Back in 2007 or so I wrote a C program to read my TP-200 serial port based GPS unit.  I was supprised to find that in 2012, the same program would compile on a Raspberry Pi Linux environment!  Go POSIX programming.

Attached is the C code file and my Makefile I use to compile it.  Use at your own risk.  It works for me, but maybe not you?  To use it, I just run  ./gps > gps_data_file-20121121.txt  and a number of lines will be written to the file each second.

In Part One of this project I developed a test circuit that proved that it would all work. In Part II, I actually redesign the bread boarded prototype into the final product.

After a quick scout around on Google, and reading a few man pages, I had a go at compiling the Netgear's WG311 Window files so that I could use it in FreeBSD 6.0.

I extracted the WG311v3.INF and WG311v3.sys files, moved them over to the FreeBSD card, then ran: 

     ndisgen WG311v3.INF WG311v3.sys 

After working through a simple wizard, and selecting the defaults, I ended up with a file called:   WG311v3_sys.ko

To get the wireless network card going, I just ran:

    kldload ndis

    kldload WG311v3_sys.ko

I knew FreeBSD 6.0 was going to be released while I was away, however, it turned out it was released the day I left  :-)

Just after I got back I downloaded Disk 1, and did a fresh install on to the 512MB CF card.  I then extracted the config files from the Subversion repository, and uploaded them.  A quick partition ghost, and I am now happily running 6.0

I am heading off on a conference soon, so I thought it was about time I had a go at Ghosting up the 512MB CF card. Using Barts Network boot disk, I made a ghost of the drive. This all went well, with no errors reported. As a safe guard, I also modified my 'config backup' script so it used scp instead of rsync. After backing up all the config files to my main server, I then uploaded them to the subversion repository for safe keeping.

After weeks of frustrating work trying to figure out the Northlight Systems Servo Controller and the base channel address, and the serial cabling issues (turned out there where none).  There where also issues with 'minicom' and how it stores some 0x00 values.