[TOS] a FOSS development textbook?

[David Humphrey]

> Has anyone looked at the possibility of creating a textbook for
> teaching FOSS development? The 'Producing OSS' book by Karl Fogel is

Chris and I have talked about this in the past, and even go so far as to 
start writing an outline.  However, the teaching and community 
involvement trumped this.  All that to say, yes, we've certainly thought 
about it.

> good, but I didn't find it to be sufficient as a textbook for the ANU
> course. It was a little too focussed on starting a project for what
> Bob and I wanted to do.

This is my issue with it as well, since our model is to take people into 
existing large open source projects/communities vs. starting your own.

This raises another issue for me.  I find the balance between 
resources/materials/curriculum/etc. and authentic open source 
experiences a hard one to keep in check.  Many people put their focus (I 
think wrongly) on having things you can hand to students (real or 
virtual).  To do this, you have to freeze your view at a certain point 
in time, or choose to focus on things that don't change.  It can also 
mean that you choose to focus on things that were, but are not now, the 
community's focus.

If this approach is taken to an extreme, you tend to see what we have so 
much of now in traditional computing education, namely, canned projects 
that can be controlled for variation with offerings.

My own view is that with open source, knowledge is embedded within the 
community itself.  What I personally want, and strive to achieve in the 
Mozilla context, is a focus that puts students into the community as 
contributors, which means insuring that they have work to do that is 
'now' and of value to the project (i.e., things the community will support).

So I am left not knowing the answer to this question: how do I best 
support them by means of materials?  Much of what we do is direct 
community involvement and communication, so we spend lots of time on 
irc, mailing lists, etc. helping with the specifics of student projects. 
    We spend lots of time introducing students to people they should 
know, helping conversations happen, and networking.  To this we add many 
labs and discussions of tools, techniques, processes.

How much of this is valuable in a book (I honestly don't know, I'm not 
saying it isn't)?  Let's say I'm going to talk about how to get your 
patch into a bug in Mozilla.  In the time I've been teaching Mozilla 
development, this has meant using CVS and now Mercurial, not to mention 
that the bugzilla flags you need to use have been totally overhauled. 
Do we focus on the things that don't change (this is what I don't want 
to do) or do we make it so relevant that, like the newspaper, it will be 
irrelevant in the next course offering?  This has led me to do much of 
my work upstream, within the Mozilla project itself: putting educational 
documents within Mozilla's documentation proper.

I am also convinced that the best way to teach open source is by full 
immersion, by which I mean using the same tools as your target 
community/communities.  Given that these communities exist via the web 
and web communication tools, do we do the right thing by not forcing the 
students into the same space, handing them a book?  I have two degrees 
in English Literature, so I'm not against books :)  But if the book is 
not the medium of open source, should it be the medium of open source 
education?

> One possibility is for us to collaborate via a site like
> http://cnx.org/, which has a reasonable structure for creating a
> creative commons textbook. It also has the advantage of being linked
> to a print on demand system, so lecturers can ask the university
> bookshop to pre-order a sufficient number of cheap hard copies before
> a course starts (on the assumption that many students will prefer a
> hard copy).

I hope that I haven't made it seem like I'm closed to this idea--I'm 
not.  When last I evaluated these ideas with Chris, it was in the 
context of two friends who didn't have enough time to do it alone. 
Maybe now in the context of TOS it makes sense to look again.

What do others think?

Dave