Hugi Size Coding Competition 12
Final Results (October 8th, 2000)

 1      Picard                  Hungary          76
 2      INT-E                   Germany          78
 3      Guillermo Sais          Mexico           80
 4      Fred                    France           81
 5      Ruud                    Netherlands      84
 6      Roland Neul             Germany          85
 6      Chut                    Hungary          85
 8      meph                    Germany          86
 9      Lawrence E. Boothby     United States    87 (83) *
 9      Blue                    Germany          87
 9      Boreal                  United States    87 (85) *
12      Aphex                   Romania          88
13      Eldar                   Estonia          90
14      gbm                     Poland           91
14      Sniper (and claw)       United States    91
14      Bonz                    Italy            91
17      Stefan                  Germany          92
17      Eduardo de la Pena      Argentina        92
19      Jibz                    Denmark          95
19      fabled                  Finland          95 (91) *
21      shr / molecoola         Hungary          96      *
22      Metalbrain              Spain            98
23      Ray Ransom              United States   102
24      Johnny L. Fencey        Germany         106
25      freaking dooz           Sweden          107
26      anti^fake               Germany         146
--      Aditsu                  Romania         107     TIM
--      BitIO / HardNull        Spain            83     INC
--      claw                    Germany          88     TIC
--      Jeff                    United States    96     TIC
--      Jonnz                   Germany         143     INC
--      Robin / NOP             Hungary          97     INC
--      rwb                     Australia       131     TIM
--      Schwendi                Switzerland     106     TIM
--      VrA                     Hungary         112     TIM

TIM: entry exceeds time-limit (90 seconds; tested with n=1000000)
INC: other error, e.g.:
     - "entry 0" or "entry" fails or gives wrong result
     - output is in an incorrect format (not recognized by test suite)
     - overflow error
TIC: entry exceeds time-limit and another error occurs
  *: error detected during Public Judgement, see next section

Public Judgement Results

In the following entries, incorrect behaviour was detected after the
release of all entries to the public:

Boreal: Extra output (00h) after CR/LF, can be fixed in this way
according to Guillermo Sais:

        ... LINE 78
        INC     BP              ; [FIX, +1 byte]
dn20:   pop     dx
        int     21h
        ;dec    dx              ; [FIX, -1 byte]
        DEC     BP              ; [FIX, +1 byte]
        jns     dn20

fabled: Uses INT 29h for output, which outputs to the console but not
to stdout as required by the rules. Smallest fix according to Guillermo
Sais:

        ... LINE 63
        db      0bfh
@@wr:   int     21h
        lodsb
        add     al, '0'
        XCHG    DX,AX           ; [FIX, +1 byte]
        MOV     AH,2            ; [FIX, +2 bytes]
        jns     @@wr

Lawrence E. Boothby: Uses ADD CL,AL in parsing. A correct version
is included in correct.zip. He submitted it after the public judgement. 

As a penalty, the size of the smallest fix and an extra byte have been
added to the sizes of these entries.

shr / molecoola: The original entry.com file gave wrong results when
executed with the parameters n=65536, 131072, 196608, ... However, the
reason was a bug in the NASM assembler; NASM assembled the instruction
JECXZ 127h to 66h E3h 0Fh, while 67h E3h 0Fh would have been correct.
The entry compiled fine after being converting to TASM-compatible code,
and by using DB 66h JCXZ 127h it was also possible to make NASM generate
the correct result. Therefore the entry was accepted without a penalty.
In the SHR directory also the fixed versions created by other competitors
have been included.

Thanks to all people for participating!
