[vox] Distributing FOSS for Windows vs only distributing Linux at events [fwd from a thread on sf-lug]

[Bill Kendrick]

So discussion came up over on the SF-LUG list when someone mentioned
creating FOSS-for-Windows CDs to hand out (for Software Freedom Day?)
and others (e.g., Rick Moen, seen replied-to below) argued that the
focus should be Linux, otherwise one's just being 'an enabler.'

Sameer Verma, who actually studies open source as part of his work
at SFSU responded with the message below.  (His long response is at
the very bottom of the post.)

I mentioned I might want to forward this to LUGOD's list, and he
encouraged me to do so, reminding me that _we_ actually participated
in some of his studies a number of years back. [*]

-bill!
[*] who found files from that study sitting in a .tar.gz the other day ;)

----- Forwarded message from Sameer Verma <sverma at sfsu.edu> -----

Date: Sun, 30 Aug 2009 19:19:26 -0700
From: Sameer Verma <sverma at sfsu.edu>
Subject: Re: [sf-lug] mastering CDs in bulk for Software Freedom Day
To: sf-lug at linuxmafia.com

On Fri, Aug 28, 2009 at 3:17 PM, Rick Moen<rick at linuxmafia.com> wrote:
> Quoting Sameer Verma (sverma at sfsu.edu):
>
>> I agree that principally, it makes more sense for a *Linux* user group
>> to hand out Linux CDs at an event such as this (BTW, does SF-LUG have
>> any plans for SFD?), and given that the event is *Software* Freedom
>> Day, and not *Linux* Freedom Day, prinicpally I would not discriminate
>> against the OS on which the Software (application) runs. IMO, it goes
>> against the spirit of some parts of the Open Source Definition.
>> http://www.opensource.org/docs/definition.php
>
> *scratches head*
>
> I'm minutely familiar with the Open Source Definition, having
> participated in OSI's licence review process for many years.  And yet, I
> cannot find any part of it, nor, as the USSC might say, "emanations and
> penumbras", that in any way suggest that it's a good thing, let alone
> obligatory, to cut CDs of MS-Windows software at one's own expense in
> time and money and hand them out.
>
> I've only been reading the OSD carefully for about a decade, so I might
> have missed something:  Does it contain some personal obligation to hand
> out software for proprietary OS platforms, above and beyond the need for
> OSD-compliant software to not be licensed specific to a product or to be
> technology-specific?
>
> (Er, no, it does not.)
>
> And, again, the fact that Sept. 19 has been declared by some guys
> running a Web site[0] to be *Software* Freedom Day doesn't make it
> necessary for desirable for LUGs to hand out software for MS-Windows any
> more than it being International Talk Like a Pirate Day makes it
> necessary or desirable for them to adopt piratical attitudes.
>
>
>> I think its safe to say that if the LUGs plan on participating, we
>> (SF-LUG?) do a CD run of Linux (which distro?) and we (SFSU) will do a
>> run of OpenEducationDisc, which SFD actually ships to teams
>> anyway...we just don't get enough to give out.
>
> Without particular objection (except what I've noted before, i.e., it's
> no skin off my back if you toil on behalf of MS-Windows users), I note
> that OpenEducationDisc appears to be (also) MS-Windows-specific, i.e.,
> to consist only of MS-Windows programs.  Sadly, the related Web page
> (http://www.theopendisc.com/education/) doesn't mention that fact.[1]
> The linked FAQ page does clarify the matter:
>
>  Q:  What are the requirements to run program XYZ?
>  A:  You need to check the website of the program where it will give you
>  exact specifications. Most of the programs on this disk need Windows XP
>  or above and a computer bought within the last few years.
>
>
>> > This particular bad argument does not improve with repetition:  It's
>> > frankly pretty obvious why giving MS-Windows users additional excuses to
>> > remain on that platform doesn't "move the user closer to software
>> > freedom", and never has.
>>
>> I actually have statistical evidence to the contrary (innovation
>> adoption of FOSS research), but I have 20 different things pulling me
>> 20 different ways, so maybe I'll share another time.
>
> Does your research indicate that handing out open source MS-Windows
> applications to users increases the likelihood of them deciding to
> run an open-source operating system, instead?  It doesn't seem, from
> your brief description that it does, which would make your contention
> somewhat non-sequitur to the point.
>
> More to the point, the truly relevant question is whether the same
> amount of money and effort applied to the actual Linux or BSD
> communities has a greater or lesser benefit.  I think it's pretty
> obvious that the benefit is greater.  Not that I have any business
> criticising anyone else's decision to help MS-Windows users -- but
> personally I would rather devote my free time elsewhere.  (I'd
> classify my, theoretically, doing it as "consulting". ;->  )
>
>
> [0] Basically, consultant Matt Oquist, in conjunction with
> Phil Harper of the OpenDisk / OpenCD effort -- though they've got
> several other folks aboard, in the several years since.
>
> [1] An alert reader might also figure out that fact, from the main
> page's citation of contents that are obviously Windows-specific, such as
> Clamwin and WinSCP.
>
>
> _______________________________________________
> sf-lug mailing list
> sf-lug at linuxmafia.com
> http://linuxmafia.com/mailman/listinfo/sf-lug
>

