spider.git
16 years agoadded cty1708 changes
Dirk Koopman [Wed, 24 Oct 2007 12:47:01 +0000 (13:47 +0100)]
added cty1708 changes

16 years agoadd cty1707 prefix changes
Dirk Koopman [Tue, 23 Oct 2007 22:13:19 +0000 (23:13 +0100)]
add cty1707 prefix changes

Also pickup version and build from PC92 K records

16 years agouse DXUtil::month for sh/contest
Dirk Koopman [Mon, 15 Oct 2007 11:29:02 +0000 (12:29 +0100)]
use DXUtil::month for sh/contest

16 years agofix regex in sh/contest for perl < 5.8.x
Dirk Koopman [Mon, 15 Oct 2007 11:25:57 +0000 (12:25 +0100)]
fix regex in sh/contest for perl < 5.8.x

16 years agoprevent PC61 going to non-spider clusters
Dirk Koopman [Sun, 14 Oct 2007 22:34:35 +0000 (23:34 +0100)]
prevent PC61 going to non-spider clusters

16 years agocosmetic change to mrtg.pl
Dirk Koopman [Wed, 10 Oct 2007 20:59:40 +0000 (21:59 +0100)]
cosmetic change to mrtg.pl

16 years agoadded PC92 stats to mrtg
Dirk Koopman [Wed, 10 Oct 2007 15:11:35 +0000 (16:11 +0100)]
added PC92 stats to mrtg

16 years agoput the version no back in 1st slot for now
Dirk Koopman [Wed, 10 Oct 2007 13:38:21 +0000 (14:38 +0100)]
put the version no back in 1st slot for now

16 years agopush non-pc9x and non-spider nodes back for routing
Dirk Koopman [Wed, 10 Oct 2007 10:45:59 +0000 (11:45 +0100)]
push non-pc9x and non-spider nodes back for routing

At the moment it appears that we can rely on non-spider nodes routing
stuff correctly for mail and rcmd. So we push those routes to the back
of the list.

16 years agopush non-pc9x nodes to the back for routing
Dirk Koopman [Wed, 10 Oct 2007 10:37:24 +0000 (11:37 +0100)]
push non-pc9x nodes to the back for routing

Currently, if a directly connected non-pc9x node advertises a route
to another node, even if that node

16 years agosend build in pc92 K
Dirk Koopman [Wed, 10 Oct 2007 07:48:43 +0000 (08:48 +0100)]
send build in pc92 K

16 years agoremove pc92 A/D first slot if == $mycall
Dirk Koopman [Tue, 9 Oct 2007 16:21:59 +0000 (17:21 +0100)]
remove pc92 A/D first slot if == $mycall

also increase the C periodic update time to 4 hours

16 years agofix new pc9x midnight rollover bug
Dirk Koopman [Tue, 9 Oct 2007 14:27:00 +0000 (15:27 +0100)]
fix new pc9x midnight rollover bug

16 years agoadded radio specific "motd_ax25"
Dirk Koopman [Sat, 6 Oct 2007 08:05:52 +0000 (09:05 +0100)]
added radio specific "motd_ax25"

16 years agotreat PC61 as PC11.
Dirk Koopman [Fri, 5 Oct 2007 17:49:35 +0000 (18:49 +0100)]
treat PC61 as PC11.

16 years agofix .gtkconsole_data creation bug
Dirk Koopman [Mon, 1 Oct 2007 17:40:19 +0000 (18:40 +0100)]
fix .gtkconsole_data creation bug

16 years agorecast the callsign gathering
Dirk Koopman [Thu, 27 Sep 2007 17:44:11 +0000 (18:44 +0100)]
recast the callsign gathering

16 years agomore README for gtkconsole
Dirk Koopman [Thu, 27 Sep 2007 17:14:04 +0000 (18:14 +0100)]
more README for gtkconsole

16 years agoadded new version of gtkconsole
Dirk Koopman [Thu, 27 Sep 2007 17:05:46 +0000 (18:05 +0100)]
added new version of gtkconsole

and admitted its existance.

