readme.help

For installation instruction view readme file in system in which you are trying to install.  This file gives
nrlolsr options and featurs as well as current state of code updates.

Nrlolsr::options -i <interfacename>[-d <debuglvl>][-l <debuglogfile>][-nrlopt][-al on | off][-h][-v][-z]
             [-config <configfile>][-w <willingness>][-hna auto|<filename>|off][-b <broadaddr> <masklength>]
             [-hi <HelloInterval>][-hj <HelloJitter>][-ht <HelloTimeoutfactor>][-hp <HelloPadding>]
             [-tci <TCInterval>][-tcj <TCJitter>][-tct <TCTimeoutfactor>][-ipv6][-ipv4]
             [-hnai <HNAInterval>][-hnaj <HNAJitter>][-hnat <HNATimeoutfactor>]
             [-hys up <upvalue> | down <downvalue> | alpha <alphavalue> | on | off][-slowdown on | off]
             [-qos <qosvalue>][-fuzzy on | off][-shortesthop][-robustroute][-spf][-minmax][-mnmport <portnumber>]
             [-link <address> up | down | default | spf <weight> | minmax <weight> | promisc <weight>]
             [-unicast on | off] [-static <time> [-recordhellohistory on | off] [-fastreroute on | off]
             [-unicasthellos on | off][-port <number>]
             [-flooding off | s-mpr | ns-mpr | not-sym | simple | ecds | mpr-cds][-smfoffdelay <delay>]
             [-unicast on | off] [-static <time>][-fdelay <MaxForwardDelay>][-unicasthellos on | opt | off]
             [-rpipe <pipename>][-smfClient <pipename>][-sdtClient <pipename>][-guiClient <pipename>]   

OPTIONS
	-i      interfacename: is device name of interface that olsr is to run on. default = N/A 

	-d      debuglvl: is value between 0-10.  default = 0

	-l      debuglogfile: is file location/name that debugging information is logged. default is stderr

        -nrlopt Sets multiple predefined parmaters which tend to work better than the standard settings.  The settings are 
                agressive and tend to work better for high mobility/congestion than more static, lower traffic networks.
                The options this sets is 

                  -robustroute -fdelay .05 -tci 2.0 -tct 8 -tcj .99 -hys down .01 -hi .25 -hj .99 -ht 12
                
                These can be overwritten by given the same command after providing the -nrlopt command.

        -al     All links option.  This option is basically sets the TC_REDUNDANCY = 2 as described in v11 olsr default = off

	-h      Display usage.
	
	-v      Display version number.
	
        -z      Use Zebra to manipulate the routing table instead of system calls.

        -config 
                configfile: is the file name which contains command line options for nrlolsr. The commands in the file will be
                as if they are passed via the command line.

	-b      Set broadcast address network. <broadaddr> is an ip address string and <netmask> is the mask length.  
                default(IPv4) = <broadaddr>  = 224.0.0.57     (was selected at random from un-assigned local multicast address space)
                                <masklength> = 32
                default(IPv6) = <broadaddr>  = ff02::705  (selected at random from the link local address space)
                                <masklength> = 128

        -w      Willingness value between 0 and 7.  7 will always be mpr 0 will never be mpr.  default = 3

        -hna    Auto will look interfaces not defined by -i interfacename and advertize network routes.  
                filename: is the file with list of HNA network addresses to advertize in the format 
                        HNA <networkaddress> <masklength>
                        HNA <networkaddress> <masklength>
                off will turn sending HNAs out off
                default(IPv4) = auto
                default(IPv6) = off (IPv6 mode does not currently have auto option)	

        -hi     HelloInterval: is hello sending interval to be used in seconds. default = .5
                (spec v8 default = 2)

        -hj     HelloJitter: is percent of HelloInterval to jitter.  Values from 0 up to 1 are valid.  default = .5
                (jitter is done in different mannor than v8 spec.  Jitter value does not effect the 
                 average HelloInterval even over moderate intervals.)
	
        -ht     HelloTimeoutfactor: is value greater than 1 which multiplied with HelloInterval defines Neighbor Hold Time.  default = 6
                (spec v8 default = 3)
                (neighbor timeouts are done according to v4 spec Vtime is not yet implemented on recieving)
                (there is an issue in how two hop neighbors are stored which makes it hard to change runtime.  As of right now two hop neighbors
                 are held for 6 seconds since last update and this value is not changed with different intervals or timeouts.  To change this 
                 you must change the value of the global variable Neighb_Hold_Time to whatever you want the hold time to be then recompile. 
                 This is being  worked on.)

