Odd Problem With Old EDU Bot

[wun]

Hey,
I'm having a really strange problem with the old EDU controller (I think two years ago -- the first one to use a PIC processor and it needed an external RC controller). I've pulled it out to use on an un-related project, but the digital inputs don't seem to be working.
I've initialized them properly:
IO1 = IO2 ... IO8 = INPUT;
I've turned off analog:
Set_Number_Of_Analog_Channels(NO_ANALOG);
I've set the rest of the pins to output:
IO9 = IO10 ... IO16 = OUTPUT;
rc_dig_out09 ... rc_dig_out16 = 0;

Then when I check the values of the pins:
printf("rc_dig_in01: %d ... rc_dig_in08: %d\n", rc_dig_in01 ... rc_dig_in08);
I get:
257 257 257 257 0 0 0 0

Things get stranger. If I ground input 1, the output becomes:
1 257 257 257 0 0 0 0

And grounding other pins change random values (pin 4 changes value 2 from 257 to 256, etc).

EDIT:
I've figured out the pattern:
Grounding input 1 will cause the value of input 1 to be 1
Grounding input 2 will cause the value of input 1 to be 256
Grounding input 3 will cause the value of input 2 to be 1
Grounding input 4 will cause the value of input 2 to be 256
And so on...
If 16 is set as an input, it will cause the value of 8 to change from 1 to 0
Every input after 8 (9 and up) are 0 and I can't get 'em to change.

It dosen't seem to make any difference if I define none of the inputs as analog or all the inputs as analog (NO_ANALOG or EIGHT_ANALOG).

I'm using the default code from http://www.ifirobotics.com/edu-rc.shtml with the most recent firmware.

I have tried simple modifications to the default code (just printf("%d\n",rc_dig_in06); (where in06 defaults to an input) and similarly strange results have occurred.

Any advice would be greatly appreciated!
Thanks,
David