutes and 54 seconds ago, saying: thinks of a good factoid. [Tue Jun 12 19:41:57 2001]
19:45:29 <Chaynez> dired, Time is time to get a watch!
19:45:29 <dired> OK, Chaynez.
19:45:29 <Chaynez> :P
19:45:34 <Chaynez> dired, time?
19:45:34 <dired> time is time to get a watch!
19:45:37 <cleverdra> hee.
19:45:40 <Chaynez> heh
19:45:51 <Trey> dired, date?
19:45:51 <dired> trey: i haven't a clue
19:45:54 <Trey> feh
19:46:02 <cleverdra> no dired, time is <reply>heh, $who, looks like \you need to get \yourself a watch\!
19:46:02 <dired> okay, cleverdra.
19:46:13 <cleverdra> hey ppl, what's the time?
19:46:13 <Trey> dired, time?
19:46:13 <dired> heh, Trey, looks like you need to get yourself a watch!
19:46:18 <cleverdra> hm.
19:46:26 <Chaynez> dired, date is something I haven't had for a while, not since the last one went horribly wrong with the shellfish incident :`(
19:46:26 <dired> OK, Chaynez.
19:46:31 <Chaynez> dired, date
19:46:31 <dired> date is probably something I haven't had for a while, not since the last one went horribly wrong with the shellfish incident :`(
19:46:37 <Chaynez> lol
19:46:39 <cleverdra> shellfish incident?
19:47:09 <Chaynez> a friend of mine.. he was at a restaurant with some friends..
19:47:18 <Chaynez> a few tables away was this gorgeous girl showing off a lot of cleavage...
19:47:30 <Chaynez> so they were havin' bets on who could get some crab meat down hear cleavage the first :P
19:48:01 <Chaynez> so they were throwin' bits of crab at her for about 15 minutes before the management ejected them
19:48:02 <Chaynez> heh
19:48:51 --- join: cleverser (jfondren@1Cust202.tnt3.florence.sc.da.uu.net) joined #forth
19:49:15 <cleverser> shellfish incident, Chaynez?
19:49:23 <Chaynez> [03:47:54] <Chaynez> a friend of mine.. he was at a restaurant with some friends..
19:49:23 <Chaynez> [03:48:03] <Chaynez> a few tables away was this gorgeous girl showing off a lot of cleavage...
19:49:23 <Chaynez> [03:48:15] <Chaynez> so they were havin' bets on who could get some crab meat down hear cleavage the first :P
19:49:23 <Chaynez> [03:48:46] <Chaynez> so they were throwin' bits of crab at her for about 15 minutes before the management ejected them
19:49:23 <Chaynez> [03:48:47] <Chaynez> heh
19:49:25 <Chaynez> :)
19:49:39 --- quit: cleverdra (Ping timeout for cleverdra[1Cust186.tnt4.florence.sc.da.uu.net])
19:50:01 --- nick: cleverser -> cleverdra
19:50:21 <cleverdra> purl, shellfish incident is (don't ask)
19:50:23 <cleverdra> err.
19:50:26 <cleverdra> dired, shellfish incident is (don't ask)
19:50:26 <dired> OK, cleverdra.
19:50:29 <Chaynez> lol
19:50:44 <Trey> cleverdra: I still haven't heard back from Sean.
19:51:20 <Chaynez> dired, time
19:51:20 <dired> heh, Chaynez, looks like you need to get yourself a watch!
19:51:27 <Chaynez> lol
19:52:02 <Chaynez> hey is this the same bot/script they got on #linuxhelp on Undernet?
19:52:21 <Trey> dired, help?
19:52:34 <cleverdra> trey - hm.  I think he gets on the net/(checks things) at 12PM EST -- at least, that's one time I definitely know that he once checked the forum.  Hopefully he's doing some important work with Enth or Flux, and is too tired to do anything else.  Sigh.
19:52:50 <cleverdra> dired, tell chaynez about infobot
19:52:59 <Chaynez> nice one, tnx cleverdra
19:59:51 <I440r> ok im back
20:00:00 <I440r> gimme that ip again :)
20:00:48 <aaronl_> <Chaynez> just point chaynez.purplecoder.com to 24.250.80.13
20:00:50 <Chaynez> 24.250.80.13
20:02:16 <I440r> whats the date today ?
