Mark Skilbeck
Mark Skilbeck
`counsel-rg` searches files for occurrences of a regexp, and then, upon selecting a particular occurrence, jumps to the appropriate location in the matching file. This is broken and jumps to...
Use posframes for overlays. There does seem to be a slight but noticable delay when displaying the posframes. Haven't investigated too much into why.
Very much a work-in-progress. Making it visible here for early commentary so that I don't go too far down a wrong path. A big sticking point is the workspace stuff...
Since quilc lives under Quil-Lang now I don't believe it should be calling-home to a Rigetti-owned location. We might also think about if/how we want to move SDK packaging and...
In #702 I changed quilc to load libraries at run-time, rather than baking-in the paths to those libraries at compile-time. The benefit of that is that libraries can exist in...
`chip-contiguous-subschedule-from-last-instructions` is recursive and, given a sufficiently deep program, can fail with an exhausted function call stack error. This can be alleviated by making the function non-recursive.
> I don't think you're forgetting anything; I think this was an accidental consequence of both the `cost-function` changes and accomplishing the original by uncoordinatedly tweaking things 'til they worked....
The "NAIVE" initial rewiring strategy tries to avoid rewiring qubits, but will if there is a 2Q instruction that cannot be enacted on the provided qubits: ``` QUIL> (print-parsed-program (compiler-hook...
See https://github.com/rigetti/quilc/pull/649/files#r452514653 where it is (now) hardcoded to use the program fidelity as the optimization strategy. Compare that with the addresser, which allows for either fidelity or duration to be...
[Eric says](https://github.com/rigetti/quilc/pull/649/files#r452563184) > It isn't hurting anyone to include this, but I'd expect the peephole rewriter to figure this out on this on its own. > Rather, what I mean...