OASIS Mailing List ArchivesView the OASIS mailing list archive below
or browse/search using MarkMail.

 


Help: OASIS Mailing Lists Help | MarkMail Help

dita message

[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]


Subject: RE: [dita] treating index-see as index-see-also


To document the discussion on the call ...

Paul returns to the most basic scenario: A user may use index-see-also
from one indexterm (the first) to another indexterm (the second). Would
there be any reason to autogenerate a return index-see-also? This could
not be done automatically because the user might (for example) view the
second indexterm as a more general term and the first indexterm as a
more specific term. Those who look up the more general indexterm would
get general references. Those who look up the more specific indexterm
would get the specific references plus an index-see-also to the more
general indexterm.

Bruce thinks this through for the case presented here. Goldfish (the
first indexterm) has both an index-see to Carp (the second indexterm)
and an indexterm entry. The index-see is therefore processed as an
index-see-also from Goldfish to Carp. A user looking at Carp would no
longer be expected to think of it as the authoritative term for
Goldfish. But we don't know whether Goldfish is an alternate, equivalent
term or instead a specific term that should be treated in the previous
paragraph. As a result, we don't know whether to generate an
index-see-also from Carp to Goldfish. In the first case, it would be
appropriate, but in the second case it would not.

So unless another argument is presented, I agree with Paul Grosso that
we cannot recommend that return index-see-also entries (from the second
indexterm to the first) be automatically generated.

Best wishes,

Bruce

-----Original Message-----
From: Grosso, Paul [mailto:pgrosso@ptc.com] 
Sent: Tuesday, October 03, 2006 11:15 AM
To: dita@lists.oasis-open.org
Subject: RE: [dita] treating index-see as index-see-also


You are suggesting that there are cases when there could be "logically
missing" index-see-also's, but couldn't this be the case in general?
Also, one person's "logically missing" might be another person's
indexing strategy.

Why should we try to fix up a "logically missing"
index-see-also in the case of an error when we don't try to fix it up
other cases?

I hesitate to add more "magic fixup" beyond something simple like "treat
index-see as index-see-also".  You run a risk that what one person
thinks is helpful turns out to be problematic for another.

paul


> -----Original Message-----
> From: Esrig, Bruce (Bruce) [mailto:esrig@lucent.com]
> Sent: Tuesday, 2006 October 03 09:59
> To: Grosso, Paul; dita@lists.oasis-open.org
> Subject: RE: [dita] treating index-see as index-see-also
> 
> It is an error if both an index-see and an index-see-also occur for 
> the same indexterm. In case this condition occurs, an implementation 
> may (but need not) give an error message, and may (but need not) 
> recover by treating all such index-see elements as index-see-also 
> elements. As part of recovery the implementation may (but need not) 
> add a corresponding index-see-also from the target indexterm of the 
> former index-see entry to the indexterm containing the former 
> index-see entry.
> 
> -----Original Message-----
> From: Grosso, Paul [mailto:pgrosso@ptc.com]
> Sent: Tuesday, October 03, 2006 10:53 AM
> To: dita@lists.oasis-open.org
> Subject: RE: [dita] treating index-see as index-see-also
> 
>  
> 
> > -----Original Message-----
> > From: Esrig, Bruce (Bruce) [mailto:esrig@lucent.com]
> > Sent: Tuesday, 2006 October 03 09:48
> > To: Grosso, Paul; dita@lists.oasis-open.org
> > Subject: RE: [dita] treating index-see as index-see-also
> > 
> > Spinning out the example further ...
> > 
> > Wouldn't the resulting index be somewhat unusable? If a user looks 
> > under Carp, they wouldn't see a reference to Goldfish, but
> some of the
> 
> > index entries are under Goldfish.
> 
> I'm not sure I understand, but regardless, I don't see what you're 
> suggesting we should do with respect to the DITA 1.1 spec.
> 
> Yes, it's possible for a user to use DITA's indexing markup in such a 
> way that the generated index is less than ideal.
> 
> We still need to define in the spec what the semantics are for all the

