Analysis APIs and support for reexecution of altered code

[Andarist]

In the process of debugging I've attempted various changes to the code to see how would they affect the app - I just didn't bother to use Replay at that stage, rerecording, losing my breakpoints etc would be a huge pain.

I feel that maybe something could be done about having to re-record replays but I totally understand that there are a lot of challenges associated with it.

---

I've also been hunting down a bug in a very generic code that runs for a lot of things but I was interested in specific invocations, related to a particular component. Hunting down stuff like this ain't ever easy - but Replay didn't make it that much easier to tackle this. I wonder if maybe it would be possible to expand on querying/filtering calls somehow. We are still bound to writing conditions based on the current closure - but with Replay maybe that limitation could be somehow lifted? You have all the info required to make it happen - but ofc how one would specify more robust conditions (inspecting a frame up on the stack or something) is also a challenging problem.

Cross-referencing call sites would be awesome.