Closed main-- closed 9 years ago
We actually don't need this I guess. Why should we? Parsing is cheap, demos are small. Non-issue.
Back then, parsing wasn't cheap, but fortunately we solved that one.
Still, the issue refers to functionality as well: Have we considered possible interactions between the two simultaneous analyses carefully? They'll probably overwrite each other's work all the time, but can this result in corrupt data? (And when we still saved the demo files, things were even funnier: We identified them by md5, but how would we handle two identical files?)
All of these are not necessarily failures or issues, just things we need to look at and keep in mind.
Cases we're handling / need to handle: