(c) Heimo Claasen / Brussels, January 1998
PRIVATISING THE PUBLIC WEB
==========================
The "Browser War"
-----------------
and how ideological components are built into Hard- and Software
----------------------------------------------------------------
The enormous differences in material preconditions for Internet accessibility
between the densely telephone- and net-connected "North" - the central regions
of the industrialised countries - and the "South" are not the theme here.
There is quite some discussion going one about that aspect, elsewhere [as well
in this Journal] and at large(1); but I draw on this debate to add another
aspect of the North/South devide, and of mechanisms of marginalisation:
Specific developments of hardware and software for Web connection, i.e. for
those who could connect at all, tend to deepen the North/South gap, and to
marginalise larger parts of the "Web public" even inside the Northern region.
The issue is not the introduction of new technology - which in any event, and
trivially, would create differentiation between those who can have it and use
it, and those who have it not (yet) - but the way it is done, and the
ideologically biased acceptance of key actors in that process to perceive of
aims or of trends in that development as of already realised, and to take the
situation of a minority (of privilegded technology "haves") as a reality of all
and as the normative standard for the whole, of the Web as such; and thereby
to exclude the "have-nots".
The thesis is that this process, as it going on now, will diminuish Web
and Web-content accessibility instead of increasing both; and that it is
marginalising, if not outright discriminating, against present Net users in
the South, and against socially weak and/or handicapped Net users in the
"rich" industrialised North (and in the rest of the world).
Secondly, that an antidote to that effect would be to maintain some least
level of accessibility of Web content for text/line mode browsers; and this
would clearly be a regulatory task in order to maintain the - originally -
"public" nature of the Web as well as of the Net. Perhaps this could even be
considered as kind of a "universal service" criterium for a least Web-"content"
quality. As there exist by now a number of well working alternatives to the
updriven hardware-dependent "graphic" browsers, this could significantly
broaden access possibilities both in terms of necessary equipment as well as
of running expenses, especially of general net charge and of individual
connection time and fees.
At the risk of redundancy, some terminological distinctions first: The "Web" is
not identical with the "(Inter)Net", despite all confusion of both terms in the
media generally. Addressing an information location on the World Wide Web
(WWW) by using the HyperText Transfer Protocol (HTTP) - and reading pages there
which are lay-outed text - with a standardised mark-up system, the HTML
(HyperText Mark-up Language), is rather a late-comer(2) among the functions
available(3) via Internet. Certainly, it has become the most spectacular
feature of the Net, with the most explosive growth of users - not least for the
trivial reason that use of electronic mail exchange had already passed its
expansion phase at that time. Still in the beginning of 1997, among the
world's "Net population" there were only 57m WWW users compared to 71m eMail
users(4).
The confusion to identify the Web as the Net has not been of a hazard either.
Despite of the fact that the two most noisily distributed Web Browsers,
Netscape and Microsoft-Internet-Explorer, only very recently acquired eMail to
use besides of the HTTP function - and still lack some of the other abilities
-, salesmen and advertising push them as "the" means of "access" to "the
Internet". Which is patently wrong, as the real conditions for access are, in
this order (and with the self-evidence of having an appropriate kind of
terminal at disposition), existence - at affordable prices - of telephone
infrastructure, and existence (or reachability at affordable telephone fee
distance, and at affordable ISP pricing again) of an Internet-connected access
point. While some, and still few, "developed" countries slowly get near the
point of factual accessibility, access is far from easily available in a large
number of others, and for most of the world's population at all.
While physical expansion of the telecommunications infrastructure will
doubtlessly continue and even accelerate(5) in developing countries, it may be
doubtful if this will follow the same pattern as in comparably homogenous,
industrialised countries with their far lesser gaps in urban-rural
infrastructure discrepancies and of high-low income distribution.
The most recent access study on the hitherto best "wired" country, the USA, may
give some hints:
"Just under three quarters of the US population had easy access to commercial
Internet service providers ... which accounted for virtually all major urban
areas in the US and some rural areas. Approximately fourteen percent of the US
population lived on the margin between easy access and none, in what might be
labeled inadequately competitive markets. These markets do not have dense
access... Approximately thirteen percent of the US population has no easy
access. The absence of a provider occurs overwhelming in rural areas, remote
areas and less dense areas. There are understandable economic reasons for this
outcome, and it is unlikely that these reasons will change much in the near
future.
