On Tue, 2007-12-18 at 10:44 -0500, Jahn, Ray (R.) wrote:

> Dear Jeffrey:
> 
> There has been a discussion (see below) by several participants on the
>  same subject.  Based on the past discussion, the need for this feature
>  appeared to arise primarily in a medium to large business
>  environment.  I do not know whether a RFE has been created for this
>  feature.  It might take more time for the OO Base community to
>  accumulate enough momentum and resolve toward such type of competitive
>  features in business operation.
> 
> Your contribution to help explain the "business" value and potential
>  competitive market pull resulting from this feature is very welcome.
> 
> Sincerely,
> 
> Ray Jahn

I've played with this idea a bit for my own 'Axis' project
( http://entropy.homelinux.org/axis ) ... note that my website is
currently down ... crappy DSL modem ...

They way I've been doing it is a little messy, but it works well for my
current needs ( read-only ).

My project is in Perl. Steps I'm using:

- Parsing the query and identify all tables involved.
- Use $dbh->column_info to inspect the fields in each table.
- Create 'temporary' tables in a local SQLite database for each table
- Query from each table / dbh and dump into SQLite, making use of pieces
from existing where clause to start with
- Query other DBs, with a MASSIVE where clause consisting of foreign
keys from prior steps, and other bits from the original where clause
- Final query of SQLite tables to pull everything together

Currently, I'm only talking to MySQL and SQLServer. And as noted
already, I'm only doing read-only stuff. If you want to do updatable
queries across multiple database servers, then I don't really think this
is a worthwhile goal. It's too messy to implement, and it only serves to
enable people to create bad applications anyway ... ie you shouldn't
have multi-table updates, even on the SAME server, and especially not on
separate servers. Yes MS Access does allow this. But every single time I
come across an updatable multi-table query, it's the source of at least
one problem.

If you want to support read-only queries across multiple DBs, then this
is a lot easier to implement, and actually useful. In the medium term
people with multiple DB servers should be looking at standardising on 1
server platform.

Anyway, if anyone is interested in working on a read-only solution that
utilises SQLite for local caching / aggregating, I've got bits and
pieces in place for it already ... in Perl. I suppose to be of maximum
use this would have to be ported to a new project, written in C, which
is beyond my abilities at this point.

Lastly, when I was enquiring on this topic myself some time ago, someone
pointed me to the Muldis::DB project ( which is also written in Perl ).

--
Daniel Kasak
IT Developer
NUS Consulting Group
Level 5, 77 Pacific Highway
North Sydney, NSW, Australia 2060
T: (+61) 2 9922-7676 / F: (+61) 2 9922 7989
email: [EMAIL PROTECTED]
website: http://www.nusconsulting.com.au

---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]

Reply via email to