if anyone has time for a code review today, can you look at https://github.com/osquery/osquery/pull/5732/files this seems like a half-measure but it does seem to work.

i'm not at a real computer, and can't check the sqlite docs. i agree this looks kinda janky, but is probably fine. the sqlite thing i want to check is about how virtual tables are invoked. i think there's a create vs call convention. we don't use that, and i wonder if the source of this bug is the lack of utilization of the create routine.

I also think the use of the static variable indicating 'first config' can be replaced with a check if the view exists or not.

yeah, i agree there's probably a cleaner if-view-exists. getting something into CI would let us refactor that

I don't know if this is related or not but I've also recently discovered this https://github.com/osquery/osquery/issues/5714