news.software.readers

[References header re-constructed]
On 2008-02-11, problems@gmail wrote:

snip

> Whisker wrote:
>> Do you mean you bypass all of Google's presentation stuff and fetch
>> individual articles using a command along the lines of this?:
>>
>> lynx
>> /groups?selm=..
>
> Yes, but threads, not articles via:
> lynx -dump >> > is separated with 2 lines: "<>...<><>" and < It's URL > >
>
>
>> Although Lynx doesn't thread, you can use the References header to
>> find the 'parent' articles of the one you're reading.
>
> Yes but the 'thread' that I get from google via lynx doesn't contain it.

That's because you aren't fetching newsgroup articles from a news-server,
you're fetching web pages from a web site - and the pages you are fetching
are designed not to include the References header.

You are of course fetching all the HTML created by Google, along with the
text of the original articles, thus using a lot of bandwidth you could
easily save by using a real news-server. Your article that I'm replying
to here, is 2909 bytes in my local 'spool'; the Google web page containing
that article (and no other articles) is 34354 bytes according to my
browser - more than 10 times as much 'bandwidth' for the same message!

To get the References header from Google, you'll need to find their
individual 'Show Original' page for each article - and then devise some way
of your own to 'thread' the resulting web pages to emulate a real
news-reader. Which seems a rather pointless exercise, given that you can
use a purpose-built newsreader program to get a real newsfeed from a real
news-server without going anywhere near Google or the web.

> And I annoy and even possibly miss posters by not being able to
> continue in the thread that I'm answering.

That's what you get for going to such great lengths to make your life so
difficult.

> OTOH [as you might have noticed] I often reply to multiple posters,
> in a single reply, which may have split into different threads, by my
> 'new branch'.

It was only by chance that I spotted this response to my previous post;
normally I rely on my newsreader to draw my attention to articles which
have any of my own MIDs in the References header, and you've broken that
conventional approach.

It's your own fault if the people trying to communicate with you can't do
it - you have quite literally lost the thread. It's the sort of thing
that will get you a bad reputation in usenet, and probably get you plonked
entirely (although the method you use to post means that you have very
nearly removed yourself from usenet as far as most people are concerned).

> BTW it would be marvelous if individual servers could search/find:
> ( GroupName, StringInSubject OR Author) .
>
> AFAIK that not part of the RFC ?
>
> Thanks for any feedback,

Why should a news-server do the work that your news-reader should be
doing? Once you've fetched the 'overview' you can search your local files
to your heart's content - it's what news-reader programs /do/.

You seem to have an account with a news-server, and a Linux operating
system, so why not use existing well-established software and methods to
read and post to usenet? Linux abounds in suitable 'stuff' designed for
the job whether your internet connection is high-speed and free of charge
or horribly slow and expensive. I'd suggest that you look at Tin, slrn,
Gnus, Leafnode, and Noffle, for a start. They're all 'free'.

--
-- ^^^^^^^^^^
-- Whiskers
-- ~~~~~~~~~~