Matthew Paul Thomas [2014-01-10 11:19 -0000]:
> This is the sort of thing that the "(by 12.04 standards)" line exists
> for on the graph. But that works only for expanding the scope of error
> reporting, not reducing it -- unless we updated 12.04 to continue
> reporting, but start distinguishing, errors from the previous session.
I got several requests to also backport this change to precise, as
this is considered as very annoying and confusing. For precise this
needs to be implemented in a different way as that still uses
ConsoleKit instead of logind cgroups, but it's certainly intended to
SRU this. Would this address this concern?
Also, crashes that happen at session logout aren't really something
that a user would ever perceive as "unreliable". I claim that counting
these as user-visible crashes was wrong in the first place.
Matthew Paul Thomas [2014-01-10 11:19 -0000]:
> This is the sort of thing that the "(by 12.04 standards)" line exists
> for on the graph. But that works only for expanding the scope of error
> reporting, not reducing it -- unless we updated 12.04 to continue
> reporting, but start distinguishing, errors from the previous session.
I got several requests to also backport this change to precise, as
this is considered as very annoying and confusing. For precise this
needs to be implemented in a different way as that still uses
ConsoleKit instead of logind cgroups, but it's certainly intended to
SRU this. Would this address this concern?
Also, crashes that happen at session logout aren't really something
that a user would ever perceive as "unreliable". I claim that counting
these as user-visible crashes was wrong in the first place.