Re: [gnome-db] recordset mysql.

[Rodrigo Moya <rodrigo gnome-db org>]

On Fri, 2005-01-21 at 16:12 +1000, Bas Driessen wrote:
> Hello,
> 
> While I am at it, I may as well finish the implementation of the
> recordset for MySQL. The Postgres recordset is build around the
> gda_data_model_hash implementation and the MySQL recordset is build
> around the gda_data_model_base implementation. For the representation
> of the data-model however, a pointer array is used. So therefore my
> question; Why isn't MySQL based on the gda_data_model_array
> implementation, where all kinds of functions are ready to use? I
> haven't looked at it at great detail yet, but is it desirable to
> implement it as a gda_data_model_array method? 
> 
> Please let me know which direction is best to go. Continue on the
> _base path or migrate to the _array path (or maybe even the _hash path
> if there is even a better reason for it). Hopefully, I can get that
> out of the way as well in the next couple of days so we then have a
> full working record set for both Postgres as MySQL.
> 
I personally find it easier to use the GdaDataModelArray instead of
*Hash, for providers that retrieve all rows, as is the case, IIRC, of
the MySQL provider. If it makes things easier, then go for it, use the
*Hash class
-- 
Rodrigo Moya <rodrigo gnome-db org>