cloning and Storable Re: [Bioperl-l] bugs on branch; tests on main trunk

[Hilmar Lapp]

On Thursday, September 4, 2003, at 12:53  AM, Ewan Birney wrote:

> If Jason/Lincoln/Hilmar all (or mostly...) liked clone() on the Root
> object then I'd have to conceed
>

My view is that it can potentially create more damage if people do 
start to copy&paste the implementation from the RE module into others. 
It seems to me that a clone() method on Root can strike a better 
balance, provided it is documented as to what the caveats are and that 
you should actually not be using this method unless you understand 
those caveats.

Copy constructors are handy at times, and setting actual attributes 
from those of the template just isn't very robust (what if a parent 
class adds an attribute? also, an inheriting class is implicitly 
burdened by having to override the copy constructor if it adds an 
attribute).

	-hilmar
-- 
-------------------------------------------------------------
Hilmar Lapp                            email: lapp at gnf.org
GNF, San Diego, Ca. 92121              phone: +1-858-812-1757
-------------------------------------------------------------