        -hp     HelloPadding: is a value between 0-1500.  All hello message sent with this option will be of at least size HelloPadding.  
                Padding of all zeros will be added to the end of all messages.  This is NOT normal OLSR operation and may cause other 
                implimentations to crash!  Only use this option when you are using all nrlolsr operating nodes.  
		
                This we added so that a "better" measure of link quality may be measured using larger packets which better reflects the non-routing
                traffic.

        -tci    TCInterval: is TC sending interval to be used in seconds. default = 2
                (spec v8 defalut = 5)

        -tcj    TCJitter: is percent of TCInterval to jitter.  Values form 0 up to 1 are valid. default = .5
                (jitter is done in different mannor than v8 spec.  Jitter value does not effect the 
                      average HelloInterval even over moderate intervals.)
	
        -tct    TCTimeoufactor: is value greater than 1 which multiplied with TCInterval defines topology tuple hold time. default = 3

        -ipv6   OLSR will operate with ipv6 addresses.  -hna auto option does not currently work with this option.

        -ipv4   OLSR will operate with ipv4 addresses.  This is the default mode.

        -hnai   HNAInterval: is HNA sending interval to be used in seconds. default = 15
                (spec v8 default = 5)
	
        -hnaj   HNAJitter: is percent of HNAInterval to jitter. Values from 0 up to 1 are valid. default = .1
                (jitter is done in different manner than v8 spec.  Jitter value does not effect the 
                     average HelloInterval even over moderate intervals.)

        -hys    up <upvalue>: where upvalue is number between 0 and 1 that defines HYST_THRESHOLD_HIGH. default = .4

                down <downvalue>: where downvalue is number between 0 and 1 that defines HYST_THRESHOLD_LOW. default = .15
                (only values for up and down make sense when up is larger than down bad things will happen if set incorrectly)

                alpha <alphavalue>:  where alphavalue is number between 0 and 1 that defines HYST_SCALING (weight of past). default = .7
	
                on: turns hysteresis on with default values.  Is on by default.
		
                off: turns hysteresis off by setting values of up down and alpha so that it does virutally nothing.
                (will have an effect if timeout values are VERY large)

        -qos    Qos value to mark OLSR packets with.  Is off by default.

        -fuzzy  on: will turn on fuzzy sighted flooding for tc and hna messages.  As of release 7.3 this is tested very little.

                off: will revert to normal full network floodin for tc and hna messages. This is the default mode of operation.

        -slowdown on: Will turn on slowdown option. TC intervals will be doubled after 3 successive sent tcs with the same addresses.
                      If the TC message to be sent would be different the tc interval is reset to the base -tci value and slowdown 
                      attempted again.  This option has not be thoughly tested esp in conjunction with the -fuzzy option.

                  off: Will revert to normal tc message sourcing. This is the default mode of operation.

        -fastreroute on: Will re calculate the routing table when changes which may effect routing are detected.  This can overburden the cpu
                         with large networks but most often will work well. This is the default mode of operation.
  
                     off: Routes are calculated every hello interval.  Only use this option if nrlolsr is using too much CPU as routing will
                          be degraded when using this option.

        -unicasthellos on: In addition to sending broadcast hello messages all neighbors currently in the one hop neighbor table will be
                           sent an identical unicast packet.  This may help maintain connectedness durring contention with some radios.
                           Use this feature with caution as it can quickly increase the amount of network traffic generated as network density
                           grows.  It will also not help delivery of TC messages.
                        
                        opt: Optionally will send unicast hellos to neighbors which do not have near perfect connectivity.  In stable networks
                             this mode will send out signficantly fewer packets than using the "on" mode.  In highly congested networks it will
                             not provide the same amout of robustness as on.

                        off: The default mode of operation only one message/packet will be sent out for each hello.
        -port <number>
                <number> is an integer which you wish to run olsr.  The default and rfc compliant port is 698.  

        -shortesthop
                Normal operation of route calculation using shortest hop paths. Is the default mode of operation.

