Mahmoud Al-Qudsi

> realpath is not a posix staple. It's not in posix at all. Sorry, I was more explicit later and said "less ergonomic or less well-designed posix or **posixy** features,"...

I think trying to (force a) move from CMake *right now* isn’t necessary - there’s no rush to move away from a dependency we already have (and have already sorted...

Starting with what I said before: we already have CMake and removing it doesn’t especially further the particular interests of the project in the same way that removing C++ from...

> (side note: it would be really cool if we found a rust-replacement for curses - we mostly use it to read terminfo, which as I understand it should be...

Note that displaying errors is possible (with color or what have you) but not "natively" via the completions machinery. See for example the `npm` completions: ```fish function __npm_helper_installed # This...

My previous `aws s3` completions are fully dynamic and query the S3 bucket to suggest bucket and path completions - does aws_completer do that?

The completion callback would need to introspect the arguments for variables that might change the success/failure outcome. I purposely didn't bother with access keys specified at the command line as...

Just to be clear, we're not *removing* the completions for `convert` and co, right?

> You'd have to set up the list of mistakes beforehand, which seems fairly awful. Maybe, but if you take into account that most are going to be infrequent, you're...

+1 mouse tracking in curses: http://tldp.org/HOWTO/NCURSES-Programming-HOWTO/mouse.html