Julian Hall
> Alternatively, we could rename the base class method `addCol`, to be similar to `addRow`, which implies low-level functionality. ... but this is tempting. There already is `Highs::addCol`, and `Highs::addVar`...
Removing `Highs::addVar` will break the API, and annoy people. However, since it can't be called from `highspy`, it might as well be removed from `highs_bindings.cpp`. We just have to make...
Closed by #1753
Sorry about this: I didn't think sufficiently when I introduced this value to HighsInfo. I don't have the time to write accessor functions - whatever they are - and we...
So, I've moved `pdlp_iteration_count` to last data member of `HighsCallbackDataOut`. If I understand correctly, inserting `pdlp_iteration_count` into `HighsInfoStruct` is OK because the C API has `Highs_getIntInfoValue` (etc) to get info...
I'm not a C programmer, so it's good to know that the struct isn't enough. I guess you might be using offsets from a pointer to it. You can't call...
Accessor functions for the callback struct members would be possible, as they access populated data members. What I need to handle with an immediate error return are calls to most...
Keeping this open as a reminder to write the access functions
An accessor function `Highs_getCallbackDataOutItem` for the C API has been written, but there is a strange set of CI failures in #1748. It seemed like a good idea to introduce...