[DGD] Recursion in recompile(), is this correct?

[Robert Forshaw]

>From: Noah Gibbs <noah_gibbs at yahoo.com>
>   DGD may call recompile() again on the parent classes
>automatically when you call compile_object() on them.
>So this'll all be pretty weird.

That is what I needed to know. I thought this *might* be the case and if it 
were, then the recursion is unnecessary.

>   This won't work.  It turns out that an object
>pointer is effectively to the latest version, always.

That I wasn't aware of. Why does the docs suggest that I should be 
destructing the object then after compiling it? (To re-quote: "If B should 
actually be recompiled (inheriting the
     new version of C from B), the driver object must destruct B"). It says 
'The driver object' not 'the driver', so presumably that means I have to 
destruct it explicitly in the recompile() function?

>When you recompile the code, the object pointers get
>updated, kinda like how when you destruct an object
>all pointers to it suddenly become nil.
>
>   Each 'version' you're talking about is called an
>'issue'.  You can't specifically touch an old issue,
>not to recompile it nor to destroy it.  Object
>pointers update automatically, so this code won't work
>the way you're hoping.

If this is true, then why does the docs say that it must be recompiled and 
destructed?

_________________________________________________________________
Tired of 56k? Get a FREE BT Broadband connection 
http://www.msn.co.uk/specials/btbroadband

_________________________________________________________________
List config page:  http://list.imaginary.com/mailman/listinfo/dgd