16 years agostart the recasting of gtkconsole
Dirk Koopman [Thu, 27 Sep 2007 16:24:55 +0000 (17:24 +0100)]
start the recasting of gtkconsole

This involves splitting up the one screen into separate ones
Adding a chat screen
Making them only popup when input arrives.

16 years agochange back to guage (but divided by 5)
Dirk Koopman [Tue, 25 Sep 2007 20:06:37 +0000 (21:06 +0100)]
change back to guage (but divided by 5)

16 years agoadd perminute modifier
Dirk Koopman [Tue, 25 Sep 2007 19:54:43 +0000 (20:54 +0100)]
add perminute modifier

16 years agofix mrtg to not use /var/* tree in debian/ubuntu
Dirk Koopman [Tue, 25 Sep 2007 19:07:35 +0000 (20:07 +0100)]
fix mrtg to not use /var/* tree in debian/ubuntu

16 years agochange the type of unit for spots
Dirk Koopman [Tue, 25 Sep 2007 18:11:57 +0000 (19:11 +0100)]
change the type of unit for spots

This is an attempt to see if we now get the correct no of spots / minute

16 years agoreset the F*****g obscount on C and K records!
Dirk Koopman [Mon, 24 Sep 2007 22:16:03 +0000 (23:16 +0100)]
reset the F*****g obscount on C and K records!

16 years agoadd changes to Changes
Dirk Koopman [Mon, 24 Sep 2007 21:12:14 +0000 (22:12 +0100)]
add changes to Changes

16 years agoTimeout PC92 nodes at the correct rate
Dirk Koopman [Mon, 24 Sep 2007 21:10:43 +0000 (22:10 +0100)]
Timeout PC92 nodes at the correct rate

16 years agomake ping reset the node obscount
Dirk Koopman [Mon, 24 Sep 2007 20:35:07 +0000 (21:35 +0100)]
make ping reset the node obscount

also add ping debugging on obscount tag

16 years agofix K crash
Dirk Koopman [Sat, 22 Sep 2007 19:21:06 +0000 (20:21 +0100)]
fix K crash

16 years agofix mrtg script to use LANG=C
Dirk Koopman [Sat, 22 Sep 2007 15:13:07 +0000 (17:13 +0200)]
fix mrtg script to use LANG=C

16 years agoK records only come from $mycall
Dirk Koopman [Wed, 19 Sep 2007 14:54:11 +0000 (15:54 +0100)]
K records only come from $mycall

The hourly C records for external nodes will be fine for
keepalive purposes.
Remove all references to Investigate.

16 years agoissue K records and start to use them
Dirk Koopman [Tue, 18 Sep 2007 19:28:17 +0000 (20:28 +0100)]
issue K records and start to use them

Also halve the frequency of main::me node C records. This
will eventually go to once a day.

16 years agonew keepalive system + timings for PC92 C recs
Dirk Koopman [Mon, 17 Sep 2007 21:44:15 +0000 (22:44 +0100)]
new keepalive system + timings for PC92 C recs

Added the new K record for keepalives.
Increased the PC92 C interval to 24 hours.
Increased dependent node C interval to 2 hours.
PC50 can reset obs.
PC50 now routed everywhere without filters except isolation,
this may change back in time, when K record is ubiquitous.

Still need to actually issue K record

16 years agoonly the first slot in PC92C should have version nos
Dirk Koopman [Sun, 16 Sep 2007 21:06:41 +0000 (22:06 +0100)]
only the first slot in PC92C should have version nos

16 years agobigger random element in PC92 ext. node holdoff calcs
Dirk Koopman [Sun, 16 Sep 2007 18:36:10 +0000 (19:36 +0100)]
bigger random element in PC92 ext. node holdoff calcs

16 years agoreduce the number of duplicate PC92 for Ext Nodes
Dirk Koopman [Sun, 16 Sep 2007 17:01:53 +0000 (18:01 +0100)]
reduce the number of duplicate PC92 for Ext Nodes

Try to reduce the no of duplicate PC92C for external PC protocol
nodes by listening for others that are out there and bumping up our
update time instead of us transmitting more or less the same info
as well.

