Andrus Adamchik <andru..bjectstyle.org> wrote:
> I don't know anything about building a book-style index, but we already
> have a mechanism to generate PDF documentation that is easily searchable
> by Acrobat or Apple Preview.app ... Just need to work on the looks (I know
> Anton, the author of PDF integration, is already looking at some of the
> style issues).
Unless Modeler and User Guide documentation is stuck into one document, it
won't help enough.
A sequential search through a pdf isn't as easy as a real index on the web
site. [Actually, I just tried out Acrobat Reader 6 and it does provide a
list of results now -- improvement].
> On a side note I was thinking of writing a Lucene (and maybe
> Cayenne)-based web application to do a trageted search of our site that
> would allow to select a search category (mailing lists, user guide,
> modeler guide, issue tracker, wiki)... Should be a simple 1-2 page
> Tapestry or even JSP app with a backend indexing. Just like with other
> things it has been postponed indefinitely for a number of years now :-)
Yeah, that'd be nice to, and would maybe be enough.
I need to write a JSF and/or Tapestry app. Maybe I'll give this one a try.
I'm still thinking a master index of some sort of the formal documentation
is best, and I'll keep looking into it.
-Mike
> > There's a lot of good information that's in the User/Developer/Modeler
> > guides, but it's not always easy to find it. Sometimes, it's almost
> > impossible to find (caching, delete rules).
> >
> > We need a master index (and automated system for building it). Anyone
> > have any ideas?
> >
> > One possiblity would cross-index every single word appearing in the
> > guide, perhaps with the a way to manually adjust the indexes (delete
> > useless words ("and", "for", "but") while combining similar terms
> > ("delete", "deleted", "deleting").
> >
> > Another possiblity would be to add and maintain a list of key words to
> > each page (or have a master list somehow combining permutations).
> >
> > Anyone used or researched such software before? Since our documentation
> > is in XML, it should be easy to coerce it into whatever format is
> > needed by the indexing software.
> >
> > -Mike
>
>
>
This archive was generated by hypermail 2.0.0 : Tue Feb 15 2005 - 11:47:44 EST