Lars Magne Ingebrigtsen | 25 Sep 15:04 2014

bash sploit

I don't think that anything on Gmane is vulnerable to the bash sploit
(no cgi-bin etc that runs bash), but I've upgraded bash on all the
servers, anyway, since there's a lot of stuff running there.

ger:~# env x='() { :;}; echo vulnerable' bash -c 'echo hello'
bash: warning: x: ignoring function definition attempt
bash: error importing function definition for `x'

(And, too.  It does run cgi-bin, because it runs cgit.)


(domestic pets only, the antidote for overdose, milk.)
   bloggy blog:
Rainer M Krug | 24 Sep 10:09 2014

Enable search in gwene groups on gmane


I am using gnus and thought, as gwene groups are under gmane, that I can
search these as well - but I can't.

Apparently, to enable search of gwene groups, two things need to be

1) Trivial fixes in nnir.el (not my words - but from Ted [1])
2) needs to support search in gwene

My question: could this be implemented in gmane?





Rainer M. Krug
email: Rainer<at>krugs<dot>de
PGP: 0x0F52F982
Gmane-discuss mailing list
Gmane-discuss <at>
(Continue reading)

era | 18 Sep 14:41 2014

Incorrect spam classification

I stumbled across a message which had been classified as spam, where
apparently it shouldn't have.

I expect to find more of these as I fool around with the spam stuff.  Please
remind me of the correct procedure for nominating articles for human review
-- I recall there used to be one, but I have been unable to find this stuff
in Google.  Some old articles imply that there used to be (or should be?) an
X-Unspam: header with a link you can visit to report the false positive, but
there isn't one now.


If this were my real .signature, it would suck less.  Well, maybe not.
Lars Magne Ingebrigtsen | 18 Sep 14:21 2014

Disk switcharoo

After procrastinating for half a year, I've decided it's too scary to
just swap out the old SSDs with the new, bigger ones.

So I've bought a new RAID controller and will be inserting that, and all
the SSDs, into a machine that has sufficient disk slots for that, and
copy over the spool.

It'll be more controlled, and downtime should be like half an hour
instead of a day.

The only machine that has sufficient number of slots is  It's currently being used as the MTA and as an
experimental search engine thing, I think.

It currently has the following disks:

/dev/sda7             2.0G  625M  1.3G  33% /tmp
/dev/sda8             7.9G  2.4G  5.1G  33% /var
/dev/sda9              12G  2.9G  8.4G  26% /usr
/dev/sda10             30G  4.1G   25G  15% /opt
/dev/sdb1             1.4T  747G  560G  58% /mirror
/dev/sdc1             1.4T  269G  1.1T  21% /index
/dev/sdd1             1.4T  1.1T  276G  79% /ispool

So four slots taken, and the mirror disks are too small.  So I'm going
to rip out those three 1.5TB disks and put in the five 1TB SSDs, and
also two new 4TB disks that can be used for search indexing and stuff.

(The machine has room for 8 disks.)

(Continue reading)

Duncan | 17 Sep 06:35 2014

Re: Does a SPAM report initiate a cancel <message-id>?

rh_ posted on Tue, 16 Sep 2014 19:33:55 -0700 as excerpted:

> Looking though this list I thought the mechanism was to initiate a
> cancel <message-id> when a spam report is made via the web interface.
> Look at comp.editor.joe for example. I've made spam reports for the spam
> but the spam remains when using nntp for that group. It's not a
> client-side problem as I've cleared my cache.

AFAIK the offending posts aren't (directly) canceled, they get cross-
posted to gmane.spam-report (don't hold me to the specific name), an 
unpublished gmane-internal list/group.

So if you set your news-client's filters to kill anything cross-posted to 
that group, the messages should disappear.

But there's a human element as well.  I'm unclear on the details but I 
believe that while the cross-post is automatic, the actual cancel would 
be from a gmane admin after review to ensure it's actual spam (to prevent 
accidents and malicious reporting of non-spam).


Duncan - List replies preferred.   No HTML msgs.
"Every nonfree program has a lord, a master --
and if you use the program, he is your master."  Richard Stallman
Steinar Bang | 17 Sep 09:38 2014

Re: Does a SPAM report initiate a cancel <message-id>?

>>>>> rh_ <richard_hubbe11 <at>>:

> Look at comp.editor.joe for example. I've made spam reports for the
> spam but the spam remains when using nntp for that group. It's
> not a client-side problem as I've cleared my cache.

Lars applied a patch to INN that removed (or tried to...) the cancelled
messages from the NOV file for the group (there should be some messages
on the patches in this group, many years back).

And the patch sort of works... sometimes... I haven't been able to see
any pattern in when it works and when it doesn't.
JLuc | 14 Sep 19:44 2014

Search broken ?

(oups sorry for posting to wrong list)

it looks like the search pages are broken on

it should be on
like for example

maybe something needs to be fixed ?
or is there a workaround ?

積丹尼 Dan Jacobson | 9 Sep 08:57 2014

.../ 404 Not Found

Why is 404 Not Found
if "This feed ( is already
subscribed as ("?
I suppose this also means one can't read it via NNTP either.
Carlos Pita | 5 Sep 15:29 2014

Policy on gwene group renaming/deleting

Hi admins,

I would like to know which is the current policy regarding (well
intentioned) requests to remove/rename dated/wrong groups in gwene. It's
preferable for this policy to be well known, so that me and others won't
pointlessly post here wasting your time and diverting your attention.

Best regards
Carlos Pita | 4 Sep 15:07 2014

Please delete gwene group with wrong name

I mistakenly created with an url that is for
economics, not politics as the name implies. Could you please delete it?
Thank you very much, Carlos.
era | 29 Aug 11:26 2014

Yahoo Groups import tool recommendations?

There is an old thread from 2006 [1] which suggests yahoo2mbox but, like
most other tools I have looked at, they stopped working when Yahoo revamped
their groups and introduced the "Neo" interface.


The least abandoned one I could find was grabyahoogroup [2] which however
did not do anything useful for me -- the latest commit from August 2013
seemed promising, but when I tried it, it seemed to try to log in etc, but
didn't retrieve anything.

$ perl ./ --messages --username me --password xxx \
      --group reuterscorpora  --verbose
[INFO][ ][ ]Started: Fri Aug 29 12:08:22 2014
[INFO][reuterscorpora][ ]Redirected to:
[INFO][reuterscorpora][ ]Redirected to:
[FATAL][reuterscorpora][ ]This shouldn't happen - capabilities section
 missing ?
[INFO][ ][ ]Finished: Fri Aug 29 12:08:26 2014


I have done some manual scraping but the stuff I am able to download from
the web interface lacks some of the information from the email message
headers.  There is a separate headers view but reconstructing a valid
message from that and the scraped content seems painful and error-prone.

I realize that this is probably exactly how Yahoo likes for things to be,
(Continue reading)