reading another group

hey

i have my config setted with groups pcre to read about 3 groups
and really in Pan when i fetch all groups from local server I see those 
3 groups
and they are getting updated every fetchnews run.

Now i want to add another group , added it to the pcre line in the 
config file
and touched and empty file in /var/spool/news/interesting.groups/
(owner setted to news (as the other files))

but fetchnews doesnt see it, groupinfo file still has no record for the 
new group.
In pan I tried reading new groups but no new groups appeared.

How can I add a group now for fetchnews to retrive ?

Thanks,
Roi

Inconsistent behavious between -P and no flags?

I post a summary of newsgroup activity to a couple of groups, this is done
by a script that puts a file in /var/spool/news.out.going with a name that
reflects the group, I've just noticed that if I run fetchnews with -P it
doesn't get posted but fetchnews does post it.

A bit of digging reveals that checkqueue expects the postable files to
contain only digits and the '-' character. I assume that without a -P
checkqueue isn't tested and any file in out.going gets processed.

But the usage states that

Setting none of the options
    -B -H -P -R
is equivalent to setting all of them, unless [-M message-id] is used.

Maybe these should be consistent and then I might have to tidy up my script!

Is there a use for a script like this being bundled in with leafnode? - I
inherited it via a chain of people who had done things with it but then
weren't involved any more with the particular newsgroup.
See
news:12231.10.8969.91.fsf@...
for an
example post.

Robert
--

-- 
Links and things http://rmstar.blogspot.com/
Robert Marshall

Patch: leafnode says active file has to be refetched but it doesn't

In some circumstances, when fetchnews can't connect to a server, fetchnews 
decides that the active file must be refetched even though it wouldn't have 
tried to fetch the active file if it had connected.  Specifically if it's a -P 
run or if the server is marked noread.
I've attached a patch for this.  I've also created a patch to make the error a 
little clearer.  Since both patches modify the same part of the program, the 
second patch will only apply after the first one has been applied.

--- fetchnews.c.orig	2005-11-08 17:54:11.000000000 -0500
+++ fetchnews.c	2005-11-08 17:56:50.000000000 -0500
 <at>  <at>  -2500,7 +2500,8  <at>  <at>  main(int argc, char **argv)
 	} else { /* reply = nntpconnect */
 	    if (verbose)
 		printf("%s: connection failed.\n", current_server->name);
-	    if (forceactive && current_server->updateactive) {
+	    if (forceactive && current_server->updateactive && !postonly &&
+		!current_server->noread) {
 		error_refetch("fetching the active list from a server failed.");
 		need_refetch = 1;
 	    }

--- fetchnews.c	2005-11-08 17:56:50.000000000 -0500
+++ fetchnews.c.new	2005-11-08 17:55:48.000000000 -0500
 <at>  <at>  -2502,7 +2502,7  <at>  <at>  main(int argc, char **argv)
 		printf("%s: connection failed.\n", current_server->name);
 	    if (forceactive && current_server->updateactive && !postonly &&
 		!current_server->noread) {

Patch: leafnode says active file has to be refetched but it doesn't

In some circumstances, when fetchnews can't connect to a server, fetchnews 
decides that the active file must be refetched even though it wouldn't have 
tried to fetch the active file if it had connected.  Specifically if it's a -P 
run or if the server is marked noread.
I've attached a patch for this.  I've also created a patch to make the error a 
little clearer.  Since both patches modify the same part of the program, the 
second patch will only apply after the first one has been applied.

--- fetchnews.c.orig	2005-11-08 17:54:11.000000000 -0500
+++ fetchnews.c	2005-11-08 17:56:50.000000000 -0500
 <at>  <at>  -2500,7 +2500,8  <at>  <at>  main(int argc, char **argv)
 	} else { /* reply = nntpconnect */
 	    if (verbose)
 		printf("%s: connection failed.\n", current_server->name);
-	    if (forceactive && current_server->updateactive) {
+	    if (forceactive && current_server->updateactive && !postonly &&
+		!current_server->noread) {
 		error_refetch("fetching the active list from a server failed.");
 		need_refetch = 1;
 	    }

--- fetchnews.c	2005-11-08 17:56:50.000000000 -0500
+++ fetchnews.c.new	2005-11-08 17:55:48.000000000 -0500
 <at>  <at>  -2502,7 +2502,7  <at>  <at>  main(int argc, char **argv)
 		printf("%s: connection failed.\n", current_server->name);
 	    if (forceactive && current_server->updateactive && !postonly &&
 		!current_server->noread) {

"last seen ... server now has ..." is the server's fault, right?

X-DCL-garcia R default

When I get a syslog message like this (it usually recurs for a few runs of 
fetchmail then stops):

Nov  1 22:31:26 garcia fetchnews[12579]: alt.religion.kibology: last seen 
article was 61692, server now has 61415-61690

does this indicate a fault on the remote server (from which fetchnews is 
downloading)?

Thanks,
Adam

"last seen ... server now has ..." is the server's fault, right?

