Copy Link
Add to Bookmark
Report
Netizens-Digest Volume 1 Number 208
Netizens-Digest Tuesday, November 17 1998 Volume 01 : Number 208
Netizens Association Discussion List Digest
In this issue:
[netz] Re: Do Internet Users exist?
Re: [netz] Is Netscape filtering sites?
Re: [netz] Is Netscape filtering sites?
Re: [netz] Re: Do Internet Users exist?
Re: [netz] Re: Do Internet Users exist?
[netz] Re: Do Internet Users exist?
[netz] Re: [ifwp] Re: Preliminary report from Boston
Re: [netz] Re: Do Internet Users exist?
[netz] Re: Is Netscape filtering sites?
----------------------------------------------------------------------
Date: Mon, 16 Nov 1998 13:39:15 -0800 (PST)
From: Greg Skinner <gds@best.com>
Subject: [netz] Re: Do Internet Users exist?
In article <19981114143938.AAD21425@LOCALNAME> Kerry Miller wrote:
>Ingo wrote,
>> Compared with a lot of more commercial networks, it worked just great. The
>> fact that it didn't work very well was because the technology was immature.
>> It might not have worked great, but the point was that it worked near the
>> practical maximum, which could not be said of other networks. Many things
>> that didn't work that good, were missing alltogether from other networks.
I think one of the things that made the Internet fun from its early
beginnings was that it was a medium in evolution, which meant if you
had ideas that you could justify to the people who ran their
respective Internet sites, you could make changes to it and
participate in its evolution. I don't think this was true of other
networks at the time, which were much more tightly controlled.
>> Well, we have to face it: (western?) Humanity got scared and
>> experimentalism was lost in the upheaval that followed. Some people still
>> have it and they are called "hackers" and are one of a crowd. This created
>> some pretty interesting results, socially speaking.
>> People have the tendency to be single-minded. That makes business people
>> forget that they are running businesses to have a life and not as an end in
>> itself. Luckily, not all people are business people.
>d) " Humanity got scared and experimentalism was lost in the upheaval that
>followed....People have the tendency to be single-minded...." Excuse me,
>but dont generalizations like these do more harm than good in *any*
>conversation? Experimentalism is not lost, and plenty of people are not
>singleminded (nor even tend to be), even if we knew how to measure such
>factors exactly.
I think you are right, but (in the context of commercialism of any
medium), the driving force tends to be collective vs individual
behavior. It's too hard to figure out what everyone wants. It's much
easier to make generalizations. Also, generalizations have shown to
be successful (from a business standpoint) in identifying how a
commercial medium can be marketed.
I'm not trying to justify making generalizations. I'm just saying
that business people can justify them on the grounds that they produce
desired results (profits to shareholders, reliable revenue streams,
etc).
- --gregbo
------------------------------
Date: Mon, 16 Nov 1998 15:56:00 -0800 (PST)
From: Greg Skinner <gds@best.com>
Subject: Re: [netz] Is Netscape filtering sites?
In article <19981115015143.AAC5450@LOCALNAME> Kerry Miller wrote:
>Tom wrote:
>> What if there are several corporate nets with root servers. Then
>> it really would depend upon whose DNS tree you use as to whether
>> something like www.cbs.com points to the Columbia Broadcasting
>> Company website or Creative Business Services web site.
This is true, and it also is one of the dangers of fragmentation.
>Is the solution simply to let the firstest have the mostest (or vice
>versa!), or to delegate sub-domain responsibilities on down the line?
>That is, why not let everybody who wants to have the *top pointer of
>www.cbs.com get together to maintain a 'menu' where the various
>different sites can be discriminated according to some small quantum
>of further information?
For a small number of hosts, this is OK. For a large number, it's not
practical. For simplicity's sake, it's better to keep everything in
the same tree.
>(Is it really so important that *all* responsibility for choice is
>handed over to some damned machine?)
It is easier from an administrative and technical standpoint. It is
politically unattractive when that choice is determined by people who
are not trusted. When the DNS was first developed, most people
transitioned to it because it was easier than maintaining a large host
table, which is the logical extension of maintaining multiple menus,
as you suggest.
- --gregbo
------------------------------
Date: Mon, 16 Nov 1998 16:26:12 -0800 (PST)
From: Greg Skinner <gds@best.com>
Subject: Re: [netz] Is Netscape filtering sites?
In article <199811142145.QAA03569@menno.com> Tom wrote:
>I'm working from memory here, so some of my facts might be
>wrong. However, adult site www.whitehouse.com was ticked
>off at Netscape because they were using some "intelligent
>process" so that when you type in just "whitehouse" it
>sends you to www.whitehouse.org rather than the usual
>method of putting www in front and .com on the end to
>build the URL out of the one word title typed in.
