Giters

PetoiCamp / OpenCat

An open source quadruped robot pet framework for developing Boston Dynamics-style four-legged robots that are perfect for STEM, coding & robotics education, IoT robotics applications, AI-enhanced robotics application services, research, and DIY robotics kit development.

Home Page:https://www.petoi.com

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

Nybble - Adding new skill: need to move zero from EEPROM to PROGMEM

Troy-Chard opened this issue · comments

commented

Not sure if this is a bug or not, but thought I'd mention and let you decide :)

Last night, while trying to write custom zero skill for Nybble, it didn't seem to work.

The fix was easy: in my local branch, inside InstinctNybble.h:

  1. I moved zero to progmemPointer[] array.
  2. Revised it to "zeroN" in skillNameWithType array.
  3. Re-ran WriteInstinct.ino to update the skills
  4. Ran and uploaded my OpenCat.ino copy.

And to gain a bit more space in Program memory I commented out hs, hs1, hs2 skills.

Now, I can just tweak the zero skill and upload the code... no need to re-run WriteInstinct. But I needed it the first time just to move zero out of EEPROM and into Program memory (if I understand right).

commented
commented

It could be what's in the repo is correct as is :)

I just noticed that, before I could use the zero skill myself to try a new behaviour, I needed to add an entry to the progmemPointer array in InstinctNybble.h.

( I saw the progmemPointer already has it in InstinctBittle.h )

Also I revised the name in the skillNameWithType array to be "ZeroN" instead of "ZeroI" ( following the example in InstinctBittle.h )

( But then OpenCat.ino got too big, so I commented out hs, hs1, hs2, entries ).

After editing my skill a few times and uploading, it didn't seem to work though (it just kept doing similar to calib pose iirc).

Then I realized I needed to run WriteInstinct.ino to rewrite the EEPROM with new settings moving zero from EEPROM to PROGMEM 💡 and then my zero Newbility worked :)

The new entries in my copy here appear as:

#if !defined(MAIN_SKETCH) || !defined(I2C_EEPROM)
		//if it's not the main sketch to save data or there's no external EEPROM, 
		//the list should always contain all information.
  const char* skillNameWithType[]={"bdI","biI","bkI","bkLI","bkRI","crFI","crLI","crRI","lyI","trFI","trLI","trRI","vtI","wkFI","wkLI","wkRI","balanceI","buttUpI","calibI","droppedI",/*"hsN","hs1N","hs2N",*/"liftedI","luI","restI","sitI","sleepI","strI","zeroN","hiI","peeI","puI","rcI",};
  const char* progmemPointer[] = {bd, bi, bk, bkL, bkR, crF, crL, crR, ly, trF, trL, trR, vt, wkF, wkL, wkR, balance, buttUp, calib, dropped, /*hs, hs1, hs2,*/ lifted, lu, rest, sit, sleep, str, zero, hi, pee, pu, rc, };
#else	//only need to know the pointers to newbilities, because the intuitions have been saved onto external EEPROM,
	//while the newbilities on progmem are assigned to new addresses
  const char* progmemPointer[] = {/*hs, hs1, hs2,*/ zero, };
#endif

( and this is just me fooling around trying my first newbility... really I should maybe be writing my own .ino. eventually :P

commented

( more background ... )

My Newbility is sad so far though...

const char zero[] PROGMEM = {  // turtle roll?   ( it's not even close yet :(
-5, 0, 0, 1,
 0, 0, 0,
  -10,-100,   5,   0,   0,   0,   0,   0,  80,  80, -80, -80, -55, -55,  55,  55,   8, 10, 0,0, // tuck
    0,  10,   5,   0,   0,   0,   0,   0,  84,  84,  80,  80,  90,  90, -90, -90,   2, 4,  0,0,  // arch
   70,  10,   5,   0,   0,   0,   0,   0, -90,   0,   0,  90,  90,  90, -90, -90,   9, 2,  0,0,  // roll - speed 100 should cause movement in one step, but, wow, hard on the servos :( 
  -70,  40,   5,   0,   0,   0,   0,   0,   0, -90,  90,   0,  90,  90, -90, -90,  30, 6,  0,0,  // Flip.A      
    0,   0,   0,   0,   0,   0,   0,   0,  30,  30, -30, -30,  30,  30, -30, -30,   2,20,  0,0, // stand?  
};

But my Nybble has protective gear to handle testing which I can peel off later pretty easy :)

image

image

commented
commented

Thanks borntoleave! :)

But, I found slight problem with latest - sketch too big ( by 32 bytes 😮 )

image

For now, easiest thing for me was to comment out the hs skills:
image

image

( I'm going to upload to Nybble later tonight and give latest skills a try :)

commented

Have you disabled the unused encoders in IRremote.h? It's about 85% after compiling on my side.

commented

I'll try that, thanks!

The problem I noticed with sketch size is in the current main branch of OpenCat :(
I just don't want people trying Nybble for the first time to hit the problem. ( sorry for the troubles my change request is causing 😮 )

commented
commented

Thanks borntoleave, that's great :)

commented

Quick follow-up... 85% is what I get too. Thanks again borntoleave :)

image