Code review comment for lp://staging/~akretion-team/openobject-server/openobject-server_5.0_patches

Hi Raphael

> I agree that the change is not small. But don't you agree it's testable?

Yes we can test it, to find all regression, but don't apply in 5.0 (see my last answer)

> Frankly using it since yesterday no issue at all. I think some of us could
> confirm it safe and after that only we merge it. What I think is that this
> is worth the effort, just saying fuck once more to all Ubuntu users for the
> next say 6 months will by no means help the product popularity...

For me, Ubuntu 9.10 is ready for Desktop, but i didn't take the risk to install a server with this version, and prefer 8.04 LTS.

> I insist, this is not like the report rewrite situation. No logic is changed
> here, it's just about swapping a lib for an other. All changes code lines
> are easily tested, available here, why not make the effort to give it a test
> and then merge it eventually?

If i listen this in geek mode, then yes we must apply the merge, but if i listen this in integrator mode, i don't want to see this in stable,

Do you have to test this with Redhat 5.X version (Only python 2.4.3) ? i have test and i doesn't work directly, for customer, we must compile ne python version 2.5, easy_install all packages, but NO CONTRACT SUPPORT FROM REDHAT :( . My reflexion is "Why Debian/Ubuntu and not RedHat/Centos ?"

Regards,