- From: Shelby Moore <shelby@coolpage.com>
- Date: Wed, 18 Dec 2002 21:29:17 -0600
- To: Mikko Rantalainen <mira@cc.jyu.fi>
- Cc: www-style@w3.org
>>>>>Ok well you write the algorithm that you propose we include in the CSS >>>>>spec, preferably compatible with the majority of CSS' users' languages >>>> >>>>End of sentence character. >>> >>>Do you mean that one would use style like >>> sentence-end-of-character: "."; >> >> No I think we were referring to a new EOS character, e.g. &eos; > >If we start putting extra characters to point out where sentence borders >are, why not just markup it correctly? For example: <s>This is a >sentence.</s><s>This is an another sentence.</s> At this point, my vote is also against sentence markup, because there is no reliable way for it to be done automatically, unless one is using a sentence parse. In which case, it might as well be in the user agent so that transformations are delayed until latest possible time, thus offering most generality and extensibility. > >That way we can use CSS we have today to style the text as we like and >we don't have the same problems we have with the <br> today. > >> In my last response, I stated that I think sentence parsers are better way >> to go. And that I bet parsers exist, even I don't have time to research >them. > >Sentence parsers sound great, but I'm afraid we aren't seeing any of >those in any browser in next N+2 years. Probably never, if the only >reason to have them is to have slight improvement in rendering. In >addition, such a parser should be able to decipher incorrect grammar and >typing mistakes to be usable in normal web pages. Grammar correction and other uses of sentence parsers (such as natural inflection text-to-speech) sounds like a great idea for a way to incentivize their inclusion in user agents. -Shelby Moore
Received on Wednesday, 18 December 2002 22:28:38 UTC