        -robustroute
                Tries to take advantage of locally known information gathered from the hellos before using "older" TC information otherwise it 
                behaves in the same fashion as -shortesthop.  This option can be used interoperably with other OLSR implimentations.  It
                picks two hop routes based on the nbr table and uses the best connected one hop neighbor for the next hop.  This option cannot
                be used with -shortesthop -spf or -minmax. Default = off
        -spf 
                Shortest path first route calculation using spf information sent around in modified tc messages.  If
                using this option make sure all nodes in the network are operating with this option as routing loops and 
                other unwanted behavior will occur. This option is not interoperable with other OLSR implimentations. Default = off

        -minmax
                Route calculations will result in paths which have the maximum minimum link metric.  The metrics
                are sent around using modified tc messages. If using this option make sure all nodes in the network are operating
                with this option as routing loops and other unwanted behavior will occur. This option is not interoperable with other OLSR
                implimentations.  Default = off

        -unicast 
                Unicast routing option.  This option when set to off will turn off all unicast routing. Route tables will not be updated
                TC and HNA messages will not be sent, and route calcuations will not be performed.  SMF (-flooding) operations and HELLO 
                messages will continue to work in this mode.  default = on

        -recordhellohistory
                When set to on olsr will write to the debug log file the history of recieved hello messages. Default = off.


        -mnmport 
                Sets the port that nrlolsr will listen to for mnm specific mac layer packets.  All mnm mac layer packets
                use only the -link process commands functionality. Default = 4999.

        -link
                The link option sets the various values for a neighbor link.  It is always followed by the address of neighbor.
		
                up:     Will bring a link up.  Some problems may result if this function is called in simulation when the neighbor 
                        is not in range.

                down:   Will take a link down and cleaning remove the two hop links.

                default: Will return the link to normal operation.
		
                spf:    Sets the link shortest path first value <weight>.  This value will be sent around to all nodes using a modified
                        tc message format if the -spf option has been used. It is an 8 bit value with a valid range of 0-255.

                minmax: Sets the link maximum minimum value to <weight>.  This value will be sent around to all nodes using a modified
                        tc message format if the -minmax option has been used. It is an 8 bit value with a valid range of 0-255.

                promisc: This option currently doesn't do anything.  It is meant to provide a count of the number of packets heard on the
                         givin link.

        -flooding
                If you are interested in simplified multicast flooding methods for MANET in real systems check out nrlsmf (works with 
                nrlolsr). Broadcast packets which are forwarding on port 699 will be treated as multicast packets.  The type of simplified 
                multicast flooding method is defined by the next option. All options use duplicate packet detection to avoid broadcast 
                storms. 
		
                off:    No flooding/forwading will take place.

                s-mpr:  Source specific multi-point relay trees will do flooding of broadcast packets on port 699. (Regular mpr flooding)
	
                ns-mpr: Non-source specific multi-point relay trees will flood broadcast packets on port 699. (All mpr nodes flood)

                not-sym: Symetric one hop neighbors will not forward packets.  Like s-mpr option only asymetric neigbhors packes
                         will be forwarded.  Allowing for faster forwarding at the cost of potential extra overhead.

                ecds:   Uses shared tree to forward data and passes that information to smf.  Decision to forward is based upon
                        algorithm proposed for MANET OSPF by Richard Ogier.  Only the essential forwarding is implimented.

                mpr-cds: Uses shared tree to forward data and passes that information to smf.  Decision to forward is based upon
                         algorithm described in inria paper "On the robustness and stability of Connected Dominating Sets"

                simple: All nodes will forward broadcast packets on port 699. 		

        -smfoffdelay
               For use with -flooding options which use the on/off paradigm.  When this value is non-zero the algorithm will pospone turning
               off relay state for the supplied amount of time.  This may be helpful in maintaining connectedness when network dynamics are 
               occuring and network state has not yet converged. 
               
               delay: A double value of the amount of delay used before turning forwarding off once the flooding algorithm determins that it
                         should not longer be a relay. Default=0
        -unicast
                This option allows disabling of unicast routing.  It is intended for use in conjunction with nrlsmf (multicast 
                routing) to lower network overhead and cpu usage.

                on: This is the default mode of operation.  Communication with nrlsmf still occurs when this is turned on allowing for
                        multicast routing with unicast routing.

