Conference vaxuum::online_bookbuilding

    I don't understand what you mean by 'double header' in hardcopy.
    Basically, you code tables the same way for online and hardcopy:
    
    <reference>(tab_one) shows this and that.
    <table>(Table One\tab_one)
    ...
    <endtable>
    
    That's standard for formal tables.
    
    For informal tables that are inline for hardcopy, you can either
    keep them inline for online if that works ok:
    
    <p>Inline table:
    <table>
    ...
    <endtable>
    
    Or, you can use the <online_popup>/<endonline_popup> tags to make
    the table a 'semiformal' table:
    
    <p>Inline table:
    <online_popup>(Table)
    <table>
    ...
    <endtable>
    <endonline_popup>
    
    The <online_popup>/<endonline_popup> tags are ignored for hardcopy.

    Does this help? If not, send me more details about your problem.
    
    Mary    

Re: 6.1 

>    Basically, you code tables the same way for online and hardcopy:
>    
>    <reference>(tab_one) shows this and that.
>    <table>(Table One\tab_one)
>    ...
>    <endtable>
>    
>    That's standard for formal tables.

That may be "standard" practice, but there's no law that says you have
to have a <REFERENCE> tag for every formal table, figure, example, etc.
However, if you don't, the online book has no hotspot and there is no
way for the reader to know that there is a table except to look under
the Tables heading in the TOC. 

Frankly, that seems wrong to me. At the very least, the bookbuild
should generate a warning message whenever there is a formal table or
whatever with no hotspot. I would much prefer if a hotspot were
generated by the presence of a formal table itself, i.e. a visible
placeholder that indicates the presence of a formal table. 

I'm also bothered by the fact that unless the reader happens to drag
the mouse cursor over a hotspot, there is nothing to indicate that
there is something special about that spot. I would prefer some sort
of permanent highlighting. 

						Barry

    A diagnostic if there is no hotspot is a good idea, and we know
    that we need to show the presence of hotspots better (that's pretty
    high on the list of future Bookreader functionality). But I wonder
    whether formal tables with no hotspots (just a generic iconic
    placeholder) might lead to ambiguity: if several tables are on the
    screen, none with hotspots (reference text), would it be difficult
    to figure out which was being referred to by the text? Use of the
    <reference> tag certainly eliminates any ambiguity either online
    or in hardcopy.
    
    Mary

I'd like to lend my support to the suggestion in .2.  Today, I managed to
see ONLINE for the first time a book I wrote.  It was a thrilling
experience, marred only by the fact that when I went looking for a couple
of tables in the book, I couldn't find any sign of them.  It wasn't until
I read this note that it occurred to me what happened to them.

In fact, I'm still not clear on how I could have made them visible.  There
was no <REFERENCE> to some of them (there are to others), and there was no
indication that a table was relevant or could have been found in any way.

I might suggest that, in the body of the book, you could substitute
something like:

     +--------------------------------------+
     |  Table 4-2, Whiff'n'Proof Samples    |
     |    appears here.   Mouse me to       |
     |               see it.                |
     +--------------------------------------+

The "icon" appearance in the text could be used as a "hotspot" to allow
immediate selection so the table could be popped up as though a normal
hotspot reference were clicked, but it would also act as an indicator
to the reader that a table "exists" at that spot.

I wouldn't object in any way to a diagnostic if there were no legitimate
hotspot, too, but I think a "flag" at the table location would also be
useful.  That way, even if several tables, none with hotspots, were on the
"page" and on the screen at the same time, there would be some indication
of their context.

If this makes sense, then please put it on the wishlist.  If not, please
help me to understand why.

Thanks,
   Jim

    By convention, the CUP writers (the largest internal group that
    we (CUP Eng) support) use the <reference> tag in conjunction with
    the formal figures, tables, and examples to generate a hotspot.
    This has always been the convention for hardcopy, although it was
    not always observed. After some discussion, writers (here in ZK,
    anyway) agreed to observe it more stringently for online books
    because, as you say, without the hotspot it's hard to find the tables,
    figures, examples.
    
    By adding a <reference> tag to each of your disappearing figures,
    you will get the hotspots. If tables (or figures or examples) are
    small and you find the hotspots intrusive, they can be coded inline
    as informal elements.
    
    You can also code them as informal elements and use the <online_popup>
    tags (documented in 'Coding Documentation Source Files for the
    DECwindows Bookreader') to create hotspots and popups.
    
    It would be possible to do what suggest with the <table>, <figure>,
    and <example> tags but that solution strikes me as more cumbersome
    (in terms of words and space used) than simple using a <reference>
    tag to get 
    
    ...text... Table 4-2 shows this and that.
               ^^^^^^^^^ 
               hotspot
    
    Also, if the <table>, etc. tags did emit such an 'icon', it would
    be redundant in the cases where a <reference> tag had been used
    to refer to the table. (And the overhead of checking for a matching
    <reference> tag -- ugh!).
    
    Does this help?
    
    Mary

Mary,

You make a good case for this.  I will (had already intended to) put
explicit references to my tables in the text nearby.  I don't want to make
them informal tables, 'cause they wouldn't have those nice titles that set
them off from the rest of the text in that case.

I still wouldn't mind seeing the "icon" appearing in the body of the book
where the table "would have appeared", even if there are <REFERENCE>s to
the table.  But it's certainly not going to be tops on *my* list of
requests for functionality.

Thanks,
   Jim