Re: [school-discuss] Developer Requesting Feedback Please

[Michael Shigorin <mike@xxxxxxxxxxx>]

On Tue, Dec 04, 2007 at 11:59:59PM -0700, Charles Coss? wrote:
> > The code isn't currently open sourced.  I was hoping to get
> > some feedback from this discussion group as to the advantages
> > and disadvantages of making such educational games open
> > source (I am new to the idea).
> All licensing issues are super-simplified if you are just
> giving stuff away, basically.  Then might as well  just use the
> GPL-v2.

Using well-known license also helps to get the process running;
but it's wise to consider the initial release license well.

E.g. a single release under *BSD license would effectively put
the software in almost public domain (these permit closing the
source of derivatives); a single release under any free software
license means that the source is out in the wild, irrevocably.

I personally would recommend "GPL v2 or later" as a safe default
(and a sane one, usually).

BTW did you look at GCompris framework?  Participating in
existing project (and probably changing the framework as well)
might be more rewarding than doing everything by hand.

> And, if you are just giving something away, then one advantage
> to making it opensource is that you can use all these great
> toolkits and not have to worry about licensing headaches.

Well, if the code is in place then this might only matter for
next-gen... but code reuse (in library form, particularly) is
great time-saver.

Another thing is that opening up the project might (but is not
guaranteed to) help with that next-gen: cleaning up what was 
a quick hack, introducing features or infrastructure that were
thought of but never got the time, and so on.

There's at least one significant moment that needs to be
considered before: just dropping the tarball or zip with sources
somewhere isn't going to make it a real opensource/free software
project, it's involvement and, well, project management that
does.

Basically, some online storage space, a web site, a mailing list
(all of these can be obtained at sf.net although their list
archives are bitter -- why not use plain pipermail?..).

One could learn something from major projects that took ages to
take off like Mozilla, OpenOffice.org, or even from XaraLX what
might be missing in an otherwise incredible decision: a large
and undocumented source tree with only a few of those who
actually know it enough to work on it (and who would rather solve
issues in cafe rather than in public); or even a closed-source
part that was hard to part with (in terms of control).

Of course, learning on mistakes should help avoiding them,
not repeating.

> A friend once told me, and i've never found an exception:
> The Best Software is Free.

I know an awful lot of "exceptions", being Linux user since 1998
and doing some consultancy on migration since ca. 2002.

That just doesn't mean I stop to look for the better pieces
and improve what I can ;-)


PS: is anyone interested in yet-another-LTSP-out-of-box announce?

Our firm participates in Russian school project to some extent -- 
that is, we have integrated LTSP5 into ALT Linux and by October
we made an installation CD.  Last week I've built beta2, and it's
going to be merged with school distribution.

It has enough difference from Edubuntu both in base distro and
LTSP implementation that it well can be either better or worse,
but it's also different (e.g. I couldn't boot our terminals with
Edubuntu 7.10beta5 while ALTSP5 boots them just fine).

-- 
 ---- WBR, Michael Shigorin <mike@xxxxxxxxxxx>
  ------ Linux.Kiev http://www.linux.kiev.ua/