[haiku-development] Re: Offer an extension to successful GSoC students
- From: Adrien Destugues <pulkomandy@xxxxxxxxxxxxx>
- To: haiku-development@xxxxxxxxxxxxx
- Date: Tue, 2 Jan 2018 16:12:06 +0100
On Tue, Jan 02, 2018 at 03:11:04PM +0100, Dario Casalinuovo wrote:
It does cost mentoring effort.
Getting some support to those students is mandatory anyway.
Also I don't know if this is a wise way to spend Haiku inc money. We
used to run the Haku Code Drive as a complement to GSoC, and eventually
the idea was scrapped and the money used for contracts for known and
skilled devs - who can work on their own without need for a mentor.
I didn't propose another HCI or an Autumn of Code, but something where the
right student can work with more autonomy, i.e. don't have an assigned
mentor, but still get help from the devs.
Now I think you are contradicting yourself.
Either the students are autonomous (obviously they can use the mailing
lists and IRC channel to get help), and this is no different from our
existing contract infrastructure. In that case there is nothing to do,
everything is already in place. But we need to improve our communication
and raise awareness of this (maybe an "how to apply" page on haiku-inc
website?)
Or, we are going for something more pro-active towards the students,
with a list of ideas, active mentoring, etc. In that case it is up to
the project to set up the ideas list (but we can use the GSoC one),
select the students who deserve it, and make them an offer. In this case
it would require more involvement of our development team, and we are
already quite busy year-round with GSoC and GCI. So it may not be that
easy to put yet another similar thing in place.
So in my case I would prefer to stay with the first option: it is
possible for people to get a contract with the inc, but I think at the
very least they would need to first earn commit access. None of our GSoC
students did this year, and in the two previous years there was no GSoC.
So, I would say this currently does not apply to anyone. But we can
think about it for GSoC 2018.
I don't know, I think getting the right student(s) involved after GSoC may
be beneficial if done correctly.
Getting students involved should happen *during* GSoC. If that doesn't
work, my conclusion is that they are not that interested anyway and were
doing it for the money (or the T-Shirt), but not for Haiku. Or maybe
there are other problems (communication again? bad mentoring?) and the
students did not get involved as they hoped. I'm afraid
throwing more money at the students is not going to improve this. We
need to make the project a great place to work and have exciting things
happening, so that talented people join us because they think the
project is interesting and going in the right direction.
This year there have been no contracts, and the Inc has allocated funds
to the coding sprint, where attending devs did not have to pay for
anything (accomodation, food, etc) and GSoC students were even refunded
for part of their travel expenses (we had Hy Che coming from Viet Nam to
attend the sprint). I think this had a very positive impact on the
project progress (several important milestones were reached during or
around the sprint) and also for the motivation of the people attending.
I wasn't motivated to get into the coding sprint anyway, but I didn't know
that. The page doesn't state that clearly, we are still in the who-know
world.
It was the first time I organized this, so there are some rough edges. I
will make a better web page next year (assuming the Inc will still be
funding this in a similar way).
NOTE: Something similar happened for GSoC selections, even if I was a
mentor and I had right to vote, *I didn't even get into the ml*, even if I
asked two times in IRC.
You can imagine how frustrating it was for me to be the GSoC org admin
and not having admin access to that mailing list so I couldn't invite
you. But this has been fixed, eventually (and you are a member of
the list since april 2017, so your issue has been solved as well).
Now that I have access to the ML management, the 2018 edition of GSoC
should go more smoothly on that side.
So it's not like Haiku inc is sitting on thousands of dollars and not
willing to spend them.
Please don't continue to tell me that, I don't want to get into this
discussion anymore. I have at least two examples of how Haiku Inc. didn't
handle the thing correctly. The problem isn't rejecting contracts, the
issue are fundamentals, *those people didn't even get a reply*. And for
reply I mean a yes or no. Someone even stated they didn't have time to
write a +/- 1 in their voting discussion.
The inc is also trying to improve. The current bylaws requires that an
absolute majority of the members agree on any decision, which, combined
with no system to kick out inactive members, means it is too easy to get
things stalled (basically, 2 members are MIA, one forgets or doesn't
want to vote, and you can never make any decision). They should rework
this so relative majority is enough, or force people to vote or exclude
them from the org. They have some other problems to solve first, however
(like not having any financial reports for the last 7 years or so). I
hope the people currently running things will manage to put the inc back
on track.
But still, I faced no problems with the funding of the coding sprint.
The event was funded in time. I know there have been some problems in
the past, I think the inc is aware of it as well, and they are working
to improve things. So let's not look too much at past problems and focus
on the present and future?
Anyhow I proposed an idea that I think can be discussed and extended to
make something good for the project.
That's fine, I'm not trying to turn it down at all costs or something.
I'm just raising some concerns on the actual cost (in terms of mentoring
resources) and the exact way it should be done (do the student apply on
their own? do the project nominate them somehow? how do we make sure the
inc replies in time?).
--
Adrien.
Other related posts:
