It says it uses a native language. What language? I hoping python because already have it installed. Anyone run on a pi?
LightBurn uses libraries that are not compatible with the ARM processor used on a Raspberry Pi. So it will not work on a Pi.
How is that possible? What libraries are you referring to? Linux is open source. I’ve never heard of closed libraries. Just curious.
In short, ARM and Intel chips have different processor architectures and instruction sets. This means that you can’t run an application compiled for the Intel architecture on an ARM computer, and you can’t run code compiled for ARM on Intel computers (not natively at least). Lightburn is compiled for Intel so it can be run on PC’s and Mac’s with intel chipsets. It runs in the linux environment on these intel boxes as well. Additionally, things like ARM CPU architecture (Generic, Cortex-A72 or ThunderX2), the implementation of OpenMP threading, the use 32-bit or 64-bit integers, etc… dictate which ARM library to use. The libraries are not necessarily “closed” its just a huge, useless, waste of time to try to support an architecture that really has no relevance in the laser realm.
@paulf8080 - Linux is open source but that doesn’t mean that everything that runs on it is open source. LightBurn, for example, is not open source and it runs on Linux just fine as we provide a pre-compiled binary version.
LightBurn uses some libraries that have open source versions but it has licensed the commercial versions of those libraries, allowing it to be used in a pre-compiled form. In the case of a few of the internal libraries LB uses the commercial, pre-compiled, version of the libraries don’t provide an ARM version. Therefore there’s no way to make an ARM version of LightBurn.
Also, just because the code of something is open source doesn’t mean it will run on ARM. If it’s python, then sure, as it’s an interpreted language. Even then, it’s not a guarantee. But something like LightBurn that’s not using an interpreted language needs to have specific ARM optimizations added to the code for it to compile.
One good example of closed Linux binaries is the official Nvidia drivers. You can install them but they are not open source. Another is actually for the Raspberry Pi itself… you cannot boot the device without a pre-compiled binary blob from Broadcom that runs the boot process through the GPU of all places.
I just was curious. I run the arduino ide on my pi to program avr cnc controllers. I run octopi to run my 3d printer. The is no x86 machine in the spare bedroom where my 3d printer and k40 are. I run my slicer on the PC in another room. Can I export gcode from lightburn on my PC? I think octoprint may work with the smoothie fw. I guess I have no business using a pi to talk to an 8 bit AVR controller.
I’m perfectly ok with people selling binaries. I had never heard of using commercial libraries. I use the gnu tools and the GPL license requires open source. I live in a different world at home. I retired from Intel as logic designer on processor chips. BTW nvinda open source drivers are now in linux distributions.
I should point out that the reason I was disappointed the lightburn didn’t run on the pi is because I was very impressed with the videos. I’m considering buying an extra x86 PC just to run it.
LightBurn is not Python, but C++, compiled to native machine code for each target platform. That makes it less portable, but roughly 100x to 1000x faster than interpreted languages like Python or Java.
The word library is what set me off, I guess. The python library is c++ and python. The CPU intensive number crunching is done in c++ and compiled by pip when added to your local library. I’m leaning towards getting an Intel NUC and installing linux. I’ve had been thinking of setting up a x86 box for a long time. This might put me over the edge.
I have LightBurn running on this:
It’s not a speed demon, but it’s usable.
Python is ‘compiled’, but to bytecode for their VM / runtime environment, not native machine code. LightBurn is written in C++ by a developer with 25+ years in the video game industry - I used to write a lot of engine and artist tool code, so I tend to focus on both speed and usability.
The python language is compiled into bit code. When you include a module the module is written in c++ and the binary is in your local library. I used the numpy module to do image processing and it was fast. Your mini PC is the exact thing I’m looking at. Thanks for the review. There’s hope now. I’m looking at the system76 meerkat with ubuntu linux.
EDIT spelled meerkat wrong. It is on sale which suckered me into ordering one.
This topic was automatically closed 14 days after the last reply. New replies are no longer allowed.