> possible markup combinations.
> 
> I'm happy to say that some combinations are errors--in fact, that's 
> exactly what I've done.  So what is it that you're want to say 
> differently than I've suggested?
> 
> paul
> 
> > 
> > In case this sort of clash occurs, to keep the index semantically 
> > coherent, the reverse see-also should be generated as well.
> > 
> >     Carp, ... Lots of references ...
> >       See also Goldfish <-- because an index-see to Carp
> was converted
> 
> > to an index-see-also
> > 
> >     Goldfish, 34
> >       See also Carp
> > 
> > This is not the case for an index-see that is not converted because 
> > the index-see could be from a deprecated term to an approved term.
> > 
> > Best wishes,
> > 
> > Bruce
> > 
> > -----Original Message-----
> > From: Grosso, Paul [mailto:pgrosso@ptc.com]
> > Sent: Tuesday, October 03, 2006 10:04 AM
> > To: dita@lists.oasis-open.org
> > Subject: RE: [dita] treating index-see as index-see-also
> > 
> >  
> > 
> > > -----Original Message-----
> > > From: Esrig, Bruce (Bruce) [mailto:esrig@lucent.com]
> > > Sent: Tuesday, 2006 October 03 08:54
> > > To: Grosso, Paul; dita@lists.oasis-open.org
> > > Subject: RE: [dita] treating index-see as index-see-also
> > > 
> > > Do we have a set of scenarios in which it makes sense to treat an 
> > > index-see as an index-see-also? I tried to construct one, and had 
> > > difficulty.
> > 
> > Sure.
> > 
> > You've got an index-see for "Goldfish, see Carp" in a topic that is 
> > referenced in your bookmap, and you generate output
> including an index
> 
> > and all is well.
> > 
> > Then you decide to reference one more topic from your
> bookmap, but it
> > happens to have an indexterm for "Goldfish".
> > 
> > So now you're generating an index where you have "Goldfish,
> see Carp" 
> > as well as a page number due to the indexterm, but it is
> incorrect to
> > have a "See" and a page number.  If you instead treat the
> index-see as
> 
> > a see-also, you would get a valid index entry.
> > 
> > I don't quite understand the rest of your message.
> > 
> > paul
> > 
> > > 
> > > The question has to do with the root cause for the clash. 
> > Is the root
> > > cause a disagreement (or unintentional inconsistency) over
> > what term
> > > to use? Is it an erroneous use in one place compared with another?
> > > 
> > > A viable scenario should show a sequence of source
> materials being
> > > processed, an intended behavior implemented by
> processing, and the
> > > resulting output. There would need to be more than one
> scenario in
> > > order to show different ways that a clash could arise.
> > > 
> > > Best wishes,
> > > 
> > > Bruce Esrig
> > > 
> > > -----Original Message-----
> > > From: Grosso, Paul [mailto:pgrosso@ptc.com]
> > > Sent: Monday, October 02, 2006 4:28 PM
> > > To: dita@lists.oasis-open.org
> > > Subject: [dita] indexterm proposed wording
> > > 
> > > Proposed additional wording for indexterm.
> > > 
> > > > -----Original Message-----
> > > > From: Grosso, Paul [mailto:pgrosso@ptc.com]
> > > > Sent: Tuesday, 2006 September 26 11:37
> > > > To: dita@lists.oasis-open.org
> > > > Subject: RE: [dita] review of index* elements
> > > 
> > > > > indexterm
> > > > > ---------
> > > 
> > > > Issue:  What if an an indexterm contains both an
> index-see and an
> > > > index-see-also.
> > > > 
> > > > Proposed resolution:
> > > > 
> > > > It is an error if an indexterm contains both an
> index-see and an
> > > > index-see-also.  An implementation may (but need
> > > > not) give an error message, and may (but need not) recover
> > > by treating
> > > 
> > > > the index-see as an index-see-also (in which case the
> page number
> > > > where the index-see-also occurred will also appear in the index 
> > > > entry).
> > > > 
> > > > ACTION to Paul:  Provide suggested wording.
> > > 
> > > Add as the final para of the first section:
> > > 
> > > It is an error if an indexterm containing no indexterm children 
> > > contains both an index-see and an index-see-also.  (Note:
> > > index-see and index-see-also elements within indexterms that do 
> > > contain indexterm children are ignored.)  In the case of
> this error
> > > condition, an implementation may (but need
> > > not) give an error message, and may (but need not) recover
> > by treating
> > 
> > > all such index-see elements as index-see-also elements.
> > > 
> > 
> 


[Date Prev] | [Thread Prev] | [Thread Next] | [Date Next] -- [Date Index] | [Thread Index] | [List Home]