                off: This setting will disable the sourcing of both TC and HNA messages.  Unicast routing calculations will also not occur.
                        Hello message will be sent and processed as normal.  Communication with nrlsmf occurs based on the -flooding option 
                        as specified.  This option should only be used when using nrlolsr as a neighbor discovery mechinism for nrlsmf and 
                        unicast routing is not desired.
        -static
                This command can be used to put nrlolsr to sleep and to wake it back up.  When sleeping olsr will close sockets and turn all
                timers off.  No route table will be modified while asleep.  No packets will be sent or processed (pcap socket remains open.)
                
                time: is the amount of time before nrlolsr will be put to sleep.  For example "-static 20" would allow olsr to run for 20 
                      seconds before going to sleep.  This allows for static routes to be established.  Valid values are from -1 and 0 to 
                      DOUBLE_MAX. -1 and 0 are special cases.  
                      "-static -1" will wake up nrlolsr and clear out any active static timer.  When woken up olsr may timeout most routes.
                      "static 0" will put nrlolsr to sleep, unless called from the command line.
                This command may be most useful when used in conjunction with a protopipe and an outside controler.
                
        -fdelay
                This feature allows for random delay in forward to help mitigate the hidden terminal problem.  Instead of having OLSR messages 
                being directly sent they are held and argrated in one OLSR packet for a jittered amount of time.  The wait time for a message, 
                given no currently held messages, is set to a random number between 0 and MaxForwardDelay.  If a message is currently being 
                held for sending the new message is appened into the same OLSR packet and sent along with the other held messages.  If the
                packet size would become too large (1500 bytes) by adding a new message, the currently held packet is sent off and a new packet 
                is produced using the new message and held for a newly generated hold time.
		
                MaxForwardDelay: is the maximum amount of time a messages is to be held before forwarding.  This time is in seconds and must 
                                 be positive.

The below commands are only availabe in non-simulated builds.  Protopipes are not supported in simulation.

	-rpipe
		This command will change which protopipe nrlolsr will listen to for further external control.  The sample protopipe example
		included in the protolib library can be used to connect and send commands to the named pipe.  The msg2msg example also included
		in the protolib examples can be used to forward messages to/from pipes and ipaddresses allowing for external/remote control.

		pipename: is the name of the protopipe. Note that in some emulated environments these should be unique for each 
				instance if control is desired.  
				Default = nrlolsr 
	-smfClient
		This command will change the name of the protopipe to send nrlsmf specific control messages.  Appropriate listening pipe settings
		must be configured on the smf side for the connection to be successful.  If a non-default recieve pipe name is used this command
		should follow any -rpipe command so the handshake will occur correctly.
		
		pipename: is the name of the protopipe.  Note that in some emulated enviornments these must be uniquely configured both
				here and on the smf instance side (smf has its own command line option for setting its controller's pipe)
				Default = nrlsmf
	-guiClient
		This command will change the name of the protopipe to send to olsrgui.  Once connected it will update the gui with its current
		recieve pipe name so the gui can connect and send control messages.  If a non-default recieve pipe name is used this command
		should follow any -rpipe command so the handshake will occur correctly.
	
		pipename: is the name of the protopipe.  
				Default = nrlolsrgui
	-sdtClient
		This command will change the name of the protopipe to send sdt specific commands.  SDT does not listen to a protopipe by default
		but the msg2msg example program can be used to forward the messages to a port/file which SDT can then consume.

		pipename: is the name of the protopipe.
				Default = sdt

Notes:  This piece of code started in ns2 at version 4 of the olsr specifications.  It has evolved and is evolving to try and keep up.
        All underlying timers are periodic and jitter is added to this periodic ticking.  This keep Intervals at numbered values and 
        allows maximum jitter. Neighbor storage is complicated and even I don't understand half of it anymore.  It works and that 
        is all I care about right now.  certain reserved fields were used in the past to do load balancing work and the code is still in 
        there for parts of it.  MIDs as defined in version 8 and 9 of the olsr spec have not be implimented.  Though some code has been 
        added so that it may be done in the future.
	
        There is a function updateSM() which writes to shared memory at location /tmp/mprsel.  Other processes can look at this memory and
        see who has selected this node for forwarding in the olsr code at the current time.

        Because of the introduction of Vtime (validity time) in Olsr packets different nodes in the same network can function together with
        differing tc hna and hello intervals.  The hysterises function does not work with Vtime as of this release.  So if nodes are going 
        to have different hello intervals make sure to use the "-hys off" option.

