Russ Allbery | 1 Jun 01:34 2004
Picon

Re: injection and moderated newsgroups

Charles Lindsey <chl <at> clerew.man.ac.uk> writes:

> No, is is clear in 99.99% of cases it is clear _in_spite_of_ the fact
> that there is no central authority. See our section 9:

More like 99%. 99% solutions *at the protocol level* are not good
engineering.  You have to deal with the other 1% too, and the current text
does not in an acceptable fashion.

In this case, there's a clear and obviously correct answer:  don't attempt
to dictate local choices, and document the fact that these decisions vary
from site to site and deal with that in the rest of the protocol.

--

-- 
Russ Allbery (rra <at> stanford.edu)             <http://www.eyrie.org/~eagle/>

Alexey Melnikov | 2 Jun 12:06 2004

Re: documents etc. (was Re: Back-references and USEAGE)

Henry Spencer wrote:

>On Thu, 27 May 2004, John Stanley wrote:
>  
>
>>If he just puts them on the website as the current discussion version, 
>>there would be no problem.
>>    
>>
>
>How about if he just puts them on the IETF draft web/FTP site as the
>current discussion version?
>  
>
> ...
>
>>It's not rocket science, Henry. It's just common sense.
>>    
>>
>
>No, it's nonsense.  Why have two levels of draft process when one suffices?
>There's no need to have "tentative" drafts and "real" ones -- a draft is
>tentative by definition.
>
I agree with Henry. New drafts should be published whenever the WG feels 
that there are substantial changes to the document. Feel free to bug me 
about publication of new revisions.

I also agree that 6 month cycle is a bit long.

(Continue reading)

Bruce Lilly | 3 Jun 03:55 2004
Picon

Re: Back-references and USEAGE

Charles Lindsey wrote:
> In <40B7D092.6040307 <at> erols.com> Bruce Lilly <blilly <at> erols.com> writes:
> 
> 
>>Charles Lindsey wrote:
>>
>>>Contrary to your claims, last year's WG discussions did NOT reach any
>>>conclusion (our Chair ceased to participate despite being asked to rule on
>>>a way forward).
> 
> 
>>That is incorrect. The Chair clearly ruled that the text in question had
>>no place in the syntax & semantics document (the definition of the USEFOR
>>document).  He also ruled that failing some credible connection between
>>"Re: " and some protocol issue (specifically excluding display), that it
>>was unsuitable for the network operations document (the second of the two
>>Standards Track documents that we were directed to produce).  No such
>>connection was ever produced.
> 
> 
> I am sorry, but you are simply not telling the truth.
> 
> There was no final ruling made.

The public record of this WG supports my claim [as I write this, the
landfield.com site has been down most of the day -- I placed a telephone
call to Kent; the information below should suffice to identify the referenced
messages when the site comes back up]:

The co-chairs set the framework, complete with milestones and a timetable:
(Continue reading)

Alexey Melnikov | 2 Jun 12:06 2004

Re: documents etc. (was Re: Back-references and USEAGE)

Charles Lindsey wrote:

>In <40B7DCFA.30204 <at> erols.com> Bruce Lilly <blilly <at> erols.com> writes:
>  
>
>>Henry Spencer wrote:    
>>
>>>Why is this different from putting it in ftp.ietf.org/internet-drafts as
>>>the current discussion version?
>>>      
>>>
>>That would work, but Charles has explicitly refused to do so when requested
>>to provide the full text of the document for discussion.
>>    
>>
>
>There is little point in publishing a full draft, whether on the landfield
>site or as an internet draft, when some text in question is under
>discussion and changing on an almost daily basis. Particularly so when, as
>in the particular recent cases referred to, the other parts of the
>document providing the context for the discussion at hand had not changed
>since the last full draft, as was pointed out at the time.
>
>I will publish a full draft whenever a relatively stable position has been
>reached, or where the interactions between various parts of the document
>have become overly complex. But where the changes are confined to one or a
>few short sections, it is better simply to publish the full text of those
>sections, as they currently stand, to this list.
>
I agree any small self-contained changes should be sent directly to the 
(Continue reading)

Charles Lindsey | 1 Jun 11:19 2004
Picon
Picon

Re: injection and moderated newsgroups

In <877jusf9wu.fsf <at> windlord.stanford.edu> Russ Allbery <rra <at> stanford.edu> writes:

>Charles Lindsey <chl <at> clerew.man.ac.uk> writes:

>> No, is is clear in 99.99% of cases it is clear _in_spite_of_ the fact
>> that there is no central authority. See our section 9:

>More like 99%. 99% solutions *at the protocol level* are not good
>engineering.  You have to deal with the other 1% too, and the current text
>does not in an acceptable fashion.

