[storm] using odbc as a backend

[James Henstridge]

On Fri, Mar 6, 2009 at 3:36 PM, Martin DeMello <martindemello at gmail.com> wrote:
> On Sat, Mar 7, 2009 at 1:09 AM, James Henstridge <james at jamesh.id.au> wrote:
>>
>> Note that there is database specific code for all three currently
>> supported databases, so it is very unlikely that a new backend could
>> be done without similar work.
>>
>> This also means that a generic "pyodbc" backend is not on the cards:
>> it would depend on what database is actually being talked to.
>>
>> It probably wouldn't be a huge effort to support PostgreSQL or MySQL
>> through pyodbc, but I'd guess that isn't what Martin is after.
>
> I was hoping that the various database's ODBC drivers would take care
> of the database-specific stuff, and allow storm to access a common
> API, as long as I didn't want to do anything too funky (just need
> simple inserts, selects, updates and deletes, for the most part).

I suppose the most important thing Storm does in the database specific
code is retrieving automatically assigned primary keys for newly
inserted rows in the database.  This is something that even the
simplest Storm apps tend to rely on, so needs to be handled correctly.

There are a few other bits and pieces, but that is the most important one.

James.