From: John Stanley (stanley@peak.org)
Date: Tue Jun 11 2002 - 02:55:09 CDT
Charles Lindsey (chl@clw.cs.man.ac.uk):
> Wow! What a tortuous line of reasoning; but most ingenious!
I'm sorry you find it complicated. It isn't, really.
>2. Step 5 says the unacceptable article MUST NOT be processed further.
>3. Therefore, Step 10, which says it MAY add other headers not already
>present, can never be reached, and so it is forbidden from applying the
>fix that would make it acceptable again :-( .
Bingo. You've got it. The injector MUST NOT continue to process an
article once it determines it is unacceptable. It does not have the
information necessary to fix it, and 8.2.2 (5) says it MUST NOT.
>Of course, it all boils down to the meaning of "unacceptable". I think
>John is trying to say that "unacceptable" is synonymous with "not
>acceptable", but English usage is not as simple as that.
Yes, dear, that's what the "un" prefix means. "Unacceptable" is the same
as "not acceptable".
> In this case, the article is "not acceptable as it stands".
Yes, that is right. Unacceptable as it stands, and the injector MUST NOT
process it further -- which would include changes that would make it
acceptable to the injector. There are two very good reasons for this
prohibition. 1. The fix may make the article unacceptable to the poster,
and poster has the right to reject any "fix". 2. The injector does not
have the information necessary to "fix" the article, so it very well may
be breaking it.
Why the hell is there a problem with this? The injector can NEVER be wrong
by saying "I cannot fix this for you, please fix it and resubmit", but it
CAN make mistakes if it tries "fixing" an article that doesn't need it or
in a way that makes it unacceptable to the poster.
> Ergo, it is "conditionally acceptable".
No, it is unacceptable. It has already failed the conditional testing.
> Ergo, it is "acceptable" (since I am sure John would agree that
> "conditionally acceptable" is a subset of "acceptable").
I do not, and I resent your continuing attempts at putting words in my
mouth.
> Ergo, it is "not unacceptable".
Yes, dear, unacceptable means it is not acceptable, not that it is "not
unacceptable."
> However, John does make a point that the wording of Step 5 is an
> overkill.
Once again, you are putting words in my mouth and failing. Step 5 is quite
correct. Nobody has suggested it be changed.
> So I have changed it.
Yet another unilateral change by the editor that is not only NOT the
result of a consensus of this group, but has never been suggested by
anyone in this group -- other than the editor, and only at the same time
he makes the change. It certainly has not been discussed.
> I don't think that actually changes the meaning of our draft in any way.
I do, and in a way that has neither been suggested nor discussed, and has
been done after the supposed "last call" for any changes.
The difference is that you are accepting an article that is unacceptable
(by failing to reject it when the problem is detected) and pretending that
you can "fix" it without the information necessary to do so or the
permission of the poster. That is NOT acceptable. You changed the rules
from "do not try to fix a broken article" into "try to fix a broken
article", and that is a significant difference.
Nor do I find your unilateral action acceptable. You've removed any
pretense that this "standard" is the product of this group. Since you have
shown you have no intention of following the consensus, using that word
only when you don't want to change something but ignoring it when you do,
you should be replaced, before your actions require complaints about the
proposal as a whole. Chairman, time to stop lurking.