#2412 BLOC First D: Buddies not disappearing even when connection does

[Zarro Boogs per Child]

#2412: Buddies not disappearing even when connection does
-------------------------------+--------------------------------------------
  Reporter:  Zack              |       Owner:  smcv                  
      Type:  defect            |      Status:  new                   
  Priority:  blocker           |   Milestone:  First Deployment, V1.0
 Component:  presence-service  |     Version:                        
Resolution:                    |    Keywords:  collaboration         
  Verified:  0                 |  
-------------------------------+--------------------------------------------
Changes (by jg):

 * cc: ypod (added)


Comment:

 In general, you should *not* try to detect failures....

 Reporting connection failures will, in fact, just cause more failures that
 *do not matter* to end users, as well as pollute the network with unneeded
 transactions.  Do *not* use TCP keep-alives.  They are *evil*.  People
 should only get failures which matter; reporting or forcing failures in
 the face of an unreliable network in advance of actual need just causes
 trouble.

 All we really care about is that when the network state changes, or a long
 time has passed, we will eventually clean up buddies.  So I would pay
 attention to NetworkManager signals, and forcibly reconnect if NM tells
 you the network state has changed.

 Note that we'll be able to get much better information from the wireless
 about presence in the future; the Marvell chip keeps quite a bit of
 information (which may or may not currently be available) that can help,
 along with algorithms that Polychronis has been working with and put into
 his activity (a good place to look for ideas and advice). He can give
 immediate advice of what is possible.

-- 
Ticket URL: <https://dev.laptop.org/ticket/2412#comment:19>
One Laptop Per Child <https://dev.laptop.org>
OLPC bug tracking system