Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss

Could converting blogs to USEnet work?

1 view
Skip to first unread message

no.to...@gmail.com

unread,
Jan 21, 2012, 1:00:08 PM1/21/12
to
We all know how blogs have degenerated USEnet,
mainly by the absurdity that you need to d/l the whole thread
to see *IF* there's a reply to your comment/query.

Which results in few multi-stage dialogs.
So the result is that 'contributors' all rush to have their
once-off shout. And it's degenerated to a Friday night
bar-shout.

Isn't it also problematic to 'hold' 8 blogs open on your
'browser' at the same time as using it for other http-fetches ?

What would happen if 10 blogs which were likely to be read
by a cluster of inetters, were 'harvested' and NNTP-served.

Let's consider the NNTP fields:
From=ok; Date=ok; Subject=ok; ... seems ok.

The threading would initially be weak/lacking because
the bloggers normally find it difficult to put that info in
their message; but as they became aware of threading
capabilities they would use it.

Probably the biggest problem would be legally, from
the owner of the harvested blogs?

Personally, I never use a full 'browser' for blog-reading.
`lynx` gets me the text, without that crap-in-ya-face.

I've never used RSS. Perhaps that already fixes the problems?

What do you think?

== TIA.





Richard Kettlewell

unread,
Jan 21, 2012, 1:06:41 PM1/21/12
to
no.to...@gmail.com writes:
> We all know how blogs have degenerated USEnet,
> mainly by the absurdity that you need to d/l the whole thread
> to see *IF* there's a reply to your comment/query.
>
> Which results in few multi-stage dialogs.
> So the result is that 'contributors' all rush to have their
> once-off shout. And it's degenerated to a Friday night
> bar-shout.
>
> Isn't it also problematic to 'hold' 8 blogs open on your
> 'browser' at the same time as using it for other http-fetches ?

That's why people use RSS aggregators.

> What would happen if 10 blogs which were likely to be read
> by a cluster of inetters, were 'harvested' and NNTP-served.
>
> Let's consider the NNTP fields:
> From=ok; Date=ok; Subject=ok; ... seems ok.
>
> The threading would initially be weak/lacking because
> the bloggers normally find it difficult to put that info in
> their message; but as they became aware of threading
> capabilities they would use it.
>
> Probably the biggest problem would be legally, from
> the owner of the harvested blogs?
>
> Personally, I never use a full 'browser' for blog-reading.
> `lynx` gets me the text, without that crap-in-ya-face.
>
> I've never used RSS. Perhaps that already fixes the problems?

The limited RSS->NNTP gateway in
http://www.greenend.org.uk/rjk/2006/newstools.html is designed with LJ's
RSS in mind, but should be readily adaptable to anyone else's.

--
http://www.greenend.org.uk/rjk/

no.to...@gmail.com

unread,
Jan 22, 2012, 3:13:45 PM1/22/12
to
Richard Kettlewell wrote:-
> The limited RSS->NNTP gateway in
> http://www.greenend.org.uk/rjk/2006/newstools.html is designed with LJ's
> RSS in mind, but should be readily adaptable to anyone else's.

Obviously the greatest [productivety eg.] improvements come from optimising
at the highest level. So to use your <blogtoNTTP> I'd want a mental picture
of how it works. We all want to avoid starting a journey which doesn't
succeed in reaching a worthwhile destination. But since it seems that we're
both on the same page, I'll try it - for a further optimisation of my inet
usage.

Perhaps you can understand my post:
Newsgroups: comp.lang.forth
Subject: Re (2): existing code to build a 'sed-like'?

I'm investigating another quirk which may interest you:
as you know, many webPublishers give you each page with the same 90% duplicated.
So [with eg. `mc`] we see that the repeated-garbage, of the lynx viewed <dump>
ends at byte 12345, and then in future we do:
<curl fetch from 12344 URL >> AccumulatingFile>

I've just tested an amazing minute `html2txt` which comes from mulinux: the
90's linux on a 1M7 fd0. I think lynx, links, elinks ..etc would crash if
the *html is mangled, with <unmatched brackets>.

The whole idea is like they had/have <advert supressors> for TV viewing.
You'd `curl` the good-text after the repeat-garbage and read it through
a `html2txt`. mulinux uses `nc` to fetch every thing. It's apparently
even more low-level than `curl`. Perhaps it's also possible to abort the
tail-part of repeated-garbage?

So you'd have a script with the 3 args, to do:
`fetchHttp head tail URL`
and then as you regularly visit certain sites you have a tuned script for each,
and then as the sites changed their format from time to time you'd easily,
retune the scripts accordingly.

Of course the trimmed down text, is no good to get your next cycle of links
from your <lynx dump>.
What I'd really like to have is a method/script which would <extract/append the
URLs to a file> while I'm reading my `lynx -dump URL` and see interesting
'links'. So while you're reading the text at the top and you see that
<link 129> looks interesting to fetch later,
" 129. <URL129>"
can automagically append the <URL129> to the file-of-URLs which will be
`lynx -dumped` next time on-line.

Thanks,

== Chris Glur.

Stefan Monnier

unread,
Jan 22, 2012, 4:31:36 PM1/22/12
to
> What would happen if 10 blogs which were likely to be read
> by a cluster of inetters, were 'harvested' and NNTP-served.

I think gwene.org is a possible answer.


Stefan
0 new messages