On Nov 30, 2007, at 3:56 PM, Mimi Yin wrote:

Back in October, I started a thread on the design list re: weird triage/recurrence/sharing bugs having to do with items being marked Unread and popping into the NOW section for inexplicable reasons.

I believe we've addressed some of the underlying causes (ie. https://bugzilla.osafoundation.org/show_bug.cgi?id=10799 : Need a mechanism to ignore certain EIM fields when determining which items move to NOW or get marked unread). But I believe there are still bugs at large.

'Notification' via automatically popping items into the NOW section is core to the personal information management and collaboration workflows. What do we need to do to make sure we've addressed this problem for 1.0 (which is in reality, less than a month away if you take the holidays into consideration)?

Can we enlist the help of the Users list in some way? Have a dedicated test session? Ideas?
Mimi

Below is a screenshot of my Office Calendar collection after 1.5 days.

Entire recurring series keep popping into NOW but no evidence of global edit
https://bugzilla.osafoundation.org/show_bug.cgi?id=10948
- See IRC QA Session item that is selected in screenshot, that's a good example of this bug. - I wonder if bug 10948 is somehow related to: https://bugzilla.osafoundation.org/show_bug.cgi?id=10795 (Byline does not match Who column)

Odd mixture of events keep popping into NOW in the Office Calendar
https://bugzilla.osafoundation.org/show_bug.cgi?id=10894

Nov 12th, 2007 Weekly PPD Meeting keeps popping back to NOW
https://bugzilla.osafoundation.org/show_bug.cgi?id=11475


The reason why this particular genre of bugs has been difficult to nail down is the simple fact that bug reports always contain the logs of the client that notices the problem on the receiving end. At this point, it's too late -- we need the logs of the client that made the change and this has been somewhat impractical when it comes to the office calendar. What we really need is a controlled environment where this problem is reproduced by two Chandlers running a particular software version, sharing one or more collections that aren't touched by anyone else, and not modified using the web UI. I also have a feeling that these bugs require time to pass so that events are auto- triaged (just a guess), which means it could take longer than a day to reproduce.

~morgen 
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _

Open Source Applications Foundation "Design" mailing list
http://lists.osafoundation.org/mailman/listinfo/design

Reply via email to