Summary: HTML index (part 3 of 3)

[Ruth Simpson <ruths -at- MCCABE -dot- COM>]

Tom Goering wrote:
> If you put your index tags on headings, or other paragraphs that read well
> as topic titles, you can get your index to produce those titles rather than
> page numbers. It's a bit convoluted, but it works:
>
> When you enter the Marker Text, end each entry with a : (this forces the
> link to be one level down from the entry, which gives you more formatting
> options)
>
> On the IX Reference page of the generated index, change the <$pagenum> tag
> at the bottom to <$paratext>. Then, on the line that defines the separators,
> change the comma to a soft return. (This results in your index producing a
> list of paragraphs instead of page numbers. That's why you need to only put
> tags on paragraphs that read well as topic titles)
>
> We had to play around with the formatting and WebWorks mappings a bit to get
> what we wanted, but it actually works fairly well.

Cassie LaBelle wrote:
> This would take some initial leg work, but beneath each entry, include
> the section name that it's in instead of the page number. For example:
>
> device driver
>      Installing Device Drivers      (put the link here)
>      Troubleshooting Device Drivers      (link)
>
> To make things quick, you could do find a replaces for some of the page
> numbers that only had one section name on the page.

Alice Preston wrote:
> I think precisely because of this kind of thing, current thinking on index
> entries is that you should add levels until a user knows precisely what
> they're going to . . . that is, see to it that each index entry has only
> one target place by fully qualifying its content.  I know, lots of work,
> huh?

Simon Bate wrote:
> I ensure that each of my entries points to only one page.  Not to get
> too much into the religion of indexing, but: as long as your index isn't
> getting too deep, it will really help the reader to know specifically
> what each page references points to.  You're not doing your readers a
> disservice.
>
> Once I generate my index from WebWorks, I then run a PERL script
> (attached) on the .HTML file that strips the numbers from the pages.
>
> You can check out one of my indexes at:
>
> http://www.pointcast.com/support/pcn/help/2.0/helpIX.html

Robin Clark wrote:
> This may be a naive suggestion, but how about making your index entries
> more specific and/or create a hierarchy?
>
> For example,
>
>         M
>         Methods of Payment / MOP
>                 Apply MOP to Item
>                 Cancel MOP
>                 Cash
>                 Credit
>                     Imprinting (Other Credit)
>                     Manual Entry of
>                     Reverse Transaction
>                 Debit
>                     Cash Back
>                 Food Stamps
>                 Process MOP
>
> etc., or
>                 C
>                 Cash
>                 Credit, Imprinting (Other Credit)
>                 Credit, Manual Entry of
>                 Credit, Reverse Transaction
>                 D
>                 Debit, Cash Back

Again, thank you listers for helping me. All of these suggestions are
valuable. I hope everybody gets as much out of the list as I did.

Ruth Simpson
Technical Writer
McCabe & Associates