On Tue, 14 Dec 1999, Bill Carlson wrote:
> On Tue, 14 Dec 1999, Geoff Hutchison wrote:
>
> > On Tue, 14 Dec 1999, Bill Carlson wrote:
> >
> > > Manipulate until I get what I want, then dump to htdig databases. It seems
> > > that a lot of people spend a lot of time getting htdig to index only what
> > > they want where they could spend very little time massaging the data in
> > > another tool.
> > >
> > > Again, I think the primary reason for a SQL backend is data access, not
> > > performance.
> >
> > In that case, would a set of tools for accessing the Berkely DB be
> > suitable? Let's say we provide a bunch of tools for dumping the databases
> > to an ASCII form, then re-loading them. Would this be sufficient for your
> > purposes?
> >
>
> For my particular purposes, that would be great. And as I said, I think
> this is the main reason many people have requested the move to a SQL
> backend. Look at how often the question of the contrib perl scripts not
> working pops up on the htdig list, people want to do interesting things
> with the htdig data other than just searching.
Now, I'm not a DB/SQL expert by any means, but...
I think that would be a good step. But it doesn't solve everything. The
site I'm developing is 99.99% dynamic. It's all in a SQL database. One
problem I know I'm going to have in the future is updating htdig. I'd
love it if I could add functions to my existing PHP or mod_perl code that
not only updates the SQL DB for page generation but the htdig DB as well.
With a SQL backend it's pretty easy and I may never have to run htdig
again. Since I use a wrapper for htsearch, I'd like to be able to run
additional queries against the database for hits htsearch returns for
things like word distancing and what not.
Also, my htdig DB is going to get REALLY big over time. Exporting it to a
file, making mods, and then re-importing it is going to be very painful.
Basically, I totally agree that the port to SQL is more for access than
for performance. If I get to a point where performance becomes an issue,
I'll scam some more hardware and do load balancing. A lot more people
know how to work with a SQL database than Berkely DB, and there seems to
me at least a much more richer toolset for SQL. And setting up a SQL DB
like MySQL isn't that difficult.
My .02
--
Aaron Turner, Core Developer http://vodka.linuxkb.org/~aturner/
Linux Knowledge Base Organization http://linuxkb.org/
Because world domination requires quality open documentation.
aka: [EMAIL PROTECTED], [EMAIL PROTECTED], [EMAIL PROTECTED]
------------------------------------
To unsubscribe from the htdig3-dev mailing list, send a message to
[EMAIL PROTECTED]
You will receive a message to confirm this.
