Chris Morgan

Yeah, that’d do it. My understanding from a couple of projects in the past that have spent quite a bit of effort on optimal high-DPI support is that for best...

I don’t recall which precise DLL calls were involved, but that looks about right (though I’ve never seen it done at the thread level, only the process level). I recall...

winit is the primary experience I’ve had of being involved with high-quality and high-compatibility DPI awareness: it started in https://github.com/tomaka/winit/pull/332 for Windows, and ended up landing in https://github.com/tomaka/winit/pull/548. The current...

Things like libpng or whatever equivalent you settle on can safely be optional, too—not everyone needs images in their documents.

It should be the other way round: this covers the entire range of regressions (`%0I`, `%-I`, `%P`, and anything else, frankly), whereas #1469 only covers `%0I` (… and it even...

Actually, case sensitive was correct; the comment is the thing at fault. I quite agree about ensuring that the parsing is done correctly; that’s something that should be fixed at...

Note that this should be an _addition_, sitting on top of what is currently there (though that too will doubtless be changed around quite a bit). I'm not certain where...

(See https://github.com/chris-morgan/rust-http/pull/23#issuecomment-30507061.)

I explicitly did not put it on crates.io and would strongly prefer it not to be put on crates.io and that any instances of it be removed. It’s obsolete and...

Scope question: would this include supporting (regardless of whether they’re included in this crate itself or not) diverse fitting algorithms, e.g. Knuth-Plass instead of the ubiquitous greedy? (Knuth-Plass isn’t suitable...