+1

Doru


On Sat, Jun 22, 2013 at 9:28 AM, Marcus Denker <marcus.den...@inria.fr>wrote:

> Very nice!
>
> Another thing we need to look at at some point is to make different VM
> build for each major version.
>
> The idea is that there is a dedicated build of the VM for "Pharo3", which
> would be frozen with the release,
> and then we start with a VM build for "Pharo4".
>
> I even want them to have the major version number is the binary, so that
> it is easy to install "Pharo3" in Parallel
> to "Pharo4".
>
> With this we free ourselves from needed the VM to be at  the same time
> forward *and* backward compatible.
> And we make sure that the VM that people use for the stable release does
> not change.
>
> Marcus
>
> On Jun 21, 2013, at 5:08 PM, Guillermo Polito <guillermopol...@gmail.com>
> wrote:
>
> Hi!
>
> We were discussing over here about it, so I decided to finally start
> something on the front of an acceptance test for our VM.
>
> The core idea is to stress the VM to test its stability and find bugs and
> regressions.
>
> So I created the ci job in [1] which does:
>  - use latest vm
>  - use 2.0 latest image
>  - loads seaside and moose on it
>  - runs all tests
>
> Right now it runs on 6 different slaves, getting different results on
> each, which we should take care about :).
>
> Also there are some things on it that should be improved easily:
>  - once 2.0 summer is released this job should use that version instead of
> the "latest" 2.0
>  - right now we are loading the development version of moose because moose
> 4.8 is not released (but will be soon :)
>
> And of course, now we have them, we can improve them (and make them run
> because not all of them do :D).
>
> Guille
>
> [1] https://ci.inria.fr/pharo/view/VM/job/PharoVM-AcceptanceTest/
>
>
>


-- 
www.tudorgirba.com

"Every thing has its own flow"

Reply via email to