I'm working on a program that will involve storing struct instances in data files, and so far have a working process of serialization, writing, reading, and de-serialization directly manipulating the Vec<u8> buffers (I understand serde is a fantastic crate, but I'm avoiding external dependencies until it's unreasonable to do so). It works in a clean, local environment (Linux), however I intend to compile to both Linux and Windows x64 targets.
My question is if there are any issues I'm not seeing now that will likely come up if this is compiled and released into the wild. Things such as ease of corruption, modification of the data by the file system, compatibility, etc., that I'm blindly walking into by using a simple struct -> buffer -> file (and then reverse) process. I don't necessarily need to know how to solve all potential issues right now, I'm just unfamiliar enough with low-level file handling that I thought it was worth a check to see if I need to dig into it or adjust course now rather than later.
2 posts - 2 participants
