[RFC] "raw_as_bytes" is bad name because "raw" is legacy name #273

Since I had been inactive for months, I forgot new spec renamed "raw" type as "str".

Naming new option as "raw_as_bytes" is bad idea.
Luckily, I didn't release it yet.

Let's rename the option name, but what name?
It should be short and clear, and based on Python 3 and new spec.

• decode
• decode_str
• str
• raw (raw=True returns bytes, and raw=False returns str/unicode)

For now, I prefer raw. Any opinion?

Or should I stop removing encoding option?
I want support only UTF-8, and I don't want to use encoding name because
I should canonicalize the name. (utf_8, utf8, utf-8, UTF-8, etc...)