Stop dividing history into conversations?

Etan Reisner pidgin at
Tue Feb 12 14:37:59 EST 2008

On Mon, Feb 11, 2008 at 10:14:55PM -0500, Mark Doliner wrote:
> On Mon, 11 Feb 2008 19:30:37 -0500, Etan Reisner wrote
> > P.S. But really, stop compulsively closing the conversation window,
> >  it doesn't help you, it doesn't help your conversation, it likely
> > just makes the windows appearance more annoying (as it likely pops
> > up in front of things you might be paying attention to), and as
> > you've noticed it makes keeping the conversation in your head more difficult.
> Doesn't it seem a bit oppressive of us to command people not to close the
> conversation window?  I kinda feel like, if someone wants to close the
> conversation window then, uh, we should probably let them.
> -Mark

If that was read as a command it shouldn't have been, my point was that
the original poster even admits that his compulsive window closing causes
him extra work because keeping the thread of a conversation swapped in is
made more difficult. Further, most window managers handle new windows in
ways that are decidedly sub-optimal (as can be seen by the constant
requests we get for not "stealing focus" on new windows, and the "hide new
IMs" preference) which means that compulsive closing makes effective and
efficient workspace management a harder task.

Nothing about our current scheme prevents people from closing the windows
as often as they want to, whereas not breaking the logs up does in fact
prevent people from having logically encapsulated log files. We also made
it a point of having logging extensible from plugins, perhaps one of the
many people who don't want per-conversation logs should actually step up
and write a better logging plugin (side note: Does the new immediate close
behaviour complicate writing logging plugins?), I will be more than happy
to include a per-day logger in pidgin.

Finally, there was a reason I made my comment via P.S. it was to indicate
that it was outside the normal realm of help/support and was intended to
be taken as guidance for why the current practice is probably, at its
root, unhelpful in many ways.


More information about the Support mailing list