On Sat, May 16, 2009 at 7:53 PM, Emmanuel Lecharny <email@example.com>
Can you be a bit more explicit, Alex ?
Alex Karasulu wrote:
On Fri, May 15, 2009 at 5:29 AM, Emmanuel Lecharny <firstname.lastname@example.org>wrote:
Howard Chu wrote:
Emmanuel Lecharny wrote:
"Presence" has established history in the protocol specs...
while applying Kiran's patch for new system index, I found that we
should fix some inconsistency. We have what we name 'Existance' index,
which is the 'Presence' index.
First, we should get rid of the 'existance' name, and replace it by
Second, we should keep only one of the 'existence' or 'presence'. Unless
I'm totally wrong, they are just the same. The question is which one
should we keep ? I would prefer keeping 'existence' rather than
presence, because it's more explicit.
As I replied to this mail in my response to your mail about the EntryCSN
(I'm a bit confused ...) :
- let's remove the exist[ae]nce and replace it by 'presence' (and not
- let's keep using a BTree for all those guys (UUID and entryCSN). We will
have to find time to experiment how many micro-seconds we may spare by using
a HTree on UUID later ;)
None of this context makes sense to me here Emm.
I guess you crossed the content of two separate threads here. I first saw all this existence and presence conversation then this CSN and UUID attribute stuff. It confused me a bit until I read the other threads.