- From: Chris Wendt <christw@microsoft.com>
- Date: Fri, 22 Aug 1997 11:58:38 -0700
- To: "E. Stephen Mack" <estephen@emf.net>, "Martin J. Duerst" <mduerst@ifi.unizh.ch>, "Jonathan Rosenne" <rosenne@NetVision.net.il>
- Cc: <www-html@w3.org>
>Side issue -- given the hugeness of the Unicode character set >and the convenience of named entities, I predict that future >versions of HTML may add new named entities. Perhaps the I don't really see any convenience in named entities altogether and certainly not in extending the range further than the current state. This is because I don't believe the future of HTML editing lies with vi/notepad/<your favorite plain text editor>. And even in the case where author uses a plain text editor to create non-system-native or keyboard-unreachable characters, the Unicode reference is not that far to look it up. I haven't seen any proof for the added "convenience" of adding more and more named entities. In the contrary this slows down the parser (you need to search a huge table) and transmission size (each character takes more bytes than any other known document encoding). This is my private opinion, not a position of Microsoft. BTW I am following up on the original issue of this thread. -----Original Message----- From: Jonathan Rosenne <rosenne@NetVision.net.il> To: E. Stephen Mack <estephen@emf.net>; Martin J. Duerst <mduerst@ifi.unizh.ch> Cc: www-html@w3.org <www-html@w3.org> Date: Friday, August 22, 1997 10:44 AM Subject: Re: Disturbing IE 4.0pp2 behavior for lang="en" At 03:59 22/08/97 -0700, E. Stephen Mack wrote: >Side issue -- given the hugeness of the Unicode character set >and the convenience of named entities, I predict that future >versions of HTML may add new named entities. Perhaps the >HTML 4.0 spec could add a section telling user agents how to >treat unrecognized named entities. > >If HTML 5.0 introduces the &foo; named entity, how should a >pre-HTML 5.0 browser treat &foo; if it sees it? Displaying >the literal sequence "&foo;" strikes me as a less-than optimal >solution. The solution I recommend is a way to include external files with entity definitions. Thus the entity names need not be part of the HTML specification. There may be some standard entity name sets defined for some lnguages. At 14:46 22/08/97 +0200, Martin J. Dürst wrote: >That mechanism is almost in place (as far as characters are >concerned; not for the glyphs, anyway). If you use an URL >in the DOCTYPE to refer to the W3C DTD, an apprpriately >constructed system will included the definitions for these >named entities, and convert them first to numeric character >references and then to the characters themselves. An SGML solution is theoretically possible, but there is no indication anyone is planning to implement it. Jonathan
Received on Friday, 22 August 1997 15:01:43 UTC