Tier 0

Notes on today's software test, attempting to connect to the SIL:

Because of network restrictions we did not have a connection between the mini and the PC, so no CoRD, no remote gui. For next time we asked for an additional IP number for the mini, and I'm hoping that would let us run the gui on the mini. For today we connected a monitor, keyboard and mouse to the PC.

We connected the dedicated cable to the SIL, and set the PC IP address as instructed (10.4.2.71, netmask 255.255.255.0, gate 10.4.2.254). There are some terminal handshake issues with CR-LF, but we were able to log in to the SIL from the terminal as expected:

• telnet 10.4.1.6 6555
• Trying 10.4.1.6
• Connected to 10.4.1.6
• Escape character is '^]'.
• 1 login user=exes_si role=instrument
• 1 A
• 1 :

This is pretty much the end of the success story :)

Using the latest version of the c++ code (in c:\exesV2), and the older version of the exes gui we were able to connect the gui to the MGT. We edited the .ini files to pint to the SIL, but clicking the 'connect' button in the gui did not work. We saw 'obsconnect' in the msg window but the button text in the gui didn't change from 'connect' in red to 'disconnect' in green as it should have. There was no feedback info in the msg window or in the part of the syslog file where the 'obsconnect' command shows up. The syslog file is long, and it may well be that a clue is buried in it somewhere but I have not yet found it.

We next tried to compile the latest version of Justin's java gui, but there is no java compiler installed on the PC and our attempt to install the Java SDK did not include one for some reason. We are now at Java 1.7, which may turn out to be a problem.

Having failed to get connected with Justin's gui, we tried to port our Java gui over to the PC. It refused to run, throwing exceptions about the notes file (even though we edited the init file to point to a valid writeable directory).

The PC is now back to its normal network connection so that Mike can CoRD in to look at the files and troubleshoot why we can't compile Justin's gui, nor run our own. This is really a bit of a detour, since what we really want to do is to run the gui on the mini. For next time we'll get the second IP, and hopefully be able to run another version of java gui on the PC a a backup.