Stefano Bagnara wrote: > > Steve Brewin wrote: > > For me, this is a prime example of why people have said we > need to discuss first, achieve consensus, then act. > <snipped> > > I'm sorry if this offend you in any way, I think this is a > needed thing > for me to keep working on James, so I did it and propose it. > > I think I did the right thing coding and not discussing > because now we > have a concrete thing to vote +1 or -1 and not an abstract discussion > about an abstract topic.
Stefano, Don't worry, you did not offend me in anyway. As you will have gathered, I don't think this is a good way of proposing a major change in a component. My main issue is purely practical. I think it seriously unwise to make changes on such a scale to any component without unit tests with sufficient coverage to validate it still works across all usage the scenarios. This is not a minor incrementl refactoring, of which I fully approve. This is a total rewrite. Without unit tests, I can't understand how we can have confidence in such a rewrite when your main argument for doing so is that you do not understand the existing code, "To me it's something like obfuscated: I feel worst than browsing decompiled code ;-)". This is not me being protective. This is me voicing me fears of treading such a path for any component in a minor (2.#) release. As I said, "We all know that major changes are prone to creating new issues". I don't think any component with so few issues open against it justifies the risk of introducing new bugs this degree of change will inevitably entail. I'm more than happy for you to prove me wrong by way of unit tests. Otherwise I think this belongs in the bucket of things to consider for 3.0. Cheers -- Steve --------------------------------------------------------------------- To unsubscribe, e-mail: [EMAIL PROTECTED] For additional commands, e-mail: [EMAIL PROTECTED]