>In this case, there's a clear and obviously correct answer:  don't attempt
>to dictate local choices, and document the fact that these decisions vary
>from site to site and deal with that in the rest of the protocol.

But, if you look closely (the third paragraph of section 7 in particular)
you will see that we don't actually dictate local choices. What we say is
that if you DO accept a newgroup message, then you MUST implement it
correctly.

However, in this case the problem is not the sites that make poor local
choices, but the fact that the consequence if those choices leads to the
propagation of unapproved moderated articles, against which the only
protection is checks made by relaying agents (and that check is only a
MAY, because you did not want me to upgrade it to a SHOULD).

Others in this discussion seem to prefer retaining the present stronger
wording in 7.2.1, so I see no consensus to change it yet.

--

-- 
(Continue reading)

Charles Lindsey | 1 Jun 11:22 2004
Picon
Picon

Re: Back-references and USEAGE

In <HyKtD6.7vw <at> clerew.man.ac.uk> "Charles Lindsey" <chl <at> clerew.man.ac.uk> writes:

>I have received the following from Dirk Nimmich which, for some reason, he
>did not post to this list. However, it seems to make the position of
>"Ant: " and its variants quite clear.

A further message from Dirk Nimmich, who seems to be having problems
posting to this list.

From: Dirk Nimmich <nimmich <at> muenster.de>
Date: Mon, 31 May 2004 20:37:43 +0200
To: usenet-format <at> landfield.com
Subject: Re: Back-references and USEAGE
Message-ID: <20040531183743.GA1487 <at> roxel.fqdn.de>
Mail-Followup-To: usenet-format <at> landfield.com
User-Agent: Mutt/1.4.1i

Charles Lindsey wrote:
> I have received the following from Dirk Nimmich which, for some
> reason, he did not post to this list.

The reason is that it wasn't intended for the public. I definitely
don't have the time to discuss things on the list; it takes me ages
to compose a message in English, and more than once I misinterpreted
things or was misinterpreted.

> However, it seems to make the position of "Ant: " and its variants
> quite clear.

Only when talking about de.*.
(Continue reading)

Charles Lindsey | 1 Jun 11:52 2004
Picon
Picon

Re: Followups and back-references

In <20040531202146.GE3152 <at> tagseth-trd.consultit.no> Eivind Tagseth <eivindt <at> multinet.no> writes:

>I don't think it is a USEAGE matter, it is one of the disadvantages of
>using a back-reference; it is not possible to replace the "Re: " string
>with a localized variant (a variant making more sense in a certain
>language) neither when creating the article or when displaying the
>article.  I happen to believe that this is a good reason for not
>adding any back-reference.

Exactly so, but the present wording makes it clear that adding a back
reference is entirely optional. So what more do you want it to say?

You seems to be arguing that, since you are not allowed to prepend "Sv: ",
other people should not be able to prepend "Re: ". Whilst that may indeed
be desirable, it is not going to happen simply because of the overwhelming
inertia behind the present practice.

>> As for "need", is the following better?
>> 
>>         ... Some
>>         reading agents take note of the Subject-header (as well as other
>>         headers) when presenting articles for display (again, see
>>         [USEAGE]) and such agents often find it necessary to be able to
>>         distinguish when such a "Re: " is present. ...

