Re: Letting google index the BTS?
On Thu, 30 May 2002, Peter Eckersley wrote:
> Has anyone else found themselves wishing that google indexed the BTS?
>
> It seems to me that it would be very useful when trying to use google as
> a problem solver, and would also help with some of the limitations of
> the existing BTS search facility.
>
> Can this be done without exacerbating spam problems?
Everything in the bts is dynamic.
The code isn't as efficient as it could be. There are times when
pkgreport.cgi doesn't have an index it could use, so it has to scan a text
file, which is slower.
Also, I've considered implementing a cache. By this, I mean when
pkgrepot.cgi/bugreport.cgi generate html, they would save it somewhere. Then,
when a new request comes in, some timestamp comparison would be done, and it
would serve the saved static page, if nothing had been updated.
The above is made complex because of helper files like Maintainers and
Sources.
I've also tried indexing the raw .log files. But it ended up making too large
of a database(I had tried some text->postgress program off freshmeat).
If we can get google interested, I can hook them into my dynamic index
generation tool, which knows which bugs get updated, and doesn't have to do a
full scan. I could then send this list off to google, to let them update just
those bugs. This would at least allow for bugreport.cgi to be indexed and
searched.
--
To UNSUBSCRIBE, email to debian-devel-request@lists.debian.org
with a subject of "unsubscribe". Trouble? Contact listmaster@lists.debian.org
Reply to: