RE: Crash recovery strategies

[Drew Adams]

> More generally, how do we feel about automatically sending crash reports
> to the FSF? Of course, we'd send reports stripped of personally
> identifiable information, as is standard practice.

I, for one, might be against that, unless the user explicitly
agrees to it ahead of time somehow, and we let the user know
exactly what (kinds of) info will be sent.

What one person considers personal information (whether or not
it identifies the person) another person might consider not to
be so.

IOW, it might depend on just what is meant by stripping.

For example, the info that is gathered automatically now for
inclusion in a bug report could be seen by a user to include
more info than s?he might want.

Is there a reason to send the report automatically?  Is it just
to avoid interacting with the user because any further interaction
might interfere with the state to be reported (in which case, take
the state snapshot before asking)?  Or is it because such
interaction could itself be problematic (in which case just do
without that particular report)?