There's tons of related work for developing unsafe code guidelines that we should all be familiarizing ourselves with. To that end, I wrote up a summary of a paper I just read, on the intptrcast model:
I will try to do the same for other papers. I encourage others to do so as well. Where applicable, I think the structure should be roughly similar to what I did in the issue above:
- Summarize in a few sentences the key idea
- Give a series of examples from the paper (or not)
- For each example, identify the key question at play, and give a concise answer of why the model answers one way or the other
But basically the idea is to give people enough background to quickly get the basic idea. I wouldn't try to summarize every detail in the paper, people can (and should) go read the PDF for themselves to learn that. (Often, that is much easier when you have a basic idea of the contents going in.)