error message multiple libraries found esp32 cam

@frogandtoad Here is a screenshot of my platform.txt. Your platform.txt file is probably in the same dir as the esp32 2.0.3 libraries etc. My path on a Mac is likely a little different but this might help. Lot's of -D's

@zander

Posted by: @zander

@frogandtoad I just posted a screen shot. I must admit I was assuming that from the old days and having seen what is in the screen shot. I wasn't trying to open up another rabbit hole, I am trying to clear my mail as I am on the w3ay to the Dentist to get put under so will be AFK most of the day.

 

No worries, I wrote and posted my previous reply without knowing you already provided a response.

Anyway, did a quick search, and it appears to me that they are somehow wrapping the g++ compiler, and posibly providing those custom switches via theor own config / make file etc... as they are not a standard part of g++ options.

Here is the help file I pipped out for the "xtensa-esp32-elf-g++.exe" command line options:

Don't get too hooked up on these custom compiler settings 😉

Cheers

@zander

Posted by: @zander

@frogandtoad Here is a screenshot of my platform.txt. Your platform.txt file is probably in the same dir as the esp32 2.0.3 libraries etc. My path on a Mac is likely a little different but this might help. Lot's of -D's

 

 

No worries, unless we know what all those '-D" options mean, then probably best to ignore them... I can almost guarantee you will solve the error before finding out what they do 😉

@frogandtoad I have an idea for a single line debug that can be turned on/off at compile time with a -D IF it still works as I remember it. I am off to the dentist shortly but will set up a test tomorrow or next day and post back results.

@zander

Posted by: @zander

@frogandtoad I have an idea for a single line debug that can be turned on/off at compile time with a -D IF it still works as I remember it. I am off to the dentist shortly but will set up a test tomorrow or next day and post back results.

No need to take that route with switches when something as simple as this will work:

// Comment (or uncomment) this line to turn our debug statements ON or OFF
// Our debugging statements will only print if this line in un-commented
#define DEBUG

void setup() {
  Serial.begin(9600);

  int value = 42;

  if(42 == value) {
    #ifdef DEBUG
      Serial.println("In the if block with value: " + String(value));
    #endif
   }
  else {
    #ifdef DEBUG
      Serial.println("In the else block with value: " + String(value));
    #endif
  }
 }

void loop() {
  // ...
 }

When you're ready for production code, just comment out the DEBUG statement, and none of the print statements will be compiled into your code, so not adding anything to its final size.

Cheers

Hi @frogandtoad and Ron @zander,

   Re -D (define) on the command line

Apologies for prolonging the agony of thread extension here ... but maybe both of you are spiralling around another curiosity, which may be of interest.

frogandtoad stated "g++ --help" doesn't mention '-D'  ... a point my machine concurs with. Likewise "cpp --help"

.. but a little Google searching suggests that "-D" is supported ... and not just in Expressif land, (so not just depending on 'specials' , which can be a slippery slope.).. e.g. 

• https://www.rapidtables.com/code/linux/gcc/gcc-d.html
• https://gcc.gnu.org/onlinedocs/gcc/Preprocessor-Options.html

The first reference includes helpful examples, the second I presume to be a 'reasonably definitive' source, given the pervasive influence of gnu.

I haven't tested, or ever used, such tricks, so maybe I am misunderstanding or misled, but the suggestion is that gcc/g++ has tricks up its sleeve that '--help' doesn't mention!

Best wishes to all. Dave

Posted by: @frogandtoad

@zander

Posted by: @zander

@frogandtoad I have an idea for a single line debug that can be turned on/off at compile time with a -D IF it still works as I remember it. I am off to the dentist shortly but will set up a test tomorrow or next day and post back results.

No need to take that route with switches when something as simple as this will work:

// Comment (or uncomment) this line to turn our debug statements ON or OFF
// Our debugging statements will only print if this line in un-commented
#define DEBUG

void setup() {
  Serial.begin(9600);

  int value = 42;

  if(42 == value) {
    #ifdef DEBUG
      Serial.println("In the if block with value: " + String(value));
    #endif
   }
  else {
    #ifdef DEBUG
      Serial.println("In the else block with value: " + String(value));
    #endif
  }
 }

void loop() {
  // ...
 }

When you're ready for production code, just comment out the DEBUG statement, and none of the print statements will be compiled into your code, so not adding anything to its final size.

Cheers

The difference is the production coder van bemodified at the compilr lrbrl and can unde the wright coditions eliminate the debug code. Sorro recoeverimg fronzFrntist

@frogandtoad True, but what if severak mudules are invovled. 

@frogandtoad Certainlt been done many/most times, give me a dy or two to show a differenrt way

@davee And simple operation shoe extensive sue

@davee

Posted by: @davee

Hi @frogandtoad and Ron @zander,

   Re -D (define) on the command line

Apologies for prolonging the agony of thread extension here ... but maybe both of you are spiralling around another curiosity, which may be of interest.

frogandtoad stated "g++ --help" doesn't mention '-D'  ... a point my machine concurs with. Likewise "cpp --help"

.. but a little Google searching suggests that "-D" is supported ... and not just in Expressif land, (so not just depending on 'specials' , which can be a slippery slope.).. e.g. 

• https://www.rapidtables.com/code/linux/gcc/gcc-d.html
• https://gcc.gnu.org/onlinedocs/gcc/Preprocessor-Options.html

The first reference includes helpful examples, the second I presume to be a 'reasonably definitive' source, given the pervasive influence of gnu.

I haven't tested, or ever used, such tricks, so maybe I am misunderstanding or misled, but the suggestion is that gcc/g++ has tricks up its sleeve that '--help' doesn't mention!

Best wishes to all. Dave

Apologies for prolonging the agony of thread extension here ... but maybe both of you are spiralling around another curiosity, which may be of interest.

Indeed, I navigated to the xtensa-esp32-elf-g++.exe (as well as gcc, g++ and c++ versions) directory based on Ron's image, to check the help file at the command line, but even that did not provide the hidden flags... quite interesting!

Thank's Dave, for your excellent researching!

@zander

Posted by: @zander

@davee And simple operation shoe extensive sue

Ron... looks like they were searching for a golden hen's tooth!

I'll have of what you're having 🤣 

Hi @frogandtoad and @zander,

re:

Posted by: @zander

@davee And simple operation shoe extensive sue

   Sorry Ron, I have come back and looked at this several times ... as someone who makes enough typos of his own ... but I still haven't been able to decypher or decrypt this one.

But, unlike frogandtoad, if it has cause, then I am not tempted to go anywhere near it. 😲 

Take great care of yourselves, my friends. 😀 

@frogandtoad @davee LOL, I was put under for some dental work yesterday and I must have tried to type something while still groggy. I have no idea either.