From: Russ Allbery (rra@stanford.edu)
Date: Wed Mar 03 2004 - 14:57:45 CST
J B Moreno <planb@newsreaders.com> writes:
> On 3/1/04 3:16 PM, Russ Allbery at <rra@stanford.edu> wrote:
>> J B Moreno <planb@newsreaders.com> writes:
>>> I consider having the article format say that the what is in the
>>> Subject is totally irrelevant and the client can do whatever it likes
>>> to it because it's unstructured to be just the sort of thing to likely
>>> cause problems -- because I see it causing problems everyday.
>> No one has proposed this.
> Actually they have -- both Lily and Stanley have said that we can't tell
> clients what to put in the Subject header....
Okay, let's see if they actually have:
> ==
> On 2/24/04 3:03 PM, John Stanley at <stanley@peak.org> wrote:
>> The header is unstructured. It is not "broken" to insert "Monstrosities",
>> but it is broken to assume some special meaning when it appears.
> ==
That didn't say anything of the sort. It says that doing so is not
broken.
> ==
> On 2/26/04 9:25 AM, Bruce Lilly at <blilly@erols.com> wrote:
>> Charles Lindsey wrote:
>>
>> [...]
>>> it is clear that inserting anything
>>> else is "broken".
>>
>> No. What is clear is that there is some disagreement about that. No
>> convincing argument has yet been made to support the assertion that
>> insertion of anything else is "broken" in any meaningful sense.
> ==
That doesn't say anything of the sort either. It says basically the same
thing as John's message above.
> Same thing, different words -- we have no right to tell clients what
> they should be putting in the Subject. If they randomly choose a back
> reference, then that's fine -- we can't and shouldn't forbid it.
Fallacy of the excluded middle. Not requiring it isn't the same thing as
not giving authors any advice. Both of those responses addressed the
claim that putting anything else in there is "broken", which is a
considerably stronger statement than "not a good idea."
Again, please observe that all three of us have agreed to language that
makes it clear to client authors what to do here. All three of us are
objecting to putting SHOULD and MUST around that language.
-- Russ Allbery (rra@stanford.edu) <http://www.eyrie.org/~eagle/>