16 years agodo some more rigorous checking of PC9x sentences
Dirk Koopman [Sat, 15 Sep 2007 17:03:55 +0000 (18:03 +0100)]
do some more rigorous checking of PC9x sentences

16 years agoadd optional INET6 capability. Need to load CPAN modules
Dirk Koopman [Sat, 15 Sep 2007 16:33:26 +0000 (17:33 +0100)]
add optional INET6 capability. Need to load CPAN modules

You will need IO::Socket::INET6 to use INET6 capability.
Also I have enabled PC92 slugging and set it to 1 minute as default.
Also I have put code in to handle empty first node slots in PC92.

16 years agoreduce the (default) length of announces to dupe
Dirk Koopman [Wed, 12 Sep 2007 21:05:45 +0000 (22:05 +0100)]
reduce the (default) length of announces to dupe

16 years agoallow PC92(A/D) to be slugged and aggregated
Dirk Koopman [Wed, 12 Sep 2007 19:43:28 +0000 (20:43 +0100)]
allow PC92(A/D) to be slugged and aggregated

Wait for a period of time before issuing pc92a or d records
and then aggregate them together as a collection of adds or deletes
in just (up to) two sentences, instead of one per event.

This has the side effect of (nearly) eliminating the "login, fire a
spot, logoff" merchants.

16 years agouse just one lastid for all pc9x sentences
Dirk Koopman [Mon, 10 Sep 2007 20:10:52 +0000 (21:10 +0100)]
use just one lastid for all pc9x sentences

This should work fine and will cure the dupe/lost talk problem

16 years agotidy blocking for windows (possibly)
Dirk Koopman [Mon, 10 Sep 2007 19:48:12 +0000 (20:48 +0100)]
tidy blocking for windows (possibly)

16 years agoadd cty-1706 and make wpxloc.raw changes
Dirk Koopman [Mon, 20 Aug 2007 15:12:26 +0000 (16:12 +0100)]
add cty-1706 and make wpxloc.raw changes

16 years agoadded CTY1705 and extended 60m band
Dirk Koopman [Tue, 31 Jul 2007 19:49:55 +0000 (20:49 +0100)]
added CTY1705 and extended 60m band

16 years agodisable use of DBI for sh/dx as default, if active.
Dirk Koopman [Sun, 29 Jul 2007 19:03:40 +0000 (20:03 +0100)]
disable use of DBI for sh/dx as default, if active.

16 years agocorrect git URL
Dirk Koopman [Fri, 27 Jul 2007 11:07:02 +0000 (12:07 +0100)]
correct git URL

16 years agoAdded Angel's changes to Messages for Spanish
Dirk Koopman [Fri, 27 Jul 2007 11:01:40 +0000 (12:01 +0100)]
Added Angel's changes to Messages for Spanish

16 years agoforce update of talk command
Dirk Koopman [Wed, 25 Jul 2007 17:18:21 +0000 (18:18 +0100)]
force update of talk command

Doing this because my update of 1.29 confuses cvsserver.

16 years agorestore talk "<call> not visible on cluster" message
Dirk Koopman [Wed, 25 Jul 2007 13:06:07 +0000 (14:06 +0100)]
restore talk "<call> not visible on cluster" message

Restore the message if the <call> is not routable. However, instead
of the old behaviour of then just stopping dead, it now sends the
message anyway as a broadcast.

