From: Seth Breidbart (sethb@panix.com)
Date: Wed Aug 11 2004 - 17:03:05 CDT
John Stanley <stanley@peak.org> wrote:
> Seth Breidbart (sethb@panix.com):
>>What is this "starts out" stuff?
>
> Ummm, is there something to this question besides the obvious? "Starts
> out" means "begins".
Why do we care about how something begins? We care about what gets
posted.
>>Is there really a difference between
>>a followup agent that starts by creating an empty file (clearly, with
>>no References header) and then appends stuff to it that includes the
>>References header, and one that starts by creating a file with the
>>References header already in it?
>
> When the definition of something is based on what it contains, yes.
Why? They both _contain_ the same stuff by the time the user gets to
edit the file.
>>We (should) care about the format of the (proto-)article, not the
>>process of generating it.
>
> By defining what kind of article it is based on what it contains, we care
> about the process.
Why? There are many different processes that lead to the same
contents. If we care (only) about the contents, why should we
distinguish among those processes?
>>If I save an article to a temp file, then use a posting (not followup)
>>agent, and insert that article as quoted, with an appropriate
>>References header, etc., then is my article a followup?
>
> Wouldn't it be nice if this kind of question didn't need to be asked? As
> in, what FUNCTION does the article perform and not just what headers are
> found in it (or not, as the case may be)?
Sure; but that isn't the point I'm trying to get at.
>>Well, it isn't a followup until I added it,
>
> And there is no reason to add it because it isn't a followup.
I want my article displayed (when displayed by a threaded newsreader)
immediately after the one I'm responding to. That's the reason I
chose to add the References header.
>>> The only thing contrary to fact is the ridiculous suggestion that we
>>> define a followup not by what function is serves but by what header
>>> it contains.
>>
>>A suggestion cannot be contrary to fact. Neither can a definition.
>
> Oh please. Try defining "apple" to be "an eight-legged member of the
> arachnid family" and see if someone doesn't tell you you've got your facts
> mixed up.
If I say that the dictionary defines it that way, then I'm wrong. If
I say that for the purpose of this discussion, assume a spherical cow,
then I'm not wrong. In the case at hand, the definition will be
whatever _we_ decide; if there were already an existing one (as there
is for "apple") then we wouldn't be having this entire discussion.
> Charles Lindsey (chl@clerew.man.ac.uk):
>>But followup agents (at least the ones whose duties are set out in our
>>present draft, and will continue to be set out in USEPRO) are not involved
>>in the generation of multipart FAQs
>
> Says who?
Statistics.
> Sorry, I didn't get the memo from you prohibiting this use of
> followup agents for this.
It isn't prohibited, so much as rare to nonexistent.
>>Which, to me, is a good argument for not regarding later parts of a FAQ as
>>"followups",
>
> Yes, you are still stuck on this idea that followups are defined as "what
> a followup agent produces" and not by the function in life that they
> serve.
A definition involving "how the poster wants them displayed" would
solve that problem.
*******
> Second, whether or not it starts with a References header, the
> poster is free to remove any such header from the article prior to
> posting, so your contention that the protoarticle starts with a
> References header does not mean it will be a followup.
So if there's a References header that the poster removes, the
article is not a followup.
But if there's no References header initially, but the poster adds
one, that doesn't make the article a followup.
That doesn't seem consistent.
Seth