10:17 < thekorn> RainCT: so the bug is: COUNT(<boo>) is working, but the 
                 sorting by timestamp is not?
10:18 < RainCT> thekorn: Right. So the question is if we want it to work (and 
if so I'll 
                fix it together with a related bug)
10:19 < thekorn> RainCT: yes, we want it to work. picking them by timestamp 
makes sense, 
                 and returning *random* results does not
10:19 < thekorn> it's also an issue in our tests that we don't see this bug
10:20 < RainCT> thekorn: OK, agreed.
10:20 < RainCT> So that this in the bug and I'll try to fix it tonight.
10:20 < thekorn> it's because we are inserting events presorted by timestamp
10:20 < RainCT> The tests are just looking at the envet ids, they should be 
looking at 
                the timestamps instead.
10:21 < RainCT> (I already swapped the order of 2 events, so checking the 
                will expose the problem)

You received this bug notification because you are a member of Zeitgeist
Framework Team, which is subscribed to Zeitgeist Framework.

  Timestamp isn't considered when choosing events by popularity

Status in Zeitgeist Framework:

Bug description:
  Currently the MostPopular* and LeastPopular* methods are choosing a
  random (*) event from within all events matching the criteria.

  Is this desired? I believe it'd make more sense to return that one
  with the highest timestamp for MostPopular* and the one with the
  oldest timestamp for LeastPopular*, or something along that lines
  (there are currently such "timestamp ASC/DESC" statements in the code,
  but they aren't working, so if we agree on this property they need to
  be fixed).

  (* Actually, it's the last one which was inserted into the database,
  but that doesn't mean anything.)

Mailing list: https://launchpad.net/~zeitgeist
Post to     : zeitgeist@lists.launchpad.net
Unsubscribe : https://launchpad.net/~zeitgeist
More help   : https://help.launchpad.net/ListHelp

Reply via email to