From: John Stanley (stanley@peak.org)
Date: Tue Apr 17 2001 - 17:48:28 CDT
Erland Sommarskog (sommar@algonet.se):
>What I am uncertain of is that they can define a header as local all
>on their own.
You just are not listening. Site B didn't do it "all on their own". Site B
didn't do it at all. But even so, why do you think you have a veto power
over this?
>And if you were not so focused on quarreling with each and everyone
>you would have noticed that what I said above is actually support of
>your position. Whether site C wants to keep the header or not is not
>very interesting, if they are even ignorant of that they are in of
>obligation of removing it.
How can you be supporting "my position" when you don't even know what it
is? Of course Site C knows it is local, they read the documentation for
the system that Site B is running that inserts it and have decided they
want to keep the header. I've said twice now that Site C knows what the
header is and wants to keep it, but it CANNOT because this draft says it
MUST be removed. Not MAY, as you said, but MUST. And if they know that
this draft says they MUST remove it, they certainly are not ignorant about
their obligation to remove it.
You want to talk about people innocently violating the standard by not
removing what they don't know should be. That is not 'my position', it is
not even close. I'm talking about people who are going to deliberately
ignore this standard because it benefits them to ignore a silly
restriction intended to deal with one header but which is being applied to
an entire class.
>This might be not what you intend, but I take it that you mean that
>this fictious nocem-header has actually been defined as local. But if
>so: where is the mistake?
In this draft, which does not have any idea what the header is or what it
does, but says that it MUST be removed when the article is tranported to
another site.
> All your arrogance side, John, we have a real problem over here.
Whose arrogance are you talking about, Mr. "Learn Swedish and Come Visit
the Real World Sometime"?
>The language could of course be made stricter to remove the obvious
>contractions, something like "strings that are known to erroneously
>be used as back-references".
Of course it could. Why don't you suggest that if this is such an earth
shattering problem over there in the "real world"? To try to slip this by
as "non-standard back-reference" when there is no such thing defined
anywhere is disengenuous and nonproductive.
Of course, we'd want to ban the equivalent strings in every language, not
just the language spoken in the real world. French, Chinese, Korean,
Russian, etc.
>> So we should stop it from happening everywhere?
>Yeah, that what's MAY means? Get real.
SV is SV. If it's a "back reference" here, it's a "back reference" there.
If it isn't a back reference here, it isn't a back reference there. That's
the point of a global standard like this one is supposed to be.
Clive D.W. Feather (clive@demon.net):
>Yes. But, from experience, I can tell you that having two normative
>sections of text saying the same thing is very dangerous. Because if they
>disagree, you're in real trouble.
A fairer summary of this draft I've never heard. Saying the same thing in
two places is bad because if they don't say the same thing you're in
trouble. But if they don't say the same thing, you aren't saying the same
thing in two places. Will someone in the real world translate this from
Swedish for the rest of us?
>At the end of the day, it's the Editor's call.
At the end of the day, it's whether this thing makes enough sense to
enough people that they bother implementing it or just ignore it. With the
section on path and the 83 new path identity separators, I'm betting on
the latter. USENET II speaks louder than words.