[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: fresh blood gets congested: long way to become DD

Brian Nelson wrote on 12/11/2005 02:39:

I know this is an old thread, but as far as I can tell, it's still an
issue. (At least I've been waiting for FD approval for over two months now.)

> It's not purely a problem with the number of AM's.  The bottlenecks I
> see are:
> 1. Some applicants are slow to respond, or give poor answers, so they
>    hog their assigned AM.
> 2. AM's get busy and are slow to respond to applicants.  Personally, I
>    frequently do AM work in short spurts, and then get too busy to do
>    any work for a month or so.
> 3. AM's are not available to have applicants assigned, either because of
>    (1) or (2), so applicants have to wait in the queue.
> 4. Then there's the infamous DAM bottleneck...

(as well as an FD bottleneck)

> A possible solution I see is, instead of assigning one AM per applicant,
> all AM's are made a part of a committee to handle applicants.

I don't think this would work. However, it would be nice if AMs could
choose to pass on handling of an applicant to some other possible AM.
For example if they suddenly are busy in their job or so. Or if two AMs
agree to share their applicants (one doing only the P&P part, the other
only doing T&S for example).
It normally won't work out right if you just send your responses to a
mailinglist and someone on there takes the responses and acts on them.
It might, however, work out if it were done using some decent issue
tracker (aka ticketing system - like requesttracker to name one).

But, from what I gathered during my NM process so far, there are
actually three noticable bottlenecks right now:
1) AM assignment (probably caused by:)
  1a) slow AM responses
  1b) slow NM responses
2) FD approval
3) DAM approval and account creation
Well, what is somewhat irritating is that there are a number of AMs, one
of which has to approve my responses (declare that i passed all his
tests), two FD members, one of which has to approve my responses
(declare that what my AM asked of me is sufficient and answered
satisfyingly) and finally the DAM (though there are two, only one is
doing actual work AFAICT) which has to approve my application (declare
that both AM and FD did their work and only passed me on when everything
was satisfactory). Note that there are three levels, each with less
staff, but each currently reviewing my full application material.

IMHO, this could be improved if the DAM (or even FD) wouldn't need to
review the complete application, but could simply check how many AMs
reviewed an application and signed it. Let me tell you how I would like
the NM process to work:

1) NM applies
2) NM gets advocated by some DD(s)
3) FD does ID/gpg signature check
4) FD assigns _primary_ AM
5) primary AM and NM go through P&P and T&S tests as today
6) primary AM signs NM's application and attaches full mail log
   (or refers to it's storage at some "random" URL), also signed
   and sends all that to some mailinglist (like newmaint?)
7) Any AM which has some free time takes the NM's application
   including the mail log and reviews it. If satisfied, he also
   signs the application. If not, he can take over the NM
  (effectively becoming the primary AM now) and ask more questions
  until satisfied at which point he jumps to step 6
8) When X potential AMs (with X being something like 3 or 4) signed
   the application, it goes back to FD
9) FD does a final review checking wether enough AMs signed the
   application. If so, the application gets signed by FD and passed
   on to DAM
10) DAM again checks wether enough AMs and FD signed the application,
    might do a brief check of the application itself and creates the

Note that (8), (9) and parts of (10) could be automated. Note also that
I think of FD members as some sort of first class AMs, so FD members
could also be the ones reviewing an application.

IMHO, this could make the whole process faster since it takes some load
off of FD and DAM, but it also could add some transparency to the whole

Anyway, just a thought.


Reply to: