Re: The ultimate luxury ?

From: Jesse F. Hughes (jesse_at_phiwumbda.org)
Date: 08/03/04 

Date: Tue, 03 Aug 2004 14:29:12 +0200

jmfbahciv@aol.com writes:

> In article <87pt68o5z1.fsf@phiwumbda.org>,
> jesse@phiwumbda.org (Jesse F. Hughes) wrote:
>>jmfbahciv@aol.com writes:
>>
>>> In article <1c4utlavuwneb.gn7j1zon6lkh.dlg@40tude.net>,
>>> Thomas Edison <tjd@privacy.net> wrote:
>>>>On Mon, 02 Aug 04 11:20:18 GMT, jmfbahciv@aol.com wrote:
>>>>
>>>>>> For Christ's
>>>>>> sake, why would you think I meant an alphabetic sort?
>>>>>
>>>>> Because you said sort.
>>>>
>>>>That's what happens when you argue with someone who knows what they're
>>>>talking about, when *you* don't have a clue.
>>>
>>> OK. Specify the key that would be used to sort the contents of the
>>> reference field. Be very specific.
>>
>>/BAH's technically right. It's not sorting on a single field exactly,
>>since one needs both the references and message-id to do the sort.
>
> You need both to do the retrieval, not the sort. The reason
> the posts are sorted is because of the order you retrieved them
> not because you sorted them. It takes very different code to
> sort the data than it does to retrive the data based on a list.

This is where I think you're simply mistaken.

At least as far as I understand my news client. Your mileage may
vary.

To check what messages are new, you query based on the servers'
article numbers, not on message ids. My client (by default)
fetches all of the new headers (up to a maximum number). The
references are used *only* for sorting the result.

My client (and I bet yours, too) doesn't know articles by message-ids
directly, but by article number. Message-id is just not used that way
by nntp servers or their clients, as far as I understand it.

What role do you think that the message-id plays? News clients do
*not* ask for articles by sending their message-IDs in normal use, as
far as I know. Does yours? Can you show me proof?

>>
>>But there's nothing very deep about this fact. We could pretend that
>>we have a data structure in which the product of these two fields is a
>>field. There's an obvious correspondence between our pretend
>>structure and the real posts.
>
> I can't flowchart that. As a coder, I would have no idea what
> you would want me to write. The only correspondence that I can
> see between the pretend structure and the real posts is an
> the heirarchy of how the contents of the posts got copied. None
> of this data was sorted given keys. (I'm not talking about
> the later sort that was done to list the posts on the TTY for
> the user to choose which to read first.)

I don't have a clue what you're talking about. The correspondence is
this. The pretend structure is the same as the real structure EXCEPT
it has an additional field REF-CROSS-ID. We sort on *that* field
according to an acceptable partial order (in the sense of my previous
post on acceptable partial orders, and adjusted for this funky virtual
field).

But, AS A CODER, you would have no reason to consider that virtual
field. It's only mentioned here for those folks that believe the term
sort must involve sorting on a single field of a record. Now, this
restriction is utterly artificial, so I'm sure that we can dispense
with such semantic nonsense and accept the term "sort" for any
computable function from a set of records to a linear order.

>
>
>>
>>As far as the order on that key field, I've sketched it elsewhere, but
>>the order I sketched needs filled in.
>
> <GRIN> Exactly. The divil's in the details.

Not here. The details are just details. The point is that "sorted
according to references" is underdetermined. You need more rules for
equivalent posts (i.e., posts x and y in which neither x nor y is an
immediate successor of any post fetched or x and y are immediate
successors of the same post).

Not a big deal. 

-- 
"It has been shown that no man can sit down to write without a very profound
design. Thus to authors in general trouble is spared. A novelist, for example,
need have no care of his moral. It is there -- that is to say, it is somewhere
-- and the moral and the critics can take care of themselves." --E.A. Poe

Relevant Pages

  • Re: Town Shaken after U S Attorneys Arrest in Child-Sex Sting
    ... First let me say - yes I see your posts. ... While i think you make a good point, that perhaps there are pedophiles ... of these chatrooms/ groups might be the sort needy for attention - any ... AB Yahoo group was removed is that they don't want to publicize the ...
    (alt.true-crime)
  • Re: Town Shaken after U S Attorneys Arrest in Child-Sex Sting
    ... Perhaps the silence that follows your posts could be considered a compliment. ... While i think you make a good point, that perhaps there are pedophiles who might prey on AB women who have real children, since women in some of these chatrooms/ groups might be the sort needy for attention - any sort of attention - and even be easily lured into putting their children into harms way, I also don't agree that his yahoo profile indicates that he is primarily interested in young girls. ... The title "Adult Babies" - to me, ...
    (alt.true-crime)
  • Re: My life..
    ... Killing someone if you are not lucky - and then nabbed by the cops ... So you have to sort yourself out so you stop being a pest. ... something useful and kick your own head into slightly better shape. ... Your posts are what this newsgroup is for. ...
    (uk.people.support.depression)
  • Re: Question concerning object-oriented programming
    ... in any sort of civil and rational discussion is futile. ... don't understand the posts. ... the "MI5 Conspiracy" spammer in my killfile and I'll try to forget he ... government about his property taxes being escorted out of the library, ...
    (comp.programming)
  • Re: Do carers get paid?
    ... However she gets no help of any sort nor is entitled to any. ... I sign my post pete ... unprovoked attacks and abuse upon me and my family has been recorded ... I am archiving every single one of your posts to me..... ...
    (uk.people.disability)
Loading