Peter Donald wrote: > > At 02:58 24/2/01 -0800, Federico Barbieri wrote: > >I can see your desire for flexibility but I stand on my point. > >Configuration and validation must be guarantee to be as clean as > >possible. I belive in this concern flexibility is very dangerous in the > >long run. > > can not be too bad or else things like the digester package from > struts/ant/other similar config systems would have seen difficulty. >
I still think those have different needs than us... that's why I was talking about the "ManageableConfiguration" wich is a name that badly such but was ment to extend Configuration and provide for example discovery... Let' call it "Script" extends Configuration Iterator getChildren(); etc Phoenix uses Configuratoin that is what it needs, ant uses Script. Isn't it accettable? > >> c. not allow child components in container config > >This is by far the best option. separated conf file for the container > >and its child enable much better pluggability, easier deployment, > >validation, separation of concerns. > > looks like this is most popular one so far ;) then there must be a good reason! :-) > > >> 3. Should validation be context sensitive? > >> ------------------------------------------ > >> > >> The more I think about this the more I begin thinking no. Context > >> sensitivity while nice is too much work in the general case. > >> > > > >God bless you! I made it! I can't belive you change your mind... what > >happend? :-) > > you used logic ;) call me Spock! :-) > I used to have running deal that anyone who could get me to change my mind > based on logic gets $10 ... that was fine up until the day I lost $80 in > the span of about 4 hours at which point I decided that deal sucked ;) > :-) > >> * disallow context sensitive (ie external) validation > > > >disallow? We can't prevent it... just not enforce it. > > well disallow it in validation phase by not implementing it and forcing > users to do it at runtime ;) ok. > > Cheers, > > Pete > Fede
