Andrew Sullivan wrote:

>On Mon, Nov 14, 2005 at 09:02:36AM +0800, Christopher Kings-Lynne wrote:
>  
>
>>Is there no way you can integrate slony with the pg build farm?
>>    
>>
>
>I suspect volunteers are welcome!  That is, AFAIK nobody has even
>approached the buildfarm people over this.  And it would require some
>significant integration work; which is, I'd say, why it hasn't been
>done yet.  There are several automated tests in the source tree, as
>well; but I'm not sure they're complete regression tests.  So are you
>volunteering to do (some of) the work?
>  
>
I have been working on getting more tests into the "new generation test
framework."

Pointedly, I have added in (not yet committed, tho) multibyte and
namespace mangling tests.

It's not all working yet, which is why I haven't committed it to CVS,
but it's on its way.

What is unfortunate, at this point, is that there are cases which we can
expect to break in particular ways; it would be good to verify that
Slony-I refuses them properly. 

For instance, if you have a Really Long cluster namespace name,
auto-created PK fields start breaking; I have added a check for 1.2
which refuses to allow the Really Long cluster name.  Unfortunately,
things we expect to break aren't nearly as easy to test :-(.

At present, we have about a dozen "duct tape" tests which cover a fair
bunch of scenarios; that probably doesn't require 12 separate test
cases, in an "automated run-them-through" system, though...