I don't think I can learn C....

[Crumpster]

@ Karlos

yep 14 ish Kb sorry finger slipped on the decimal point
(I obviously can't count, did I cover it ok?  :-P )

Thanks fo the correction.

@ Athiest

Not on an A1 yet I'm afraid, crappy works P2 233Mhz with redhat 7.3 on it.

Cheers

Crumpy

[xeron]

Compiled with GCC for the Win32 console the exe is 23,468 bytes.

[Crumpster]

Tickly Wrote Thus:

Compiled with GCC for the Win32 console the exe is 23,468 bytes.

Good old Windows Bloat!!  :-D

[Atheist]

Crumpster wrote:

Good old Windows Bloat!!  :-D

Is it just me not understanding how these things work, or is that somwhat excessive? I wonder if it still relies on a DLL being present?

Also, as I wondered out loud in another thread, how would going to a 64 bit CPU instruction set affect the size?

AmigaOne! "What the DLL you talkin' 'bout, fool?", Mr. T?

[takemehomegrandma]

That's very cool!  :-o

[Karlos]

Atheist wrote:

Also, as I wondered out loud in another thread, how would going to a 64 bit CPU instruction set affect the size?

AmigaOne! "What the DLL you talkin' 'bout, fool?", Mr. T?

Here

[Aragorn]

 4788 bytes on sparc if the -s switch is used

[pVC]

6216 bytes with SAS/C with default options on 68k AmigaOS

[xeron]

I get the same size with SAS/C, however if I size optimize it and turn off some other things I can get it down to 5k, but it then gurus 8000 0004 after running Oh well, 6k's still not bad

[danamania]

3968 bytes on 68k linux. 3 cheers for 680x0 .

3.875Kb sounds even better!

-rwxr-xr-x    1 dana     dana         3968 Apr  3 04:44 a.out

[Karlos]

Ok everbody,

How big is the pure object file generated on your systems?

The size of the stdio / startup stuff has a lot of impact on some systems.

To see the true code generation size, just look at the unlinked object file. Should be interesting :-)

[pVC]

Object file is 1128 bytes with SAS/C.

[danamania]

mine comes out to:

-rw-r--r--    1 dana     dana         2060 Apr  3 05:34 xmas.o

2060 bytes. I know little about gcc and any optimisations, but that's with 2.95 on debian woody on a 68040.

dana

[Karlos]

Compiled under Watcom 11.0c as a Win32 character mode executable (speed rather than space opts)


02/04/2003  20:32                  881 mystery.c
02/04/2003  20:36               31,232 mystery.exe
02/04/2003  20:36              113,664 mystery.ilk
02/04/2003  20:36                   19 mystery.lk1
02/04/2003  20:36               22,729 mystery.map
02/04/2003  20:36                   89 mystery.mk
02/04/2003  20:36                  669 mystery.mk1
02/04/2003  20:36                1,496 mystery.obj
02/04/2003  20:36               16,599 mystery.sym


As you can see, the object file is only 1.4kB :-)

-edit-

Check out the winbloat! 30kb of sh*t linked just to make the exe :-D

-edit-

My guess is

1) the original text is compressed with something like LZ or dynamic huffman style compression, encoded into the big string in the middle

"@n'+,#'/*{}w+/w#cdnr/+,{}r/*de}+,/*{*+,/w{%+,/w#q#n+,/#{l,+,/n{n+\
,/+#n+,/#;#q#n+,/+k#;*+,/'r :'d*'3,}{w+K w'K:'+}e#';dq#'l q#'+d'K#!/\
+k#;q#'r}eKK#}w'r}eKK{nl]'/#;#q#n'){)#}w'){){nl]'/+#n';d}rw' i;# ){n\
l]!/n{n#'; r{#w'r nc{nl]'/#{l,+'K {rw' iK{;[{nl]'/w#q#\
n'wk nw' iwk{KK{nl]!/w{%'l##w#' i; :{nl]'/*{q#'ld;r'}{nlwb!/*de}'c \
;;{nl'-{}rw]'/+,}##'*}#nc,',#nw]'/+kd'+e}+;\
#'rdq#w! nr'/ ') }+}{rl#'{n' ')# }'+}##(!!/"

2) The binary lookup for the encoding is stored in the smaller string

"!ek;dc \
i@bK'(q)-[w]*%n+r3#l,{}:\nuwloca-O;m .vpbks,fxntdCeghiry"

3) main() is defined as a recursive extraction algorithm (similar to some huffman decoders) and character output.

4) This is an oversimplification since main() is called internally several times and tested conditionally with operator ? :...

Anybody else got any ideas?

This thing is cool :-)

[Karlos]

danamania wrote:
mine comes out to:

-rw-r--r--    1 dana     dana         2060 Apr  3 05:34 xmas.o

2060 bytes. I know little about gcc and any optimisations, but that's with 2.95 on debian woody on a 68040.

dana

Given that youre exe was only a kilibyte larger, you sure do have a good startup/stdio lib :-D
Probably just a stub that relies on OS calls deeper in?

[Piru]

The size of the stdio / startup stuff has a lot of impact on some systems.

Yes indeed. So why not skip stdio & startup?

Here is what I did:

#include
#include

#include
#include

#define putchar(x) FPutC(Output(),x)

struct DosLibrary *DOSBase;

void entry(void)
{
  struct ExecBase *SysBase = *(struct ExecBase **)4;
  DOSBase = (struct DosLibrary *) OpenLibrary("dos.library", 37);
  if (DOSBase)
  {
    main(1,NULL,NULL);
    CloseLibrary((struct Library *) DOSBase);
  }
}

//#include
.... rest of the original code...

Then compile with size optimization, link without startup code, and strip resulting exe:

68000, AmigaOS (SAS/C 6.58):
mystery                           1048 ----rwed Today     23:32:37

PowerPC, MorphOS (gcc 2.95.3):
mystery                           2200 ----rwed Today     23:32:42

Next step: Optimize the generated asm code. :-)