The point of this is to make it easier to maintain both branches.
DXProt (DXProtHandle has already been copied) route/* DXChannel
DXCommandmode and console.pl have been either copied wholesale,
where necessary, modified to use the old Msg based networking stack.
This is for Jim AD1C really. But "set/debug cluster" will put out
lines like:
CLUSTER: node N0VD-7 added
CLUSTER: user AD6KH added
and
CLUSTER: user AD6KH deleted
CLUSTER: node N0VD-7 deleted
and represent this node's view of which nodes and users are
available on the cluster as a whole.
Add the ability to limit the no of connections an incoming user/node
has. If a node/user is already connected elsewhere more than the maximum
no of times then this incoming connection is refused.
This is what Changes says:
Change the route finding algorithm completely. No more recursion. No more
tree searching. It now gives you answers even on a partial cluster map. Oh
and the answers are correct, instead on completely random.
Also completely remove RouteDB from the equation.
Also change sh/newc to default to node map rather than node+user map.
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.
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...
This is an attempt to cut down on the number PC92Cs that are flying
around. It does this be storing the last PC92C and only sending
it (which should be ignored by other people if broadcast) and
a PC92A. It does mean that the PC92C could be quite out of date
but a fresher will be along soon(ish)
I have also increase the pc92 update time to one hour and
randomised it to one hour - upto 1/4 of the update time.
It will also do each node separately, rather than all together in
one big gob.
2. strip the top off each character in any text field of a Spot for duping.
3. Try to make the routing tables more accurate for users that login to
more than one node.
4. Make PC50s come out in one heap on all channels every 14 mins, instead of
on the 14th minute in the connection time for each channel. This should
reduce (slightly) the dups that are dumped.