20:02:17 --- quit: cleverdra (Leaving)
20:02:24 <Chaynez> 12th
20:02:29 <Chaynez> 13th in another hour or so though
20:03:46 <aaronl_> not here :)
20:05:14 <I440r> do this
20:05:25 <I440r> nslookup chaynez.purplecoder.com purplecoder.com
20:05:40 <I440r> that says use purplecoders name server to look up that address
20:06:25 <I440r> i dont know if im supposed to do this or not but i dont know why not :)
20:06:31 <I440r> let me know if the IP changes
20:07:08 <I440r> by doing the above nslookup you force propogation of the name too
20:07:24 <I440r> have all ur friends do that too and it will propogate nicely :)
20:07:55 * aaronl_ does it
20:08:12 <aaronl_> BTW, my name server has it now :)
20:08:21 <I440r> :)
20:08:57 <I440r> see if i get any nasty letters off my isp hhe
20:09:00 <Chaynez> aight nice one
20:09:01 <I440r> my isp is secondary
20:09:08 <Chaynez> all my friends ain't usin' Linux :)
20:09:09 <Chaynez> heh
20:09:13 <aaronl_> I440r: ?
20:09:17 <Chaynez> but it's cool
20:09:21 <aaronl_> oh hehe
20:09:22 <I440r> they normally charge for doing name service but because i asked if i can do my own i dont get charged :)))))
20:09:27 <aaronl_> Heheh
20:09:34 <Chaynez> :)
20:09:38 <aaronl_> my ISP runs my secondary nameservice too
20:09:40 <aaronl_> (i run primary)
20:09:42 <I440r> they can nslookup from windows too if they have it
20:09:50 <I440r> im promary for my own domain
20:09:50 * Trey has always run his own DNS.
20:11:13 <I440r> it solves alot of problems ill tell u :)
20:12:35 <aaronl_> yeah
20:13:47 <aaronl_> ugh c is yucky
20:13:54 <I440r> c sux
20:13:56 <I440r> nuff sed
20:14:05 <aaronl_> i wish it was lower level
20:14:12 <aaronl_> so i didn't have to guess about things like alignment
20:15:15 <I440r> reverse is a bitch tho
20:15:20 <I440r> do u do ur own reverse ?
20:15:26 <aaronl_> no, my isp does
20:15:28 <aaronl_> (fortunately)
20:15:31 <I440r> yea
20:15:42 <I440r> do they subnet their ip range to sub class c networks ???
20:15:47 <aaronl_> yes
20:15:48 <I440r> that REALY makes it a bitch
20:15:51 <aaronl_> it does :/
20:15:58 <aaronl_> but they're great to work with
20:16:06 <I440r> but it IS POSSIBLE if they do it smart...
20:16:16 <I440r> i couldnt ask for a better isp
20:16:17 <aaronl_> ya
20:16:18 <aaronl_> me too
20:16:34 <aaronl_> my ISP hosts dns and lets me run servers
20:16:37 <I440r> do you use ipchains or iptables ?
20:16:39 <aaronl_> and they're very reliable
20:16:40 <aaronl_> so no complaints
20:16:41 <aaronl_> iptables
20:16:42 <I440r> i realy need a firewall
20:17:03 <I440r> every time i try implement one i get a very nice "blocks everything" firewall
20:17:04 <I440r> grrrr
20:17:12 <aaronl_> So I HATE dealing with incompetent c coders
20:17:13 <I440r> do yo uuse ipmasqurading ?
20:17:20 <I440r> thats 99% of them
20:17:20 <aaronl_> LOOK at these functions
20:17:21 <aaronl_> yes i do
20:17:27 <aaronl_> static int
20:17:27 <aaronl_> test_bit (char *buf, int bitno)
20:17:27 <aaronl_> {
20:17:27 <aaronl_>         char c, m;
20:17:27 <aaronl_>         c = buf[bitno / 8];
20:17:28 <aaronl_>         bitno = bitno % 8;
20:17:30 <aaronl_>         bitno = 7 - bitno;
20:17:32 <aaronl_>         if (!bitno)
20:17:34 <aaronl_>                 m = 1;
20:17:36 <aaronl_>         else {
20:17:38 <aaronl_>                 m = 2;
20:17:40 <aaronl_>                 while (--bitno)
20:17:42 <aaronl_>                         m *= 2;
20:17:44 <aaronl_>         }
20:17:46 <aaronl_>         return c & m;
20:17:48 <aaronl_> }
20:17:49 <I440r> ugh
20:17:50 <aaronl_> I replaced that SHIT with:
20:17:54 <aaronl_> #define test_bit(buf,bitno) ((buf>>bitno)&1)
20:18:00 <aaronl_> apparently the person who wrote it had never heard of shifts :P
20:18:07 <aaronl_> (LOOK at that while loop!!)
