- From: Andy Seaborne <andy.seaborne@epimorphics.com>
- Date: Mon, 29 Aug 2011 18:20:58 +0100
- To: Richard Cyganiak <richard@cyganiak.de>
- CC: public-rdf-wg@w3.org
On 29/08/11 16:17, Richard Cyganiak wrote: > On 29 Aug 2011, at 13:26, Andy Seaborne wrote: >> The content type suggested is: >> >> text/ntriples+turtle >> >> but text defaults to ASCII. > > It defaults to ISO-8859-1 to be precise. it is all a bit messy: http://www.rfc-editor.org/rfc/rfc2046.txt section 4.1.2: """ The default character set, which must be assumed in the absence of a charset parameter, is US-ASCII. """ but the example is ISO-8859-1 ISO-8859-1 and UTF-8 diverge. and as you note text/html: http://www.rfc-editor.org/rfc/rfc2854.txt [text/html] section 6 discusses the mix up > > So it would need the same handling as Turtle and N3, which require “;charset=utf-8” appended to the media type. > > There is ongoing work in IETF to change the default for text/*: > http://tools.ietf.org/html/draft-melnikov-mime-default-charset-01 > > But I don't know anything about the status of that work, or how likely it is to become official, or at what timeframe. > >> N-triples is less readable than Turtle and more directed to machine processing. >> >> application/n-triples+turtle >> >> seems more appropriate and allows the default charset to be UTF-8. > > RFC 2046 has this to say about the text/* types: > > [[ > In the absence of appropriate interpretation software, it is reasonable to show subtypes of "text" to the user, while it is not reasonable to do so with most nontextual data. > ]] > > While it says this about application/*: > > [[ > This is information which must be processed by an application before it is viewable or usable by a user. > ]] > > Also, application/* subtypes unknown to an implementation MUST be treated as binary data. > > In practical terms, opening text/* in a browser causes it to be displayed, while opening application/* causes it to be downloaded. (At least in my browser of choice; I don't know how others behave.) > > There's mixed precedent. JSON uses application/*, HTML uses text/*, CSV uses text/*. :-| > > Best, > Richard > > >> >> Andy >> >> On 24/08/11 00:52, Gavin Carothers wrote: >>> Hey folks, >>> >>> Have added an initial draft of N-Triples Prime (name to be agreed on >>> later) to a branch of the Turtle specification. >>> >>> http://dvcs.w3.org/hg/rdf/raw-file/n-triples/rdf-turtle/index.html >>> >>> Branch is named "n-triples". >>> >>> Issues: >>> The RDF WG has not settled on a name for N′, N′ is used to distinguish >>> it from the RDF Test Cases N-Triples. ( Should I say "N\u2032" ) >>> The RDF WG has specified N-Triples Prime to allow UTF-8 characters in >>> IRIs, literals and blank node identifiers. Readers with an opinion >>> about whether or not N-Triples should be ASCII-only may wish to >>> comment. (huh, that is poor language and not correct copied from old >>> issue, will fix later) >>> Current grammar doesn't deal with comments correctly. (For any value of correct) >>> Current grammar doesn't deal with triple lines that end in EOF rather >>> then EOL. (neither did the old one) >>> >>> Cheers, >>> Gavin >>> >> >
Received on Monday, 29 August 2011 17:21:29 UTC