MicroPython on ESP32 with SPIRAM support
Re: MicroPython on ESP32 with SPIRAM support
@on4aa
Thank you for your comment an suggestion.
I'll try to collect all the licensing information for used third-party componnents/libraries and also add the licensing information in all of my source code where it is missing.
As far as I know, all the code used in this repository is open source, but not all with MIT license.
The one exception I'm aware of is the part of the esp-idf (low level WiFi/BT drivers) which is closed source.
It is an "issue" with all esp-idf based application.
The only solution, if you want to make fully open source application (in this case the MicroPython firmware), is to not include/link the binary blobs in the firmware (→ Component config → ESP32-specific → No Binary Blobs) in which case Wi-Fi/Bluetooth will not work.
If you are aware of any specific licensing issue, please feel free to report here or on GitHub repository.
Thank you for your comment an suggestion.
I'll try to collect all the licensing information for used third-party componnents/libraries and also add the licensing information in all of my source code where it is missing.
As far as I know, all the code used in this repository is open source, but not all with MIT license.
The one exception I'm aware of is the part of the esp-idf (low level WiFi/BT drivers) which is closed source.
It is an "issue" with all esp-idf based application.
The only solution, if you want to make fully open source application (in this case the MicroPython firmware), is to not include/link the binary blobs in the firmware (→ Component config → ESP32-specific → No Binary Blobs) in which case Wi-Fi/Bluetooth will not work.
If you are aware of any specific licensing issue, please feel free to report here or on GitHub repository.
Re: MicroPython on ESP32 with SPIRAM support
@loboris
Thank you for your clarification. This inspires a lot of confidence.
As you suggest, it would indeed be wise to list the items that are not licensed under the MIT terms in a file in the root, together with their license type.
This would allow users who intend on using your software in a commercial setting, to take appropriate actions: e.g. include the GNU GPL license and publish the GNU GPL affected source code.
Thank you for your clarification. This inspires a lot of confidence.
As you suggest, it would indeed be wise to list the items that are not licensed under the MIT terms in a file in the root, together with their license type.
This would allow users who intend on using your software in a commercial setting, to take appropriate actions: e.g. include the GNU GPL license and publish the GNU GPL affected source code.
Serge
Re: MicroPython on ESP32 with SPIRAM support
loboris, looks like the RMT can be tickled to do input capture. Any chance that might be on your development timeline?
Re: MicroPython on ESP32 with SPIRAM support
Does Loboris port include GPL code somewhere?
Re: MicroPython on ESP32 with SPIRAM support
Answering to myself here, yes it seems it does. Thanks to viral nature of GPL my understanding is this makes the whole project effectively GPL licensed.
Re: MicroPython on ESP32 with SPIRAM support
RMT is used by the get_pulses() method of the pycom branch to capture input pulses.
Re: MicroPython on ESP32 with SPIRAM support
Thanks loboris!
Re: MicroPython on ESP32 with SPIRAM support
By consequence, M5Stack is affected as well.tuupola wrote: ↑Sun Jan 28, 2018 6:09 amAnswering to myself here, yes it seems it does. Thanks to viral nature of GPL my understanding is this makes the whole project effectively GPL licensed.
Serge
Re: MicroPython on ESP32 with SPIRAM support
Looking into the GPL search @tuupola provided there are three areas of concern:
I would also strongly prefer the MIT license - at least for compatibility with upstream Micropython (let alone that GPL is near-impossible to introduce to my professional environment). If @loboris agrees to this goal - and it's his code, he can do as he pleases! - then I believe these files need to be removed from the repository.
1 should be relatively easy, it appears to be a random number generator. 2 is libquickmail, a C library to allow sending of emails. Although large, libcurl could potentially be used as a replacement (?). 3 is interesting; it looks to be an implementation of HW I2C for the ESP32 by Pycom. This would need to be rewritten though, presumably, it is wrapping up the I2C primitives provided by the ESP-IDF. In the meantime, software I2C could be used.
Some of the above could be wrong; I've been particularly careful to avoid looking at the code in detail.
- random.h/.c
- quickmail
- machine_hw_i2c.c
I would also strongly prefer the MIT license - at least for compatibility with upstream Micropython (let alone that GPL is near-impossible to introduce to my professional environment). If @loboris agrees to this goal - and it's his code, he can do as he pleases! - then I believe these files need to be removed from the repository.
1 should be relatively easy, it appears to be a random number generator. 2 is libquickmail, a C library to allow sending of emails. Although large, libcurl could potentially be used as a replacement (?). 3 is interesting; it looks to be an implementation of HW I2C for the ESP32 by Pycom. This would need to be rewritten though, presumably, it is wrapping up the I2C primitives provided by the ESP-IDF. In the meantime, software I2C could be used.
Some of the above could be wrong; I've been particularly careful to avoid looking at the code in detail.