launchpad-dev team mailing list archive

[Robert Collins <robert.collins@xxxxxxxxxxxxx>]

On Wed, Sep 29, 2010 at 11:10 AM, Elliot Murphy <elliot@xxxxxxxxxxxxx> wrote:
> On Sun, Sep 26, 2010 at 6:41 PM, Robert Collins
> <robertc@xxxxxxxxxxxxxxxxx> wrote:
>> We can even 'run' postgresql in this fashion, and have it return the
>> DB name to use.
>
> This would be fantastic - having to configure postgres *at all* in
> order to get a launchpad dev system running is a barrier to the drive
> by contributor IMO. I believe I've already pointed you at
> implementation details for firing up an ephemeral instance of postgres
> (which I originally stole from jeroen).

Yes, you did, for clarity here - I'm not proposing to start
*postgresql* dynamically [yet, the contract does permit it though]
simply for performance reasons.

It takes minutes to build a configured, empty, LP database.

It takes seconds to clone the template test database to make an empty
working copy of it.

So, the fixture I'm interested in is one which creates a unique db
name and does the existing clone-template-to-test-name step, rather
than a full bring-up-postgresql-from-scratch.

We may want that full facility later, and it should Just Work when we
want it. (At which point I'll be looking at the particular code we
need to trigger that - I haven't as yet actually looked at it).

-Rob