John,
JWNI> For example, we want users to be able toexchange email in their
JWNI> native language.
At 5:15 PM +0100 2/1/04, Iljitsch van Beijnum wrote:
I think most of them already dothat...
JWNI> The tacit assumption is that all parties to the message
JWNI> have thesame native language. Accepting that assumption, I
JWNI> generallyagree.
This particularly example was intended to provide an example that I
was pretty sure we already had solved, so folks could ease into the
concept of non-technical requirements thinking.
(I certainly was not intending to be clever about complexity of
multi-native discussion contexts.)
However...
JWNI> We want them to be able to have addressesthat are
JWNI> expressed in their native charactersets.
At 5:15 PM +0100 2/1/04, Iljitsch van Beijnum wrote:
This is very dangerous anyone having anon-latin email
address can't expect to receive mail from people whodon't master
this character set. S
JWNI> Well, now you've gone and made the tacit assumption
JWNI> explicit,Iljitsch. Should it be possible for a sender using a
JWNI> romanizedscript to address a message destined for a receiver
JWNI> using an orientalscript (to pick one specific situation)?
With the vast majority of the world using non-Roman characters we need
to be careful about what uses of Roman characters are imposed on them,
including for addresses. As Iljitsch notes, there are significant
issues with the mixed-native situations. However, this entire example
was also intended to refer to something that, perhaps, was already
solved, with IDN. If it hasn't been, we need to understand what
user-level capabilities are missing or inadequate.
JWNI> At 5:15 PM +0100 2/1/04, Iljitsch van Beijnum wrote:
JWNI> Ok. The main thing that I desire from anew mail system is
JWNI> more control over the mail I receive, and betterfeedback over
JWNI> what I send. Because I run a business, I need to be ableto
JWNI> receive mail from people I don't know. However, I want to be
JWNI> sureit's actual _people_ sending me messages
I thought this text from Iljitsch was quite excellent. It is succint
and includes the often-missed requirement for spontaneous
communications. It is stated entirely in human terms, so that we can
follow a path for developing a solution, or perhaps find that the path
is too expensive. But at least we can debate and agree on the goal.
For example, I'm sure everyone would agree that they would like more
control over the mail they receive. However not everyone wants to be
burdened with configuring that control. Also, it is not automatically
clear how much of that control is reasonable to put into an
infrastructure. (A scheme which pushes back to the sender, through
mechanisms embedded in the email infrastructure, would be an example
of this concern.) And so on.
JWNI> How can receivers increase their confidence that
JWNI> messagesthey receive are specifically intended for them?
Your phrasing sounds pretty good, but I'm not sure how well it
actually works. We start getting into debates about "specifically
intended". But, then, debating such a matter of semantics is exactly
why we need to have the discussion at this level.
d/
--
Dave Crocker <dcrocker-at-brandenburg-dot-com>
Brandenburg InternetWorking <www.brandenburg.com>
Sunnyvale, CA USA <tel:+1.408.246.8253>