Re: Retiring as an Application Manager
MJ Ray <email@example.com> writes:
> Marc 'HE' Brockschmidt <firstname.lastname@example.org> wrote:
>> [...] Your AM work was below average and created more work for
>> everyone else down the line. [...]
> In short, the above is a symptom of a misstructured NM process.
> Current NM tests the AM and some myths way too much. NM should be
> about testing the applicant. The AM's knowledge (or lack thereof)
> should not enter into it until their last step, perfoming a summative
> test before passing the profile to FD. If they can't apply the test
> properly, that should become obvious and there should be a clear way
> to deal with that, not left to hints and then locking AM accounts.
Sorry, but that's simply not true. We do rely on the AM to do the most
important part of the NM process by gathering all data needed to assess
an applicant's knowledge and skill. Later steps are solely based on
the report submitted by the application manager, so it should contain
all needed information.
Debian has at no point defined rules what a DD needs to know - as a
common ground, we have some very basic philosophical issues . All
technical skills depend on the interest of the NM. If an applicant is
interested in doing web service maintenance, he will have to show his
knowledge in that area (by packaging such software, explaining common
security problems in such applications, ...), while an applicant
interested in porting will need to display different skills.
I believe that this dynamic system is much more useful than sitting down
and creating one long list of topics that all applicants need to know
about, never going too deep into one subject - OTOH, it enforces
application managers to know about all of these topics enough to ask the
right questions (and actually check them), so doing the NM checks
becomes more than applying a pre-fabricated test.
 All applicants are required to give a signed statement that they
know and agree to uphold the social contract