20:18:11 <I440r> hehehe
20:18:18 <aaronl_> he also wrote an inverse_bit function
20:18:30 <aaronl_> that set a bit n to 0 if it was 1 and 1 if it was 0...
20:18:33 <aaronl_> he did not use XOR
20:18:37 <I440r> hehehehe
20:18:41 <aaronl_> instead, he anded it with a bitmask
20:18:42 <I440r> so   
20:18:45 <I440r>  a ^= b
20:18:48 <I440r>  b ^= a
20:18:48 <aaronl_> i know!!!
20:18:50 <I440r> a ^= b
20:18:57 <I440r> would confuse him :)
20:18:59 <aaronl_> yeah
20:19:03 <aaronl_> i dont think he knows what xor is
20:19:05 <I440r> erm i missed the ;
20:19:07 <I440r> bleh
20:19:20 <aaronl_> Oh yeah, the best thing was, he didn't really NEED the inverse_bit function
20:19:26 <aaronl_> since he was writing code like this:
20:19:26 <I440r> look on cdrom.comn for some realy nifty functions
20:19:40 <I440r> like to get the # set bits in a byte
20:20:05 <I440r> or a routine to bit reverse a buyte...
20:20:07 <aaronl_> if (should_be_on && test_bit(buf, n) != 1)
20:20:12 <I440r> obfuscated but nifty :)
20:20:18 <aaronl_>        inverse_bit(buf, n);
20:20:21 <I440r> dont need the =1 part
20:20:34 <I440r> if (blah && bit)
20:20:36 <I440r> \{
20:20:41 <aaronl_> else if (!should_be_on && test_bit(buf,n) == 1)
20:20:42 <I440r>   bit_is_set();
20:20:42 <I440r> }
20:20:47 <aaronl_>     inverse_bit(buf, n);
20:20:50 <aaronl_> crazy freak
20:21:06 <I440r> hhehe
20:21:18 <aaronl_> basically, he tested the bit, then he tested it again, and tepending on that result he either anded it with the complement or or'd it with the bitmask :P
20:21:21 <aaronl_> well, enough of that
20:21:31 <aaronl_> heh i've seen some pretty fucked up coders in my day
20:21:35 <aaronl_> and i blame most of it on C
20:21:38 <I440r> u get used to it :P
20:21:41 <I440r> yup
20:21:46 <aaronl_> like, this one person didn't know how a linked list works
20:21:59 <aaronl_> so he made a transversal function in O(N^2)!!!!!
20:22:00 <I440r> erm.. thats basic shit
20:22:39 <I440r> 6 new spam messages
20:22:40 <I440r> grrr
20:22:42 <aaronl_> for (i=0; i != length(LinkedList); i++) { Node *p = get_nth_node(LinkedList, i); operate(p); ... }
20:22:45 <aaronl_> heh very stupid
20:23:10 <I440r> :)
20:23:23 <aaronl_> does forth have binary operations?
20:23:42 <I440r> have you ever seen a 10 page function with a single switch statement where each case was a 10 case switch statement where each case was a 10 case switch statemetn ????
20:23:43 <I440r> i have
20:23:45 <I440r> argh!
20:23:50 <I440r> yes
20:23:51 <I440r> and
20:23:52 <I440r> or
20:23:53 <I440r> xor
20:23:53 <I440r> not
20:23:56 <aaronl_> cool
20:24:10 <aaronl_> anyway i disagree with you on "C style"
20:24:14 <aaronl_> like, the post to linux-kernel
20:24:17 <I440r> and 2* isalways done in asm as a SHIFT in forth
20:24:19 <Trey> Forth has n-ary operators.
20:24:32 <I440r> linus didnt like that post :)
20:24:35 <I440r> hehe
20:24:41 <aaronl_> i think that embedding the switch statment is GOOD because otherwise you have a new function and you have to branch for no reason!!!
20:25:04 <aaronl_> otherwise you would have to call it, even though there would only be one source of the call
20:25:11 <I440r> i NEVER use the c switch stattement
20:25:13 <I440r> AT ALL
20:25:18 <aaronl_> do you ever use c? ;-)
20:25:36 <I440r> there are ways of doing it that dont cost any more time or space but save alot of vertical source space
20:25:40 <I440r> yes i use c
20:25:50 <aaronl_> like what ways of doing in?