"(This result) sheds light on the canard that the spread of the Internet was
reducing distances within the US. It appears that, as of the Spring of 1997,
this was clearly true for urban areas and false for many remote rural areas in
the US."(6)
In less (telephone-)developed countries - i.e. in most of the "Third World" -
the "best" coverage is far worse, and correspondinly, the costs of access there
are out of this world. A recent price sheet in Kinshasa/DRoC offered a full
internet access account for US$ 5,650: per month. Monthly salaries of
academically trained people there, if ever paid, hardly reach 300 US$.
The core of the argument is a very economical one: Both in terms of network
band width, and in terms of end user accessibility, the quantitative volume of
bits and bytes to transfer, and thus costs - direct (connect time) or indirect
(higher bandwith of the network) - makes the decisive difference for the
quality to afford Net and Web access or not at all. In this respect, the www
browser development of the Netscape/Microsoft type introduces a new
accessibility obstacle: these browsers can only be run on the latest generation
of end-user hardware; and their use - or the design of Web sites for their
use - implies hundred- to thousandfold transfer volumes compared to the same
(text) content in basic, text-browser readable format, (HTML marked-up or not).
Reasonable users even in Northern "rich" societies already switch off
downloading "images" by their web browsers, thus considerably diminuishing
connection time (and fees) - if they find the hidden, and most certainly
unexplained switches of the programs to do this. Notwithstanding that, the
sheer mark-up of text content with HTML to render it for "graphic" screen
presentation by those browsers - and naturally, the (text) HTML mark-ups in
downloaded pages cannot be switched off, even if the browser is told to ignore
them -, bloats a simple information item to a multiple of its text bytes
volume(7).
The aspect of reasonable cost-efficiency can be enlarged as one of a broader
social condition of accessibility, with the nearly trivial argument that in the
course of technical innovation there is always a - more or less considerable -
time lag before some new instrument is available for more but a small part of
its potential, or even intended, user population; whatever the reason may be
for this or that group to have access earlier than all the rest. But in
dealing with this rather special instrument of "information" access, i.e. of
participation in one of the core processes of social organisation, such a time
lag may have dire consequences for the coherence of the social system and for
the wellfare of its participants, especially if "all the rest" is large, and
the information accessible by only that specific "instrument" is salient.
In addition an actually large, and potentially much larger, group of Web users
is entirely shut off from access of "graphical" presented web sources for
technical reasons. All existing devices for Braille or speech output depend on
text screen reading. This will remain so for foreseeable time, even if there
is research ongoing - and publicly subsided such - into the development of
high-cost hardware to re-convert screen output of multi-bytes, to pixel-image
converted characters which had been created, and transmitted into the
reproducing browser software, as dead simple one-byte letters of text.(8)
Nevertheless it shows that even "the market" has discovered that field and
considers it of some business potential.
Thus there are three material reasons - telecomm and PC technology
infrastructure, cost-efficiency and social preconditions, and
technical-physical conditions for the handicapped - to insist on some "least
accessibility" criterium of text-mode accessibility. There is an additional,
institutional reason, closely related to, but diferent from, the material-
social one of "affordable" access:
Perhaps commercial sites may do what they want in aiming to close out specific
groups of sightseers. However, even that argument is disputable, in reference
to the similar "public" (factual) situation and (legal) status of a shop
window in an urban street. At least this "interface" of the private enterprise
shop is considered public, and nobody would imagine any rules licensing the
right of a passer-by to look at it; on the contrary, "public order" rules set
clear limits to the private rights of the shop owner to what may or may not be
put up at display.
It seems sufficiently evident that "public" information, by definition and by
reason of being issued from public instances, should always allow general
accessibility of its "salient" content - which is, in any event, of a text
nature.
Under these four aspects, the unreserved adoption of the "new" - proprietary,
"graphic" presentation based - browser technology by key actors in the
"information system" appears as a doubtful development. More so in that it
transports, and is transported by, an ideological bias for which I do not
hesitate to use the polemical name of "techno-racism".
Hard evidence will follow, but some of the terms used have to be qualified
again:
Key actors in introducing new Web technology - and somewhat different to the
(earlier) development of the Internet(9) - are rather one-sidedly distributed
in the web of net-participants. The lack of real interactivity between
(appropriately termed) end-users and the system managers(10) - the right term
again in all relevant aspects here - is [evident | discussed elsewhere [in this
issue]]. The technology to use is almost completely prescribed by the
minority of key actors, and there is at best a highly indirect, and not
representative, feedback from all the rest of the Web "members" - actual,
potential or intentional ones - via "the market" of PC equipment and browser
software. It will be shown though that the employment of sales data, in order
to justify lopsided decisions and conditions for Web access, is part of a
ideologically biased assessment of the net-membership and of that of "all the
rest" of "end"-users.
