RE: Re[2]: FW: acronym in title...

> The risk depends on how the items are matched. In Chales' 
> examples, if the author of that page edits the title, the 
> annotation still starts at character 19, but no longer at the 
> right word.

Note that using pointers that rely on the character position of a word
in a sentence means that localisation of the data potentially becomes a
nightmare.  As the text changes, every single pointer will need to be
changed for every single language involved - in a moderately to heavily
annotated site this could introduce a huge overhead in schedule and cost
that equates to a serious barrier to international deployment.

RI

============
Richard Ishida
W3C

tel: +44 1753 480 292
http://www.w3.org/International/
http://www.w3.org/People/Ishida/



> -----Original Message-----
> From: Martin Duerst [mailto:duerst@w3.org] 
> Sent: 14 March 2003 20:09
> To: Al Gilman; ishida@w3.org; w3c-i18n-ig@w3.org; 
> public-i18n-geo@w3.org
> Cc: w3c-wai-gl@w3.org
> Subject: Re[2]: FW: acronym in title...
> 
> 
> At 18:07 03/03/13 -0500, Al Gilman wrote:
> 
> >At 04:37 PM 2003-03-13, Martin Duerst wrote:
> >>At 09:57 03/03/13 -0500, Al Gilman wrote:
> 
> >>>I think that the one thing I should add right away is that 
> there is 
> >>>an option using "annotation" techniques to leave the 
> attribute as is 
> >>>and introduce a higher-quality equivalent through a 
> structure which 
> >>>refers to the attribute.
> >>
> >>I agree that this is an option in principle, but for the 
> things we are 
> >>considering (<span xml:lang='...'>, ruby,...), it seems like an 
> >>enormous overhead.
> >
> >And there are more things that need doing than you are considering.  
> >And somewhere along that line the preference for how to do 
> it may flop 
> >over.
> 
> Yes, definitely. The more something applies in more than one 
> place, for example, and the more it is part of one specific 
> rendering rather than the 'original' text, the more sense it 
> makes to put it offline rather than inline.
> 
> 
> >It's not clear which way the overhead is greater.  A 'span' on each 
> >instance could be much more size overhead when compared with 
> a single 
> >glossary entry. But the latter is more risky because it is more 
> >indirect.
> 
> The risk depends on how the items are matched. In Chales' 
> examples, if the author of that page edits the title, the 
> annotation still starts at character 19, but no longer at the 
> right word.
> 
> 
> >I am trying very hard not to prejudge the choice, here, because in a 
> >migration strategy for an incompatible change I believe that 
> there is a 
> >place for more than one way of doing things, and that we should not 
> >rush to judgement on how it is to be done.
> 
> My current judgement is that we need both:
> - Better ways for inline annotation, not limited due to the fact that
>    some things are in attributes when they shouldn't.
> - New (or better established) ways to link to glossaries and similar
>    stuff.
> 
> 
> >As I said at the end of the day Wednesday, what I am hoping the 
> >community will develop is better discernment about when to use which 
> >manner of annotation, inline and offline.  Markup is inline 
> annotation.
> 
> I think the community implicitly already has quite a good 
> understanding of that distinction. On average, and for purely 
> practical (but very
> important) reasons, inline markup seem to have performed much 
> better in most cases, but there are good cases for offline 
> stuff, too (in particular stylesheets).
> 
> Regards,   Martin.
> 
> 
> 
>   
>

Received on Tuesday, 18 March 2003 07:37:33 UTC