Open Stack

On Tue, 2011-09-27 at 19:56 -0500, John Dickinson wrote:
> Meetbot's summaries are precise, but they have a non-human flavor to
> them, and they are only as good at the moderator for the meeting. 

Ah, interesting! The way I read this sentence is that the moderator can
improve the quality of the meetbot summaries without having to rely on a
third party to sift through the logs and write a third report. 

I suggest to proceed by identifying the best and worst examples of
meetbot summaries, and ask moderators to get improve the best examples
and stay away from bad summaries. 

I can read the latest ones and see if I spot patterns. Anybody that
feels strongly for this please do the same and send your analysis to the
list for further discussion.

cheers,
stef