On 30/04/2013 02:35, Marvin Humphrey wrote:
I don't know of any other reference counting mechanisms with such a behavior. For that reason, I have a mild preference for a novel name.
+1 for CAPTURE/RELEASE, then.
For String, though, it seems like one internal encoding matching the primary host encoding ought to suffice.
This should work. But for the Python bindings, it would mean to use the Python encoding for every Clownfish string, even if it isn't passed to the Python side at all. And what about the on-disk index? Since it probably should always use UTF-8, most strings passed to and from Python have to be reencoded anyway. So we could as well use UTF-8 for all Clownfish Strings.
Nick