>I still find it very awkward.  How about:

>	Some reading agents are able to recognize the presence of a
>	back-reference (or, actually _guess_ that a back-reference is
>	present), and will handle the article different than usually.
(Continue reading)

Charles Lindsey | 1 Jun 11:29 2004
Picon
Picon

Re: Back-references and USEAGE

In <20040531184532.GD3152 <at> tagseth-trd.consultit.no> Eivind Tagseth <eivindt <at> multinet.no> writes:

>* Charles Lindsey <chl <at> clerew.man.ac.uk> [2004-05-31 11:39:17 +0000]:

>> In <20040529003448.GB836 <at> tagseth-trd.consultit.no> Eivind Tagseth <eivindt <at> multinet.no> writes:
>> >Ok, I agree, the term "recognize" is wrong.  But I still don't see the
>> >damage in _guessing_ if the start of the subject is (or rather: was) a

>> Would s/recognize/detect/ help?

>No, the point is that it is not possible to tell if the subject of an
>article contains a back-reference or not (depending on the definition of
>a back-reference of course), or if "Re: " is written by the user, having a
>completely different meaning than a back-reference.  The newsreader
>can only _guess_.

Ah! In that case they have found an "apparent" back-reference. How about
the following:

   It would be wiser for any followup agents which detect apparent non-
   standard back-references such as "Re(2): ", "Sv: ", etc. to refrain
   from prepending anything further, but other attempts to mend that
   problem are likely to do more harm than good.

--

-- 
Charles H. Lindsey ---------At Home, doing my own thing------------------------
Tel: +44 161 436 6131 Fax: +44 161 436 6133   Web: http://www.cs.man.ac.uk/~chl
Email: chl <at> clerew.man.ac.uk      Snail: 5 Clerewood Ave, CHEADLE, SK8 3JU, U.K.
PGP: 2C15F1A9      Fingerprint: 73 6D C2 51 93 A0 01 E7 65 E8 64 7E 14 A4 AB A5

(Continue reading)

Alexey Melnikov | 2 Jun 00:50 2004

Re: Back-references and USEAGE

Nick Boalch wrote:

> Charles Lindsey wrote:
>
>>>>   4. Even the presence of "Re: " at the start of a Subject may
>>>>      occasionally be misleading, because it might have been
>>>>      deliberately placed there by a perverse poster rather than having
>>>>      been generated automatically by a followup agent.
>>>
>>> I don't think there's really any call for the word 'perverse' 
>>> there... a Subject-header starting with "Re: " is perfectly 
>>> legitimate, after all.
>>
>> Yes, perhaps that it a bit OTT, although it does represent the 
>> actualité.
>> I don't think anyone has ever seen a Subject starting with "Re: " unless
>> it has been autamatically generated, or it was a poster genuinely trying
>> to join an ongoing discussion, or it was some show-off trying to be
>> "clever".
>
>
> That's probably true, but by our own draft's definition it is an 
> entirely legitimate way to start the header, so I think we should tone 
> down the sentence to 'intentionally placed there by a poster as an 
> integral part of the Subject-content' or some such wording. 

This sounds reasonable to me.

Alexey

(Continue reading)

Bruce Lilly | 2 Jun 16:08 2004
Picon

Re: injection and moderated newsgroups

Charles Lindsey wrote:
> In <40B7D35D.4020807 <at> erols.com> Bruce Lilly <blilly <at> erols.com> writes:
> 
> 
>>If I understand Russ correctly, he is saying that it is *NOT* clear what
>>*the* moderation status of a group is, precisely *because* there is no
>>central authority.
> 
> 
>>Do you agree with that or not?
> 
> 
> No, is is clear in 99.99% of cases

Please cite the source of the "99.99%" claimed statistic.  Or admit that it
is yet another unsubstantiated claim.

>  it is clear _in_spite_of_ the fact that
> there is no central authority. See our section 9:
> 
>    ......... Usenet is a prime example
>    of an Internet Adhocratic-Anarchy; that is, an environment in which
>    trust forms the basis of all agreements.  It works.

We are not discussing whether or not Usenet "works" for some unspecified
definition of "works".  We are discussing text which is either suitable
for a Standards Track RFC or not suitable.

A putative "standard" that doesn't address 100% of the requirements
that *that* document places upon would-be compliant software is not
(Continue reading)


Gmane