Matthew Fernandez

Apologies for the late response. It seems none of us were actually on the notify list for this repository. You are referring to the leading underscore, I take it. Yes,...

seL4/seL4#2 relates to headers inside the kernel. These are clearly part of the implementation, as @lsf37 has said, so underscore-prefixed identifiers are allowed in this context. The headers referenced in...

In the case of generated headers, I agree that something more thorough is required, but for these static headers I think what's there is adequate. Even for generated headers though,...

This repository itself doesn't need any guard prefixes, as everything in here is an application. That is to say, any author of a header here knows exactly what will and...

Instead of removing the reserved prefixes, these guards should be replaced with `#pragma once`. This could be done in the generated header as well. If you want to submit a...

Surprisingly an intermediate point in some database normalisation work (5e1c86b655c860512f8a37a180800c34fcb80069) is already showing a 77% speed up in database construction. If we continue on this trajectory, going to an amalgamation...

Can you give an example of something other than Cscope itself that consumes the Cscope database format? Its format has a number of shortcomings which directly motivated _not_ reusing the...

It's certainly possible. A Python script using pycscope would probably be reasonably straightforward for this.

I guess this was resolved out-of-band. Please re-open if you’re still having problems.

IIUC this actually affects all non-ASCII unicode. Or maybe the following is a separate problem? ```sh # under X xdotool key U201c # types “ # under Wayland ydotool key...