getting a hold on (all) running adaptors?

Hi,

Is there a way to get to the instances of (all) running adaptors?
In Pharo/Squeak, this is done by:  WAAdmin defaultServerManager adaptors

I'm porting SeasideTesting to Gemstone, and I need an entry point to send an HTTPRequest to for the simulated browser testing. In Pharo, Squeak and Visualworks, that happens by getting a hold of  an adaptor.

cheers
Johan

On 11/01/2010 01:27 PM, Johan Brichau wrote:
> Hi,
>
> Is there a way to get to the instances of (all) running adaptors?
> In Pharo/Squeak, this is done by:  WAAdmin defaultServerManager adaptors
>
> I'm porting SeasideTesting to Gemstone, and I need an entry point to send an HTTPRequest to for the simulated browser testing. In Pharo, Squeak and Visualworks, that happens by getting a hold of  an adaptor.
>
> cheers
> Johan

Johan,

I had an earlier version of SeasideTesting ported to GemStone and I
wired the tests to use the Hyper adaptor ... FastCGI wouldn't have been
much use.

So I would guess you'd want to do a similar thing for Seaside 3.0 ...
right now the Swazoo adaptor is the only one that makes sense.

For GemStone this kind of "assumed image state" is not natural ... with
GemStone you can actually have a couple of gems running with Swazoo
adaptors and a couple of gems running with the FastCGI adaptors (I've
done it, but it takes a little bit of behind the scense monkey business)
and since the tests will be running in a vm independent of what you are
using for Seaside it become necessary to wire in a logical default or
expect the user to have made a decision along the lines of:

   WAGemStoneRunSeasideGems default

but of course as I mentioned before FastCGI isn't a logical option.

Dale

Hi Dale,

Thanks for the pointers. The port is coming to life.
Still some work though: 52 run, 31 passes, 0 expected failures, 3 failures, 18 errors, 0 unexpected passes

;-)

cheers
Johan

On 01 Nov 2010, at 21:39, Dale Henrichs wrote:

> I had an earlier version of SeasideTesting ported to GemStone and I wired the tests to use the Hyper adaptor ... FastCGI wouldn't have been much use.

On 11/04/2010 08:07 AM, Johan Brichau wrote:
Very Cool!

Dale