Mailing List Archive

GLEP51 (kbase): Resurrection!
Hiya fellow (GDP) Developers;

Sven and me have been talking about kbase lately, since we have both
the desire to revive that project. We're certain that this
gdp-oriented project would make all of our lives a lot easier. That's
why we're asking you for help.

We need input on how kbase should be working, since regular
implementations such as mindmeld[1], sciret[2] and others are just not
good enough for our use. That's why we need to find what we really
want; bundle all of our ideas and get to the point of writing our own
implementation, in coordination of the infrastructure project, if no
suitable alternative can be found.

The original glep[3] contains a set of requirements we need to need to
go over. One of our issues would be how the search engine should work.
At first Sven thought of having "natural language" as a requirement of
the search strings. Of course this could become inefficient or too
hard to implement. As such, this might be better off as an option
rather than a requirement.

If we're thinking about the functions of our implementation, we should
also have a thought on how we want to maintain, create, edit, delete,
etc our pages. Will we be using our GuideXML system in addition to
comments stored in a database? Or will we completely store everything
into the databases? Since each page would be maintained (perhaps
through a maintainer/herd-alike system), it would almost be a
requirement that the maintainers have absolute knowledge of GuideXML
if the topics are to be stored in GuideXML format. Perhaps we can
enhance Beacon[5] (the guideXML wysiwyg editor)(written by Anant aka

Please share your thoughts.

Kind Regards,
Sven and Dimitry