Flashing a A-110 firmware on A-100 does fail pretty early on.
1) make A100 think its A110. Will it work? Does it check things too deep? (cpukeys etc)
- Might not be possible, but if it would work, it would make everything complete.
- Changing /etc/config, /etc/kinfo and hostname from 402 to 403, will make it complain about wrong size. Perhaps a better idea is to change the firmware to say 402 in its version: c403dPOPk15
- Unfortunately my first tests of this complained about wrong flash size, this turns out that you have to do a stepping flash before you can do newer firmwares, so this could be re-visited. For now, I have made a 403 firmware be 402.
2) pull out ucode from A110, and add to A100 firmware. Not too hard, but tedious. Have all keys, and CRC code already.
- Apart from ucode, other parts of the software might check for model number to enable/disable 2ch mix-down. So we might have to do partial 1) anyway.
3) pull out ucode and attempt to get it to initialise after A100 has booted up.
- Unlikely replacing ucode after boot will work, and may still need part of 1) to work.
4) Make an A-110 firmware be A-100 firmware.
- That worked rather well. DTS works.
So, while we are changing the firmware, I could add telnet, and even load FUSE. Can I replace NTFS with NTFS-3g automatically? Perhaps not. Replace "mount" with script, but sys_mount() I can not.
Status of DTS
The very simple test of making an A-100 become an A-110 worked well, somewhat. I could use it, and I got DTS decoding to 2 channels working. However, since A-110 is HDMI-1.3, that part of the ucode is also loaded. If you then try to use HDMI (which I do not) the A-100 will hang.
The trick would then be to attempt to keep DTS decoding, but not use HDMI-1.3. But since HDMI-1.3's new feature is mostly that it handles Audio in HDMI signals, I do not know if that is possible.
I did have success in just moving the audio_ucode over to A-100 firmware, but the Gaya GUI does not give you the option to select Analog for DTS. Work is then needed to figure out what parts would be needed, finding differences between the firmwares etc. I do not currently have time for this, I would welcome other hackers to attempt it.
If you also found a NMT device with DTS decoding but only HDMI-1.1 that would also be feasible.
Related to this, I have ucode for CSS, we should also try to get that to run.
Well, you need to make a firmware using the audio_ucode_dts, that is the most important part of the puzzle. After that, change setups.cgi for the gui selection, but just use mono directly for quicker tests.