        I have noticed that a good amount of cpu time is used by the code.  I will  be addressing this issue in the future.  If the cpu is
        being overworked (use top) I would recomend using a slower hello and tc timer. (v8 spec 2 and 5)  Because of this speed issue I 
        have commented out the DMSGs.  Feel free to uncomment them as you wish to get a feel for what is going on.

  Release 4.8 (ipv4 only)
        Future  work:
                I will be adding ipv6 support.  (done)
                Fix Vtime Htime and LinkCodes in packets. (Vtime is done, Htime sent but not used, LinkCodes sent correct but not finished) 
                Develop code to support windows platforms and ns2. (up next)
                Clean code?!? I don't think so! Though I will try to comment better where needed.
                I would like to link my nrlolsr directory to the protolib code instead of putting it in the directory.(done)
	
  Release 5.0 (v6 support!)
        Future work:
                Speed up the code.  Mainly by replacing the linked list with something faster.
                Get hysterises functioning with the Vtime and Htime correctly.
                Add code to automaticly find hna's when ipv6 mode is turned on.
                Maybe add MID packet support for use with more than one wireless interface.

  Release 5.1 
                small bug fixes
  Release 5.2 
                Important bug fixes.  Fixed bug in which asymetric links would sometimes not time out correctly.  Found and fixed bug in which 
                default routes added by HNAs would not update correctly when node moves away (linux only problem).  
                Added command line quality of service marking option.  Packets can now be marked with a given qos value.  How that value is used
                is the users choice.  Also included is Makefile.arm-linux.  You will need a cross compiler to make a binary.  I have only confirmed
                that it works under open zarus.  It does not work under normal zarus os.

  Release 5.3   
                More Important bug fixes.  Fixed bug in ttl and hop count fields in both hna and tc messages.  They now decrement and count up.
                Fixed problem with changing timer intervals.  Old code does not change intervals via the command line correctly.
                New feature added.  Debug lvl 2 now outputs hna information as well as link state information.  This information will be used 
                by the upcomming cmap visulization tool.  Debug lvl 1 will only output the links though both std/cmap and jmap should be able 
                to deal with the extra info.

  Release 5.4  
                Changed way hna information is outputed so that all information is outputed not just what is being used.  
  Release 6.0 
                Windows support!  Moved code over to newer protolibkit library.  Will allow for easier platform portability.  Cleaned up the 
                code a bit.  Removed older make routing table functions that were not being used. Also added a few more comments.  Found and 
                fixed bug in hystersis method which would cause unstable links in some cases.  Messages are no longer sent using broadcast 
                but a multicast address in ipv4 mode.  Broadcast support may or may not work using the command line option.  

  Release 6.1 
                Fixed minor typo bug which would cause 6.0 not start correctly when attempting to use the -ipv6 option.

  Release 7.0beta
                Code now works with both ns-2 and WinCE.  Inststallation instructions are located in the respective directories.  Testing
                needs to be performed to find bugs which may have been introduced while porting to new operating systems.

  Release 7.1 
	        Code changes to protolib updated nrlolsr to work with those updates.

  Release 7.2 
                Added support to opnet 10.  Fixed routing bugs in core code which would incorrectly calculate routes in some cases.  Fixed
                bug in unix ipv4 mode which shared memory code would overwrite variables not belong to it.  Fixed bug in selecting mprs, old
                code could ignore some one hop neighbors in certain situations.  Various protolib updates, major one being IPv6 fix for unix.
  Release 7.3 
                Fixed a bug which would sometimes allow for empty tc message to be sent out and in the process, in rare occations, cause
                crashing.  Added new options for setting link metrics and routing using that information.  Both shortest path first and
                maximum minimum path routing is now possible.  Other minor bugs fixed.  Mnm project specific mac packet parsing added to
                change the link metrics on the fly added.  Also added fuzzy sighted flooding as an option.  When turned on it will flood
                tc and hna messages only a certain distance into the network depending on the time that the hna/tc is sent.
  Release 7.3.1 
                Fixed ProtoManetKernel so that nrlolsr now works with ns-2.27.

  Release 7.4  
                Fixed bugs with fowardwarding section of code.  One had to do with a bracket being too high up and the other had to do with
                the ttl factor.  Also fixed tc mssn wrap around code which was old and had a typo.  Switched how fuzzy sited option works.
                Add another -slowdown option which when turned on will attempt to slow down the tc message if the neighborhood is logically
                static.  WARNING both -slowdown and -fuzzy options are not currently specified anywhere and have been tested very little 
                use at your own risk.

