Re: Gnus: How to catch up huge newsgroups?

[Ted Zlatanov <tzz@lifelogs.com>]

On Mon, 03 Nov 2008 20:10:17 -0500 Joe Fineman <joe_f@verizon.net> wrote: 

JF> I run Gnus v5.10.6 under Emacs 21.3.1 under Windows XP (alas).
JF> Because Verizon has truncated its service, I recently signed on to a
JF> latitudinarian news server (Motzarella).  I find that four of the
JF> newsgroups I want to read come up with backlogs on the order of 30
JF> million postings.  When I try to do catchup (C) on one of them, it
JF> stalls & eventually gives an out-of-memory message.  (The first time,
JF> when I let it try for a longish time, it ended up giving interminable
JF> beeps & hanging both Emacs and Windows; I had to do a hardware
JF> shutdown.)

JF> What does one do in such a situation?  I see nothing about it in info
JF> or Google or the Gnus FAQ.

I've never seen this, but it should be fairly easy to edit the
newsrc.eld file and catch up from there.  Make sure you're not running
Gnus.  For example I have:

("nntp+news.albasani.net:comp.risks" 3 ((1 . 148)) ((seen (118 . 148))) (nntp "news.albasani.net"))

Which says that 118-148 have been seen.

You may run into problems with very large numbers as far as Emacs is
concerned.  There's no easy fix there but you should report it to the
Gnus mailing list nevertheless so you can get some answers and maybe
someone will work on it.

JF> I have tried twice to post this query on gnu.emacs.gnus (a moderated
JF> group), but it has not appeared, nor have I received any reason for
JF> its rejection.

I'm not sure what's up with your news server--that group is not
moderated on mine, but you can try the ding mailing list <ding@gnus.org>
instead.  That's the usual place to talk about Gnus development, and
it's accesible through gmane.org as well if you prefer NNTP.

Ted