20:25:57 <aaronl_> can you show me what it would be like in asm?
20:26:18 <aaronl_> heh, i guess you're right
20:26:27 <aaronl_> because for every case you have to branch anyway...
20:26:28 <I440r> ok lets say you have a switch statement for different input keys
20:26:30 <aaronl_> just C hides that
20:26:32 <aaronl_> ok
20:26:55 <I440r> and each for each key you run a different function
20:27:04 <I440r> i would do something like the following
20:27:17 <I440r> let me think this out a second.... been awhile hehe
20:27:20 <aaronl_> hehe
20:27:38 <I440r> typdef struct
20:27:38 <I440r> {
20:27:45 <I440r>   char key;
20:28:06 <I440r>   void how do you do a pointer to a function here again;
20:28:07 <I440r> }
20:28:15 <aaronl_> ok
20:28:23 <I440r> that woudl bbe void function how do you do this again ???;
20:28:24 <I440r> then
20:28:27 <I440r> er 
20:28:31 <I440r> } options;
20:28:32 <I440r> then
20:28:49 <I440r> options vecdtors[] =
20:28:55 <I440r>  key, vector
20:28:57 <I440r>  key, vector
20:28:59 <I440r> }
20:29:04 <aaronl_> cdecl> declare f as pointer to function returning void
20:29:04 <aaronl_> void (*f)()
20:29:13 <aaronl_> UGLY syntax :(
20:29:21 <I440r> that sort of thing, make an array of key,vectors pairs
20:29:26 <aaronl_> ya i see what you mean
20:29:33 <I440r> i did that in my bot...
20:29:35 <aaronl_> and then u call the function member of the stuct
20:29:38 <I440r> yes
20:29:39 <aaronl_> right/
20:29:40 <aaronl_> ok
20:29:48 <I440r> the source code to my bot is in /pub
20:29:52 <I440r> not finished but working
20:30:00 <I440r> its been ages since i did any of this :P
20:30:06 <I440r> but u get the ida
20:30:07 <I440r> also
20:30:31 <I440r> like i said, the source for my loop is alot smaller than a 10 case switch statement 
20:30:34 <I440r> case a of
20:30:35 <I440r>  alaqeoighs
20:30:36 <I440r> askfhlksahf
20:30:37 <I440r> laksdghlkf
20:30:37 <I440r> asdlfalksf
20:30:39 <I440r> case b of
20:30:40 <I440r> a;lkfpsaf
20:30:41 <I440r> asdlasdj
20:30:42 <I440r> lsadfkj
20:30:42 <I440r> aslkdfhhf
20:30:47 <I440r> bullshit visual clutter 
20:30:47 <aaronl_> mmmkay
20:31:01 * I440r hates c :P
20:31:05 <aaronl_> you know what is even moer evil?
20:31:07 <aaronl_> inline asm
20:31:16 * aaronl_ pulls out some that he wrote
20:31:42 <aaronl_>            __asm__ __volatile__("
20:31:42 <aaronl_>                 loop: 
20:31:42 <aaronl_>                 subq #1, %1 /* Dec loop counter */
20:31:42 <aaronl_>                 bmi out /* Terminate loop? */
20:31:42 <aaronl_>                 clr.l %d6
20:31:43 <aaronl_>                 move.b (%2,%1), %d6 /* Get normal char */
20:31:45 <aaronl_> .......
20:31:52 <aaronl_> "
20:31:53 <aaronl_>                 :
20:31:53 <aaronl_>                 : "a" (rot_str), "d" (len), "a" (norm_str)
20:31:53 <aaronl_>                 : "d6", "cc", "memory");
20:32:02 <aaronl_> that is so fucked!
20:33:08 <I440r> :)
20:33:43 <aaronl_> i forget what your stance on commenting was...
20:33:45 <aaronl_> i remember it came up
20:34:24 <I440r> the more you comment your code the better, 
20:34:25 <I440r> well
20:34:35 <I440r> as long as you comment it properly, not just like....
