Re: my annual complaint

[Thomas Lord]

On Mon, 2009-04-13 at 21:32 +0200, address@hidden wrote:

> I don't agree with your negative asessment of the GSoC program on
> several grounds; but I do not really have time nor inclination to
> discuss this now.

That's fine.


> Regardless of that, there are two questions I'm going to ask you. First,
> do you really think that the GNU project can realistically gather the
> $50000 for a GSoC-like program? (Ignoring the various GNU projects
> participating as organisations on their own -- if you want all of these
> to go for an alternative program too, the sum required would be more
> like $300000 or so...)


I don't have a crystal ball so I can't tell.  
All that could happen is to try.

$50,000 is less than 10% of what the FSF raised in
at least one recent year I checked.

$300,000 would represent a huge increase in
funds raised but I don't think it is the FSF's job
buy all of those other projects back from Google.

$50,000 is not an easy amount of money to raise, no
matter how you slice it but I would *guess* that
the FSF could do it *if* it prepared by articulating
a clear case for why it is desirable and what will
be done with the funds.

It would take some careful thought.  For example,
GSoC has the *practical* benefit of imposing 
very little administrative cost on the FSF.  An
in-house program would consume extra resources.

By the way, what do you think of the format of
GSoC?  Is that the best way to do it?   Aside 
from the Evils of Google generally and the 
surveillance and viral marketing to students
specifically, the format is something else I don't
like about GSoC.   It is too easily gamed by those
in it for money rather than a good student experience.
Even for those in it for a good student experience
it is a lousy structure.

I think that students miss out by not taking an 
internship in an actual workplace.  I think students
miss out by not having a larger number of older
people around keeping an eye on their experience.
I think when students go off the rails in a GSoC
project it is frustrating and tragic that they
aren't in an environment where people around them
can try to get them back on track.   Thinking back
to good internships I experienced when I was younger:
much of the value of the experience was outside the 
main work whether it was socializing with more senior
people in the field, or talking "around the water cooler",
or joining a group to attend a lecture and discuss it
after - that kind of thing.   GSoC instead rewards
students for learning to be semi-anonymous worker
bees with an Internet API rather than well rounded
participants in an intimate professional culture.


> Second, do you really think that without the popularity of Google in
> general and GSoC in particular, we could find many good student
> candidates?

That would depend on what the FSF proposed to do with
the money (and on how well it advertised the opportunities).


> I know that for the Hurd at least, the GSoC project is *extremely*
> beneficial, and unless you can propose a realistic alternative, dropping
> out is indiscutable.

I guess that I'm old fashioned because where I
come from, way back in ancient history, what you
just said there is a sure sign of a very suspect
internship program.

The perfect summer internship gives the student
the opportunity, the tools, and access to the 
teaching to *possibly* and *with a little luck*
accomplish something great.   The same internship
tries to put a little gentle pressure on the student
to work seriously and at least try to be useful.
But there are higher priorities than those:

A good internship program gives a student a chance
to begin to see and experience the profession they
are working towards and to understand it better
(something GSoC utterly fails at).

A good internship program gives a student a chance
to develop professionally and as a human being.

The very last thing one wants to hear about an 
internship program - at least where I come from - 
is that it is important to the success of the
project into which the student is hired.
It is supposed to be the student's chance to
shine, not a chance for the host project to form
a reliance on the student.

In fact, if a manager says "We don't have enough
workers here and we're almost out of budget.  Let's
ask the firm to give us a student intern," then
that is a red flag warning that (a) that manager 
should not be allowed within 20 yards of a student
(b) that project is in trouble, quite possibly through
mis-management.

Where I come from, when people tell stories about
"internships from hell" the story often includes 
elements about managers who decided to rely on interns
for something critical.

A good internship program is supposed to be a 
professional duty (of the host organization) - 
the main goal is to "pitch in" on education.

A good internship program is supposed to be a
way to share the wealth.  E.g., a successful
project has been coming in under budget or the
revenues from it are doing surprisingly well
so there's a choice:  give everyone a tiny bonus
or buy a nice pool table for the break room
or take everyone on a field trip to wine country
for a "team building retreat" (get drunk and have
fun) or ..... nah, let's host an internship.

Alice over there has a free enough schedule and
a good bs-cutting demeanor to be the student's 
boss.  Bob, down the hall, has a couple of months
before his XYZZY project gets crazy busy and he's a 
great teacher to stick his head in every now and
again.  4 of the folks on the team went to the 
university down the street where we can recruit
so there's plenty of shared culture around.  The
student can minimally just help fix simple bugs
in the new FrooBaz driver but if they look like
they can handle it, give them a shot at the 
SpingBrt module (and, if they blow it, Carol will
get to it in September).

I don't mean internships are supposed to be
ecstatic and painless experiences.  Not at all.
Perhaps the student will run afoul of Difficult 
Dave who doesn't like anyone.  Maybe the student will
waste too much time listening to the ramblings of Eccentric 
Elaine. Those are realistic working conditions
and in this case - the student's first exposure
to such things - there are other adults watching
out of the corners of their eye to give some
gentle guidance and help the student navigate.

None of that is present in the GSoC format.

Saying "project X will have trouble making progress
without the cheap labor of students" is pretty 
much the opposite of what I think of as a good
internship program.

And what about reporting?   If there's to be 
reporting to a third party about the performance
of an intern it's supposed to be (in the stone
age I come from, at least) to the teachers closest
to the student in the student's academic life - 
teachers being the professionals in the situation
who are supposed to use those reports to the benefit
of the student.  How in god's name we get from there
to a system of reporting that amounts to tattling
to a prospective future employer is beyond me....
Hey, give a bad report about your student in GSoC
and, gee, maybe you're hurting their future employment
opportunities even though the real reason for the 
fail was that the student's parents were going 
through a messy divorce which you would have known
if you interacted with the student other than on the
"X-dev" mailing list.

I believe you when you say that asking the Hurd
project to not use GSoC is beyond discussion.
I'm sure it is.

That makes me sad.

I should just be resigned to it:  let's just
redefine our profession being a kind of anonymous
human-hosted web service with a mailing list
API that you feed in sub-market (and sub-living) 
wages and get out needed code.

Maybe these kids are learning everything they
really need to know, after all.

-t





> -antrik-
> 
>