byte limitations

[Orillion]

After testing iago's binary packet buffer he wrote for his Javabot, it became apparent that adding bytes with high values seemed to be impossible so I ran some quick tests only to discover this error

Code Select Expand


*** Semantic Error: A byte value must be an integer value in the range of -128..127

NOTE: I achieved this error from just compiling a primitive byte with a value of 255.

From my understanding this means we'd never be able to add a byte with a decimal value higher than 127. This would make adding 0xff for example, impossible. Any thoughts on this problem?

[iago]

Hmm, I think a simple change can fix that.  Just store it as an integer or long until it gets put into a byte, then mask it to put it into the byte.

[Etheran]

After testing iago's binary packet buffer he wrote for his Javabot, it became apparent that adding bytes with high values seemed to be impossible so I ran some quick tests only to discover this error

Code Select Expand


*** Semantic Error: A byte value must be an integer value in the range of -128..127

NOTE: I achieved this error from just compiling a primitive byte with a value of 255.

From my understanding this means we'd never be able to add a byte with a decimal value higher than 127. This would make adding 0xff for example, impossible. Any thoughts on this problem?

why wouldn't 0xff work?  in a signed variable 0xff = -128 am I wrong?

[Skywing]

After testing iago's binary packet buffer he wrote for his Javabot, it became apparent that adding bytes with high values seemed to be impossible so I ran some quick tests only to discover this error

Code Select Expand


*** Semantic Error: A byte value must be an integer value in the range of -128..127

NOTE: I achieved this error from just compiling a primitive byte with a value of 255.

From my understanding this means we'd never be able to add a byte with a decimal value higher than 127. This would make adding 0xff for example, impossible. Any thoughts on this problem?

why wouldn't 0xff work?  in a signed variable 0xff = -128 am I wrong?

I think the problem they are speaking of is having to use signed constants (this being inconvenient and a pain).

[Kp]

in a signed variable 0xff = -128 am I wrong?

Yes.  0xff = 255 = -1.

[Etheran]

in a signed variable 0xff = -128 am I wrong?

Yes.  0xff = 255 = -1.

err yes of course, that would be 0x80 = -128 and 0x7F = 127

[Orillion]

Yeah the problem i think stems from the fact that java automatically signs everything. Although we have gotten around that to a certain extent

[j0k3r]

Er, sort of off topic, what part of '0x80' determines that it's negative 128?

[Banana fanna fo fanna]

The sign bit, or the first bit in the number.

[iago]

The sign bit, or the first bit in the number.

Technically, it's the last bit.  I would call it the left-most bit

[Banana fanna fo fanna]

Yes, left-most.

[Etheran]

and to find your signed value if the sign bit is set, you take the two's complement of the number:
~x + 1

example:

1101 1001 = 0xC9

take the two's complement:

0010 0110 + 1 = 0010 0111 = 39
add the sign and you end up with -39

[iago]

hmm, I don't see what your problem was.  I just tested it out with the following code:

Code Select Expand

   public static void main(String args[])
   {
       Buffer buf = new Buffer();

       buf.add((byte)250);
       buf.add((byte)251);
       buf.add((byte)252);
       buf.add((byte)253);
       buf.add((byte)254);
       buf.add((byte)255);

       buf.add((byte)-1);
       buf.add((byte)-2);
       buf.add((byte)-3);
       buf.add((byte)-4);
       buf.add((byte)-5);

       System.out.println(buf.toText());
   }


And it came back with the (correct) results:

Code Select Expand

Buffer contents:
fa fb fc fd fe ff ff fe fd fc fb                        ...........
Length: 11