Nor is the "group" of key actors homogenous and monolithic - factions and
frictions among them (cf. the legal procedures of the US government, and the
anti-trust inquiry by the EU Commission against Microsoft, or the rather
strong signal by the World Wide Web regulatory W3C consortium at the
introduction of HTML-4 in the beginning of this year) leave realistical
chances of alternative Web developments, indeed open for the public.
And there are indications of diverging individual and organisational
sensitivieties too. Some Web "operators", to take the Eurospeak term,
seem to be aware of the problems. The BBC "Newsroom" website already cited
at least does not shut out text browsers, even if it employs some HTML
mark-ups rather incorrectly and unreflectedly(11), and the "Guardian"
newspaper's site is well accessible in text mode too -, while others seem
either ignorant or arrogant; the evenly British "Independent" newspaper's site
is just a pixel-ony presentable, black hole, and some of the European
Commission's "public" servers bluntly brush off visitors with the disdainful
message "Your browser cannot use frames !"(12).
Such individual variations of operators' decisions get another quality if they
constitue a common pattern of a significant part of the key actors, and under
the condition of unequal footing of Net/Web participants.
And this then is what happened within the seemingly "technological" development
of Web site operating in the last two to three years. The data basis is
sketchy and anecdotal but seems nevertheless sufficient to support the quality
of the argument.
The point in question is how many of the actually used PCs are able to run
those later versions of the two web browsers, Netscape or Microsoft Internet
Explorer, which have "proprietary" properties which can be checked by www site
servers, and thus allow site operators to close out non-compliant clients.
This capacity, together with some "proprietary" - and non-HTML compliant -
particularities of the two browsers, make the difference of accessibility or
not; and there is a clear "cutting date" given by the marketing of the INTEL
80586 "Pentium" main processor chip and the concurrent market introduction of
Microsoft's "Windows 95" operating system (and a revision of the parallel
"Windows NT" program): other PCs based on INTEL CPUs cannot run usefully, if at
all, the "newer" Netscape/Microsoft browser versions. This date is for all
practical reasons the end of 1995. Thus all machine stock of an earlier date
is "outclassed" if www-site operators use those specific browser
properties.(13)
One of the problems for assessment of real Net use consists of the way most
statistics about the whole field are gathered - these are almost always based
on sales data or vendors' publicity, and there are hardly any independent and
valid surveys of actually used equipment to find. There is an obvious
incompatibility, for instance, between the claim of Microsoft, to have its
newest version of its "Windows Operating System" installed "in 85 % of the
world's computers", and another statement which finds that "67 % of corporate
desktops (are) still running DOS" or the earlier DOS+"Windows" set-up which is
uncompatible with the "new" browser versions(14).
There are rather congruent estimates that at the mentioned "cut-off" date -
end of 1995/beginning of 1996 -, worldwide about 130m individual PCs were
running, of which ca. 60m in the USA. These estimates had been based on
sales figures - and an "average replacement time of 2.3 years"(15): which in
turn, was based on highly selective surveys of "office" equipment use. Less
macro-biased, and more empirical surveys of the (then beginning) debate on
recycling problems of computer scrap however, rather consistently concluded
that the average "real life time" of PCs is in the range of 4 to 5 years(16);
which would mean a significant larger stock of effectively used PCs of earlier
generations.
Neither do sales data - and many of the "computer density" or "computer
penetration" calculations based on such - take account of the booming market
of second-hand equipment; this in spite of the fact that even "big players" on
the PC (production) market, like Siemens-Nixdorf, more and more perceive of
re-use circuits for re-taken hardware with new sales(17). To a considerable
degree, replaced PCs of the earlier generations are (re-)exported too. Almost
the entire - and booming - West African PC market consists of second-hand
equipment(18) from Europe, as state-of-technology units are outpriced in terms
of purchasing power for even better-of indivuiduals, for institutions and
businesses there.(19)
Taking this into account, and assuming that all of the yearly new sales of
PCs in 1996 and 1997 (in the region of 70 and 90 million units, respectively)
would be of the higher-end segment (which they are not), and all would be
equipped with the MS-"Windows" operation system (which they are neither), it
would leave still a majority part of the existing technical base of PCs being
inapt to run the "newest" browser generation.
