[EAS]The Right to Tinker

pjk pjk at design.eng.yale.edu
Sat Aug 3 20:56:27 EDT 2002

Subject:   The Right to Tinker

[This has been sitting in my EAS-INFO prospects folder for a while,
because I have mixed feelings about it. Probably too long and
self-indulgent. On the other hand, you're not going to get anything
from me for about three weeks now, during which I'll have a
technology-free (except for my digital camera) time in the
Northwoods of Maine, then followed by catching up with a monster
email backlog.

So here it is. Have a good rest of the summer.  --Peter

|  Peter J. Kindlmann     |  Prof.(Adjunct), Director of Undergrad.  |
|  Dept. of Elect. Engrg. |  Studies and the Morse Teaching Center   |
|  Yale University        |  tel.(203)432-4294, fax (203)458-3803    |
|  New Haven, CT 06520    |  email: pjk at design.eng.yale.edu          |
|        http://www.eng.yale.edu/EE-Labs/morse/about/pjk.html        |
Dear Colleagues -

This mailing could also be called "the changing ways of owning
technology" My long-ish, and partly auto-biographical, comments
below are prompted by four recent articles.

-- "Edward Felten, a professor at Princeton University, argues that
the 'freedom to tinker'--the right to understand, repair and modify
one's own equipmentÑ-is crucial to innovation, and as valuable to
society as the freedom of speech."

-- Technologies that inhibit end user innovation, starting with
digital rights management (further explained next).
http://www.nytimes.com/2002/07/04/business/04SCEN.html (free sub req'd)

-- The scary implications of TCPA and Palladium, by Ross Anderson, a
computer scientist at the University of Cambridge.

It's not just your imagination -- planned obsolescence has become a
firmly  entrenched marketing strategy for most makers of electronic
equipment. "We  joke that we design landfills," says a senior
industrial engineer at  Pentagram Design, which builds portable
devices and computers for companies  like Hewlett-Packard. In the
past year Dell Computer has slashed warranty  periods from three
years to one, and Apple's iPod digital-music player  features only a
90-day warranty. Sony requires buyers to fill out a lengthy 
questionnaire in order to qualify for a full year of support on a
Clie  organizer -- otherwise they get only 90 days. At the same
time, companies  are making it more difficult to get items repaired,
even if the customer is  willing to pay for it. Many PDAs from
companies such as Handspring, Palm  and HP have built-in
rechargeable batteries that generally can't be  replaced without
shipping the whole unit off to the manufacturer -- a  "feature" that
lands many of them in the rubbish heap. Part of the problem  is a
much shorter new-product cycle, which has sharply reduced the amount
 of time allotted for testing. The result is that things break much
more  often. Add to that today's much lower prices combined with
prohibitively  high repair costs, and many customers just opt to
replace anything that  breaks. (Wall Street Journal 16 Jun 2002)
http://online.wsj.com/article/0,,SB1026764790637362400.djm,00.html (sub req'd)


Starting around 1962 and until about 14 years ago I would not buy a
major piece of technology unless I could also buy a service manual
for it. That applied to cars, VCRs, TV and Hifi equipment. (Most
kitchen and laundry appliances were then still self-evident enough
so that one could figure them out by inspection.) Even oil furnaces
came under my purview after a memorable single-digit winter night in
1974, when my house and greenhouse lost the oil furnace's heating
because of negligent servicing by a 'professional'. I fixed it that
same night and have been doing it ever since. 

Unless I felt I had the information needed to fix an item myself, or
to determine what needed fixing before agreeing to have someone else
do it, the terms of ownership were unacceptable to me, akin to
another mortgage.

Side benefits accrue from really owning something with insistence on
a comprehensive understanding of its function. For a start, the
purchase is likely to be based on more informed research, and a more
perceptive installation will minimize maintenance requirements. And
when a washing machine wasn't obliging enough later with a desirable
wash/rinse cycle temperature combination, I redesigned it to give it
one. Much else in my house is custom-designed or customized for
enhanced function and reliability, from the solar hot-water system,
to various house automation functions and the electric fence for my
wife's garden.

Dizzyingly frequent technology upgrade/replacement cycles now
distract us from the fact that much technology can have a very long
functional life, especially with just a little maintenance. I have
an electronic indoor/outdoor thermometer (the "indoor" is the
greenhouse) with vacuum-fluorescent display that has been operating
continuously since 1975. During that time I replaced one integrated
circuit and one capacitor. My stereo system's Quad electrostatic
speakers still please me greatly. They were bought in 1969. The
solid-state amplifier that drives them is going on 20 years. 

My list is much longer, but I won't bore you with further details.
In short, there is an alternate universe out there, unsuspected by
most, where things can last five or even ten times longer under
terms of technically informed ownership. Most of the material I see
discarded is still completely functional. Maybe a switch needs
replacing, or an ego needs to be scaled down to a slower CPU speed.
There is also the psychological comfort of not feeling helpless
vis-a-vis technical malfunction. For most of my technology, domestic
or otherwise, I have "fall-backs," "spares." If our life and work
require us to live in machines, we had better understand them.

In this story about myself I'm skipping altogether the issues I
don't subscribe to, of technology as fashion accessory, as
life-style statement. Those are really the biggest reasons for our
rapid technology turn-over. It has created an economy that is now
addicted to it.

