Product name: ABRA

searchable, sortable and sharable named resource list
'bookmarks' expanded to local, cgi and dbi resources
associated with category, keywords and description for searching and sorting

all users for very inexpensive version run as a service
webmasters and some professional users for installed software

Initially provided as a service rather than an installed program.

Allow users to build category hierarchies and create active information objects 'AbOb's within them.
Allow users to sort and search these objects in various ways.

An 'AbOb' can be some static text, reference to a local file on the user's machine, URL to a remote file, URL to a CGI interface, or reference to a database thru DBI.

Each 'AbOb' has a user-defined name, optional keywords and description associated, and can belong to one or more categories.

A search can be simply within the names and keywords the user has defined, within all the contents of the URL's, or as a broker to a selection of the CGI and DBI interfaces the user has configured.

Abra can be used for research on the Internet, to create a public 'portal' type site, or for personal organizing and filing.

Allow users to register and create demo resources for free, which last a short period of time unless paid for. Provide scalable service starting very low cost, order of $1/week by Paypal, higher for more resources.
Sell the installed program $1-5K.

bootstrap This is potentially a large project; try to create small useful parts first, free for certain period then charge low fee. Simple organizer and personal search resources may be useful enough that people will be willing to pay a few cents for it.

Major parts of the work include

• Category Manager
  allow user to add/delete categories, name and attach keywords
  allow mapping of one category to another
• Resource Selector
  allow user to add URL's of specific sites or cgi forms or to pick existing category from dmoz to index
  options to Name this category, Map to known category
• Data Source Manager (internal)
  used by both resource selector and query handler,
  keeps internal track of all data sources defined, input & output templates
• Query handler/translator
  for searches, accept user input and translate as needed
  hardest part here is mapping multiple inputs, don't need to do this in the first version
• Request engines: Webglimpse for sites, CGI broker, DBI, other
  actually perform searches/queries
• Output merger & formatter
  using known output templates, gather & merge all results into
  a single list of "hits". May still indicate source of each.
  hardest part here is probably parsing the cgi results, I have an idea for automating this.
  also use output templates to give flexible view into the hierarchy, may create 'MyPage' for front page with combined specific parts

webglimpse.net currently gets ~ 50,000 hits/week, will prominently link there

By being our own best customer - use the product to create search objects that are useful to webmasters, and publicize those. (I already use WG whenever I want to search a site not currently searchable, like gnome.org or the mSQL docs - I just create an index for myself)

By making free limited service - say, with only capabilities to create categories with static text and possibly URLs, but not searchable.
Offer free service to nonprofit and certain projects.

It is possible to make income from such software; Webglimpse search engine already sells significant amount and this is something related.
But, there is no guarantee; so best thing is if we make something at least useful for ourselves.

See PPL Licensing The basic idea is that, anyone who contributes to the development will be paid in proportion to the number of lines of code used in the product, minus some administrative overhead. It is up to the project leader(s) to accept or reject code offered. Contact Golda Velez if interested.

---