And there are certainly more quantitative limits, trivial such - evidently, not
every PC is used for, and equipped with, outside connections -, and more
difficult ones to assess like user preferences and replacements of operation
systems after the units have been sold. One source quoted a factual ability to
run those two browsers with a rate of as low as 5 % of the West European PCs in
use mid-1997(20).
From recent surveys there are approximations only of the sort and type of
factual PC use. A background study which seems to be the major data base for
the EU Commission's "information technology policy" does not even bother to
report the precise sample bases or other essential information to assess the
quality of some of the cumulative accounts done, and the various tables are
hardly comparable. Where this is possible, some interesting differences
appear. Thus the number of housholds said to be equipped with PCs and modems
is given at nearly half that much more than the number of subscriptions to (a
selected number of commercial) online services(21): even if a significant part
of the modem-equipped "non-subscribers" would use telecommunication of their
PCs only for specific connections (for instance, between home and workplace
network), this could be read as a clear indication for the importance of
"other", i.e. non-commercial use of Internet functions.
Sampling of the type "How often did you use ...(Internet; GSM; CD-ROM;
etc.) during ...(the last month e.g.)" in Northern countries delivered rather
seldom relevant data either. A German market survey found for end-1997,
"Windows-95 running with half of the PCs (of the persons interviewed), an older
Windows (version) on one quarter of the PCs, 11.3 % run pure DOS Applications,
3 % are Macintosh (Apple PCs), and 4 % do not know what operating system their
PCs are using."(22) This is quite obviously incomplete, be it solely for the
reason that UNIX systems - with one of the strongest markets in Germany
(estimated to around 10 % even in individual computer use) - do not even appear
among the data.- And most certainly there's none of such surveys at all for
vast regions of the world like, say, Africa or India.
But all these inclomplete data do indeed allow for the conclusion that a
sufficiently large part of the "web public" is shut off from access of
information sites which make use of specific, non-standard properties of the
two, duopolistic browsers.
With already some 19.5m WWW "host sites" operating(23), it appears as even more
impossible to assess this side of web performance quantitatively; it does not
seem necessary either, in that respect that a - more or less large - part of
these hosts are not, and do not intend to be "public". For instance, it's
their own ridicule if some of the larger computer distributors in Germany and
France allow access to their publicity web sites only for those PC/web users
who already are equipped with the last fashion gadgets, thereby excluding
precisely those potential clients eventually willing to "upgrade"(24).
The declared intentions of public instances however, to let the Internet, and
the World Wide Web in particular, become a primary purveyor of general and
public information, and specificly of their own "public" information - which
then would be less accessible by alternative means - do indeed call for some
critical regards on how this is done, especially with their very own "shop
windows".
US government and other public instances pass the test comparably well; which
is perhaps not so astonishing as there is something like the "Americans with
Disabilities Act", and there's a certain public sensitivity for - or better,
against - monopolist market manipulations(25).
Public instances in Europe fare much worse, and particularly the European
Commission's servers - the very EU Commission of which (part of) the leading
figures perceive of their role as the pioneers of technological innovation and
broad public use of Internet facilities: Several of the EU-Commission's
servers are not accessible except with the latest vintage of Netscape or
Microsoft browsers.(26)
Asked about that, the "Information Resources Manager" of the puclic
information service of the EU Commission showed complete ingnorance and
incomprehension for accessibility problems: "It is a fact that Netscape
and Internet Explorer control over 90 % of the market for Browsers while the
line mode alternative, Lynx, accounts for approximately 1 % of the market. It
is therefore very difficult for the Commission to justify the additional
expense of modifying its databases to make them fully usable by such minority
users in the market. ... If the market decides that only Netscape and
Internet Explorer will survive, it is not the European Commission which will
alter this situation."(27)
An answer giving ample reason for sarcastic commentary, not to speak of the
fact that the EU Commission's official even got his facts wrong; what is at
issue here though, is the mechanism with which a key actor in a major public
institution decides in a biased way on the steering of the technical conditions
for access of public information.
