Very interesting questions to which probably no one can give accurate
Together with colleagues, I have studied a number of international
networks/Information Highways for a number of years and would hate to
attempt to hazard a wild guess at figures, particularly % values, for the
No one is even able to agree on the current size of the Internet, or the
rate of growth. User figures published during the last three months range
from 3 million to 100 million users. Growth rate estimates (year on year
growth) range from 40% to 260%. The only way to produce reasonably accurate
figures is to poll major organisations using the Internet and estimate what
% of the total user base they might represent. Even that produces figures
which may be imperfect given the nature of the Internet community but would
probably be considerably more accurate than many of the very wild guesses
published to date.
The behaviour of the Internet community is also unpredictable. For example,
there are several guesses published as to the number of people using the
GreatCircle BBS. Someone at GreatCircle presumably knows how many
subscriber accounts exist but it would be difficult to predict how many are
really active, or for that matter what consitutes real activity. It is also
very difficult to estimate how many people indirectly participate, in that
Firewalls notices are circulated through other BBSs and where someone
visibly participates, they may be representing an unkown number of other
Given that the user base and type range is so poorly understood, the base
figure against which % calculations does not really exist. It is therefore
highly unlikely that anyone could produce reasonably accurate % figures for
firewalls, even if the firewall itself was a very easily measured product.
The firewall itself is not a product but a concept. Taking the C&B
definition, any form of gateway barrier could be described as a firewall if
you remove any claim that a firewall provides any guarantee of protection.
If you require 'guaranteed total protection' in the description, the
firewall does not even exist.
Most firewalls are 'roll your own' designs built by users, or consultants,
frequently using tool kits and a range of hardware products. Therefore,
traditional market research methods are also unlikely to produce accurate
figures. For example, a particular router vendor might provide information
on the number of units he ships each month and estimates can be made of his
market share to produce reasonably accurate figures for the output of all
router vendors. From a series of snap shots it would be possible to produce
growth charts, but that still does not produce accurate figures for the
number of firewall because not every purchaser will limit himself to a
single router per firewall, not every firewall will use a router of that
type, and many routers may be used internally and not truly be Internet
To: Vern Paxson; Dave Crocker
Cc: firewalls; jsq; schwartz
Subject: Re: How many firewalls & what IS one?
Date: 27 January 1995 18:28
At 10:03 1/27/95, Vern Paxson wrote:
>> 1. They have heard an estimate that 8% of those attached to the Internet
>> have firewalls. This was a second-hand comment and could not be traced
>> its source. So, my question is: Does anyone know a legitimate estimate
>> and can they cite how it was derived?
>Mike Schwartz did a study of whether the rate at which sites are adding
>firewalls is outstripping the rate at which the Internet is growing:
>The basic conclusion of the study is that the Disconnection is not
>at least not yet - new sites are coming onto the Internet significantly
>than sites are walling themselves off behind firewalls. But the fact that
>these rates are different means that a number like "8%" can at best reflect
>a single snapshot, and isn't constant.
I haven't read the article yet, but I've really got to wonder about the
relevance of an article written 4 years ago on this particular topic...
Things have changed a LOT in the last 4 years with respect to the Internet.