I know of only one error situation in the TIP03 soft ware which could cause a major crash. If the input data overruns the buffer in disconnected mode, there is not a failsafe interlock on the buffer ring pointers. Usually this can only happen if one of the TNC's is connected to a computer, which can generate lotsa lines in no time. A typical scenario is that the computer is sending out a long file, and the transmit buffer is full, and then a problem on the link causes the boards to disconnect after 32 retries on the channel. The TIP goes into disconnected mode, the computer is given a Control-Q, and a wrap-around occurs which is not provided for in the code. You can eliminate this problem by having the TIP discard any lines of text it receives if it is not connected. I didn't see any easy way to to throttle back the host in disconnected state.
I've also seen VADCG's go crazy if they are not properly shielded. Very strange stuff can happen if RF gets into the board. Any VADCG that is not mounted in a metal box with RF filtering is suspect.
Hope that might help you resolve the problem. Keep up the good work.
73, Hank
Hank,
Depends on your definition of life, to be sure , no actually I guess we could blame it on the summer doldrums too. You should be getting a newsletter shortly and the next thing on our list is the annual Personal Computing Fair where we will be giving the usual demonstrations and talks.
Plans include getting on 20m with the Collins crystal controlled rig, putting a TNC on a local multiport BBS and an RTTY gateway. After that there will be the satellites to consider. We are still hoping to see a station node developed, as well as a broad band dedicated radio design.
Stuff is still selling steadily, although we must have over 6 months inventory of some parts at the current rate. We've sold nearly fifty of the TNC parts kits and another fifty plus of the modems. How many TNCs have you had made up from the artwork Don sent you? We still have about 150 left from our last order. Wouldn't mind moving them out before 2708s become completely passe.
Now that there is some agreement on the standardization of the protocol we should get the various TT versions rationalized so that bugs common to all have been removed and if there is any reason to have more than one that they all have a full set of features. Is any further work being done on the .04 TT software? VE7CVS's revision is supposed to go in the mail this week. Enough for now.