This is a somewhat long answer, but it addresses a few different parts
to the problem at hand.

Diffusion and adoption of an innovation (FOSS as an innovation) has
been studied for around 50 years and has a well established record
across various fields such as agriculture, healthcare, technology,
etc. In the realm of technology, some of the topics that have been
studied are: facsimile, Internet, E-mail, E-commerce, WWW, streaming
video (my dissertation's proposal document is still up at
http://verma.sfsu.edu/profile/phd.proposal.pdf) and of course, FOSS.
Here's a link to one of my papers:
http://aisel.aisnet.org/amcis2005/429/

Most of the work done in this area leads back to Everrett Rogers
(http://en.wikipedia.org/wiki/Everett_Rogers) and Frank Bass
(http://en.wikipedia.org/wiki/Frank_Bass). There are two stages to
this process: diffusion, and adoption. Diffusion is when the knowledge
about the innovation spreads via various channels (usually
word-of-mouth and mass media, propagated by the network effect as
well) which eventually leads to adoption (or rejection). Adoption
implies significant acceptance and use. The S-curve that represents
stages of adoption are broken into:

* innovators
* early adopters
* early majority
* late majority and
* laggards.

I would estimate that people like Rick Moen (and others of his
profile) would be in the innovators category, and bulk of the mailing
list would fall under early adopters and maybe spill over to early
majority. Populations to be addressed by events such as Software
Freedom Day are the early majority going through the persuasion stage.
We still haven't seen the late majority and laggards yet. Note that to
be a laggard, one has to acknowledge that the innovation exists in the
first place. The population at large is largely oblivious to FOSS. See
more at http://en.wikipedia.org/wiki/Diffusion_of_innovation

Once diffusion happens effectively, the decision to adopt gets
influenced by various attributes. This is where a radical concept
takes hold. The decision to adopt is very strongly influenced by
*perception*. The *perceived* attributes play a stronger role than the
*real* attributes. For instance, it does not matter how good the
technical merits of Compiz or Beryl may be, the decision to adopt (by
the population at large) will be influenced more by a Youtube video of
a spinning desktop cube! This is the same reason that posters of
cowboys on horses sell more cigarettes and scantily clad women
standing next to hotrods sell more hotrod magazines. I am sure you can
think of oodles of advertisements that have characters that have
nothing to do with the real attribute of that product or service.
Perception is key. Swallow that pill, and you'll be good at marketing
to the masses. Hey, don't balme me. I don't make the rules!

Some of the key perceived attributes are:

Relative advantages. Sample question: "What relative advantage would I
get when I use OpenOffice over Microsoft Office?" Answer: "Well, you
could print PDFs directly from all the office apps without installing
Adobe Acrobat!". At this point, waxing poetic about the FOSS merits of
OOo, availability of source code, etc. wouldn't help much with
adoption.

Compatibility (with current work environment). Sample question: "Can I
open Microsoft Word files? How about Excel?" Answer: "Yes, not only
can you open files, you can save the files back into their original
format of doc or xls." At this point, explaining the ODF vs OOXML
debate and why they should save only in the ODF format wouldn't help
the cause.

