I’m about to be something of a jerk, and I’d like your help in doing it.
I have noticed a common pattern in a lot of software correctness research, which is that they often follow the following format:
- Oh god software, it’s everywhere.
- And it doesn’t work very well at all, amirite?
- This is only going to get worse.
- We’re basically all doomed.
- Therefore in order to fix this we propose the following.
- We have made an improvement.
- To a category of tool that literally nobody uses.
- Happy to help.
- Love, The Authors.
What are your “favourite” examples of this genre? I’m particularly interested in examples from software testing research but I’ll take anything vaguely correctness related.
The “That nobody actually uses” part of it is also somewhat inessential – I’m mostly interested in good examples of the motivation section, and in many ways having examples from things people actually use is quite helpful because it gives me papers I can point to without feeling too mean about it!
Answers accepted in comments, emails, tweets, whatever.
Edit to add: I do not mean to suggest that such papers are necessarily bad papers. Many of them are interesting and intelligent research.
I’m also not interesting in being mean to individual papers or authors (I originally approved a comment which I have since retracted because I decided it was too far in that direction).
What I want to call into a question is our priorities. We have all of these introductions which claim that what we’re trying to do is solve these real world problems, but if we’re not addressing the fundamental question of why is nobody using our stuff then maybe we need to admit that that’s not actually what we’re trying to do.
As a pure mathematician I’m perfectly on board with a program of pure research, but as a software developer I am much more interested in the question of what a research program that actually cared about these questions would look like.