On 2010-08-17, at 17:44, P T Withington wrote: > Don's comment from http://www.ietf.org/rfc/rfc1738.txt implies that [] would > not be permissible. > > Quoting from elsewhere in that same spec: > >> Other characters are unsafe because >> gateways and other transport agents are known to sometimes modify >> such characters. These characters are "{", "}", "|", "\", "^", "~", >> "[", "]", and "`". > > So, I think we want (), and whatever Max saw is some browser or something > being hyper-conservative and encoding () even though the spec specifically > says it does not need to.
Later in the spec: > On the other hand, characters that are not required to be encoded > (including alphanumerics) may be encoded within the scheme-specific > part of a URL, as long as they are not being used for a reserved > purpose so, it is within the spec to "over-encode". But surely this is a moot point, because, other than making the URL less legible to a human, when the server/application see the URL it will have been decoded and should just work.