20:34:46 <I440r>  mov ax,0  ; move 0 into ax
20:34:48 <I440r> ugh
20:35:00 <aaronl_> Hrm. gcc produced better code using a 'long long' than using a char[]. I think i'll go with the long long, especially since it will produce MUCH better code on some archs
20:35:04 <aaronl_> i agree with you completely
20:35:07 <aaronl_> and not with linus
20:35:21 <I440r> write and tell him :P
20:35:45 <I440r> linus believes that if you cannot understand teh code simply by reading the code youdo not belong in the code
20:35:51 <aaronl_> long long is pretty morbid tho
20:36:03 <aaronl_> i think comments should be able to help someone who _almost_ can follow and wouldn't otherwise
20:36:46 <I440r> i think that if you comment it properly and enabled some totally clueless coder to follow your code you have done him a great service
20:36:59 <aaronl_> ya me too
20:37:02 <I440r> aned you might even make him learn a little and eventually become a good coder
20:37:10 <aaronl_> i mean, it can even help experts
20:37:30 <aaronl_> if you write bizarre code that's highly optimized it helps to have some simple idea of wtf it's doing
20:38:00 <aaronl_> Ok
20:38:06 <aaronl_> i have a simple assignment for you
20:38:10 <aaronl_> i wanna see if you can beat gcc ;-)
20:38:16 <aaronl_> very simple function....
20:39:31 <I440r> sure
20:39:35 <aaronl_> I have a lower longword and an upper longword in ecx and edx respectively
20:39:42 <I440r> ok
20:39:47 <I440r> upper in edx
20:40:24 <aaronl_> i want to test a certain bit, where 0==last bit of upper word and 63==first bit of lower word
20:40:51 <aaronl_> and put either 1 or 0 in eax depending on if it's set
20:40:51 --- join: adu (andrew@adsl-63-201-89-121.dsl.snfc21.pacbell.net) joined #forth
20:41:46 <I440r> ok you want to move any one of the 64 bits into eax then
20:41:58 <aaronl_> umm ya, i guess
20:42:00 <I440r> where is the parameter
20:42:11 <I440r> i.e. the bit #
20:42:26 <aaronl_> oh, its in eax
20:42:51 <I440r> and is it expressed as a bit number or as a bit mask
20:42:58 <aaronl_> for some reason gcc puts a JUMP in this function!!! i bet you can do better
20:43:00 <aaronl_> bit number
20:43:07 <aaronl_> e 0==last bit of upper word and 63==first bit of lower word
20:43:33 <I440r> u got your bit numberings backwards :P
20:43:46 <aaronl_> umm, other way is actually easier for me :-)
20:43:57 <aaronl_> if you can write code that's >= the speed and efficiency ;-)
20:44:25 <aaronl_> ya in fact i think it shouldn't matter
20:44:44 <aaronl_> so go with 0== first bit lower longword ... 63 == last bit upper longword
20:45:07 <I440r> mov eax,1
20:45:11 <I440r> xchg eax,ecx
20:45:17 <I440r> shl eax,ecx
20:45:26 <I440r> oh 
20:45:28 <I440r> erm no
20:45:29 <I440r> shit
20:45:36 <aaronl_> heh
20:45:38 <I440r> this is a little more complecated
20:45:47 <I440r> i gotta see which register im testing :)
20:45:48 <I440r> ok wait
20:45:51 <aaronl_> gcc does:
20:45:54 <aaronl_>     shrdl   %edx, %eax
20:45:54 <I440r> gona write this out by hand :)
20:45:55 <aaronl_>     sarl    %cl, %edx
20:45:57 <aaronl_> hehe
20:45:59 <aaronl_> thanks
20:46:02 <I440r> \nononon
20:46:06 <I440r> dont show me
20:46:08 <aaronl_> ok ok :)
20:46:10 * I440r not lookming...
20:46:13 <aaronl_> type /clear ;)
20:46:38 <adu> is that intel?
20:46:45 <I440r> 0 - 15 is edx, 16 to 31 is eax... yes ?
20:47:12 <aaronl_> umm, whatever. i don't care :-)
20:47:17 <aaronl_> it's actually taken off the stack ...
20:47:19 <adu> or at&t syntax
20:47:28 <I440r> hrm that shrdl im not familiar with :P
20:47:41 <aaronl_> 16 up on the stack is lower longword
20:47:42 <I440r> hrm 
20:47:46 <aaronl_> 12 up on the stack is upper
20:47:54 <aaronl_> 8 up on the stack is bit number
20:48:15 <I440r> erm wait
20:48:18 <I440r> im confused
20:48:24 <aaronl_> screw the registers
20:48:24 <I440r> edx = high dword
20:48:26 <I440r> eax = count
20:48:31 <I440r> wheres low dword again ?