Lesser souls among web site operators sometimes combine ignorance with
stupidity and/or arrogance - a dip into the full basket of daily available
examples:
"It't time to join the end of the 20th. century".(28)
"My isp ... told me I could not use a DOS browser. One guy at [another ISP]
even told me there was no such thing. When I told him I possessed it, he told
me I was wrong. When I tried explaining that I was a blind computer user, he
told me I probably couldn't use a computer, so I shouldn't worry about using
the Web."(29)
And most certainly, as with other exemplary perceptions of "superiority", the
image is mirrored by some of those excluded:
"Actually I can't understand the reasons of people wich are living in West
Europe or America with good financial potencials for not buying a better
computer than an old XT,286 or 386!!! ... I'm a Bulgarian, we're living in
a hart situation now, so it's normal for me to have money problems and to
use such a prehistoric machine." (30)
The interplay of these types of attitudes, together with the strong push of
hardware and software salespeople, creates a similar process of what is known
as "self-fulfilling prophecy" elsewhere. And it appears as a specificly
aggravating condition that key actors in "public" instances, decisive for the
Web development, fall for it. This is particularly alarming since, as the "Web
Accessibility Initiative" notet, "[it] is part of the traditional role of
government sensitizing the key players - content providers, in this case - to
the needs of an important minority population with special needs." (31)
Another governing instance, the very standardising body of the World Wide
Web, the WWW-Consortium (W3C), indeed has come out strongly in favour of
text-based accessibility criteria, in its latest release of the HTML-4.0
specification. A specific chapter there on accessibility, and repeated
references to accessibility conditions in almost all the subsections of the
HTML specification, constitute a normative appeal to reason and to maintain the
openness of the Web.(32) Use of text-mode browsers is recommended there as a
quality test criterium of HTML layout for WWW use.
At the simultanous launch of the International Program Office of W3C, dedicated
to the Web Accessibility issue, one of the constituants resumed: "The W3C's Web
Accessibility Initiative presents an unprecedented opportunity to bring
together web authors, browser developers and internet infrastructure architects
to address the issue of access to the next generation web in a way that would
not otherwise be possible... For the first time we have an opportunity to
address access for people with disabilities - and those who are older or have
low bandwidth systems or are using smaller devices - in a systematic and
effective manner... Given the rate at which the web is being integrated into
education and employment settings, this development is essential for real
access to next generation education and employment environments for people with
disabilities and those who will be older."(33)
It leaves us with the need to insist on compliance by some of the other "key
actors".
=====================
Notes:
=====================
(1) Uwe Afemann, Internet - another panacea to solve the world's problems ?
http://www.rz.uni-osnabrueck.de/rz/special/misc/inet-3w/inet-3w.htm
(2) First standards for HTTP were established (by ITU) in 1992, for WWW using
HTML in 1993 (by the predecessor of the now standardising body, W3C or World
Wide Web Consortium).
After sufficiently many "Internet Service Providers" (ISP), i.e. points of
access to the Net by individual users, had installed HTTP - and after
sufficiently many ISP had established and offered dial-up access publicly -,
the year of 1994 is now considered as the break-through date for the Web.
(3) Historically and functionally, HTTP has merely been a specific,
standardised set of commands for file retrieval according to the File Transfer
Protocol (FTP) long existing; only the (usually) coupled application of a
program which interpretes the mark-up (in HTML) of the text file thus
retrieved, constitutes the specific category of (Inter)Net use. The basic
functions of Net use are:
Telnet (use of the basic Internet protocol, TCP/IP, and a packet
driver on the individual client side, of a simple BBS-like,
quasi-direct telecommunication connection)
SMTP (Mail send/receive), later POP3
NNTP (Newsgroups messages retrieval)
FTP (file transfer and retrieval)
Gopher (an FTP based, menu-driven file retrieval system depending
on servers responding to the specific client demands; has
been marginalised by HTTP/WWW use - and server set-ups -
despite its faster and much less cumbersome file/content
transfer capability)
Finger, and Ping/ICMP
IRC ("chat" via keyboard, originally as a mode of Telnet use)
Server functions: -- Time (& Daytime synchronisation)
(ISP or backbone) -- Address/Domain name resolving
-- Routing (the most important of all,
costituing the "net" property)
HTTP (and additionally, on the server side, CGI) i.e. processing
of WWW addresses and client/server demands; all additional
task of WWW/HTML, including processing of audio-visual
elements, are strictly browser dependent presentations of
HTTP-transported files on the client side.
(4) http://www.mids.org/press/pr9701.html
(5) The whole industry, and even regulatory bodies like the EU Commission or
the ITU, expect unbroken optmistic increases of lines and channels to be
available, and of exponential growth of transfer volumes (in terms of content
as well as monetary value). Some scepticism seems appropriate though - taking
the example of mobile telephones, the "population" of users was almost
completely recruited from the one already connected to conventional telephone
networks. Real new recruitment concerned economical elites in infrastructure
deprevated regions with comparably high popolation density - capital cities
in some African countries, Eastern Europe - where conventional telephone
systems have broken down.
