-- troels knak-nielsen <[email protected]> wrote (on Thursday, 12 February 2009, 03:56 PM +0100): > On Thu, Feb 12, 2009 at 2:25 PM, Matthew Weier O'Phinney > <[email protected]> wrote: > > Most issues that suggest changes merely for the sake of OO purity will > > be ignored in favor of _real_ issues/feature requests. There's simply > > very little, if any, benefit to investing the time to make the changes > > in most cases. > > So good design is not a priority? > > Sorry for the trolling, but really ...
I'm saying that refactoring solely for the purpose of OO purity is not a priority. I'm not saying that good initial designs are not a priority, as they should be, nor am I saying that refactoring that has benefits to maintainability, extensibility, and performance will not be considered. But if the sole purpose is to achieve some sort of OO nirvana, then it's a pointless exercise. Refactoring simply to get OO purity typically will introduce BC breaks, while simultaneously doing nothing to change the intended behavior of the component. What purpose does this serve for end-user developers? Typically, not much, if any -- but it does provide plenty of headaches while they refactor their code to fit the new API. We *will* be doing some refactoring for 2.0, in places it makes sense. The internal team as well as some contributors have identified areas that could be improved in design. However, even then, changes will be evaluated to see what benefits they will offer to end users, to maintainers, and for performance; if there is no significant justification, we will likely leave the code as is. -- Matthew Weier O'Phinney Software Architect | [email protected] Zend Framework | http://framework.zend.com/