20:48:33 <aaronl_> do whatever
20:48:37 <aaronl_> it's actually all on the stack
20:48:42 <aaronl_> gcc was moving them into specific registers
20:48:46 <aaronl_> but use your own ...
20:48:48 <I440r> aha
20:48:49 <I440r> ok
20:48:51 <I440r> well
20:49:02 <aaronl_> tho, it shouldn't be on the stack if i was using asm
20:49:06 <I440r> well discount the pops used byu gcc to set the registers
20:49:15 <aaronl_> why don't you just pick whatever registers you want and document it/
20:49:15 <aaronl_> yeah
20:49:17 <I440r> im jhust going to assume that eax is the low word
20:49:20 <I440r> edx = high
20:49:21 <aaronl_> ok
20:49:22 <aaronl_> great
20:49:27 <I440r> and ecx = count
20:49:29 <aaronl_> s/word/longword, tho
20:50:09 <aaronl_> What is SFENCE?
20:50:16 <aaronl_> i'm finding all these instructions i dunno in my book...
20:50:43 <aaronl_> o, i see what shrd is
20:50:54 <aaronl_> "Dbl precision shift right"
20:51:03 <I440r> yes
20:51:19 <I440r> sfence ?
20:51:23 <I440r> i know wht fense is
20:51:32 <I440r> wait 
20:51:34 <I440r> thunkin
20:52:15 <aaronl_> it says "store fence"
20:52:33 <aaronl_> hmm. i don't like how xor is named EOR on m68k
20:53:03 <I440r> am i allowed to use self modifying code ??? :)
20:53:44 <aaronl_> Sure ! :)
20:54:23 <aaronl_> tho what gcc does is pretty straightforward. i didn't get it until i looked up shrd, but that instruction is exactly what we need...
20:54:40 <I440r>    mov b[foo+7],cl   ; set shift count
20:54:55 <I440r>  jmp foo   ; this is needed to flush prefetch
20:54:59 <I440r> foo:
20:55:05 <I440r>   shrd  edx,eax,0
20:55:08 <I440r>  ret
20:55:10 <I440r> done
20:55:30 <aaronl_> wow
20:55:38 <aaronl_> does that put 1 or 0 in eax?
20:55:44 <I440r> oh
20:55:48 <I440r> youshould prolly do an 
20:55:51 <aaronl_> hehe
20:55:51 <I440r> and eax,1
20:55:52 <aaronl_> yeah
20:55:56 <I440r> before the return
20:56:00 <aaronl_> also, what's the point of shrd  edx,eax,0
20:56:07 <aaronl_> isn't the last arg the one of how much to shift?
20:56:22 <I440r> it shifts the 64 bit edx = high and eax low by the immediate cound of 0
20:56:22 <aaronl_> oh, i see
20:56:25 <aaronl_> n/m
20:56:31 <aaronl_> self-modifying :) i like it ;)
20:56:33 <I440r> the 0 is overwritten by the shift count you pass
20:56:44 <I440r> i know... im a heretic
20:56:47 <I440r> but i admit it :)
20:56:56 <aaronl_> isn't there a form that uses CL implicitly?
20:56:56 <I440r> if ur gona use that i would test it if i were you
20:57:05 <aaronl_> hehe
20:57:47 <aaronl_> anyway, here's what gcc did (without stack stuff stripped, in INTEL syntax just for u)
20:57:55 <I440r> erm wait it might not work
20:57:57 <aaronl_> test_bit_64:
20:57:57 <aaronl_>     push    %ebp
20:57:57 <aaronl_>     mov %ebp, %esp
20:57:57 <aaronl_>     mov %ecx, DWORD PTR [%ebp+16]
20:57:57 <aaronl_>     mov %edx, DWORD PTR [%ebp+12]
20:57:58 <aaronl_>     mov %eax, DWORD PTR [%ebp+8]
20:58:00 <aaronl_>     shrd    %eax, %edx, %cl
20:58:02 <aaronl_>     sar %edx, %cl
20:58:03 <aaronl_>     and %ecx, 32
20:58:05 <aaronl_>     je  .L2
20:58:07 <aaronl_>     mov %eax, %edx
20:58:09 <aaronl_> .L2:
20:58:11 <aaronl_>     and %eax, 1
20:58:13 <aaronl_>     pop %ebp
20:58:16 <aaronl_>     ret
20:58:57 <aaronl_> i don't get why it jumps like that
20:59:46 <aaronl_> brb
20:59:52 <I440r> it wont work
21:00:09 <I440r> shrd wont allow shifts of beyond 32 bits with an immediate shift count like that
21:00:10 <I440r> grrr
21:01:27 <I440r> but that gcc code is fscked....wait :)
21:04:55 <I440r> well my idea would have worked if shrd reg,reg,immediate wasnt crippled
21:05:45 <aaronl_> hehe
21:07:32 <aaronl_> have you ever coded on a 64bit arch?
