On 3/23/06, Jeff Lester <[EMAIL PROTECTED]> wrote: > I just want to add I build enterprise level web-applications as well. I can > see both sides here. Basically Shannon is making a suggestion to tweak > something that works on his level, obviously there are going to be different > organizations that use different methods, but ultimately, designing an ORM > such as reactor so that its auto-generated API names are customizable is > both highly desirable, and extremely complex in terms of dynamic coding.
FWIW, I actually think adding a configuration parameter that specifies the record suffix string (and the iterator suffix string) is a useful idea that would indeed solve Shannon's / Kurt's problem. My issue here is twofold: - persistent criticism of a framework's philosophy - insistence that someone's pet tweak is trivial These two things are common to so many discussions I've seen around frameworks and they really run counter to the idea of consistency and coherence in frameworks. A frameworks *always* involves tradeoffs - you cannot expect a framework to do everything *your* way. > The end question is, how hard is it to implement this into the configuration > as an IteratorPostfix / RecordPostfix variable setting? If it's not that > difficult where is the harm in implementing it if only to get more people on > board? I'd be willing to help however needed. Every single configuration option makes a framework harder to learn and more fragile. It's a dangerous tradeoff. Just because it's a "good idea" doesn't mean it should be implemented. As I say above, I think this *is* a good idea but I'm not sure whether it should be implemented... -- Sean A Corfield -- http://corfield.org/ Got frameworks? "If you're not annoying somebody, you're not really alive." -- Margaret Atwood -- Reactor for ColdFusion Mailing List -- [email protected] -- Archives at http://www.mail-archive.com/reactor%40doughughes.net/