Complexity. Sample question: "How hard is it to use OpenOffice?"
Answer: "Oh, its quite easy. Its very much like Microsoft Office. Very
easy to install. If you don't like it, all you have to do is uninstall
it. Here's a CD. Hope you like it". Handing a CD is one step closer
than handing out a download URL.

Observability (observing someone use FOSS in the field for everyday
work). Sample question: "Wow, how did you print that PDF? Do you have
Adobe Acrobat?" Answer: "No, its built-in as a feature in OpenOffice."
Observability goes a long way in helping with adoption. I use Ubuntu
on my Thinkpad X300, and while the laptop does a terrific job of
suspend/resume, I make sure I plug into a VGA projector in class and
reboot the machine so that students get to see the bootup sequence and
watch the Ubuntu/GDM login screen for a minute or two. They also see
me use OOo, Firefox, GNOME, etc. in every class. Does it work? You bet
it does. Approximately 20% of my students come by my office asking
about OOo or PDFCreator, Ubuntu.

There are a few other attributes like trialability, volntariness,
image, etc. but you can read about it at
http://en.wikipedia.org/wiki/Diffusion_of_innovation

So, as per diffusion and adoption research, to push people in the
knowledge and/or persuasion stage to adopt FOSS, its very important
that we address not just the relative advantages of FOSS over what
they already use, but to address the compatibility with existing
systems (which in all likelihood will be a Windows environment).
Excluding Windows users defeats the purpose. After all, we don't need
to persuade a Rick Moen or a Christian Einfeldt or even a Alex Kleider
(much more recent adopter)! In fact, if I remember correctly,
Christian was a Windows user, when one of his friends introduced him
to FOSS. Many people also do not have the option to install a new OS
on their work machine, even though they can install apps. Why should
we exclude them? Many people use Windows because of a favorite game or
application they need for work. Why exclude them? I personally dual
booted between XP and Redhat/SuSE/Lycoris/Debian/Ubuntu for five years
in my work environment. I still have Windows XP in a VM on my laptop
because sometimes I have no choice but to use Windows at work. I am
thankful to those who helped me along the way to make a complete
switch. Except for the Windows XP VM, I have *nothing* that runs a
proprietary OS. Even my car music system runs on Linux!
(http://www.phatnoise.com/)

My other point about the spirit (not just the letter) of the Open
Source definition (http://www.opensource.org/docs/definition.php) was
that items 5, 6, 8, 9 and 10 convey a sense of non-discrimination,
which goes both ways. After all, writers of code for Firefox or
OpenOffice did not have in mind that they would or would not help
someone on a non-free OS. Heck, they even produce and release binaries
for Windows - something the license does not require, but they still
do so. Why? Because adoption happens gradually. Compatibility has to
be addressed if you want to build critical mass.

Furthermore, if one thinks of the entire system divided into layers such as:

--------------------
   Applications
--------------------
 Operating System
--------------------
        BIOS
--------------------
     Hardware
--------------------

Helping people with a free OS but not a non-free OS is discriminating
on the basis of their OS of choice. Any Adobe Flash users on this
list? You might be next! Why stop there? How about the BIOS? How many
list members are guilty of running a proprietary BIOS? At least my
OLPC XO-1 runs on OpenFirmware! I am willing to bet that most of your
computers run on a proprietary BIOS. Should we stop helping you? Maybe
bill you by the hour for supporting a tyrannic BIOS industry?

I agree with david at sterrit.com that its a spectrum of users out there.
It does begin with an itch for Firefox or PDFCreator, or WinSCP (hey,
help people transact securely!) and goes on from there. In closing
I'll say that while I do believe that a LUG should focus on Linux as
its primary vehicle (might I go so far as to say No BSD? That's
another thread >8-)~ ) not addressing the "unwashed masses" results in
"preaching to the choir". I know I don't need any conversion myself.

cheers,
Sameer
-- 
Dr. Sameer Verma, Ph.D.
Associate Professor, Information Systems
Director, Center for Business Solutions
San Francisco State University
http://verma.sfsu.edu/
http://cbs.sfsu.edu/
http://is.sfsu.edu/

----- End forwarded message -----

-- 
-bill!
Sent from my computer