Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]

Groups > comp.lang.python > #106881 > unrolled thread

Is threading better on Usenet or gmane?

Back to article view | Back to comp.lang.python

Contents

  Is threading better on Usenet or gmane? Grant Edwards <invalid@invalid.invalid> - 2016-04-11 14:24 +0000
    Re: Is threading better on Usenet or gmane? Random832 <random832@fastmail.com> - 2016-04-11 10:41 -0400
      Re: Is threading better on Usenet or gmane? Grant Edwards <invalid@invalid.invalid> - 2016-04-11 14:55 +0000
        Re: Is threading better on Usenet or gmane? Grant Edwards <invalid@invalid.invalid> - 2016-04-11 15:58 +0000
          Re: Is threading better on Usenet or gmane? Grant Edwards <grant.b.edwards@gmail.com> - 2016-04-11 17:16 +0000

#106881 — Is threading better on Usenet or gmane?

I've been reading c.l.p on Usenet for many, many years.  There has
always been a certain abount of thread breakage (presumably due to
broken e-mail clients and/or the list<->usenet gateway), but it seems
to have gotten worse lately.

Has anybody noticed whether the threading is less broken if one reads
the list on gmane?  Everytime I decide to try to do a direct
comparison, I can't find enough thread breaks to tell if there's a
significant difference.

-- 
Grant Edwards               grant.b.edwards        Yow! Somewhere in Tenafly,
                                  at               New Jersey, a chiropractor
                              gmail.com            is viewing "Leave it to
                                                   Beaver"!

[toc] | [next] | [standalone]

#106883

On Mon, Apr 11, 2016, at 10:24, Grant Edwards wrote:
> I've been reading c.l.p on Usenet for many, many years.  There has
> always been a certain abount of thread breakage (presumably due to
> broken e-mail clients and/or the list<->usenet gateway), but it seems
> to have gotten worse lately.
> 
> Has anybody noticed whether the threading is less broken if one reads
> the list on gmane?  Everytime I decide to try to do a direct
> comparison, I can't find enough thread breaks to tell if there's a
> significant difference.

I've already outlined under what circumstances threading is likely to be
broken (before the recent changes):

| For users reading by the mailing list, Usenet users' replies to
| Mailing List users will be broken (but their replies to each other
| will be fine). For users reading by Usenet, Mailing List users'
| replies to each other will be broken (though all replies made via
| Usenet or to Usenet users will be fine).

gmane is essentially the same as reading it by email as far as this
issue is concerned.

What the new change (see Mark Sapiro's recent posts on this topic)
should fix is the first half of that. It doesn't necessarily do anything
about the second half:

Email User A posts message B@email.com, this becomes C@python.org (the
new change makes this look like a reply to the nonexistent B@email.com)
Email User D replies with E@email.net, this becomes F@python.org (but
still says it's a reply to B@email.com)
Usenet User G replies (to A/B) with H@usenet.com, this is a reply to
C@python.org, but (with the new change) acknowledges B@email.com as the
"grandparent article", so mailing list users should still see it under
B's thread.

So, even though D's message is a reply to A's message, it won't show up
that way on usenet. G's message should show up fine on both.


Everything would be perfect if everyone posted to Usenet and read by
email/gmane.

[toc] | [prev] | [next] | [standalone]

#106884

On 2016-04-11, Random832 <random832@fastmail.com> wrote:

> I've already outlined under what circumstances threading is likely
> to be broken (before the recent changes):
>
>| For users reading by the mailing list, Usenet users' replies to
>| Mailing List users will be broken (but their replies to each other
>| will be fine). For users reading by Usenet, Mailing List users'
>| replies to each other will be broken (though all replies made via
>| Usenet or to Usenet users will be fine).
>
> gmane is essentially the same as reading it by email as far as this
> issue is concerned.
>
> What the new change (see Mark Sapiro's recent posts on this topic)
> should fix is the first half of that. It doesn't necessarily do anything
> about the second half:

OK. So recent fix should fix all gateway-related thread breakage for
mailing-list readers (which includes gmane).  I had read most of the
thread on that change, but was still confused about exactly what was
getting fixed.

Thanks for clarifying it.

-- 
Grant Edwards               grant.b.edwards        Yow! I'll eat ANYTHING
                                  at               that's BRIGHT BLUE!!
                              gmail.com            

[toc] | [prev] | [next] | [standalone]

#106885

On 2016-04-11, Grant Edwards <invalid@invalid.invalid> wrote:

I just finished checking a very recent thread containing 67 articles
by pointing slrn at news.panix.com for the Usenet version and at
news.gmane.com for the mailing-list version.

Both servers appeared to show the same set of 67 articles (that alone
is pretty good) though the tree structures were rather different.

Then I checked each article to see if its parent article was availble.

On the Usenet side, 35 of 67 articles had unavailable parent articles,
which is pretty bad compared to non-gatewayed groups.

On the mailing-list side 15 of 67 had unavailable parent articles,
which is a bit shabby by general usenet standards, but probably pretty
decent for gatewayed groups.

How representative that thread was is left as an exercise for the
reader.

-- 
Grant Edwards               grant.b.edwards        Yow! If Robert Di Niro
                                  at               assassinates Walter Slezak,
                              gmail.com            will Jodie Foster marry
                                                   Bonzo??

[toc] | [prev] | [next] | [standalone]

#106887

On 2016-04-11, Grant Edwards <invalid@invalid.invalid> wrote:
> On 2016-04-11, Grant Edwards <invalid@invalid.invalid> wrote:
>
> I just finished checking a very recent thread containing 67 articles
> by pointing slrn at news.panix.com for the Usenet version and at
> news.gmane.com for the mailing-list version.
[...]
> On the Usenet side, 35 of 67 articles had unavailable parent
> articles,
[...]
> On the mailing-list side 15 of 67 had unavailable parent articles,

FWIW, an automated test on the 1000 most recent articles showed this:

                         Usenet   Gmane
        No References      69      104
   Unavailable Parent     357       34

Due to moderation/filtering, the set of articles from the two servers
probably isn't identical.

For posts that aren't follow-ups, no references is normal, but I'm
surprised by the size of the difference in number of articles with no
references header.

An unavailable parent means that there _was_ a refrences header, but
the most recent reference didn't map to an available article on the
server in question.

In general, the situation appears to be significantly better on Gmane
-- even if we assume that the excess number of "no reference"
ariticles on Gmane indicates those articles are broken and should have
references.

-- 
Grant Edwards               grant.b.edwards        Yow! Let me do my TRIBUTE
                                  at               to FISHNET STOCKINGS ...
                              gmail.com            

[toc] | [prev] | [standalone]

Back to top | Article view | comp.lang.python

csiph-web