[PLUG-TALK] Open Source Mini-essay

Keith Lofstrom keithl at kl-ic.com
Fri May 6 17:11:34 UTC 2005


A local group of software developers plans to discuss open source
software today.  I submitted this in lieu of attending (sorry, invite
only).  I thought it might entertain PLUG folk:
------------------------------------------------------------------------

I am really torn between attending your meeting, and attending the POSSE
meeting at the same time.  POSSE will have representatives from OSU's
open source online archive, and someone from the Beaverton Open Technology
incubator, discussing open source.  I will probably go there simply
because the time commitment is smaller and it is less crowded.  So
I will throw in a few thoughts here.

I am managing an open source project;  you can look at www.dirvish.org ,
and the magazine articles it points to, to see how some of it works at
ground level.  At the large scale, it is a "movement", but in detail it
is essentially a way of getting software made that forgoes both the 
major costs and the occasional benefits of creating proprietary software.
I have a *lot* of opinions on the subject, and could belabor people at
length about them, but in essence open source is a participatory
experiential process.  Until you've participated, and seen the strengths
and the weaknesses first hand, it is hard to understand why it works,
and for what.

What fits the open source way of doing things?  Anything that needs
public scrutiny from a wide variety of viewpoints.  Any process that
some programmers use (and programmers are people first, and use most
processes).  What doesn't fit?  Anything that you want to hide
because it is ugly and needs to stay that way.  Anything that is
used entirely by non-programmers, or whose programmer-users can't
afford to contribute (i.e. CAD software, typically used by people
with no time to spend improving it).  Open source is like a dominant
gene - over time, it tends to displace recessive proprietary projects;
it is common for open source to gain user share, it is rare for 
proprietary software to win users back.  

Most of the roundtable members have an academic CS viewpoint;  from
that viewpoint, open source "doesn't work".  It is not designed or
planned very much;  typically it just evolves in a Darwinian way.
A good discussion of why this works can be found in Eric Raymond's
"The Cathedral and the Bazaar":
   http://www.catb.org/~esr/writings/cathedral-bazaar/

Understanding how an unplanned system made from unreliable parts
can grow into a highly reliable system is a problem for creationists,
politicians, and some academic engineers.  Nontheless, it works, 
even if the successes are hard to explain.

Again, I could hold forth for hours, and you probably don't need
that at the roundtable.  So let this tiny essay be my contribution.

--------------------------------------------------------------------------

Keith

-- 
James Shore - Titanium I.T. LLC - Successful Software
Upcoming presentation: "Offing the Offsite Customer"
Oregon Graduate Institute, May 10th. http://tinyurl.com/9vucu

phone: 503-267-5490
email: jshore at titanium-it.com
web/blog: http://www.jamesshore.com


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

-- 
Keith Lofstrom          keithl at keithl.com         Voice (503)-520-1993
KLIC --- Keith Lofstrom Integrated Circuits --- "Your Ideas in Silicon"
Design Contracting in Bipolar and CMOS - Analog, Digital, and Scan ICs



More information about the PLUG-talk mailing list