Re: timescale of 'search-linuxkb'

[Aaron Turner <aturner@linuxkb.org>]

On Wed, 15 Dec 1999, Hugo Rabson wrote:

> 12/16/1999    - use get-cnn.tcl's html-parsing code as the basis for
>                 a LinuxKB search page parser
> 12/20/1999    - v0.1 close to completion (simple searches only, so far)
> 12/27/1999    - v0.2 close to completion (simple/advanced searches)
>               - v0.2 API documentation near completion
>               - feature freeze for Xmas :^)
> 01/03/1999    - v0.2 program and API complete
> 
> v0.3 will have:-
>     - drilldown?
>     - elementary, tk-based UI
> 
> v0.6 will give applet-writers as many of the search engine's capabilities as
> possible. It will grow with the search engine, hopefully reaching v1.0 by
> March.

Sounds good.  Definately keep me in the loop with the API.  I can make it
look almost like however you want.  For now, just look at the
implimentation in the advanced search form (note that there is currently a
incorrect implimentation of the Limit Docs to internal/external which will
likely be fixed sometime early next week.)
 
> Desktop applets will talk to search-linuxkb using the API (which will be
> well-documented). I've written very few UI-based apps for Linux, so I'm
> hoping other programmers will write the applets & use search-linuxkb as
> their API. (I'll write an anyway: I need the practice.)
> 
> I was going to write search-linuxkb in tcl, which would have allowed for a
> tcl/tk interface very easily. However, I have more experience with C, so
> I'll probably use that. I'm still in two minds about it, actually.
> (Comments/suggestions?)

My guess is that something in C would be more useful for GTK/QT
developers, but that's rather uninformed since I can hardly rememeber how
to do 'hello world' in C or C++.  Maybe Marc, our inhouse C++ expert has
an opinion :)
 
> To retrieve the search engine's web page, I'll use Lynx or wget.

Sounds perfect.  Code reuse.   Lynx is probably installed on more systems
by default and for this sort of thing easier to work with in my
experiance.

> Eventually, there will be an RPM for search-linuxkb.

Like Jason said.... drool.

> I'm working on the html-parsing code now; please find attached 'get-cnn.tcl'
> tarball to see how it is used. (The script needs wget.) It's nowhere near
> finished but it'll give you some idea of what I'm doing & how. Get-cnn and
> search-linuxkb will probably be developed in parallel, using a lot of common
> code. Eventually, I'll figure out what Microsoft did to Hotmail & then I'll
> fix hotmole too: it was the only way AFAIK that MSN users could read their
> email under Linux.

Ah that's right... I need to create another version of Search.pm to output
like your API spec.

> IMO, LinuxKB is a really cool project. :-) I'm looking forward to
> contributing.

Glad to have you on board!

--
Aaron Turner, Core Developer       http://vodka.linuxkb.org/~aturner/
Linux Knowledge Base Organization  http://linuxkb.org/
Because world domination requires quality open documentation.
aka: aturner@vicinity.com, aturner@pobox.com, ion_beam_head@ashtech.net