[osis-core] OTCite: final call

[Steven J. DeRose]

At 13:36 -0700 2003-10-20, Scribe wrote:
>Guys,
>	I understand and respect the idealistic argument for 'pay
>attention to the pickin' attribute value and you'll get it right'.
>
>	I also know that we are a very practicle group seen, for instance,
>by our catering to the provide 'levels' of OSIS compliance.
>
>	Like the group, while I agree that people should pay attention to
>all attributes and should be 100% OSIS compliant, I also know that they
>will not all start there.
>
>	Rendering all <note>'s the same, <title>'s the same, even <q>'s
>(currently, just possibly ignoring the 'block' type) the same, gives a
>basic, OSIS conformance that I believe should probably be good 'level'
>distinction.
>
>	My point is not to disagree that <q type="otCite"> would be
>acceptable in a perfect world.  Just that the PRACTICLE result of someone
>not paying attention to the attribute will render misleading and confusing
>information to the user.
>
>	I also liked Chris' point (obviously because it gives concession
>to a solution I favour), that although we have <name>, we have given
><divineName> for a case that happens just so much in our target text its
>frequency alone might have warranted it.
>

I have mixed feelings on this one. I'm reluctant to add another 
element, though at least this is one that we had once before.

On the other hand, I'm not convinced by arguments re. the use of 
attributes -- even CSS lets you condition your styling on attributes 
(conceivably OEB doesn't support that, and I know browsers have some, 
uh, variation....). I think Troy's point above about misleading 
rendering from an engine that did ignore the attributes is 
significant -- but I guess I don't see displaying an (alleged) OT 
quotation with quotation marks would be so bad.

On the other hand, I agree completely with whoever pointed out 
recently that <q> does not equate to having quotations marks. Besides 
the block quote case, there is a lot more complexity too: nested 
quotes use different characters, in patterns that differ 
substantially from locale to locale. Handling of quotes in a 
multi-paragraph speech or turn in dialog differs with country, 
language, and probably even publisher; and I know at least Spanish 
does weird things when quotes are interrupted by "he said" and such 
things. In other words, quotes are just not simple. If some apps are 
willing to sacrifice all those nuances, sacrificing the nuance of 
formatting OTquotes in a special way rather than with quotation marks 
doesn't seem like a big deal.

On the other hand (as Tevye said, "No! I am running out of hands!"), 
the argument that OTquote is a natural notion in the cognitive 
lexicon of translators seems important.

So I guess that lands me about dead center re. adding OTquote. I do 
think that, whatever we do, we need to document it so everybody does 
it the same way.

S
-- 

Steve DeRose -- http://www.derose.net
Chair, Bible Technologies Group -- http://www.bibletechnologies.net
Email: sderose@acm.org  or  steve@derose.net