X-DCL-garcia R default

When I get a syslog message like this (it usually recurs for a few hours 
then stops):

Nov  1 22:31:26 garcia fetchnews[12579]: alt.religion.kibology: last seen 
article was 61692, server now has 61415-61690

I assume it indicates an error on the server from which I'm downloading -- 
is that correct?

Thanks,
Adam

Patch: leafnode says active file has to be refetched but it doesn't

(I tried to send this to the list on Tuesday and yesterday, and neither of 
them appear to have gone through -- Is the list down?)

In some circumstances, when fetchnews can't connect to a server, fetchnews 
decides that the active file must be refetched even though it wouldn't have 
tried to fetch the active file if it had connected.  Specifically if it's a -P 
run or if the server is marked noread.
I've attached a patch for this.  I've also created a patch to make the error a 
little clearer.  Since both patches modify the same part of the program, the 
second patch will only apply after the first one has been applied.

--- fetchnews.c.orig	2005-11-08 17:54:11.000000000 -0500
+++ fetchnews.c	2005-11-08 17:56:50.000000000 -0500
 <at>  <at>  -2500,7 +2500,8  <at>  <at>  main(int argc, char **argv)
 	} else { /* reply = nntpconnect */
 	    if (verbose)
 		printf("%s: connection failed.\n", current_server->name);
-	    if (forceactive && current_server->updateactive) {
+	    if (forceactive && current_server->updateactive && !postonly &&
+		!current_server->noread) {
 		error_refetch("fetching the active list from a server failed.");
 		need_refetch = 1;
 	    }

--- fetchnews.c	2005-11-08 17:56:50.000000000 -0500
+++ fetchnews.c.new	2005-11-08 17:55:48.000000000 -0500
 <at>  <at>  -2502,7 +2502,7  <at>  <at>  main(int argc, char **argv)

leafnode2 - parse problem with newnews/newgroups

Hi,

I have been running leafnode-2 for a couple of months now, mainly
because I needed the possibility of adding local groups.

I have had to patch it a bit, mainly because newnews and newgroups
directives not are similar enough, as assumed by the donew_common
routine.

In addition there some minor cosmetic changes for unused variables etc.

I have attached the patch ...

Keep up the good work!

/Niels

--

-- 
Niels Baggesen --  <at> home -- Århus -- Denmark -- niels@...
The purpose of computing is insight, not numbers  --  R W Hamming

diff -ru ../leafnode-2.0.0.alpha20050914a/fetchnews.c ./fetchnews.c
--- ../leafnode-2.0.0.alpha20050914a/fetchnews.c	2005-08-10 12:15:49.000000000 +0200
+++ ./fetchnews.c	2005-10-27 22:29:45.099743651 +0200
 <at>  <at>  -2042,8 +2042,10  <at>  <at> 
     struct sigaction sa;
     static int forceactive;	/* if 1, reread complete active file */
     struct serverlist *current_server;
+#ifdef COMPILE_BROKEN

Re: Patch: leafnode says active file has to be refetched but it doesn't

Brian Sammon schrieb am 2005-11-10:

> (I tried to send this to the list on Tuesday and yesterday, and neither of 
> them appear to have gone through -- Is the list down?)

It was, it's back up, and it's now hourly checked by cron if it's still
up, so that root gets a mail if the mailman qrunner isn't up.

> In some circumstances, when fetchnews can't connect to a server, fetchnews 
> decides that the active file must be refetched even though it wouldn't have 
> tried to fetch the active file if it had connected.  Specifically if it's a -P 
> run or if the server is marked noread.
> I've attached a patch for this.  I've also created a patch to make the error a 
> little clearer.  Since both patches modify the same part of the program, the 
> second patch will only apply after the first one has been applied.

You're right - and that was a good catch. Thank you, I have merged your
patch for leafnode 1.11.4.

Kind regards,

--

-- 
Matthias Andree

Re: Inconsistent behavious between -P and no flags?

Robert Marshall <robert@...> writes:

> I post a summary of newsgroup activity to a couple of groups, this is done
> by a script that puts a file in /var/spool/news.out.going with a name that
> reflects the group, I've just noticed that if I run fetchnews with -P it
> doesn't get posted but fetchnews does post it.

> A bit of digging reveals that checkqueue expects the postable files to
> contain only digits and the '-' character. I assume that without a -P
> checkqueue isn't tested and any file in out.going gets processed.

Your observation is correct, -P believes there were nothing to post
unless there is at least one file that matches [0-9-]+

Given that leafnode-2, as an alpha version, makes no warranties about
its internal spool layout, and users aren't supposed to hack the spool,
I think...

> Setting none of the options
>     -B -H -P -R
> is equivalent to setting all of them, unless [-M message-id] is used.
>
> Maybe these should be consistent and then I might have to tidy up my
> script!

...consistency would be nice, but what for? The inconsistency is
actually an optimization. (Yes I know about premature optimization being
the root of many evils.)

If you want consistency, how about using a small script (Perl, Python,