[gb2world]

QT's process of using their customers to test releases is nothing new. You can search the forum for every release of a xx.0 branch and find in the rants many helpful suggestions from this community for improvements in testing. You see constant advise about avoiding xx.0 branches. It is also clear the 4.5.5 release is equivalent to what would have been xx.0 release in the past - so all the advise about avoiding xx.0 branches was not applied, and the frustration is now magnified.

It is also true that this architecture is not easily upgraded or testable. Since XC-Next is not ready in terms of functionality, and even if it were, there is not an easy migration path from X-Cart. Therefore, you have a large captive audience stuck on this branch, which although good in a functional sense, does not use modern software development techniques (bad for testing and upgrading).

Since QT is committed to this branch because of its large captive audience - please don't continue to ignore your problems with test before release. If you don't like any of the suggestions you find here - please come up with your own that works with your current processes. The more transparent your solution is to your customer base, the better. But please don't continue to ignore the problem, because it can be solved on X-Cart Next.

Here are more suggestions for you:
- Use development releases - 4.5.5.dev. This keeps people focused on production away, and ranting will decrease. It gives the 3rd party mod developers opportunities to test and update their mods. All the people watching for the next release can try it, understanding what it is - development, not fully tested.
- Encourage early adopters to test dev releases by offering something. In the past, I suggested offering points, but these are gone. You could give them transferable licenses. Isn't cleaning up a dev release and making it a solid production release worth a few transferable licenses? The only valid test is running this stuff out in the field on real carts. It does not appear you have a tough enough internal test environment. With some incentive, people will be motivated to help you test your releases, and lead to true production releases.


--