  Release 7.5
                Fixed a few minor bugs.  One would cause a crash when hna routes were removed, in linux, with debuging 3 or higher turned on.  
                Added protoPipe support.  You can now use protoPipe (non-simulation only) to send command line options to olsr while olsr is 
                running.  Check out the protoPipeExample in the protolib directory for an example of how another program can talk to NRLOLSR.  
                The NRLOLSR server pipe is named "nrlolsr_server"  There will be an option to change this name durring startup as well.  
                Also added is extra debuging/network information output, for linux, of neighbor information send to stdout.  This output will
                only be printed with debug level 1 or higher.  I have done some rough compile and run testing to clean the code back up for 
                release but there may be a 7.5.1 version comming out soon if people catch something I missed.
 Release 7.6
                Methods added to support Simplified Multicast Flooding for both real world and ns2 simulations.  Added interface support 
                for nrlsmf, NRL's version for Simple Multicast Forwarding in MANETs, through piping.  The NRLOLSR default server pipe name 
                has changed from "nrlolsr_server" to "nrlolsr".  For ns protoManetKernel and nrlolsrAgent have been changed quite a bit 
                to support smf behavior in the ns environment. 
 Release 7.7 
                New gui!  The gui works under linux, windows, and windowsce. Nrlolsr now interfaces with nrloslrgui.  Methods and a new 
                pipe were added to interface with the gui. Because of the gui interface some minor modifications were done with the command 
                line options (they can now be changed durring operation). Modification to duplicate detection method to allow for correct 
                performance.  This was a previously undiscovered bug.

 Release 7.7.1
                Added -hp which allows the user to set a minimum hello size.  This should only be used with an all nrlolsr network as it may
                cause other OLSR implimenations to break.  

 Release 7.7.2 
                Added -unicast option which allows user to turn off unicast routing.  This option when set to off will turn off all tc and hna messages.
                Routes will not be calcuated and will not be added to unicast routing tables.  Hello message will be sent and smf -flooding options will 
                continue to work if used.
 Release 7.8beta
                Fixed some bugs dealing with asymetric links in small networks which would break routing in certain cases.  Fixed bugs with protopipes
                in windows which allows for olsrgui to now work.  Fixed bug in which interval times would not change correctly via gui commands. Added 
                fastreroute option to allow for less cpu usage with standard operation in large networks.  Many updates to opnet build.
 Release 7.8.0
                Fixed bug introduced between 7.7.2 and 7.8beta which cause neighbor link state to become unstable.  A delayed fowarding feature was added
                which enables agragation of multiple messages into one packet and can may help mitigate hidden terminal problems which arise when
                flooding TC messages.  A new command line option to disable unicast routing was also intoduced to allow deployment of multicast only
                enabled networks without unneeded overhead of a unicast routing protocol.  Changed the reserved field in the olsr packet header to 
                a packet seqno to reflect the latest draft.
Release 7.8.1
                Fixed ns2 bugs related to duplicate tree dependancies on changes of protoTree.
Release 7.8.2
                Added a the -unicasthellos option to allow better neighborhood stability durring contention.
Release 7.8.3 
                Modified robustroute and added documentation on its usage.  Also tweeked use unicast hello option to include the opt option which will
                only send unicast hellos to neighbors with less than .99 connectivy values (only send to less than near perfect links). Added -nrlopt
                which sets multiple preset options at once.
Release 7.8.4
                Added the static command which will shut down olsr after a given amount of time.  It will keep discovered routes up and continue to
                listen on protopipes for more commands.
Release 7.8.5
                Added a port option to allow operation on ports other than 698.  Also added the port reuse to allow for operation on certain 
                emultion systems.
Release 7.8.6
                Added protopipe support for sdt3d.  Documented pipe options for external control of nrlolsr using protopipes and methods for
                optional control of other NRL software gui/smf/sdt

Release 7.9.0
                Added initial support for route manipulation though the use of Zebra. -z option added and functional on linux distributions.
Release 7.9.1
                Added a config file option to allow file based configuration of nrlolsr.
Release 7.9.2
                Updated the base route algorithm to include the hello links.  Also headers are only included for a given link type if the link type exists. (credit MITLL)