So what happend about 14 years ago, as I mentioned at the start? The
technology "food chain" had evolved enough to integrate complex
technical functions into ever large forms. A clock radio would be
embodied mostly by one large integrated circuit. Service manuals,
their implications always of service "in the field," replacing the
bits that go bad among the bits that are still good, started to
become extinct. In the words of the statement intended to keep the
uninformed from hurting themselves, for the most part there now were
really "no serviceable parts inside." Field service technicians,
potentially important partners in the design evolution of products
(e.g. see Julian Orr's unique ethnography of Xerox field service
technicians "Thinking about Machines") became deskilled in all but
instances of the most complex equipment. On the level of consumer
goods insides, my sense of technical informedness didn't get me very
far any more. It was around 1990 that I last managed to fix a VCR.

The good thing about large scale integration was a significant
improvement in reliability. Fewer circuit packages means fewer
connections in the world of solder joints on circuit boards, fewer
discontinuities between materials of different thermal expansion
coefficients, fewer things to break. So there is "Intel Inside" not
just in computers but even inside some microwave ovens and toasters,
but "no serviceable parts." 

Integrated circuit design was automated, became accessible widely
with the advent of Mead and Conway's 1980 "Introduction to VLSI
Systems", a sea change was often likened to publishing. The old
proprietary ways of designing integrated circuits ("chips") were
likened to "only publishers being empowered to write books." The
Mead and Conway methodology, and its evolution since, allow
engineers to cast their circuits into integrated form in one place,
and have another place maybe half-way around the world actually
fabricate it on silicon from their recipe. Engineer "authors" are
empowered in their own right, it was said, and can chose among
multiple "publishers." That beneficent ideal wasn't quite realized
that splendidly, but in any case it never included the technically
informed consumer as "author."

Since then even the automated integrated circuit technology has
yielded to a newer form of authorship, via software. Microprocessors
can be pervasively applied, in rather standardized form, in a
variety of products whose actual function is defined by
application-specific microprocessor software. The software may be
stored in a separate chip, accessible to a new generation
technologically adroit owners, who replace a car's sedate engine
control program with one suitable for hotrodding, or reprogram their
Furby or Sony's Aibo robotic dog.

When I redesigned, "tinkered with," "hacked," my GE washing machine,
that was between me and the terms of warranty, not that they even
still applied. When someone reprograms Aibo and posts the results on
the Web, Sony gets annoyed and may sue. In fact, products made
smarter by computing create a new era of higher-profile
opportunities for computer-savvy owners. And thus develop the
tensions between today's "tinkerers" and manufacturers who are now
trying to prevent tinkering by going back to non-reprogrammable
"cast in silicon" product implementations. Under what many consider
unenlightened business models, those manufacturers feel that
tinkering will cost them profits.

That, combined with the typical sense of insecurity about technology
in organizational contexts, leads to the well-known adage "Nobody ever
got fired for buying Microsoft," even if the software doesn't really
do what you need. That has become an accepted form of common
suffering. But do something more unusual, build your computer from a
kit to an even greater extent than Dell's customization, in pursuit of
very specific application advantages, and you may have trouble getting
funding from institutional sources. In a land of techno-helplessness
you may become a leper if you know too much and act on it.

So finally on to the three recent articles with which I started this
mailing, articles about the contribution of users to innovation.
Edward Felten of Princeton is passionate about the freedom to tinker
and thus innovate, and spent last academic year at Stanford, working
with Lawrence Lessig, the law professor who is equally passionate
about the freedom of ideas. (I've mentioned before his recent book
"The Future of Ideas".)

The NYT article discussed the restrictions manufacturers are
planning to build into cell phones, CD players, computers, printers,
to keep you ever more captive to their circumstances. In effect they
want to license you the use of those products much as the terms of
most computer software are only licenses for use, not ownership at

Finally the TCPA (Trusted Computing Platform Alliance) item from
Cambridge refers to an Intel-led initiative which will embed Digital
Rights Management into your next PC and entertainment electronics.
Also named the "Fritz chip" (after Senator Fritz Hollings of South
Carolina, who is working tirelessly in Congress to make TCPA a
mandatory part of all consumer electronics.) A few quotes illustrate
how deeply this could enfringe on your ownership of a device.

> The fundamental issue is that whoever controls the Fritz chips will
> acquire a huge amount of power. There are many ways in which this
> power could be abused, and Intel has refused to answer questions on
> the governance of the TCPA consortium. 
> One of the worries is censorship. An application enabled for TCPA,
> such as a media player or word processor, will typically have its
> security policy administered remotely by a server. This is so that
> content owners can react to new piracy techniques. However, the
> mechanisms might also be used for censorship. 
> For example, the police could get an order against a specific
> pornographic picture of a child, and cause the policy servers to
> instruct all PCs under their control to search for it and notify
> them if it were found. As another example, the scientologists have a
> record of getting courts to give them injunctions against their
> critics. In future, if they can convince a court that a certain
> document should be banned, they might also get an order against a
> policy server.  

As for me, tinkering with software-enabled products, which are
becoming the norm, is sadly not my forte. I'm unlikely to modify the
operating system of my digital camera. I can only affirm my sense of
ownership by adroitly researched initial choice, expectations of
reliability (an increasingly troubling area--see the 4th story at
the beginning), and a suitably wide range of functions to allow my
use to grow. Confining forms of ownership, questionable reliability,
restricted "licenses," are situations I will avoid altogether
whenever possible.

As I mentioned before, in instances where I rely heavily on a
technology, I try to have fallbacks, some redundancy, maybe two of
them. If I get version n of something, I usually keep (n-1)
functional as a fallback. I have different ways of getting email,
several different computers (all of which I maintain myself, spare
hard drives and modems on hand), I keep my hard drives carefully
backed up, etc.

And, oh yes, my oil furnace will get its annual service later this
summer and run as efficiently as ever.


More information about the EAS-INFO mailing list