Thorsten Scherler wrote:
>> This was confirmed in response to my recent suggestions to create all >> new features in separate branches that can be integrated (and >> released) as soon as they are stable. > Merging different branches that are have to go to the core are *not* > possible, merging will be nearly impossible (too many possible > conflicts). So why was this decided in the first place? Is it really impossible or does it just require a different approach? > Besides the danger of: > El jue, 29-09-2005 a las 10:46 +1000, David Crossley escribió: >> Branches tend to become islands of lone development, ... Hmm ... This point was made before and I didn't object. Now I do: What is different about a feature development in a branch that I (for lack of time or interest) _choose_ not to look at before it is proposed for integration and a concurrent development of several features (current situation) that a significant number of committers cannot follow anymore because they simple get lost in the complexity of several things happening at the same time. Please do realize that other people do not have your level of understanding of Forrest so carrying on at the level of complexity you will simply loose their input. And that in my view is no longer a matter of choice of their part so I'd clearly prefer the first way. -- Ferdinand Soethe
