Arseny Kapoulkine

I think I might have seen this with slightly earlier Mesa versions; you need the absolute latest commit from Mesa repository (or, latest as of a week or two ago...

fwiw I haven't had time to look at the PR yet, and probably won't have a chance in the next couple of weeks - sorry about that. I believe there...

List of CMake issues that would need to be covered by this test: #383 #389 #390 #393

I'm generally in favor of an optional-like interface. This depends on #15 since right now "invalid" state is not detected by pugixml. Also I'm not sure if it makes sense...

The change itself makes sense but it would be nice to apply the same change to xml_node::set_value as well as xml_text::set.

Unsure, likely later this year.

memory_resource_adapter seems a bit too complex - std::pmr::memory_resource seems to guarantee alignment for max_align_t which should be enough to satisfy the alignment requirements. Do we need to spend extra memory...

> I cannot detect this error case from someone deliberately writing &lol9; If you want to perform custom entity expansion yourself, you need to disable `parse_escapes` during parsing. After this...

strconv_escape doesn't have the context pointer of any kind. For it to store the error, it would need to get an extra pointer that needs to be propagated all the...

> Which makes me thing that providing an overload which accepts bool is too wide: it can be any pointer at least. Would be nice to constrain it somehow. This...