>The concern to this group was the comment that when
>Netscape does this, they are, in effect, filtering
>out sites and restricting ability to get to all of the
>sites available, and "bypassing DNS." I thought it
>was humorous that they talked about bypassing (Netscapse
>response was that if you type in the full URL you
>still get there).
Netscape is correct from a technical standpoint. From a political
standpoint, it raises questions of equal access, because the default
behavior takes one to a presumably "popular" site.
However, assuming "whitehouse" means www.whitehouse.com raises the
same questions, however, when it is the default behavior.
My preference would be for the software to assume nothing. Let the
default behavior be needing to type the whole URL. If this is not
satisfactory, a configurable menu specifiying some heuristics
(e.g. assume all names end in .com) could be employed.
This lets the user make choices according to their needs, and
localizes change to the user (ie. does not require that the DNS itself
be changed to emit different responses based on external
information).
Also, it should be made clear that these are indeed heuristics, so
they may not always work or may use more resources than just
specifying the URL in full.
Interestingly enough, Netscape has this feature for specifying
proxies. The default is to assume no proxies. I guess they are using
statistics to determine that most one word entries were really
requests to go to www.the-one-word-entry.com. :) Here, we see how
statistics can influence how software is written, that might work for
a large majority, but not work in all cases. :)
>For discussion: what about DNS servers that would list
>some sites and not others. Would this indeed restrict
>people since you would have to go back to knowing the IP
>address rather than a DNS name?
Yes. In addition, due to differing security policies, some sites
might not be able to communicate at higher protocol levels, even if
they had IP level connectivity, because reverse queries (a query for
the host name given the IP address) might not work.
- --gregbo
------------------------------
Date: Tue, 17 Nov 1998 22:10:47 +0100 (CET)
From: Ingo Luetkebohle <ingo@devconsult.de>
Subject: Re: [netz] Re: Do Internet Users exist?
On Mon, 16 Nov 1998, Greg Skinner wrote:
> I think you are right, but (in the context of commercialism of any
> medium), the driving force tends to be collective vs individual
> behavior. It's too hard to figure out what everyone wants. It's much
> easier to make generalizations. Also, generalizations have shown to
> be successful (from a business standpoint) in identifying how a
> commercial medium can be marketed.
Generalization of opinion does not necessarily represent a majority
opinion, it only represents an opinion that is not likely to be opposed by
the majority.
That is a *big* difference.
- --
Ingo Luetkebohle / 21st Century Digital Boy
dev/consulting Gesellschaft fuer Netzwerkentwicklung und -beratung mbH
url: http://www.devconsult.de/ - fon: 0521-1365800 - fax: 0521-1365803
------------------------------
Date: Tue, 17 Nov 1998 13:56:41 -0800 (PST)
From: Greg Skinner <gds@best.com>
Subject: Re: [netz] Re: Do Internet Users exist?
Ingo Luetkebohle <ingo@devconsult.de> wrote:
>On Mon, 16 Nov 1998, Greg Skinner wrote:
>> I think you are right, but (in the context of commercialism of any
>> medium), the driving force tends to be collective vs individual
>> behavior. It's too hard to figure out what everyone wants. It's much
>> easier to make generalizations. Also, generalizations have shown to
>> be successful (from a business standpoint) in identifying how a
>> commercial medium can be marketed.
>Generalization of opinion does not necessarily represent a majority
>opinion, it only represents an opinion that is not likely to be
>opposed by the majority.
I never said anything to the contrary. In fact, I agree with you. My
point was that generalization of opinion has proven to be an effective
means of determining how a mass medium can make money.
- --gregbo
------------------------------
Date: Tue, 17 Nov 1998 21:51:44 -0400
From: kerryo@ns.sympatico.ca (Kerry Miller)
Subject: [netz] Re: Do Internet Users exist?
>>>generalizations have shown to
> >> be successful (from a business standpoint) in identifying how a
> >> commercial medium can be marketed.
>
> >Generalization of opinion does not necessarily represent a majority
> >opinion, it only represents an opinion that is not likely to be
> >opposed by the majority.
>
> I never said anything to the contrary. In fact, I agree with you. My
> point was that generalization of opinion has proven to be an effective
> means of determining how a mass medium can make money.
