Re: slots request

[Nikos Mavrogiannopoulos]

On Mon, Apr 9, 2012 at 10:28 AM, Bastien <address@hidden> wrote:

>> Thus we can safely say the amazing should be at least 10, and
>> recommended be 14?
> This suggests the implicit rule that, if GNU gets 10 slots for
> amazing projects, we will have to give one slot per project.

Why is this a problem? Couldn't each project have one amazing
proposal? If some projects do not have such proposals they should
grade them as 4.

> I suggest mentors review all projects (as we are supposed to do,
> right?), rate them 5 if they find them "amazing", and below five
> otherwise.

I disagree with that. I am a mentor for the gnutls project. How am I
supposed to grade a lillypond or octave proposal? How could I grade
something that someone else will  supervise? I believe the people who
do the actual supervision should have the last  word, because this SoC
is ought to be fun for all parties. Moreover, there are issues like
the following.

1. "Elliptic curve optimization" vs "DTLS heartbeat implementation".
Which is more amazing from the other? Even as the one who proposed the
projects I have a hard time putting such stamp on them, as all student
proposals look very good. For someone outside gnutls they might sound
meaningful projects though.

2. wget proposal, vs octave proposal vs lillypond proposal vs bison
proposal. How would you compare them? There are no objective criteria
and there can be many subjective. E.g I am not interested in music so
lillypond might not take a good grade from me which is unfair.
Moreover I have missed all discussion on the relevant mailing lists
between mentors and students.

> "Amazing projects" will be the ones with only-5 votes.  If we
> have less than 5 amazing projects, let's get to another threshold,
> and pick up projects with lower notes.
> What do you think?

I agree on the idea, but not on the implementation. I believe the ones
who agreed to mentor, or other mentors who are somehow related to the
project should vote. The people of a project should be allowed/trusted
to set a proposal on their project as "amazing", otherwise big
projects with many mentors could set aside the smaller ones (and might
not even be intentional).

Otherwise we must set some concrete objective goals for mentors to
vote, which is not feasible given the deadline (and might still not be
fair because the number of mentors varies from project to project).

regards,
Nikos