We’ve run out of space (132kb) and while we are investigating memory saving coding techniques, this is a battle we’ve been facing for some time and I do think that for the long term health of my project that we explore a long term option fo maybe using a chip that has more than 132kb of usable application space. I noticed that the device-os project has peripheral files for STM32F4 (Cortex M4) chips and some of these have 2mb of flash instead of one. I’m more of a web engineer so I’m curious on others opinions how hard it would be to compile the particle OS with more space allocated to application space (on a larger chip) and flash onto something like a Cortex 4M while maintaining things like OTA, etc.
Even if you could (not saying you can’t), that chip would not be granted cloud access as it is not a Particle device.
I agree that refactoring applications to save space is a constant pain with the 132Kb. How about keeping the P0 for the Cloud comms and OTA for the majority of your application but transferring part of the data processing onto a co-processor which has more RAM and FLASH but just focuses on managing a sensor? I have done this by making some sensors “smart” with a small 8-bit ATMEL MCU and also others which have CV applications where they have a built in M4.