>
Again, referring to the linkage between aggregate usage -> hard/soft ware
design -> individual learning or expectation, isnt the relevant question
therefore how to *balance* commercial/ generalized patterns with the
freedom to discover for oneself? As long as there are substantial alternative
means to a given end, the balance tends to take care of itself, but the closer
a system gets to hegemony (the DNS being the case in point), the more it
has to have *built in* protections for both aspects -- in short, the educational
value of the *structure (never mind the content) of the internet deserves to be
preserved.
kerry
------------------------------
Date: Tue, 17 Nov 1998 21:52:16 -0500 (EST)
From: Ronda Hauben <ronda@panix.com>
Subject: [netz] Re: [ifwp] Re: Preliminary report from Boston
Karl Auerbach <karl@CaveBear.com> wrote:
>Although I was not pleased by much of what I heard at the meeting, I do
>feel that we need to let the board have time to consider what was said and
>to react to it.
>From various side conversations during breaks at the meeting, I got a
>positive feeling that these issues are not black and white in the board
>members' minds and that some, and perhaps even a majority, might, upon
>reflection of what was said, move to make improvements.
But what it seems is that there is a timeline that was made up
outside of the Board's doing and that they were in general chosen
for their lack of knowledge about the Internet and of the
matters involved in the privatization of the essential functions
of the Internet.
And none of the board members noted trying to learn about how they
were selected nor why.
Thus it doesn't seem that any of them care about what is really
going on behind the scenes.
It seems as if the clock is ticking on the prepared calendar
of events and we are invited to be present, but until one
has some understanding of what is happening behind the scenes
and by whom, the clock will just continue ticking and the
calendar of events will continue to unfold as planned.
One only needs to remember that in July in Geneva, Ira Magaziner
said that he had to have a proposal with the consensus of the
Internet community behind it by September 30 or else a new agenda
would play out.
There was *no* such proposal by September 30, but it didn't
matter because the U.S. governments' (the IANA) proposal to
privatize a piece of itself just continued to be promoted
in October.
The Department of Commerce said that there had to be some
consensus on certain issues for ICANN to get their approval.
But they have already begun the transfer of contracts of IANA
and NSI(is that also already being transferred?) to ICANN.
So none of their conditions have any reality,
they are merely stated to avoid people realizing what is actually
happening.
The Nov. 14 Saturday ICANN Meeting showed that many people are
upset with what the U.S. Government and ICANN are doing. But
there is *NO ONE* at either of these institutions to hear what
people are saying.
Unless someone or something begins an inquiry into what is going
on, it will just continue going on.
Why would those who are to directly benefit from billions of dollars
of assets being passed their way object to the continued momentum?
They can take a little dissent, that's what they were put on the
board to make possible. If a newspaper article dares to really
present what they are doing, they will attack the reporter. But
in general they will avoid getting any but in house publicity.
This is a done deal, with those who are doing the dealing behind
the scenes, and those they put out front willing to brave a few
complaints to keep their hands on the pot of gold.
Ronda
ronda@panix.com
Netizens: On the History and Impact
of Usenet and the Internet
http://www.columbia.edu/~hauben/netbook/
in print edition ISBN 0-8186-7706-6
------------------------------
Date: Tue, 17 Nov 1998 18:54:59 -0800 (PST)
From: Greg Skinner <gds@best.com>
Subject: Re: [netz] Re: Do Internet Users exist?
In article <19981118015608.AAD11179@LOCALNAME> Kerry Miller wrote:
>Again, referring to the linkage between aggregate usage ->
>hard/software design -> individual learning or expectation, isnt the
>relevant question therefore how to *balance* commercial/ generalized
>patterns with the freedom to discover for oneself? As long as there
>are substantial alternative means to a given end, the balance tends
>to take care of itself, but the closer a system gets to hegemony (the
>DNS being the case in point), the more it has to have *built in*
>protections for both aspects -- in short, the educational value of
>the *structure (never mind the content) of the internet deserves to be
>preserved.
*sigh*
Like I've been saying all along, the DNS does not have influence over
anyone. It is the way the DNS has been operated for the past few
years, which saw the rapid transition of the Internet from a
newly-emerging commercial medium to a mass medium, that has created
the conditions that make it seem as if it has influence over people.
In fact, used with the original intent of its developers, all these
name conflicts would be moot.
I don't know how it works in other operating systems, but in Unix one
can modify one's name resolving environment to search static host
tables before doing DNS lookups, if one wishes. And one can also
install a caching DNS server on one's machine, and configure it to
search any root servers it wants. So, there is already the option for
learning that you feel is necessary, for those who use Unix systems.
I imagine if this becomes a significant concern, more people will use
Unix rather than other OSes that do not have this feature, or other OS
manufacturers will make this an option. I also imagine there are
people who will not want to go to the trouble, once they realize what
a pain it is to maintain a bunch of static mappings in a large,
distributed, non-static Internet.
- --gregbo
------------------------------
Date: Tue, 17 Nov 1998 23:01:26 -0400
From: kerryo@ns.sympatico.ca (Kerry Miller)
Subject: [netz] Re: Is Netscape filtering sites?
