Re: [apps-discuss] Fwd: I-D Action: draft-yevstifeyev-ftp-uri-scheme-04.txt

["Martin J. Dürst" <[e-mail]>]

On 2011/07/16 10:24, Frank Ellermann wrote:
> On 15 July 2011 06:02, "Martin J. Dürst" wrote:
>
>> Please read the text after the
>>     reserved    = gen-delims / sub-delims
>> ABNF rule in RFC 3986.
>
> Okay, but it doesn't explain the "lost"<national>  in RFC 1738:
>   national = "{" | "}" | "|" | "\" | "^" | "~" | "[" | "]" | "`"

At that time, various versions of ISO 646 were still in use (or at least 
feared to still be in use) which would have meant that these codepoints 
would have shown up as different glyphs in some environments (e.g. 
äöüÄÖÜß in the German version of ISO 646). As time moved on, this kind 
of concern grew smaller and smaller.


> This "forbidden" set was later reduced to<unwise>  in RFC 2396:
>   unwise   = "{" | "}" | "|" | "\" | "^" | "[" | "]" | "`"
>
> Skipping RFC 2732 introducing IPv6 literals using "[" and "]"
> RFC 3986 does not explain that some unencoded US-ASCII char.s
> are never permitted in URIs: "{" | "}" | "|" | "\" | "^" | "`"

Well, yes, it's the spec for URIs, not for "what's not allowed in URIs".


> And the "new" unencoded<gen-delims>  "[" / "]" are only allowed
> for "IPvNot4" literals in the<host>  part of the<authority>,
> no matter what any specific URI schemes such as RFC 6068 or
> specific fragment specifications such as RFC 5147 might wish.

That "[" / "]" can only be used for IPv6 addresses and the like is 
definitely somewhat inconsistent. If there ever were an update of RFC 
3986, I can imagine that to be changed (but that's only a guess).

> Unrelated, it is rather annoying that the IE and Chrome folks
> have not yet implemented RFC 5147, this nice feature should
> take the proverbial "five minutes" for a developer.

Does that mean that it's implemented in Firefox and Opera? That would be 
great to hear. For Chrome, you could submit a bug or patch to webkit, I 
guess.

Regards,    Martin.