We are all end-users, the question is whether the work here should be more
responsive to the needs of typical end users.
We have interminable discussions premised on the bizare assumption that the
typical end user cares more about whether he has a /48 or /56 than whether his
or her network works reliably, without fuss, supports the applications they
want to use and does not make them artificially dependent on a particular
service provider.
Equally you can be sure that in any discussion of a proposed change to the
Internet that there will be someone who will come up with the equivalent of
'but if we phase out use of UUCP that will cause systems which people rely on
in remote parts of Africa to fail', and you can be sure that the person raising
the objection 1) has never been to Africa, 2) has zero personal contact with
anyone in Africa who has the issue he claims to be critical, 3) is entirely
ignorant of the constraints under which Internet management in Africa operates,
4) has some 30 year old UUCP installation that they want to keep running for
personal reasons.
We used to have the same issue with accessibility. Some people would bring up
accessibility in a similarly insincere fashion. By this I mean using
accessibility as a debating maneuver without any real interest in meeting needs
of disabled persons.
That is rather harder today, first there are quite a few people in the IETF who
have direct experience in those areas, either as a user or developer of
accessibility technologies, second there is a whole accessibility effort in W3C
that focuses on the issues directly.
So when folk object that using PKIX logotypes in certificates rasies
accessibility issues with blind and partially sighted people I have a lot of
resources I can draw on. I talk to the people concerned, take their input into
account and come up with a (slightly) revised proposal.
I don't think it is actually very difficult to become an advocate for the
ordinary user. Just decide to become 100% intolerant of network administrivia.
When friends or relatives ask for help setting up their systems ask why the
assistance is needed and how the network could be changed to make that help
unnecessary.
Sit in on some usability tests for real products. It is quite interesting
watching someone who claims tho be proficient in the use of certain network
infrastructures actually using them to achieve what should be routine tasks.
The User Interface itself may be out of scope for a protocol but the need for
user interaction is not. In particular a protocol must provide the information
necessary for the UI to function. A very large number of UI issues are really
caused by poor architecture, lack of though given to error conditions is a
major cause. In PKIX the protocols make the bizare assumption that every
Internet user wants to become a trust engineer. This is better than the
original assumption that PEM was based on but still too far away from the idea
that everyone wants to have the choice of who they delegate their trust
management tasks to.
-----Original Message-----
From: Jeroen Massar [mailto:jeroen(_at_)unfix(_dot_)org]
Sent: Wednesday, September 19, 2007 6:30 PM
To: Stephane Bortzmeyer
Cc: ietf(_at_)ietf(_dot_)org
Subject: Re: Representation of end-users at the IETF (Was:
mini-cores (was Re: ULA-C)
Stephane Bortzmeyer wrote:
On Wed, Sep 19, 2007 at 12:50:44AM +0000, Paul Vixie
<paul(_at_)vix(_dot_)com>
wrote a message of 32 lines which said:
in the IETF, the naysayers pretty much kick the consenting adults'
asses every day and twice on sunday. and that's the real problem
here, i finally think.
Time to have a formal representation of end-users at the IETF?
What is defined as an 'end-user'?
You, me, the rest of the people, are all end-users IMHO.
That we might have quite a bit more knowledge on how things
work and that we might have some connections to people so
that we can arrange things, is nothing of an advantage over
people who are not technically inclined (or how do you put
that nicely ;)
The point is that those people don't know better and as such
they also don't know what is possible and what they are missing.
Eg, if you tell somebody "oh but I have a /27 IPv4 and a /48
IPv6 at home and I can access all my computers from the
Internet wherever I am", they will be going "and? why would I
need that". The typical lay-man end-user really couldn't care
less, as long as their stuff works.
The only people really noticing problems with this are
hobbyists and most likely the gaming crowd trying to setup
their own gameserver and finding out that they are stuck
behind this thing called "NAT".
P2P people, thus quite a large group of people using the
Internet today, have their tools to nice NAT tricks, thus
these won't notice it.
And for the rest of the population the Internet consists of
http:// and https:// if they even recognize those two things,
thus most likely only "www" and "email", the latter likely
only over a webinterface...
Which group do you want to 'involve' in the IETF and more-over, why?
Last time I checked the IETF was doing protocols and not user
interfaces.
Greets,
Jeroen
_______________________________________________
Ietf mailing list
Ietf(_at_)ietf(_dot_)org
https://www1.ietf.org/mailman/listinfo/ietf