# Strategy to stream large gcode file from serial port

I'm connecting to a printer that's using Marlin 1.0.3 dev firmware through the pyserial Python package.

I want to stream large gcode files into the printer and I guess that the buffer has a limited size. How do I know when the buffer is full, so I can wait until a command is processed to send a new one?

I've tried reading a big file, sending all the commands and checking for the printer responses in another thread. I didn't get an error or something that indicates that the buffer is full. I've also didn't found anything about this in the official docs.

It's my approach correct? Is there something I'm missing?

Some of the code:

def check_for_responses():
while(True):
print(response)

async def stream_gcode():
await printer_connection.open_connection('COM7')
asyncio.sleep(2) # wait to initialize
file = open("test.gcode", 'r')
for line in file:
command= printer_connection.format_command(line)
printer_connection.serial.write(command.encode())

asyncio.new_event_loop().run_until_complete(stream_gcode())

• Please explain how this is relevant to 3D printing. Your code excerp is too tiny and your explanation does throw up more questions for me: What is the problem, what is your expected behavior, where does it throw errors? Nov 17, 2018 at 14:48
• I simply need to figure out a strategy to stream commands to the printer. I've added the code just as an example, but my question is more general. I don't think that reading the entire file and just sending it to the printer would be a good idea. Neither sending the commands one by one waiting for the response of each. Nov 17, 2018 at 15:27
• that... seems like Stack Overflow stuff. Programming the interface from 0 I mean. we are usually not programming interfaces here, we just use them... Nov 17, 2018 at 15:36
• Ok, I will remove the question and ask there then. I thought this was the correct place. Anyway, do you know how the printer buffer works or where to find information about it? Nov 17, 2018 at 16:23
• The question in itself is ok, but we need more heft to be able to answer it here. You could though get faster help on SO about programming Nov 17, 2018 at 19:47

Well, after some research I found something about Buffering on the RepRap documentation. I didn't see it before.

Works like this:

1. You send a command. An "ok" response it's received immediatly, that means that the command was successsfuly added to the buffer.
2. Then you can send another command. When the buffer is full, the "ok" response will be delayed until a new command it's received.

Some commands, for example M190: Wait for bed temperature to reach target temp will cause that the "ok" response will be delayed until the command finish his execution. That means that the printer won't accept more commands in the meantime.

I still have to check how many commands the buffer can save approximately, and if it's convenient to fill it entirely. I guess not, because let's say you want to pause the print. If you added a lot of commands to the buffer, you'll not able to pause it until all of them finished.

This thread on the RepRap forum was also useful.

• Have you figured out how to tell how many the buffer currently holds? I'm trying to do the same but I'm not sure how to synchronise this. Thanks! Aug 27, 2021 at 20:51