(6) Shane Greenstein, Universal Service in the Digital Age. Conference paper,
Symposium on The Impact of the Internet on Communications Policy,
Harvard Information Infrastructure Project, Dec. 4-5, 1997
http://ksgwww.harvard.edu/iip/iicompol/Papers/Greenstein.html
(7) Rather ridiculous examples may be found, of all places, at the "text_only"
www-sources for the BBC's "Newsroom", the textual presentation of news
features of the British BBC World Service radio(!). (Wrong) use of the
HTML-"Table" elements, of colour attributes and the like, bloat the text of a
simple news feature from 4,304 bytes to a 49,734 bytes HTML "document" to be
downloaded. A sober - and appropriate - HTML mark-up would hardly add more
than some hundred bytes to the original. Ever so, a random sample of HTML
marked-up text items from that same source showed an increase of a quarter of
the byte volume by the HTML-marks alone, and and increase of more than 300
per cent in volume between the the sheer text content of items - certainly
line- and paragraph-formatted for good readability on a screen - and the
final HTML-marked up page. (cf http://www.news.bbc.co.uk/text_only.htm)
(8) Another example of inherent irrational development may be seen with other
recent technological innovations, fax and Optical Character Recognition (OCR).
The motive to develop OCR and its primary use for humans, it would seem, should
be the access of the blind to written sources. It is not; its "economical"
raison d'etre is to re-transform printed characters which originally had been
stored and presented as bytes and transformed to print style, into stored bytes
again. Almost amusing is the process where "word-processing" equipment and
software is used to produce paper sheets which then are transmitted from a fax
machine to another only to be read in by OCR hard-&-software there again in
order to be "processed", read and stored.
A global economical assessment might be rather damaging, not to speak of
incidental ecological charges to our limited resources (bloat of paper
consumption in general, and of that of the environmentally dangerous fax paper
in particular). The perfectly appropriate technology of the first-generations
PCs would do the job probably for a fraction of total "costs", and without
the additional equipment for OCR. It may be reminded that these are real,
social costs: resources spent on the means of organisation of our societies
are taken from the available surplus value of productive work, and they are
lost for other, meaningful employment, even if some (minor) parts of the
economy - say, paper, fax, and scanner producers - may have their particular
profit from it.
(9) ARPA, the US-military network of linked (computer-)information sources,
certainly had the function to serve a power-centered and hierarchical
institution originally - even if this was a "public", but nontheless
secrecy-obsessed one. Its very contruction however, as a completly
decentralised, non- (or even anti-)hierarchical system made it nearly "by
nature" a public thing. Once opened, the public of that time accepted it as
such and contributed, again in a totally decentralised way, to its development
and expansion. Much of what followed then may be interpreted as a
permanent attempt to catch and cage the swarming birds (cf. the debate on
privacy and censorship, concisely resumed in that respect in Katja
Diefenbach, Kontrolle, Kulturalisierung, Neoloberalismus, in: P.Schultz, Ed.,
Netzkritik. Berlin: Edition-ID-Archiv 1997)
(10) A tentative classification of "key actors" could look like this:
-- owners/administrators of physical infrastructure:
-- basic infrastucture -- telcos
-- backbones/nodes (big) servers
-- ISPs=interfaces -- (smaller/local)servers
-- regulators:
-- on infrastructure (basic/technical telco functions)
-- on material access -- (at first) by public instances:
administrations, universities,
libraries
-- on their material support
(i.e., budget authorities)
-- on physical inter-operablity (ITU)
-- on software interoperability (ISO/W3C,ITU,...)
-- on "content" (public legislators/judiciary)
-- hardware/software producers -- for intermediary use (ISPs/backbones)
-- terminal/end-user use (machines as well
as applications)
(11) The "" element of HTML mark-up has in all versions clearly been
defined for what is says, the ordered row/colums presentation of (variable but
volume restricted) tables, and has been recommended to not be used for
general page-layout tasks. This is arguably the most abused mark-up tools of
all though, and it plays havoc with screen readers for Braille output. But
this is very much a problem of professionalism of web site designers, and a
task for supervision by their superiors.
(12) Which is patently wrong - "FRAMEs" can be read by even the most primitive
versions of text/line browsers sequentially, so this should not be used as a
pretext to bar access to their contents.
