RE: [ba-ohs-talk] open source question
Kevin's post is terrific, Bravo! (01)
It rings true. (02)
Open Source is more than a gimmick for propagating OHS, here's a few reasons
why: (03)
1. Robust software development of useable, useful code - if your code is
likely to be exposed to the whole world, you will take more care with it
than you would if you knew that no one was ever going to look at it. (04)
2. Diverse software development approaches win out over monolithic "one
mind, one way" approaches - OHS facilitates collective knowledge
augmentation by recognizing that many people collaborating may have many
different perspectives, many different ways of understanding and many
different preferred modes of viewing and learning. And yet, it is still
possible to collaborate, and it is important to facilitate the collaboration
of diverse minds. Open Source offers many approaches to providing useful
functionality for different users, perspectives etc. (05)
3. Diverse software functionality results - if someone wants to do something
to provide additional functionality, they can do it. There is not
bottleneck, no gate keeper, no one saying: "Did you get the committee to
agree that we want to provide access for (eg) Australian aboriginal
tribespeople who can communicate geographic directions only by singing in a
language spoken by 100 people?". If you want to do it, and think it should
be done, you can go ahead. (06)
4. Supports the evolutionary co-development of OHS by many independent
minds - provides survival (& support and future development) of the fittest
software in the most direct manner: If you write it and it is well used and
you get tons of positive feedback (& maybe even consulting contracts and
speaking gigs), you are more likely to support it and continue to upgrade
it. If no one wants to use it, well no one wants to use it, and you probably
won't even want to do much more work on it. (07)
5. I think it is possibly the fastest way to source and identify celebrity
software developers and celebrate their prowess in a wider audience.
Everyone benefits from this because great software developers are harder to
find than hen's teeth. It's even harder to match great software developers
with the great projects that they can develop marvellously beyond compare.
Open Source lets' great developers follow their passion and let's the world
benefit from the results. Fast. (08)
The interesting corollary question is: Should everything in the OHS be Open
Source? (09)
I believe the answer is no, it does not have to be. (010)
-----Original Message-----
From: owner-ba-ohs-talk@bootstrap.org
[mailto:owner-ba-ohs-talk@bootstrap.org]On Behalf Of Kevin Keck
Sent: Monday, February 18, 2002 8:57 PM
To: ba-ohs-talk@bootstrap.org
Subject: Re: [ba-ohs-talk] open source question (011)
on 2002/02/18 10:36 AM, Johannes Ernst at jernst@r-objects.com wrote: (012)
> [...]
> I mostly hear the price argument in the context of OHS, but that
> can't be all of it -- just changing the list price in different
> countries, or for different purchasers (e.g. schools) would solve
> that issue as well, and does not necessarily imply Open Source. So
> what is it about OHS that requires Open Source? I appreciate your
> insights ... (013)
As you seem to imply, it's not clear that open source is a requirement so
much as it's clear that it's an easy win along a number of dimensions. But
the importance of such wins shouldn't be underestimated. (014)
Andy Grove proposes the "10X" rule of thumb, the estimation that a
potentially disruptive technology needs to be at least ten times better than
the incumbent in order to overcome market momentum and the edge created by
Metcalf's Law (network effects). Since a factor of ten is quite a handicap,
the challenger is well-advised to expoit every gimmick at their
disposal--and as gimmicks go, Open Source is a pretty good one.* (015)
But more critically, it's important to understand that this "10X" hurdle is
not just objective but is subjective: the consumer must perceive the 10X
benefit. And for reasons I haven't found well articulated yet, but which is
readily confirmed by seemingly everyone who's ever tried to market software
to people, it is usually impossible for people to make such an estimation of
the value of a piece of software in any other way than to try using it. This
is more than just observing that there's a learning curve and switching
cost--Doug appreciated and sketched the theory for those pretty well quite
some time ago. The point I'm trying to emphasize here is that not only can
people not use the tool effectively until they've had a chance to learn it,
but they can't even judge its merit until they've had a chance to use it,
first-hand. It's hard to remember (or admit) now, but back in '94 or '95
almost nobody could appreciate Mosaic or the WWW based on any description or
explanation of it, or even based on a live demonstration--nearly everybody
had to sit down at the keyboard and mouse and experience "surfing the web"
first-hand in order to fully comprehend its utility and significance. On
paper and second-hand it seemed so mundane, esoteric and/or simplistic; only
by sitting down and using it could anyone be expected to fully "get it",
that it was indeed a critical mass of functionality which constituted a new
"killer app". (016)
This is why AOL has always spent so much money distributing disk(ette)s, and
why Netscape gave their early browsers away for free, is so people could be
given a chance to try them out, risk free. Many more people are willing to
risk wasting some time trying out a new product they'll turn out not to like
than are willing to pay even a couple bucks up front for an unproven
product. This is for perfectly rational reasons having more to do with
discouraging con-artists than with the relative value of time and money to
the consumer. (017)
And OHS's largely collaborative focus only amplifies the need for
minimal-risk trial, because in order for anyone to genuinely try using it
they'll need to have collaborators using it with them, all of whom would
need to endorse the risks of money, time, and potential vendor lock-in
associated with trying out a proprietary product. (018)
Furthermore, the improvement to productivity will be greatest between
collaborators with the fewest other tools or mechanisms for collaboration at
their disposal (such as geographically-dispersed, informally affiliated
groups with little budget for clerical and administrative assistance) and
who are less worried about missing deadlines than they are about maintaining
sustained co-participation despite such resource limitations. In other
words, the easiest users to recruit would be among the very most difficult
groups of people to win as paying customers. (019)
Of course you know this first-hand: I don't imagine you chose Yahoo! Groups
to host the semanticweb-business list because you think it offers the best
user experience; I instead suspect you chose it because it is free, easy,
familiar, reasonably reliable, and ubiquitous, and doesn't appear inferior
to the alternatives by a factor of ten. ; ) (020)
In summary, you're right, no, it's not really so much about price/cost. But
before you'll be able to make any case that people should expect to pay for
it, you'll first need to convince them that they really want it. (021)
At least, that's my take on it.
--
Kevin Keck
keck@kecklabs.com (022)
* Flame Deflector: Please forgive the cynical choice of words, I'm just
trying to convey a point about marketing which is completely orthogonal to
the actual merits of Open Source, about which I was trying to say absolutely
nothing. (023)