On Sun, 16 Sep 2001, Jim wrote:
> Last month, there was some mention of putting together a knowledge
> base for the mod_perl 2.0 documentation. I'm getting to the point at
> work that I can start throwing some time to it. Here are some
> thoughts I'm having regarding the design (programming target).
>
> I am able to provide hosting at least during development. I'll put
> out a URL when we get something together.
That's great!
Once it's working, we may start using it for all ASF projects and then
it'll be easy to host it on ASF machines.
I've just stumbled upon SuSe knowledge base -- very nice:
http://sdb.suse.de/en/sdb/html/index.html
> I think the first thing we need to do is figure out what we want the
> knowledge base to do. I would like to see this be a general purpose
> application that happens to do the job well enough for our use with
> mod_perl docs as well as other unrealted areas. It is because of the
> general use of the application that I may able to devote time at work
> to the project.
>
> So... in general, my take on a KB is that it provides the following:
>
> o Easy access to the right document at the right time
> This is for customers and customer support
> o Easy management of documents
> This is for the developers and writers
>
> These two items could easily be built as two independent interfaces
> around a common backend. I'd like to take this fall semester and
> develop the backend so we can work on front ends in the spring.
>
> The backend will need to support the following operations:
>
> o Indices of documents - keywords, title, author, etc.
> o Management of documents - insert, delete, modify, classify
> (setting keywords, ...)
Also check http://thingy.kcilink.com/modperlguide/, which is a special
version of the guide that generates a page per section rather than having
huge pages. So it should be able to take a document and split it into
factoids/sections and while keeping them in relation, like the URL above.
So when you search, each item gets searched separately, but when you view
a section, you can still easily jump to the adjusent sections in the
original doc.
> o Ranking of documents for search results
>
> o probably other things I'm not thinking about at the moment
>
> Documents will need to be understood in a variety of formats such as
> docbook and pod. Probably use pod as the initial format since that is
> what the mod_perl 2.0 docs will be written in.
and we have some docs in modperl-docs cvs rep already. Not much yet, but
it's the beginning.
> We can use AI::Categorize to form the initial keywords for a document.
> The system can learn as documents are added.
>
> We might want to include revision control someday -- perhaps use CVS,
> which might allow for an easy CLI. Otherwise, we can create an
> XML-RPC interface and design some CLI tools around it. Or both.
> This would be in addition to any web-based interface we might have.
>
> I'll try to spit out some more thoughts as they come. Please feel
> free to contribute. Now's the time to get the wish list put together.
Cool!
One thing I'd like to add is that the system should be capable of
scratching its database and rebuilding everything from scratch. Consider
some documentation being used as a source, getting modified, or some items
become absolute and get removed, some items get added. Somehow the system
should be able to handle this.
_____________________________________________________________________
Stas Bekman JAm_pH -- Just Another mod_perl Hacker
http://stason.org/ mod_perl Guide http://perl.apache.org/guide
mailto:[EMAIL PROTECTED] http://apachetoday.com http://eXtropia.com/
http://singlesheaven.com http://perl.apache.org http://perlmonth.com/
