David Machaj

Results 15 comments of David Machaj

I tried to cover that suggestion with this portion: `In some cases all access to class fields happen before the first suspension point, so these accesses are safe. However, it...

That seems fraught with peril. If you want the outer object to be able to destruct then it is probably best to have the coroutine _not_ be a class method....

@hsutter this is distinct from the one about reference parameters because the concern is with the `this` pointer. If the lifetime of the current class instance is not ensured then...

(My changes to BinSkim are still in testing. I fixed a bug and it produced an additional warning that should have been in the original bug report. I added it...

I am doing some additional testing of my BinSkim change so I had the linker command-line for these binaries handy. Curiously some of the arguments are duplicated. I don't think...

My immediate problem is that installing the drivers and integration tools will show the "Do you trust this publisher?" UI for Red Hat, even when running the installers in `/quiet`...

Nevermind, I now see that there is natvis support. I just didn't see an issue for it so I assumed it was missing. It is here: https://github.com/biojppm/rapidyaml/blob/81b2685f069c17e75f975008e38ba699b7eab252/src/ryml.natvis#L8.

I got the natvis wired up in my project. Unfortuantely the support seems incomplete. For example, I am using `ryml::ConstNodeRef` for a readonly YML file. The [existing natvis](https://github.com/biojppm/rapidyaml/blob/81b2685f069c17e75f975008e38ba699b7eab252/src/ryml.natvis#L8) has `ryml::NodeRef`...

That's fair. If I do decide to take this on it would be very helpful if there is some sort of overview of the primary ryml types (perhaps a comment...

I tried poking at this a bit today. The existing natvis provides partial coverage. There is a missing set of parenthesis that is preventing [children] from working on NodeRef. Also,...