Re: Initial Contact
Andreas Metzler <ametzler@downhill.at.eu.org> wrote:
> You are twisting Andrews argument, the applicant's "job" is not the NM
> application, but his work for Debian.
He twisted my argument, so I think it's fair. In a similar way, does
the above mean that you are saying that your NM process didn't need you
to do any work?
> [...]
> Iirc this idea was already presented about two month's ago, there was
> some talk on this list.
I must have missed or forgotten that. Who was going to do it?
[...]
> I think I've already posted this somewhere: Giving the applicant his
> login is a very important step, because it allows direct uploads to
> Debian, and therefore big possibilities for error or even malicity(?).
I'm not saying that NM is too difficult. It has to be rigourous, for
the reasons you state. Isn't enforcing that rigour the AM's task?
Rather, I am saying that NM seems to move too slowly, maybe because
the applicant gets insufficient feedback, and that there seem to be
too many inconsistencies in the process.
An aside: I am involved with teaching a course with a three-stage
assessment process for portfolios. The candidate does the work, then it
is marked by one of classroom tutors and referred back if necessary. Then
it is sent for moderation by another tutor at the institution, who sends
back inadequate work with two referrals: one to the candidate and one
to the original marker. If the marker's recent error rate is too high,
then they are offered guidance and to resubmit referred work. Finally,
the work is sent for assessment by someone outside the institution.
If they are happy with it, it's passed. If they are unhappy, it is
referred back to the institution moderator to deal with as before.
If that institution is having too many referred back, they are offered
guidance and so on, or prevented from assessing the qualification and
other sanctions.
Now, this process is not perfect and there are bugs in the actual
implementation, but I think the basic design is sound. At the moment,
the AM is behaving like the classroom tutor and DAM is like the external
assessor, but I think we only have one person trying to do the job
between them? So, maybe we need people to help the front-desk? How do
we solve that?
Well, back to the course for the answer. The institution assessors are
normally the more senior classroom tutors. So, I suggest that we look
at which AMs have handled quite a few applicants and are well-regarded
by tbm and elmo, and ask them to look at the reports of other AMs and
endorse or comment on their reports. Normal safeguards to make sure
that people don't have to check their own work, etc.
I think someone else has suggested something similar to this after my
earlier rambling, but I've only just found a good experience that makes
me agree with it. The new bit is that an eye will need keeping on AMs
to make sure their reports are accurate enough.
> The person who holds the position to actually hand out these accounts
> needs to be ultimately trusted by the Debian project. Currently we have
> only one account manager, James Troup, *imvvvvvvvvvvvvvvho* for two reasons:
> * most developers think that process works "reasonably well"
Most developers do not need to go through NM and seem happy to make
jokes about "being stuck in DAMnation" instead of investigating
publicly and potentially being seen as unpopular.
> * nobody with the same trust level as James Troup has stepped forward
> and said "I think the NM process needs another DAM, let me do it."
I think joey is a deputy DAM, but, as I have already said, I am not
convinced that the problem is insufficient people in any one group, nor
that any of these people are defective, least of all James Troup.
There are some odd things in the way applications are handled just now,
from the perspective of someone going through NM.
> Just for reference: I finished my application about six weeks ago, iirc
> it took me about five months, but my memory might be wrong.
You got through quickly, as I think I've said to you before ;-)
Can you roughly give the length of time each stage took for you?
MJR
Reply to: