[Bf-committers] Google SOC stuff

[Ian Gowen]

I was going to write up some recommendations based on my experience
last year, but Tim just said everything I was going to and more.

The first two weeks I coded full blast but after that I really lost
steam. Since I was more or less on my own, I didn't have anything or
anyone getting behind me and pushing me forward. The blog/progress
report/Sunday meeting thing would have really helped me I think.

I'd also like to see a little bit more dedication from mentors and
from the Blender developers at large- correct me if I'm wrong, but
aren't booleans and the fluidsim the only two projects from last year
that have been integrated into the main tree? It's a little
discouraging to get paid for something that won't even get used.

Anyway, I hope we all can learn from last year's experience. Good luck
to all those who are applying :)

-Ian

On 4/16/06, Tim Wakeham <tim.wakeham at gmail.com> wrote:
> Ok after my experience with SoC last year, I have a few ideas for this time
> round.
>
> Firstly, as Ton/Chris pointed out, most of the participants last year had
> trouble with self directed work from home. There are a few things that I
> think could improve this situation.
>
> a) proposals should have a mandatory timeline and deliverables section, with
> detailed project outline, which can later be worked to, with mentors help.
>
> b) a web form, or blog page should be created that participants must fill
> out like a diary, detailing what they achieved and what they need help with,
> also good for mentors trying to get a grasp on how to help the student.
>
> c) as already discussed, absolutely mandatory participation at the sunday
> meetings where the students should be required to have their blog up to
> date, and their project in line with the proposal timeline.  This also gives
> room for devs and mentors to put pressure the participant if they aren't up
> to scratch.
>
> I disagree with the notion of providing a list of projects and/or enforcing
> that only they can be applied for.  Had that been the case last year, I have
> the feeling we would not have a fluid engine now.  It limits innovation and
> inclusion of projects which could potentially propel Blender forward.  It
> also stops people applying if there isnt something there which interests
> them.
>
> What I do think would be a good idea is to have everyone considering
> entering a propsoal for Blender, to be involved in an online forum.  This
> would be a good place for prospective participants to throw around ideas and
> get feedback from devs on feasibility, changes, project timelines, what to
> write in their proposals (in general terms of course),  and a whole other
> myriad of things that could definitely be beneficial to both the people
> selecting candidates - seeing their participation, eagerness, experience,
> etc etc - and the candidates themselves.
>
> We definitely need a stricter review of projects and the participants before
> projects are settled on, this was painfully evident last year when more than
> one project never even made it off the ground.  I dont see why an online
> interview and/or submitted resume would be unreasonable except for the
> obvious review process rerquired of the organisers.  That said, limiting
> participants to only those with past experience kinda defeats the purpose,
> and should probably be avoided.
>
> I think this year our mentors also need to be far more involved. I'm not
> saying they weren't great last year, but I heard from more than one other
> person that they were having problems contacting mentors or timezone issues
> or whatever.  I understand life always gets in the way of best made pans,
> but in essence the mentors need to make the same committment that the
> participants do.
>
> In the lead up to SoC, and after the 'start gun', I really think there
> should be a couple of coder lead workshops to help familiarise people with
> the Blender codebase, where to find things, how to get it compiling
> efficiently, how to get it compiling debug builds, style guidlines, what to
> do/what not to do....and all the rest, you get the general idea.  I think
> this is a really important thing that was sorely missed by more than one
> participant last year, and they are not necessarily totally obvious things
> regardless of your experience and skill level.  I don't advocate spoon
> feeding, and I definitely encourage self directed learning, since I
> understand how powerful that is from my experience last year, but it is very
> hard to get started if you arent intimately familiar with Blender's code.
>
> Efficiency is a hard nut to crack, especially when there is no face to face
> contact with mentors, peers, organisers, etc etc, and no real
> responsibility, sure money is a big enticement, but without someone
> overseeing the whole thing there is no accountability.  Which is where
> stricter control and oversight from mentors comes in, and again with peer
> criticism through the meetings and blogs.  I'm sure there are other measures
> that could be taken, too.
>
> On the whole, I think we have all learned a great deal from the way SoC
> operated last year, and I'd call upon any of the other participants last
> year to put forward their feelings regarding organisation.
>
> -Tim
>
> _______________________________________________
> Bf-committers mailing list
> Bf-committers at projects.blender.org
> http://projects.blender.org/mailman/listinfo/bf-committers
>
>
>