full and half projects

[Karsten Wade]

Swiping an idea from Ruby SOC, what do you all think about having 1/2
projects that are smaller in scope, designed to run in 1/2 the time of
a full project, and have 1/2 of the stipend?

== full project ==

* ~10 weeks
* Full scope
* $5000

== half project ==

* ~5 weeks
* Smaller scope
* $2500

[Aditya Patawari]

+1
I like the idea. This will give opportunity to take up smaller projects and
will rescue those who have certain commitments during the holidays.

On Wed, Apr 21, 2010 at 8:43 PM, Karsten Wade  wrote:
> Swiping an idea from Ruby SOC, what do you all think about having 1/2
> projects that are smaller in scope, designed to run in 1/2 the time of
> a full project, and have 1/2 of the stipend?
>
> == full project ==
> ~10 weeks
> Full scope
> $5000
>
> == half project ==
> ~5 weeks
> Smaller scope
> $2500
>
> --
> name:  Karsten 'quaid' Wade, Sr. Community Gardener
> team:                Red Hat Community Architecture
> uri:               http://TheOpenSourceWay.org/wiki
> gpg:                                       AD0E0C41
>
> summer-coding mailing list
> [e-mail]
> https://admin.fedoraproject.org/mailman/listinfo/summer-coding

[Matt Domsch]

On Wed, Apr 21, 2010 at 10:13 AM, Karsten Wade  wrote:
> Swiping an idea from Ruby SOC, what do you all think about having 1/2
> projects that are smaller in scope, designed to run in 1/2 the time of
> a full project, and have 1/2 of the stipend?

+1 here, though I would expect the full amount of effort for any
mentor of such a project.  There may be less to do, thus less code
review, but helping get people involved in the larger community,
communicating, providing advice, doing the mid- and end reviews etc.
would remain unchanged.

[Karsten Wade]

On Wed, Apr 21, 2010 at 10:32:51AM -0500, Matt Domsch wrote:
> On Wed, Apr 21, 2010 at 10:13 AM, Karsten Wade  wrote:
> > Swiping an idea from Ruby SOC, what do you all think about having 1/2
> > projects that are smaller in scope, designed to run in 1/2 the time of
> > a full project, and have 1/2 of the stipend?
>
> +1 here, though I would expect the full amount of effort for any
> mentor of such a project.  There may be less to do, thus less code
> review, but helping get people involved in the larger community,
> communicating, providing advice, doing the mid- and end reviews etc.
> would remain unchanged.

+1 to that, too!

We'll have to put that in to the description.  The producing part of
the project (code, content) is the only part of the project that is
cut in half.  The how-to-be-part-of-a-FOSS-project is going to take
the same effort.

- Karsten

[Toshio Kuratomi]

On Wed, Apr 21, 2010 at 08:13:53AM -0700, Karsten Wade wrote:
> Swiping an idea from Ruby SOC, what do you all think about having 1/2
> projects that are smaller in scope, designed to run in 1/2 the time of
> a full project, and have 1/2 of the stipend?
>
> == full project ==
> ~10 weeks
> Full scope
> $5000
>
> == half project ==
> ~5 weeks
> Smaller scope
> $2500

One word of warning here -- even if the student has gotten acquainted with
the project before the start of the five weeks, there's still some amount of
time that they spend getting a feel for stuff.  For instance, they may know
who they're supposed to talk to and what the source code looks like but they
may still not know how best to submit code for review.  This kind of stuff
will be fairly constant whether it's a full or half project so,
proportionally, you have less time spent coding in the half project.

Another thought around this -- have we considered a formalized bounty system
for small tasks?  I think it takes quite a bit more work to organize since
these need to have clearly defined deliverables, be a small enough chunk
that you're never left wondering if the bounty is being fulfilled by the
person who took it, and have enough of them that the students looking at
them have adequate choice to choose from.  But they seem to me to be
a better method of organizing small tasks since they're very much about
providing a deliverable rather than working on a feature specification where
there's a question of whether partial work still deserves to be paid for.

I think the google highly open participation contest took this form:
  http://code.google.com/opensource/ghop/2007-8/

-Toshio

[Karsten Wade]

On Sun, Apr 25, 2010 at 11:11:30AM -0400, Toshio Kuratomi wrote:
> One word of warning here -- even if the student has gotten acquainted with
> the project before the start of the five weeks, there's still some amount of
> time that they spend getting a feel for stuff.  For instance, they may know
> who they're supposed to talk to and what the source code looks like but they
> may still not know how best to submit code for review.  This kind of stuff
> will be fairly constant whether it's a full or half project so,
> proportionally, you have less time spent coding in the half project.

That's a fair point ... we might want to visualize half projects as
being less than 1/2 of the coding to make up for this.  Say, 1/3rd the
coding of a full project?
> Another thought around this -- have we considered a formalized bounty system
> for small tasks?  I think it takes quite a bit more work to organize since
> these need to have clearly defined deliverables, be a small enough chunk
> that you're never left wondering if the bounty is being fulfilled by the
> person who took it, and have enough of them that the students looking at
> them have adequate choice to choose from.  But they seem to me to be
> a better method of organizing small tasks since they're very much about
> providing a deliverable rather than working on a feature specification where
> there's a question of whether partial work still deserves to be paid for.

If you haven't seen OpenHatch.org, take a look.  They are putting
together a cross-project system for tracking such small/bite-sized
projects.  (It's also tied in to reputation building and such, so it
has value for us and the student beyond just the getting things done.)

Some of us have been talking with OpenHatch on a few fronts:

* Getting a method to identify and scrape easy fix bugs from
  bugzilla.redhat.com

* Similar for small project needs from Fedora Hosted.

* Manual insertion of project pieces that made be or add up to a
  summer coding project.

There wasn't time to use OpenHatch for this season, but I'd like to
see something happening perhaps by the fall for the next season.

- Karsten