Newcore status
dr
After a long, long, long time of seeming like nothing is happening, kismet-newcore
is actually beginning to be functional!

<pre>
dragorn@DRD1812 ~/src/kismet-newcore $ ./kismet_server -c pcapfile,foo.dump,foo 
Reading from config file /usr/local/etc/kismet.conf
Created Kismet UI TCP server on port 2501
Allowing clients to fetch WEP keys
No specific sources named, all sources defined in kismet.conf will be enabled.
Channel hopping enabled in config file
Channel splitting enabled in config file
Disabling channel hopping, no enabled sources are able to set channels.
Dropped privs to  (1000) gid 100
Source 0 (foo): Opening pcapfile source interface foo.dump...
Inserting basic packet dissectors...
Starting GPSD client...
ERROR: TCP client connect() failed: Connection refused
ERROR: Could not create initial connection to the GPSD server
ERROR: GPSD Reconnection not enabled, disabling GPS
Creating network tracker...
ERROR: Netracker failed to read SSID cache file version, cache file will be replaced
ERROR: Netracker failed to read IP cache file version, cache file will be replaced
Detected new network '' BSSID 00:0E:35:76:53:47
Detected new network '' BSSID 00:50:F2:76:E6:C4
Detected new network 'MSHOME' BSSID 00:50:F2:7B:D1:06
</pre>

Thats newcore bootstrapping, building the basic components, doing 802.11 frame
dissection, and basic network aggregation.  GPS works too, though not in that
example.  Commandline parsing and config file parsing are all modularized into
the components that handle them directly, as is the main select() loop
handling.  This means core files are smaller (kismet_server.cc is 494 lines of
code in newcore, vs. 3100 lines of code in stable), which means -compiling is 
quicker and takes less ram-.  

Of course things aren't done yet, major portions of functionality (logging,
alert processing, most packet sources) aren't ported to newcore, but having a 
running framework is key for getting the rest of the work finished.