16 years agoshutdown quicker by not manipulating routing table
Dirk Koopman [Wed, 25 Jul 2007 11:45:44 +0000 (12:45 +0100)]
shutdown quicker by not manipulating routing table

16 years agoTry to improve PC93 discrimination between talk and chat
Dirk Koopman [Tue, 24 Jul 2007 15:38:08 +0000 (16:38 +0100)]
Try to improve PC93 discrimination between talk and chat

16 years agoprevent empty callsigns on talk
Dirk Koopman [Mon, 23 Jul 2007 20:13:16 +0000 (21:13 +0100)]
prevent empty callsigns on talk

16 years agofix and simplify dup announces
Dirk Koopman [Mon, 23 Jul 2007 18:16:45 +0000 (19:16 +0100)]
fix and simplify dup announces

16 years agofix (and simplify) the spot dup checking code.
Dirk Koopman [Mon, 23 Jul 2007 18:12:53 +0000 (19:12 +0100)]
fix (and simplify) the spot dup checking code.

16 years agoprevent hop counts of -1 being generated.
Dirk Koopman [Fri, 20 Jul 2007 14:41:53 +0000 (15:41 +0100)]
prevent hop counts of -1 being generated.

16 years agoprotect send_local_config from processing undef nodes
Dirk Koopman [Fri, 20 Jul 2007 08:27:45 +0000 (09:27 +0100)]
protect send_local_config from processing undef nodes

Don't understand how this is possible but if it happens much more
I am going to have to do some serious digging...

16 years agocater better for 0 pingint in links cmd
Dirk Koopman [Thu, 19 Jul 2007 11:33:29 +0000 (12:33 +0100)]
cater better for 0 pingint in links cmd

16 years agoMake set/isolate work in BOTH directions
Dirk Koopman [Wed, 18 Jul 2007 16:04:46 +0000 (17:04 +0100)]
Make set/isolate work in BOTH directions

Since the beginning I have had set/isolate be unidirection, so that I could
see the far node's config, but he could not see mine. This is too confusing
for yer modern sysop and I have now made it such that only the local configs
are remembered, regardless of what is sent back to me. I.e. it is now
symmetric and bidirectional.

16 years agoRemove dupefile before restarting node.
Dirk Koopman [Tue, 17 Jul 2007 21:53:22 +0000 (22:53 +0100)]
Remove dupefile before restarting node.

Also provide the clear/dupefile command to do it at runtime.

16 years agofix another possible(?) error on broadcast p92 update
Dirk Koopman [Thu, 12 Jul 2007 20:53:28 +0000 (21:53 +0100)]
fix another possible(?) error on broadcast p92 update

16 years agoAdd help for sh/band and improve sh/band command
Dirk Koopman [Thu, 12 Jul 2007 16:20:49 +0000 (17:20 +0100)]
Add help for sh/band and improve sh/band command

Allow sh/bands to have arguments so people can query
a band or region.

16 years agochange node disconnection code to improve node clearouts
Dirk Koopman [Thu, 12 Jul 2007 14:54:04 +0000 (15:54 +0100)]
change node disconnection code to improve node clearouts

16 years agoallow store nodes that have a dxchannel
Dirk Koopman [Wed, 11 Jul 2007 18:46:39 +0000 (19:46 +0100)]
allow store nodes that have a dxchannel

16 years agoremove $Id$ strings from everywhere that I can find
Dirk Koopman [Wed, 11 Jul 2007 15:41:51 +0000 (16:41 +0100)]
remove $Id$ strings from everywhere that I can find

16 years agoimprove speed of sh/c/n (a bit)[probably].
Dirk Koopman [Wed, 11 Jul 2007 15:15:52 +0000 (16:15 +0100)]
improve speed of sh/c/n (a bit)[probably].

16 years agoadd #9000 as a default group and tidy up send_chats calls
Dirk Koopman [Tue, 10 Jul 2007 16:27:11 +0000 (17:27 +0100)]
add #9000 as a default group and tidy up send_chats calls

16 years agoremove rspf, add pc10 eph_dup, tidy up eph_dups
Dirk Koopman [Mon, 9 Jul 2007 21:34:16 +0000 (22:34 +0100)]
remove rspf, add pc10 eph_dup, tidy up eph_dups

16 years agoAllow PC10s to be converted to PC93 and be processed
Dirk Koopman [Mon, 9 Jul 2007 16:45:30 +0000 (17:45 +0100)]
Allow PC10s to be converted to PC93 and be processed

The PC10 code converts all input to PC93 where all talk
processing is done. It was being called from the wrong
channel.