21:07:36 <aaronl_> there this would be really simple
21:08:48 <I440r> yea it would have :)
21:11:00 <aaronl_> i think 64bit is overkill for most things
21:11:11 <aaronl_> i bet it would be nice for high-precision timing, or databases...
21:11:20 * Trey waits impatiently for AMD 64bit CPUs.
21:16:15 <I440r> :)
21:22:58 <aaronl_> it will be hell to align stuff on 8-byte boundries, and stuff
21:23:07 <aaronl_> itanuim has 128 integer registers tho
21:23:09 <aaronl_> crazy
21:27:39 <Trey> If you have a machine with 64bit addresses, you probably have at least a GB of memory to throw around.
21:27:49 <aaronl_> tru
21:29:18 <I440r> erm that depends
21:29:32 <I440r> just because you have the address space doesnt mean yhou have the addresses :P
21:29:32 <aaronl_> Gar
21:29:36 <aaronl_> I need a htonll() function
21:29:49 <I440r> :)
21:30:07 <aaronl_> i'm being an idiot
21:30:16 <aaronl_> you know those bit offsets i was storing for the 64bit int?
21:30:29 <aaronl_> well, the int is being recieved from the network
21:30:36 <I440r> :)
21:30:45 <aaronl_> so the offsets are totally bogus on little endian
21:31:01 <I440r> hehehehehehe
21:31:17 <aaronl_> right?
21:31:23 <I440r> :)
21:31:31 <I440r> prolly :P
21:31:34 <aaronl_> i mean, i should just ntohll() it when i get it in from the network
21:31:41 * aaronl_ digs up a function to do that from his personal library
21:32:10 <aaronl_> ah
21:32:13 <aaronl_> /* Host to network 64-bit */
21:32:13 <aaronl_> static u_int64_t hton64 (u_int64_t x)
21:32:13 <aaronl_> {
21:32:13 <aaronl_> #ifdef M_LITTLE_ENDIAN 
21:32:13 <aaronl_>   if (sizeof(long) == 8) /* also evil */
21:32:14 <aaronl_>     return htonl (x);
21:32:16 <aaronl_> # ifdef __bswap64
21:32:18 <aaronl_>     return __bswap64 (x);
21:32:19 <aaronl_> # else
21:32:22 <aaronl_>   {
21:32:23 <aaronl_>     /* We demand a shrubbery. */
21:32:25 <aaronl_>     union { u_int64_t ll; u_int32_t l[2]; } w, r;
21:32:28 <aaronl_>     w.ll = x;
21:32:29 <aaronl_>     w.ll = x;
21:32:32 <aaronl_>     r.l[0] = htonl (w.l[1]);
21:32:34 <aaronl_>     r.l[1] = htonl (w.l[0]);
21:32:36 <aaronl_>     return r.ll;
21:32:38 <aaronl_>   }
21:32:40 <aaronl_> # endif
21:32:41 <aaronl_> #else
21:32:43 <aaronl_>   return x;
21:32:45 <aaronl_> #endif
21:32:47 <aaronl_> }
21:32:51 <aaronl_> eh i dont know how that comment got in there :P
21:33:11 <aaronl_> how would i compute a bit offset for a big endian quadword?
21:33:37 <Trey> ?
21:33:39 <aaronl_> like, "on a big endian system i know this would be 63 bits to the right, what would it be on this pathetic little endian sys?"
21:33:48 <aaronl_> in a quadword
21:35:43 <Trey> Bits as they sit in memory, or bits when you take the uint64 as a whole?
21:36:25 <aaronl_> the latter
21:36:34 <aaronl_> i know how it sits in memory -- it is in big endian format by definition
21:40:24 <Trey> Once you get it in a register, you can do all the bit tests you want.
21:40:37 <Chaynez> Hey Mark... you here?
21:40:38 <Chaynez> or anyone?
