After some debugging of the solar charge controller, power seems to be working now.
Then I used an AVR dragon to set the MCU fuses (F7 96 FD)
I then checked that the crystal was working with an oscilloscope and found 16MHz. Then used the Dragon to put Optiboot in it from the zip found in v6.2 releases (thanks WesfW for keeping this up to date).
After that I had to hack my RPUadpt board so the RJ45 was facing up, it allowed me to bootload with an Uno over a long distance (RS485).
http://epccs.org/indexes/Board/RPUadpt/
Next I setup board.txt, and variants/irrigate7/pins_arduino.h files for this board.
https://github.com/epccs/epccs-avr
Then tinkered with the new interface to the Arduino IDE's Boards Manager, which I struggled with somewhat but now that it is working I have to say wow Arduino, that is totally cool thank you so much and just in time for me to tinker with.
https://github.com/epccs/epccs-board-index
So far, I have only started debugging but hear is the firmware
I looked at a lot of the ATMega128P setups and may be able to offer at least one improvement which is analogPinToChannel(). I expect analogRead(A1) or analogRead(A0+1) or analogRead(1) to all read the same channel and think that can be done using analogPinToChannel(), e.g. on irrigate7 it is
#define analogPinToChannel(p) ( (p) < NUM_ANALOG_INPUTS ? p : ( (p) >= 24 && (p) <= 31 ? (p) - 24 : -1) )