Maxence, hello.

On 25 Nov 2011, at 13:04, Maxence Guesdon wrote:

> Thanks for your explanation. If I understand correctly:
> - there is smal additional cost (memory and cpu) due to copying,

Well yes, but no more than would be expected anyway in normal operation.  A 
GC-based implementation would have object creating and copying being very cheap 
anyway.

> - sharing is prevented.

I'm not sure what you mean here.  When the wrapper layer copies a librdf object 
it's done via a librdf function call, as Lauri mentioned.  In many cases this 
will be implemented by incrementing a librdf-internal reference count, which 
means very little cost; and this wouldn't prevent sharing.

The problem I mentioned in the other message came about because one of these 
librdf copy operations was accidentally a shallow copy rather than a deep one, 
and if I recall correctly may have been only a documentation bug.

All the best,

Norman


-- 
Norman Gray  :  http://nxg.me.uk

_______________________________________________
redland-dev mailing list
[email protected]
http://lists.librdf.org/mailman/listinfo/redland-dev

Reply via email to