16 years agoimprove the fix slightly to avoid false positives
Dirk Koopman [Sun, 8 Jul 2007 15:17:37 +0000 (16:17 +0100)]
improve the fix slightly to avoid false positives

16 years agoput in temp fix for pc10 talk crashes
Dirk Koopman [Sun, 8 Jul 2007 15:08:34 +0000 (16:08 +0100)]
put in temp fix for pc10 talk crashes

16 years agotry to make sure that local nodes take precedence over external ones
Dirk Koopman [Sat, 7 Jul 2007 10:34:28 +0000 (11:34 +0100)]
try to make sure that local nodes take precedence over external ones

There seems to be a problem with localnodes sometimes being overridden by
data from external PC92 data. Usually (but not always) data for local
PC19 nodes being overridden by data from other PC92 with the same node
connected.

Also put in a frig for the sysop announce problem.

16 years agofix pc93 -> pc12 conversions
Dirk Koopman [Fri, 6 Jul 2007 20:18:13 +0000 (21:18 +0100)]
fix pc93 -> pc12 conversions

This should stop pc93 conversions which are sent out a) being redisplayed
and b) as SYSOP. It may also stop all these strange chat messages to
unlikely chat group names.

Also rewritten chat import routines for new regime.

16 years agohandle old style spider version strings
Dirk Koopman [Fri, 6 Jul 2007 16:47:39 +0000 (17:47 +0100)]
handle old style spider version strings

16 years agotart up version no reporting in sh/node
Dirk Koopman [Fri, 6 Jul 2007 16:43:21 +0000 (17:43 +0100)]
tart up version no reporting in sh/node

16 years agoFix the over midnight ID rollover calculation
Dirk Koopman [Fri, 6 Jul 2007 09:01:18 +0000 (10:01 +0100)]
Fix the over midnight ID rollover calculation

Also use a big enough past age limit to cope with inactive nodes
that only output a config record and nowt much else, so that they
don't time out.

Add an 'id too far in the future' compared with current systime
check.

16 years agomake sure the node C record is actually sent
Dirk Koopman [Thu, 5 Jul 2007 21:49:47 +0000 (22:49 +0100)]
make sure the node C record is actually sent

16 years agofix unwanted dupe notifications on looped PC12
Dirk Koopman [Tue, 3 Jul 2007 22:16:08 +0000 (23:16 +0100)]
fix unwanted dupe notifications on looped PC12

16 years agofix typo in links
Dirk Koopman [Mon, 2 Jul 2007 17:16:59 +0000 (18:16 +0100)]
fix typo in links

16 years agoadd iso and filter info to links command.
Dirk Koopman [Mon, 2 Jul 2007 17:05:46 +0000 (18:05 +0100)]
add iso and filter info to links command.

16 years agoadd the possibility to search for a string in sh/cmd
Dirk Koopman [Mon, 2 Jul 2007 14:30:35 +0000 (15:30 +0100)]
add the possibility to search for a string in sh/cmd

16 years agoadd sh/cmd_cache
Dirk Koopman [Mon, 2 Jul 2007 14:14:29 +0000 (15:14 +0100)]
add sh/cmd_cache

16 years agotry to make set/isolate more bombproof.
Dirk Koopman [Fri, 29 Jun 2007 13:29:13 +0000 (14:29 +0100)]
try to make set/isolate more bombproof.

Add a set/wantpc9x (default = on) command. It is likely
that (apart from mistakes) you will want to unset it mostly.
If either this flag == 0 or the node is isolated then pc9x
will not be offered on the PC18 and PC9x's will be ignored.

