When I was looking at the oem software file format and program, I noticed a few things.
1. in the .dat saved file, they only store the "main" block. i.e., without the ident.
2. they seem to only check the ident on the fly at clone time
3. they don't store the "aux" block, but this is read and written back on the fly. (Probably because there can contain radio specific, i.e., calibration fields that shouldnt be shared across radios.)

With that in mind, 
1. should we put the ident at the end of the memory map? (It might break some backwards compatibility unless extra logic is taken)
2. should we consider handling the aux block as sort of "volatile"?

-Jens


On Oct 1, 2014, at 7:06 AM, IZ3GME Marco <iz3gme.marco@gmail.com> wrote:

...mmm... another idea: it can be simplier to just compress the 12 byte 
ident down to 8 bytes in the memory map, this avoid changing the memory 
(and file) format.

Just e cent more ;)
73 de IZ3GME Marco

On 01/10/2014 11:53, Jim Unroe wrote:
Marco,

It isn't just the lengh of the file. Since the "ident" that is returned
is written to the .img first and the rest of the radio's memory follows
it, all of the memory is shifted 4 bytes.

So...

#seekto 0x0B08;
struct {
  u8 code[5];
  u8 unused[11];
} pttid[15];

for a "normal" BaoFeng UV-5R style radio would need to be...

#seekto 0x0B0C;
struct {
  u8 code[5];
  u8 unused[11];
} pttid[15];

For this "odd" class of UV-6 and UV-8 radios.

I think there is a similar problem when writing to the radio. CHIRP
currently does this...

    # Main block
    for i in range(0x08, 0x1808, 0x10):
        _send_block(radio, i - 0x08, radio.get_mmap()[i:i + 0x10])
        _do_status(radio, i)

It starts a 0x08 (right after the normal 8 byte "ident"). I think it
would need to start at 0x0C (right after the 12 byte "ident").

Jim

On Wed, Oct 1, 2014 at 4:52 AM, IZ3GME Marco <iz3gme.marco@gmail.com
<mailto:iz3gme.marco@gmail.com>> wrote:

   Just a couple of suggestions:
   - for the communication you can try reading 12 bytes lowering timeout so
   if only 8 are received read will bail out quickly enough
   - i would expand the structure to the longer size to have to deal with
   different sizes only when accessing ident field (i guess ident is only
   used during clone)

   my two cents
   73 de IZ3GME Marco

   On 01/10/2014 01:53, Jim Unroe wrote:
Hi all,

I am working on two issues, #1225 (UV-8) and #1707 (UV-6), that after
doing research seems to indicate both issues are caused by the
   same thing.

The problem is that after the "magic" is sent to the radio to
   initiate
cloaning, the radio returns a 12 byte ident when every other Baofeng
radio that I know of returns an 8 byte ident. I have temporarily
   edited
the driver to accept 12 bytes and CHIRP downloads from the radio. But
since the ident is 4 bytes longer, all of the structures are off
   by 4 bytes.

After saving the downloaded data to a .img file, I can use a hex
   editor
to remove 4 bytes from the begining of .img file, save it to another
name and load it back into CHIRP and it is fine.

So I am looking for some help to figure out how to have CHIRP to be
tolerant of the 8 byte and 12 byte idents and also 4byte offset data
structures.

Got any ideas?

Thanks,
Jim KC9HI




_______________________________________________
chirp_devel mailing list
chirp_devel@intrepid.danplanet.com
http://intrepid.danplanet.com/mailman/listinfo/chirp_devel
Developer docs: http://chirp.danplanet.com/projects/chirp/wiki/Developers


_______________________________________________
chirp_devel mailing list
chirp_devel@intrepid.danplanet.com
http://intrepid.danplanet.com/mailman/listinfo/chirp_devel
Developer docs: http://chirp.danplanet.com/projects/chirp/wiki/Developers