00:00:00 --- log: started forth/03.03.07
00:05:19 --- join: Kitanin (~clark@SCF61185.ab.hsia.telus.net) joined #forth
00:06:35 <Robert> Hi Kitanin i:)
00:06:37 <Robert> er
00:37:27 --- join: gilbertdeb (~gilbert@fl-nked-ubr2-c3a-29.dad.adelphia.net) joined #forth
00:47:35 --- quit: gilbertdeb ("Client Exiting")
01:06:51 --- quit: Kitanin (Read error: 104 (Connection reset by peer))
01:19:37 --- join: Serg_Penguin (~Z@nat-ch1.nat.comex.ru) joined #forth
01:25:33 --- join: Speuler (~Speuler@mnch-d9ba4bb6.pool.mediaWays.net) joined #forth
01:27:01 --- part: Jim7J1AJH left #forth
01:35:55 --- quit: Serg_Penguin ()
03:02:13 --- join: Kitanin (~clark@SCF61185.ab.hsia.telus.net) joined #forth
03:39:25 --- quit: Speuler (Read error: 60 (Operation timed out))
04:05:30 --- quit: Kitanin (Read error: 104 (Connection reset by peer))
04:26:31 --- join: Speuler (~Speuler@mnch-d9ba4bb6.pool.mediaWays.net) joined #forth
05:45:49 <Speuler> 'morning
06:05:07 --- join: Kitanin (~clark@SCF61185.ab.hsia.telus.net) joined #forth
06:08:05 --- join: Herkamire (~jason@wsip68-15-54-54.ri.ri.cox.net) joined #forth
06:25:15 --- join: gilbertdeb (~gilbert@fl-nked-ubr2-c3a-29.dad.adelphia.net) joined #forth
07:06:13 --- quit: Kitanin (Read error: 104 (Connection reset by peer))
07:28:12 --- quit: gilbertdeb (Read error: 60 (Operation timed out))
07:56:07 --- join: Kitanin (~clark@SCF61185.ab.hsia.telus.net) joined #forth
08:27:13 <Herkamire> do I have this right:    : tuck ( a b -- b a b ) swap over ;
08:27:26 <Herkamire> or is tuck supposed to do something else?
08:28:47 <TreyB> That works.
08:38:51 <Herkamire> cool. just making sure it wasn't supposed to be    : tuck ( a b c -- c a b c ) or something
08:40:00 <Robert> :)
09:01:55 <Herkamire> I'm having so much fun writing this forth
09:02:08 <Herkamire> the basic words are so easy to implement
09:02:24 <Herkamire> I haven't done any flow control words yet though
09:06:14 <Speuler> in principle, flow control words are easy too
09:07:01 <Speuler> just need to watch out for correct branch offset
09:08:14 <Speuler> and, if you have your flow control words to check structure, there's some fiddling with else 
09:08:31 <Speuler> and then
09:10:58 <Speuler> +loop runtime can be non-trivial
09:11:58 <Herkamire> last go round I only made one flow control word:  0;
09:12:08 <Speuler> efficient implementations use the overflow bit for signalling end-of-loop
09:12:12 <Herkamire> I made it exit if TOS was 0 (and not drop it)
09:12:37 <Speuler> like  for...next
09:13:54 <Herkamire> what platform are you talking about?
09:13:57 <Speuler> does> will be fun
09:14:01 <Herkamire> I'll just use the count register for for/next
09:14:49 <Speuler> not about a specific platform. a number of forths support for..next.  
09:14:57 <Speuler> cmforth does, gforth does
09:15:17 <Herkamire> I mean for overflow signalling end-of-loop
09:15:32 <Herkamire> I'm definately not going to implement do/loop
09:16:06 <Speuler> eforth also has for next
09:16:15 <Speuler> bye
09:16:25 <Speuler> nope, don't mean to "bye" here
09:16:33 <Speuler> meant to "bye" eforth :)
09:17:07 <Speuler> no nested loops, it is ?
09:17:27 <Speuler> or will you push/pop the count register ?
