- From: Martin Duerst <duerst@w3.org>
- Date: Wed, 30 Apr 2003 11:50:27 -0400
- To: Bjoern Hoehrmann <derhoermi@gmx.net>, public-iri@w3.org
Hello Bjoern, Are you suggesting that these examples, or similar ones, are added to the draft? Or just checking your understanding of the draft? As it's not clear whether you are actually asking for some changes, I'm not (yet) opening an issue. (see more comments below) At 05:23 03/04/29 +0200, Bjoern Hoehrmann wrote: >Hi, > > Some examples: > >[1] http://www.example.org/~björn (IRI in XML) >[2] http://www.example.org/~bjo%CC%88rn (URI in UTF-8 text/plain) >[3] http://www.example.org/~bjo%CC%88rn (IRI in UTF-8 text/plain) >[4] http://www.example.org/~bjo-rn (IRI in a finctional legacy > encoding where "-" maps to > U+0308) I think I know what you mean with example 4, but may be extremely misleading. A hyphen is always a hyphen, and should not map to "-". If you want to say that the relevant octet maps to U+0308, that's okay, but it may be better to write the original as http://www.example.org/~bjo<2D>rn to show that this is about the octet <2D>, rather than the character '-'. Another problem with that example is that '-' is a valid character in URIs, and using that for something else creates additional problems. Regards, Martin. >These identifiers should be interpreted as or converted to IRIs and then >converted (back to) URIs. > >IRIs: > >[1] http://www.example.org/~bjo<U+0308>rn >[2] http://www.example.org/~bjo<U+0308>rn >[3] http://www.example.org/~bjo%CC%88rn >[4] http://www.example.org/~bj<U+00F6>rn > >URIs: > >[1] http://www.example.org/~bjo%CC%88rn >[2] http://www.example.org/~bjo%CC%88rn >[3] http://www.example.org/~bjo%CC%88rn >[4] http://www.example.org/~bjo%C3%B6rn > >Right?
Received on Wednesday, 30 April 2003 12:09:06 UTC