Re: [F-Script-talk] [ANN] F-Script 1.3.1 Release Candidate

[Philippe M. <pm...@ac...>]

Le 30 janv. 06 =E0 04:06, Ken Ferry a =E9crit :

> On 1/29/06, Todd Blanchard <tbl...@ma...> wrote:
>> It is possible to say 'anObject setFoo: aValue' and 'anObject foo' to
>> an NSManagedObject if the object's entity has property
>> 'foo' (basically it maps setFoo: to setValue: forKey:'foo' and foo to
>> valueForKey:'foo'
>
> So is there any reason we're only doing this for NSManagedObjects?  Is
> it too evil to want to get at normal object ivars in this way as well?
> :-)

Even mentioning it is evil ! :-)

Well, the goal with NSManagedObject was to re-objectify the whole =20
damned thing and make it easier to use it with F-Script's array =20
programming model. But I don't want to introduce too much magical =20
behavior. And it is already possible to directly manipulate ivars =20
from F-Script using KVC (storedValueForKey: and =20
takeStoredValue:forKey: generally work). On the other hand I'd like =20
to have a graphical browser showing instance variables. There was =20
something like that in GDB on NEXTSTEP 3 and it was terrific for =20
debugging.

> I usually use FScript for exploration and debugging, so I'm generally
> not too concerned with encapsulation.
>
> The release candidate is working great for me so far!

Thanks for testing it.

Philippe Mougin=