Re: AM Report for Week Ending 08 Dec 2002
On Wed, Dec 11, 2002 at 06:58:17PM +0100, Martin Michlmayr wrote:
> I think I know most AMs fairly well and I get some information about
> NMs from their advocate. I'm fairly happy with the current
> situation, I don't really see how check boxes would help.
Well, if the current situation works for you, then I guess everything
I originally said "it became clear to me that there's a much larger
degree of manual intervention than what I expected". The question is
then what did I expect? My point of reference is the process of
reviewing papers for conferences. In that case the chairman gets a
list of papers to review and a list of potential reviewers. The papers
have keywords which classify it in very large bins. Same thing for the
reviewers. The problem reduces to a sorting one, and you can do 80% of
the work in an automated fashion. The other 20% are the cases where
there's not enough reviewers avaiable or there's no match between paper
The NM assignment problem is not fundamentally different. Ideally
you'd match NMs to AMs with the same interests in order to get a good
evaluation of the NM. You say that the current system let's you do
this matching accurately and efficiently. Fine.
> And now you're arguing against an automatic NM assigment system or
What I said was:
> The process should be a FIFO most of the time: the first person
> submitting an application should be the first person getting an AM
> [...] For some time I thought AM-assignment was a completely
> automated process
If the NM process was really a FIFO, the assignment would be mostly
automated. For most cases NMs would get an AM assigned as soon as
there's one available who is capable of processing the application.
But I'm thinking about the global picture. This bit was something I
happened to mention in my original email.
Obviously it's impossible to have a FIFO because the duration of the
process is not constant, but the point is that what we have now is, for
a casual observer, almost random.