There are plenty of literature on this grand question which keeps a researcher afloat. For example see Michael Mitzenmacher’s blog or Jason Eisner’s detailed description. However, reading and summarizing paper is lot of task and also, while talking about some result, one has to remember things. It may not be possible to remember the whole content of a paper, even not the proofs or any other detail. So, then, I guess it is better to remember only 3 grand questions on any paper:
- The model.
- The main result, forget about the proof and its correctness. If it is a published material, it has already been reviewed by many people, whom you can trust.
- The shortcomings, this is important if one is interested in working further in problems. And rest assured, EVERY paper has shortcomings even though they claim not to be.
It really makes the job concise, doesn’t it?
Conflict and Cooperate 