Greg,
> >How do you see it as not being in the tree? Why isn't it the same
> >reaasoning as applies to the various TLD servers? (I'm not
> >suggesting that the menu admin becomes a registry.)
>
> Perhaps I misunderstood what you wrote earlier. I took your earlier
> statements to mean that you wanted multiple roots to be maintained,
> and have menus available for people to select which roots they want
> based on some external criteria.
>
My 'people' are the surfers, trying to find www.cbs.com; your people are
looking for a place to register their sites -- let's call em Consumers and
Producers. OK, Consumer couldn't care less how the net works or where
the registry is kept, while Producer thinks its an advantage to have CBS
associated with his operation. I submit that the prevailing notions that
a) C *wants* to go directly to the site selected, and
b) P *needs* a unique URL
are simply *assumptions* of how the net 'ought' to work. There is nothing in
the protocols that supports either one - and in fact, they are only two sides of
the same premise -- that there will always be enough names to go around --
that has created an artificial shortage in name-space. This is what is driving
the trademark issue, which itself is cited in the White Paper as the *first*
reason why the DNS should be privatized.
So, I say, eliminate the shortage. ICANN registers the IP numbers, DNS
points each number to a CBS *menu page*, where each P who wants the
name can chip in (proportional to its net-profits ;-?) to keep the host up, and
each C can look over a very finite list of links, choose Candian Bull Shine
over Coholumbia Broadcasting Service, etc., and go on from there. So what
if the DNS lookup is not a one-to-one map? Who cares?
(1) Basically, as far as I can see, I'm only reinventing the .net domain;
(2) to dispatch C in the right direction, the menu host should have its
own IP lookup table, so it doesnt get referred back to itself, but is that an
awful chore?
(3) How many www.cbs.com's are there likely to be? Good question,
and I'll be happy to think about the *politics of it further, if the hardware is
sound.
> >But doesnt your earlier argument apply equally well here?
> > " Here, we see how statistics can
> > influence how software is written, that might work for a
> > large majority, but not work in all cases."
>
> >When the responsibility is given over *by default* to the machine,
> >people are less inclined to exercise that responsibility themselves.
>
> It's not quite the same. In the case of how Netscape is writing their
> code, they are using a human factors statistic (how likely it is that
> people will want to access some company's or brand's web site) to
> optimize a particular type of access method (a URL). In the case of
> the DNS at large, the system was developed to generalize a large
> number of access methods (any service that needs to resolve some
> network name to some other network resource, usually an IP address).
> For a name space that is growing exponentially anually, maintaining
> manual mappings, even if they are mappings to root servers, does not
> scale. Furthermore, the mapping of network name to network resource
> is not static.
>
> >> [*When the DNS was first developed*], most people
> >> transitioned to it because it was easier than maintaining a large host
> >> table, which is the logical extension of maintaining multiple menus,
> >> as you suggest.
> >And now (even if that's what I was suggesting)?
>
> See above. It is even more necessary now to have DNS due to the
> explosive growth of the Internet.
>
Perhaps it's been growing because certain fundamental limitations have been
ignored.
> I don't know what your background in computer science is, but I think
> if you took some time to read the RFCs that describe the algorithms
> and protocols of the DNS, you would come to the conclusion that it is
> the most optimal way of doing resolution of network names to network
> resources in a large, distributed, dynamic environment. The problems
> that have occured are due to policies that were set in an era when it
> was possible for the root servers to be maintained by one organization
> because any contention for names was easily resolved by the parties in
> question. Lots of people named their computer "frodo", "bilbo", etc.,
> but it was not an issue because they were considered distinct names
> within the DNS tree (bilbo.lcs.mit.edu, bilbo.cc.purdue.edu,
> bilbo.cs.ucla.edu, etc.). The problems we are having now are due to
> the fact that people are overloading the SLD space with well-known
> names for convenience and branding. The DNS was not designed for this
> purpose. It can certainly be used this way, but it does not perform
> as well. It will perform even worse if there is an overload of TLD
> space, because it will effectively make the root servers a modern
> version of the host table, and introduce the type of scaling problems
> that caused the DNS to be created in the first place.
>
> --gregbo
Obviously you've got more CS that I do, but I agree, performance is as clear
an argument as we can have for getting trademark issues out of the net
admin, but most of what you're saying seems rather beside the point here.
The single host table solved one problem, but the name feeding frenzy is a
worse one, imo. If the net is so great at facilitating communications, maybe
its time to see if it can keep distributed tables together.
>
> p.s. since the message I quoted was sent only to me, my response goes
> to you only.
My fault.
kerry
------------------------------
End of Netizens-Digest V1 #208
******************************