However, the - "legal" and standardised - procedure allone of the web site
server to identify the type of client (the name/type of the browser used by a
visitor) raises a number of more fundamental questions of privacy, personal
data identity/determination and more (not to speak of "proprietary" properties
of servers to use "proprietary" qualities of a client's browser to manipulate
the client's machine, like "cookies").
Of quite another nature is the argument about how far a Web designer can or
may determine the complete end-product of "presentation" on a client's screen.
The intention to make it come out "best" - in terms of presumptions of
the clients' browser capacities - is equally legitimate as the users' choice of
browser programs (or ability to choose/employ such). Under condition of equal
footing this would not produce a problem of basic accessibility, as both
sides could balance advantages of "better" screen presentation for secondary
attributes.
(13) There is a factual, but slowly less rigid link between hardware equipment
and software employment specificly in respect of the basic operationg system
with INDIVIDUAL PCs. Because of the INTEL CPUs' memory addressing limits and
modes, PCs of the pre-80386 type could not use alternative operating systems to
the Microsoft-DOS or MS-Windows(=DOS) type (though there had been and there are
alternative, MS-DOS compatible OSs, which by-pass those problems -
but the later NS/MS browsers never were adapted to them). Apple, with its
"Macintosh" the main alternative of hardware-plus-OS stand-alone PCs, saw a
steady decline of its market share, from about 15 % of NEW PCs sold in the
beginning of the '90ies to about half of that in later years (but less so in
absolute numbers), and in any event the NS/MS browsers were always late to be
adapted to the Apple-Macintosh OS. IBM's own, new "OS/2" only came off the
ground, and rather slowly, with the "Pentium" CPU generation. The UNIX-type OS
was designed from the beginning for "big" computers, using individual PCs only
as (relatively "dumb") end-user terminals, connected mostly (at first) via
Local Area Networks (LANs) of the Ethernet type and pertinent additional
hard- and software (from another near-monopoly, Novell) - but basically any
DOS-run PC could serve in this environment, and connected to the large
university networks, still to-day quite a large number do, but these are
unaccounted for in any statistics. LINUX, a completely non-commercially
developed UNIX-type (and compatible) OS however, has gained steadily ground on
the user-side, but again this is mostly unaccounted for in the usual
"market"-statistics of PC-sales with an OS installed (users just replace that
with the nearly cost-free LINUX); UNIX/LINUS does require at least the
past-1992 generation of INTEL 80386 CPUs however.
All this boils down to a rather steady share of non-MS-DOS equipped "sales of
new PCs" in the range of 15 % of the total; with an overall increase of that
total. Only in that respect the publicity claim of Microsoft of having 85 % of
PCs "installed" with their OS (and until now, all "Windows" versions are MS-DOS
based too) may be true. It does not say anything about how and with what OS,
and if, with what "Graphical User Interface" (GUI, for instance MS-"Windows)
these PCs then effectively are run - a rather recent report, based much on
Microsoft/CH's own data, concludes that "about half of all" PCs in Switzerland
are run under MS-"Windows" and MS-"Windows-NT" (WochenZeitung, Zurich: 8.1.98).
According to the French Microsoft office, after a first year of "Windows-95"
sales "43 % of new PCs sold" were equipped with that software. (De Morgen,
Brussels: 23.8.96)
(14) The first was quoted from Microsoft publicity in a commentary of the
"Gaurdian" newspaper at the occasion of the anti-trust procedure then in course
in the USA, the second quotes the then newest update of the publicity page of
Caldera, an independent DOS operation systems producer; both date from the last
week of December, 1997.
(15) Giaco Schiesser, conference paper (Luzern, 26.5.95); in:
WochenZeitung, Zurich, 2.8.96; and in: H. N. Mueller (Ed.), Umwelt und
Kommunikation, Bern 1996. Schiesser and other authors quote EPA (US
Environment Protection Agency) papers and those from a German Bundestag enquete
commission (parliamentary enquiry in preparation of recycling legislation for
electronic equipment scrap) of 1995/96.
(16) Marco Soldera, Oeko-Computer, Gebensdorf CH), 1995 (cit. "C'T" 10,
Hannover: October 1996, p.104). Andreas Grote, Oekobilanz, in: "C'T" 12,
Hannover: December 1994. Compaq Vice-President Kurt Dobitsch, in: "C'T" 10,
Hannover: October 1996 p.90.