16 years agofix typo in /ex in chat mode
Dirk Koopman [Mon, 25 Jun 2007 21:19:45 +0000 (22:19 +0100)]
fix typo in /ex in chat mode

16 years agoadd comment about chat to Changes
Dirk Koopman [Mon, 25 Jun 2007 21:16:07 +0000 (22:16 +0100)]
add comment about chat to Changes

16 years agoadd continuous chat mode (like talk) contchat
Dirk Koopman [Mon, 25 Jun 2007 20:19:19 +0000 (21:19 +0100)]
add continuous chat mode (like talk)

16 years agofix dates on Changes file
Dirk Koopman [Mon, 25 Jun 2007 18:51:23 +0000 (19:51 +0100)]
fix dates on Changes file

16 years agofix longstanding possibility of a crash on new user.
Dirk Koopman [Mon, 25 Jun 2007 18:32:29 +0000 (19:32 +0100)]
fix longstanding possibility of a crash on new user.

16 years agofix announces
Dirk Koopman [Mon, 25 Jun 2007 17:04:59 +0000 (18:04 +0100)]
fix announces

This was caused by double duplication, in turn because I
have moved to having everything done by the pc93 handler.

16 years agoreally REALLY fix period C records
Dirk Koopman [Mon, 25 Jun 2007 14:17:19 +0000 (15:17 +0100)]
really REALLY fix period C records

16 years agoreally make sure a C record is send and stored!
Dirk Koopman [Mon, 25 Jun 2007 13:45:26 +0000 (14:45 +0100)]
really make sure a C record is send and stored!

16 years agomake sure that periodic C updates come from node
Dirk Koopman [Mon, 25 Jun 2007 10:27:35 +0000 (11:27 +0100)]
make sure that periodic C updates come from node

16 years agoextend 4m in bands.pl to 70631
Dirk Koopman [Sun, 24 Jun 2007 20:45:17 +0000 (21:45 +0100)]
extend 4m in bands.pl to 70631

16 years agoconnect up the discontinuous trees and route better
Dirk Koopman [Sun, 24 Jun 2007 20:15:23 +0000 (21:15 +0100)]
connect up the discontinuous trees and route better

This involves making sure that all the external nodes get fixed
up even though their parent node does not yet exist (because no
config record has come in. It is safe to create these, in
anticipation of the C record coming in somewhen.

In the process of this, obscount reseting now happens much more
frequently (which I hope means correctly).

Allow Route::findroutes to find routes to these disconnected
nodes that are in the routing table but have no parent that is
yet anchored to a dxchan. Simply remember the dxchan that that
C record came in on. It doesn't matter if it gets overwritten by
another node, if it came in that way, there must be a route.

In any case, chances are, the first dxchan to get the C record
is the best and the deduping will (mostly) prevent it being
overwritten. When the linking C record comes in, it is all
irrelevant anyway...

16 years agoimprove debugging of obscounting
Dirk Koopman [Sun, 24 Jun 2007 09:44:48 +0000 (10:44 +0100)]
improve debugging of obscounting

16 years agoremove all $Id$ strings
Dirk Koopman [Sun, 24 Jun 2007 01:17:43 +0000 (02:17 +0100)]
remove all $Id$ strings

16 years agotarget eph_dup better
Dirk Koopman [Sun, 24 Jun 2007 00:47:37 +0000 (01:47 +0100)]
target eph_dup better

16 years agoremove redundant PC10 handling after PC93 conversion
Dirk Koopman [Sun, 24 Jun 2007 00:24:16 +0000 (01:24 +0100)]
remove redundant PC10 handling after PC93 conversion

16 years agomake Route::findroutes the basis of all routing
Dirk Koopman [Sun, 24 Jun 2007 00:03:03 +0000 (01:03 +0100)]
make Route::findroutes the basis of all routing

16 years agoexplicitly ignore just hoptable.pl
Dirk Koopman [Sat, 23 Jun 2007 23:46:21 +0000 (00:46 +0100)]
explicitly ignore just hoptable.pl

16 years agoadd *.pl (to catch hoptable.pl et al)
Dirk Koopman [Sat, 23 Jun 2007 23:37:02 +0000 (00:37 +0100)]
add *.pl (to catch hoptable.pl et al)