09:17:44 <Herkamire> yeah I think I'll push/pop the count register to/from the rstack
09:17:53 <Herkamire> only down side is that I need an unloop
09:18:04 <Speuler> do loop  is not a big deal
09:18:19 <Speuler> as opposed to +loop
09:19:46 <Speuler> Herkamire: problem with +loop is that it needs to work for negative loop increment as well
09:20:56 <Speuler> most efficient way i know is to use the overflow at  $7FFFFFFF <> $80000000 for end-of-loop detection
09:21:02 <Speuler> works both ways
09:22:23 <Speuler> problem is, you can't use  =  to compare loop index and loop limit
09:22:48 <Speuler> or    11 0 do  2 +loop   never terminates loop
10:05:28 --- quit: Kitanin (Read error: 104 (Connection reset by peer))
10:24:56 <Herkamire> I'm using 0; together with tail recursion to do loops
10:25:25 <Herkamire> 0; drops and exits if TOS is zero. otherwise it does nothing
10:26:42 <Robert> I'm using the following, maybe your variant would make things a bit more convenient..   is0 drop ; then
10:27:17 <Herkamire> yeah. basically I stuck that in a definition
10:27:44 <Robert> :)
10:28:00 <Herkamire> basically:   : 0; dup 0 = if drop rdrop exit then ;
10:28:40 <Herkamire> or, depending on your implementation, maybe:   s/rdrop/rnip/
10:29:09 <Herkamire> I can't define it using if though, because I don't have if (yet?)
10:29:27 <Herkamire> I might use this in place of if. we'll see
10:38:57 <Herkamire> I like this:   : mod 2dup / * - ;
10:39:54 <Robert> Mmmmm...speed.
10:44:14 <Herkamire> ssshhh
10:44:20 <Robert> ;)
10:44:33 <Herkamire> yeah, I know, I could do it in asm
10:44:35 <Robert> http://www.azillionmonkeys.com/qed/sqroot.html <-- mmm.. speed.
10:44:45 <Robert> That's something for IsForth.
10:44:46 --- join: Kitanin (~clark@SCF61185.ab.hsia.telus.net) joined #forth
10:44:48 <Robert> Hi Kitanin 
10:45:01 * Kitanin picks jaw up off floor.
10:45:03 <Kitanin> Hi!
10:45:10 <Herkamire> PPC does sqrt in one instruction
10:45:27 <Herkamire> Hi Kitanin 
10:45:38 * Herkamire picks Kitanin off the floor.
10:45:46 <Robert> Herkamire: Really? Nice :)
10:45:54 <Kitanin> SCO...  sued...  IBM...  for...  patent...  infringement...
10:46:30 <Herkamire> Robert: yeah :) it looks like I'll have to find cool algorithms for sin, cos, atan etc though :)
10:46:43 <Herkamire> Kitanin: on what?
10:46:59 <Robert> Herkamire: Is that integer or floating point square root we're talking about?
10:47:38 <Kitanin> Apparently, IBM couldn't _possibly_ have made Linux work as well as it does without stealing UNIX knowledge from SCO.
10:48:20 <Herkamire> humph
10:49:16 <Kitanin> Yeah.  I think we're going to be able to roast whole herds of cattle over the smoking hole that Caldera/SCO is about to become.  :-)
10:49:22 <Herkamire> Robert: floating point
10:49:35 <Robert> OK... can FPUs on PC do that?
10:49:52 <Herkamire> I don't know
10:50:06 <Robert> And...if you do want integer square root, how fast can you convert integer->float->integer to get that?
10:53:52 <Herkamire> float<-->int is one instruction (fast) but  it slows the pipeline down if you use it right away (5-9 cycles??? I forget)
10:54:20 <Herkamire> my knowledge is way out of date. I'm sure it's better now
10:54:46 <Robert> Neat...
