Evan Buswell said: > I'm really not saying something very profound here, so I'm a bit > confused by the sarcastic response. For certain things it's pointless > and inefficient to parse something and then deparse it later. It's not > like you're gonna put UTF-8 parsing into cat.
This brings you into encoding detection and either assumptions or guessing. I just can't believe you're willingly dragging yourself into it. I really don't want some special support for utf-8 in cat, because my local files generally have the same encoding with my terminal, and I won't lose much resources when I have to cope with differently encoded files. But it has nothing to do with network-interfacing format, where any spec ambiguity directly translates into either loss of resources (bandwidth, time, processing power) or loss of clients. That's not to mention that you may have another encoding specified elsewhere in your delivery chain (eg. in http headers). -- Dmitrij D. Czarkoff
