02:08:57: -!- Doxin has quit (Remote host closed the connection).
11:55:27: -!- Doxin has joined #corewars.
12:49:10: -!- Manji has joined #corewars.
12:49:15: <Manji> hi there
12:49:34: -!- Manji is now known as Guest27838.
12:50:01: <Guest27838> Hi again ;D question about sal: Is it down forever?
12:51:50: <Guest27838> second question: Has there ever been a quickscanner/qscanner decode engine that DOES NOT use a mathematical relation between scanned positions?
12:52:27: <itsy> Hi Guiest27838
12:53:14: <itsy> SAL has moved to http://sal.discontinuity.info and seems to be working at the moment.
12:54:15: <Guest27838> oh sry didnt know that. thx
12:56:00: <itsy> The first quickscanners were something like:
12:56:17: <itsy> SEQ somewhere, somewhere+step
12:56:33: <itsy> MOV #somewhere, found
12:56:48: <itsy> SEQ somewhere_else, somewhere_else+step
12:56:57: <itsy> MOV #somewhere_else, found
12:57:51: <Guest27838> Sorry, no I mean: There is no mathematical relation between the scanned positions.
12:57:55: <itsy> But quickscanners which use either addition or multiplication to decode the location are more faster
12:59:24: <itsy> Well you could have a quickscan which scans a table of locations. The table could be completely random.
13:00:37: <itsy> By the way, there's a tournament coming soon... Look out for the announcement next month :-)
13:01:10: <Guest27838> suppose you know where the decode engine were jumped from. and also suppose ONLY the seq instruction hold the position and no table (table would mean wasting another half of the quickscanner). then the problem arrises that the decoding take quite a lot of steps (at least 7 instructions)
13:01:35: <Guest27838> oh a tournament. Will check it out although I have exams coming >.<
13:02:54: <itsy> The deadline will be about 3 months after the announcement, so plenty of time.
13:03:20: <itsy> Otherwise you'll have to wait for the Halloween Tournament :-P
13:04:12: <Guest27838> the decoding takes so long because you cant just copy one scanned position with offset and if it failed add a certain amount(except you would use a table with the differences).
13:10:40: <itsy> Hmmm... if you had SNE a,b / SEQ c,d and you knew something was scanned at a,b,c or d then you could decode which in 3, 4, 7, or 8 cycles depending which. (average 5.5) Faster might be possible, I'm not sure.
13:55:23: <itsy> johnkw: how can I get a link fixed on koth.org? Have you got access? KOTH@SAL moved to http://sal.discontinuity.info
16:20:54: -!- Guest27838 has quit (Quit: ChatZilla 0.9.90.1 [Firefox 27.0/20140127194636]).
20:10:28: -!- fiveop has joined #corewars.
20:34:00: -!- fiveop has quit ().
21:18:00: -!- Manji has joined #corewars.
21:18:24: -!- Manji is now known as Guest13537.
21:18:50: <Guest13537> hi
21:19:58: <john_metcalf> Hi :-)