Error while trying to install a micropython distribution package using upip

[sara]

Hello,
I am trying to use upip to install micropython-tinyrtc-i2c distribution package and it gave me the following error:

Code: Select all

Error installing 'micropython-tinyrtc-i2c': list index out of range, packages may be partially installed.

I'd like to know how to solve it!
Thanks in advance!

[pythoncoder]

upip installs libraries located on PyPi (https://pypi.org/) or in a special repository for the MicroPython library. The micropython-tinyrtc-i2c package doesn't seem to exist on PyPi. To install it you need to copy the program files from the GitHub archive to your device. To clone the GitHub library to your PC issue

Code: Select all

git clone https://github.com/mcauser/micropython-tinyrtc-i2c.git

From your PC you can then copy the Python files to your hardware using whatever IDE you are using or a command line tool like rshell.

[sara]

Hum... I see, thanks!
So, I am gonna try to use mpy-cross to compile the .py file to .mpy file and then I'll use the ampy (https://github.com/scientifichackers/ampy) to send it to the board!
best regards,
Sara

[Roberthh]

Unless you run into an memory error on running your code, you do not need to precompile it. Just copy the python files to your board.

[sara]

ok, but, only for curiosity, answer me a doubt please:
Doesn't this procedure optimize the memory usage?
best regards,
Sara

[pythoncoder]

When you import a .py file the compiler runs on the target. This consumes RAM and may fail if the file is too big relative to the target's RAM. Cross-compiling removes the need for the compiler to run on the target and avoids this error.

The RAM used by the compiler should be reclaimed by the garbage collector once the compiler has run. So, if the import proceeds without error, the runtime benefits of cross-compiling are doubtful. There may be a reduction in RAM fragmentation.

Note that in both these cases the Python bytecode produced by the compiler exists in RAM. This can be avoided by using frozen bytecode where the bytecode is located in flash as part of a firmware build. This can save a lot of RAM at runtime.