21:40:46 * Trey heads for bed
21:40:51 <Chaynez> What's the command to find out how much space a dir and all sub-dirs/files is takin' up?
21:40:58 <Trey> du 
21:41:04 <Chaynez> tnx
21:41:05 <Chaynez> that's the one
21:41:14 * Trey is away: sleeping [23:40]
21:42:50 <I440r> or df
21:42:59 <I440r> well df is for the whole drive
21:43:06 <aaronl_> heh, so i learned two's complement a few days ago
21:43:08 <I440r> du says what teh dir and sub dirs are using
21:43:11 <aaronl_> it's so cool!
21:43:15 <I440r> :)
21:48:13 <adu> how goes it?
22:20:42 <aaronl_> want to see something scary?
22:20:44 <aaronl_> #define ENTRY(x,y) {(x!=-1)?
22:20:45 <aaronl_>                         (63-((WORDS_BIGENDIAN)?
22:20:52 <aaronl_>                              x:
22:20:52 <aaronl_>                              (n%8)+8*(7-(n/8)))):
22:20:52 <aaronl_>                         -1, 
22:20:53 <aaronl_>                     y}
22:29:53 --- join: edrx (edrx@200.240.18.76) joined #forth
22:38:40 <I440r> edrx!
22:38:40 <I440r> nite ppl, ive gotta go zzz
22:38:45 <edrx> hey, long time no see you!
22:38:48 <I440r> :)
22:38:49 <edrx> 'night
22:38:55 <I440r> yea i was on a contract job
22:38:59 <I440r> now im between :P
22:39:02 <I440r> looking for a new one
22:39:17 <edrx> nice
22:40:11 <edrx> I'm trying to finish an article (it's about 25% done :( ) and I should be going to Montreal in january (math-related stuff, not computer-related)
22:40:31 <edrx> these are my main news, I think
22:40:36 <aaronl_> night
22:40:49 <I440r> teach me mafz ;P
22:40:59 <I440r> r = (1/r) + 1
22:41:10 <I440r> solve that for r :)
22:41:17 <aaronl_> heh
22:41:43 <I440r> which number is exactly 1 more than its own reciprocal 
22:41:43 <edrx> r^2 = 1 + are
22:41:51 <edrx> r^2 = 1 + are
22:41:57 <edrx> r^2 = 1 + are
22:42:05 <I440r> nope
22:42:11 <aaronl_> (1/r) = r/2. Solve that for r.
22:42:15 <aaronl_> edrx: having xchat problems? :)
22:42:16 <edrx> ? there's something translating my "r" to "are"...
22:42:17 <I440r> 1.61803499 or something
22:42:30 <aaronl_> edrx: Settings->Setup...
22:42:32 <aaronl_> Err
22:42:33 <edrx> aaronl_: you got it :)
22:42:37 <I440r> its xchar
22:42:43 <aaronl_> Settings->Replace...
22:42:48 <aaronl_> I HATE that xchat "feature"
22:42:52 <aaronl_> its what i like least about xchat
22:42:55 <aaronl_> i always turn it off first thing :P
22:43:03 <I440r> me too
22:43:12 <I440r> except for the teh -> the
22:43:14 <I440r> heheh
22:43:28 <edrx> great - deleted, thx
22:43:39 <aaronl_> heheh
22:43:54 <aaronl_> I440r: what was your bigest forth program?
22:44:12 <I440r> hrm
22:44:27 <I440r> the long wall coal mining shit
22:44:44 <I440r> roof support system
22:44:57 <aaronl_> wtf was that?
22:45:00 <aaronl_> they wrote a roof in forth?
22:45:01 <I440r> :)
22:45:05 <I440r> yes :P
22:45:15 <I440r> ill explain next time
22:45:18 <aaronl_> k
22:45:25 <I440r> troo tired to run error hgehe
22:45:31 <I440r> : sleep   
22:45:36 <I440r>    bed go tuck
22:45:43 <I440r>    light off ; immediate
22:45:47 <aaronl_> :)
22:45:49 <I440r> (tm)
22:45:51 <I440r> :)
22:46:19 <I440r> what would :) do in forth :)))
22:46:42 <aaronl_> what would it do?
22:51:21 <edrx> you need to define it
23:12:49 --- quit: edrx ([x]chat)
23:22:26 --- quit: Chaynez (Rebooting)
23:35:09 --- quit: adu (IRCStep)
23:59:59 --- log: ended forth/01.06.12