10:55:07 <Herkamire> I'm not describing it well. I just know that there's a stall if you convert and then want to use it right then
10:55:27 <Herkamire> I've never used the FPU in asm
10:56:11 <Herkamire> my forth has . (dot) again :) my life is complete
10:56:21 <Herkamire> ;)
11:09:39 <Kitanin> Okay, I give up.  How did you lose dot in the first place?
11:17:49 --- join: I440r (~mark4@1Cust116.tnt1.bloomington.in.da.uu.net) joined #forth
11:17:55 <Herkamire> hehe
11:18:42 <Robert> Hey I440r 
11:18:51 <Herkamire> I stripped my kernel down to 8 words, and wrote an assembler in forth
11:19:02 <Robert> I440r: Please, don't use that old version of my pong in IsForth, you're embarrasing me :
11:19:06 <Robert> :)
11:20:19 <I440r> lol
11:20:25 <I440r> no im not - YOU are :P
11:20:28 <I440r> email me a new one :)
11:20:33 <Herkamire> heh 
11:20:35 <Kitanin> Ah, okay.  I figured you'd just managed to break it somehow.  :-)
11:21:19 <Herkamire> Kitanin: in a manner of speaking. though deleting such words as / was pretty deliberate :)
11:21:39 <Herkamire> not to mention DUP
11:30:05 <Kitanin> You eliminated dup?  Occam would be proud.
11:31:34 <Robert> I440r: Haha
11:31:40 <Robert> I440r: Sure.. :)
11:34:13 <Robert> The fastermail.com address?
11:34:40 <Herkamire> I could take drop out of the kernel too, I only call it once before I can assemble it.
11:36:28 <Herkamire> I could get the kernel down to 4 words (plus interpret) pretty easily
11:58:02 --- join: Soap` (~flop@202-0-42-22.cable.paradise.net.nz) joined #forth
12:09:43 --- part: Jaster left #forth
13:05:15 --- quit: Kitanin (Read error: 104 (Connection reset by peer))
13:12:41 --- join: Kitanin (~clark@SCF61185.ab.hsia.telus.net) joined #forth
13:42:38 --- quit: I440r ("Reality Strikes Again")
14:07:47 --- quit: Herkamire ("weekend! here we come!")
14:33:17 <OrngeTide> http://www.amresearch.com/v5.html ... hrm. an LGPL forth for linux. i never heard of this company
14:33:36 <OrngeTide> it's written in forth. it uses gforth to compile itself 
14:33:49 <Robert> Hey OrngeTide 
14:33:57 <Robert> Long time no see :)
14:35:13 <OrngeTide> hey. :)
14:36:12 <OrngeTide> oops. gotta go again. work has been hectic. :(
14:39:51 <Robert> :/
14:39:53 <Robert> See you
15:29:44 --- join: Klaw` (anonymouse@ip68-4-155-247.oc.oc.cox.net) joined #forth
15:29:44 --- quit: Klaw (Read error: 104 (Connection reset by peer))
15:29:51 --- nick: Klaw` -> Klaw
15:43:38 --- quit: Kitanin ("Client killed by developer sick of answering stupid questions.")
18:38:20 --- quit: Soap` (Read error: 104 (Connection reset by peer))
19:08:40 --- join: karingo (karingo@33.portland-03-04rs.or.dial-access.att.net) joined #forth
19:10:37 --- join: Soap` (~flop@202-0-42-22.cable.paradise.net.nz) joined #forth
19:58:08 --- quit: karingo (Read error: 60 (Operation timed out))
20:16:51 <Speuler> 'morning
20:55:25 --- quit: Soap` (Read error: 54 (Connection reset by peer))
21:08:26 <Speuler> robert awake ?
22:10:59 --- quit: Speuler (Read error: 110 (Connection timed out))
22:12:38 --- join: Speuler (~Speuler@mnch-d9ba4867.pool.mediaWays.net) joined #forth
22:35:38 --- join: Soap` (~flop@202-0-42-22.cable.paradise.net.nz) joined #forth
23:23:16 --- quit: Speuler (Read error: 60 (Operation timed out))
23:59:59 --- log: ended forth/03.03.07