The "Ideal" User Assistance Application?

[Scott Prentice <sp@leximation.com>]

Hi...

I've had a number of conversations with various people on this subject 
.. initially I was very excited about the idea of developing an 
XML-based UA delivery app that would provide all of the features lacking 
in the other available apps. This using some common schema and internal 
XSLT parsing that would generate the topics and navigation (TOC, index, 
??) on the fly. Although I still think this would be an interesting 
project, I think it may have some shortcomings when used for medium to 
large projects (namely performance). There is something to be said about 
the current paradigm (used by most UA apps) of defining the topics, and 
navigation elements as static files that reference each other as needed, 
then wrapping that up in a nice package. This eliminates the need to 
choose one data model as the "UA Standard" and moves back to the concept 
of generating the (XHTML?) files from your source using XSLT or some 
conversion process.

Along these lines, a goal that I can see for this group is to come up 
with the specs for a new platform independent, open source, user 
assistance application. This UA app would provide all of the features 
that content developers expect and rely upon as found in the current 
Help delivery options, but implements the features properly and 
completely. Some of the features I'd like to see are ..

- This UA app should be natively compiled (C++?), requiring no external 
components
- Supports both compiled and uncompiled content delivery from local 
and/or remote locations
- Familiar content navigation elements such as TOC, Index, Search, and 
Bookmarks
- Content filtering based on user type (or other metadata)
- Provision for context-sensitive topic access
- Popups and other developer-defined window types
- Built-in update mechanism for both compiled and uncompiled content options
- Support for all (many?) languages
- more I'm sure ...

In my mind, it is very important that this UA app be as self-contained 
as possible. Otherwise, it is very difficult for content developers to 
be able to rely on their content being rendered properly (or at all). 
This is a problem for current UA apps that are Java-based, as well as 
those that rely on the user's default or installed browser.

It seems that with Microsoft's recent announcements at the WritersUA 
conference .. (1) "Vista Help" will not be available to content 
developers; (2) WinHelp will not be supported in Vista; (3) They are 
looking for community feedback as to what we want for UA delivery .. 
this is an idea time for these discussions to be coming about.

Not to take this too far, but I could see leveraging mozilla for the 
content rendering, and clucene as the search engine. This could probably 
become a mozilla.org project, since really, this is just a very custom 
web browser. Mozilla provides the environment for plugin development, 
and already has some nice update capabilities.

What do you think?

...scott

Scott Prentice
Leximation, Inc.
www.leximation.com
+1.415.485.1892