(17) "C'T" 4, Hannover: April 1996, p.126
(18) IPS, cit. De Morgen, Brussels: 30.7.96
(19) In that respect, even "pro-wiring" biased estimates by institutions like
the International Telecommunication Union (ITU) might lag behind in their (low)
assessments of Third World, and specificly African, PC-densisity.
(20) www.datatestlab.com, cit. De Morgen, Brussels: 11.7.97; requests for
details resulted however in that the data given was said to have come from a
Reuters news item, and could not be traced more precisely.
(21) Commission of the European Community, Imformation Market Observatory:
European Information Trends; http://www2.echo.lu/imo/en/trend96/ (publicly
available since Nov. 1997 only).
The accumulated calculation for "Hausholds with (PC and) modems", and for
"European online services: Numbers of Subscribers" are given for -
'000 Hh./w. MODEM[1] '000 Subscribers[2] (Online Serv.)
France 170 124 5
Germany 2 000 1 435 3
Italy 213 60 1
Great Britain 682 570 3
(total) (3 065) (2 189)
Sources given: [1] Jupiter Communications
[2] Financial Times
Typically, "Online Services" exclusively named commercial providers and
did not consider semi-public access (e.g., universities, libraries) or
publicly available, non-commarcial access accounts (eg. the rather large
networks like APC in GB, Comlink in Germany); not to speak of those
interlinked BBS networks which, like the FIDO or Zerberus networks, at that
time offered already all Internet functions but WWW access (and some of them
even that).
(22) Academie-Data, Essen: Press release (december 1997).
(23) Afemann (cf. Fn.1), quoting "Net Wizard",
http://www.nw.com/zone/WWW/report.html
(24) This is rather absurd particlularly in Germany where the mail-order
market for the substantial private/individual use of PCs is much more
developed than in France or Belgium, where PC sales to a much higher degree
are oriented towards the "office" use.
(25) From a usenet discussion, Tue, 15 Jul 1997:
"Subject: Re: Discrimination of non-illiterates
>> Installing WINDOWS in an office is illegal... it violates the
>> "Americans With Disabilities Act" in that though the pretty
>>pictures for the illeterate are nice... The blind can't read
>>them and have to have a TEXT based system such as DOS.
>
>Is this real (or wishful thinking) ? Has there been a (test) case ?
At the moment it's both. The law is real, but nobody has tested it
yet. So far companies like LOTUS (makers of 1-2-3) have managed to
buy off the victims of WINDOWS in their offices far as I know.
But considering the damage that windows is doing to me (Server carpal
tunnell made much worse by using that stupid electronic rodent) I'd
love to see it happen.
Oh, yes... There are a few attornies on INTERNET and most of them
(all the ones who have responded in fact) feel it might be a good
case. Emphasis is on the word MIGHT however."
Note: There are precedences of disabilities-related, or outright disabling
cases. The "Financial Times" for instance had nearly a third of editorial
staff affected by RSI, had to pay heavy indemnities for some permanent cases
and was forced to replace the used corporate system with ergonomically harmful
screen hardware (and screen presentation software) and keyboard equipment.
Cases like these are considered relevant for later procedures on disabl*ING*
conditions.- Germany has a constitutional provision (article 3, para 3 of the
"Grundgesetz", a directly appliccable clause against discrimination on
grounds of disabilities or handicaps) but despite of some 12 to 15 thousand
blind computer users, there had been no testcase (yet).
(26) An example e.a., is the EU Commission server for press releases,
http://europa.eu.int/en/comm/spp/rapid.htm. The European (ministerial) Council
(http://ue.eu.int/angl/newsroom.htm) sends text-only using clients to the
commercial site of Netscape to download and install that browser first in
order to access the (pure text, and all text) content of its official
communications.
(27) Denis McCann, Information Resources Manager, Spokesman's Service,
European Commission: Note on Access to RAPID, Brussels 21 Apr 1997
(28) Webmaster of a Flemish regional government host site in Belgiun, 12 Oct
1997.
(29) Usegroup posting, 01 Nov 1997.
(30) Usegroup posting, 20 Apr 1997.
(31) Web Accessibility Initiative (WAI) Briefing Package, February, 1997.
http://www.w3.org/pub/www/Disabilities
(32) W3C Recommendation on HTML-4.0, 18 Dec 1997:
http://www.w3.org/TR/REC-html40-971218
(33) Press Release, W3C Launches international Program Office for WAI.
Washington D.C: 22 Oct 1997 (http://www.w3.org/WAI/)