CompuServe Thread

#CIS Problems

11 messages in this thread
#103217From: Greg WhiteJan 12, 1988 11:49 PM
Don, I've been having problems with CIS and 2400 baud that may coincide with the new forum software that went online a while back. Maybe you can offer some help. I've been using 2400 baud since CIS equalized the cost with 1200 baud and have had no problems until recently. As will happen at 2400 baud, I occasionally get a short burst of noise characters, maybe 6 or 8. This has never caused any problems before. However, now what happens is that after these noise characters, the system (CIS?) apparently sends a ^P. On my screen I see something like "83/.';][P ^P". It is always a ^P, never anything else. This really causes a problem when I am doing a RTN on the message base as it throws me back to the function prompt and my reading of the threads is messed up. As I seldom get noise at 1200 baud, I don't have the problem there. I have tried 2 different term packages (Diga and Aterm6.1), differents Amigas, different modems, and different phone lines. Same problem. I can, however, call in thru a Telenet node and do not see the problem. But this costs more and is slower. I am presently dialing directly into a CIS node in Houston TX. I have talked with CIS people and they can offer no help. They said they checked the system out and found no problems. Can you (or anyone else reading this) offer any solutions? Thanks. Later, Greg
#103224From: Don Curtis/SYSOPJan 13, 1988 12:22 AM
Greg, I, and a few others are having exactly the same problems, althou I don't always get the ^P, that also happens. I've been bugging NetWork about it and they are supposedly checking things out here in Denver. One common point amongst the rest of us who are experiencing the problem is that we are on direct CIS 2400 baud lines and using CTS 2424ADH modems. 2 of us are here in Denver, and one other is in New York City so it's apparently some sort of problem with the new 2400 baud nodes (it's not the system software). For all of us, it started up about a month ago. We all have the problem ONLY on CIS direct lines, not on TynNet lines. Here in Denver, we have 2 2400 baud nodes, one is clean 99% of the time, the other one is dirty 99% of the time…unfortunately, the poor modem is the only one you can dial into…the other only becomes active when the first is full. I'm going to pass your message on to NetWork as another example of the problem. They'll eventualy figure out what's going on and fix it. Don
#103369From: Greg WhiteJan 13, 1988 9:51 PM
Don, It looks like we are all in the same boat as I too have a CTS2424ADH modem. I have tried a friend's modem (also a CTS) and, of course, his did the same thing. Is it possible we are seeing some sort of 'buffer full' handshake problem as I noticed that it only happens if I am dumping text at pretty much full speed? Doesn't matter if I am dumping the message base or one of the DLs. However if I just sit on a function prompt, I can sit there for 5 mins without seeing a single noise character. If I do a text dump I get the noise burst every 10-15 secs. I also can download a 20K file on the same line and not get a single error so it doesn't seem like your normal noisy line.C As you said, I don't always get the ^P either. I guess I get it about 50% of the time. Later, Greg
#103389From: Don Curtis/SYSOPJan 13, 1988 11:27 PM
Greg, Well…that's good news and bad news. Good news that the problem seems confined to a specific brand of modem…bad news that we own them! Actualy, the problem is probably not the modem itself…as you know, it works just fine with other services (TymNet, local BBS's et al) and here in Denver, if we are luck, and one of the nodes happens to fill up, we get switched to a secondary 2400 baud node and that node is perfectly clean. What it is looking more and more like is that somehow or other, the CIS modem starts to fall back to a lower speed for some reason or other and then picks back up to 2400 baud. It's not like I'm losing characters, but rather that some characters are getting sent as garbage…the same thing that I've seen happen at work if a terminal and computer port (on our UNIX machines) are at different baud rates. I've got a few ideas to test…I'll let you know if I find anything out. Don
#103521From: Greg WhiteJan 14, 1988 7:43 PM
Don, I called and talked with Tod at CTS this afternoon to discuss our our problem to be find out if he had heard anything about it. He was not aware of this specific problem but said he would call CIS and talk with them about it. He has been hearing about some strange problems concerning some new UDC (or UDS?) modems that CIS is testing. He also wonders if this could be connected with problems in the MNP protocol that these new modems apparently havY as an option. Although I can not link the change directly with this problem, CIS recently (the last month or so) switched to a new phone exchange (227 to 462). I don't remember a direct connection, but the time frame is certainly similiar. Can I assume that they also switched to new modems at the same time? Tod said he would call me back as soon as he finds out something. When he does, I'll pass it on. Later, Greg
#103266From: Steve AhlstromJan 13, 1988 10:51 AM
GreG, I'm experiencing exactly the same problems on one particular 2400 baud node (DEM). Out of curiousity, what type of modem are you using (please don't say CTS! The only people I know with this problem are using CTS modems).
#103370From: Greg WhiteJan 13, 1988 9:52 PM
Steve, It looks like we are all in the same boat as I too have a CTS2424ADH modem. I have tried a friend's modem (also a CTS) and, of course, his did the same thing. Is it possible we are seeing some sort of 'buffer full' handshake problem as I noticed that it only happens if I am dumping text at pretty much full speed? Doesn't matter if I am dumping the message base or one of the DLs. However if I just sit on a function prompt, I can sit there for 5 mins without seeing a single noise character. If I do a text dump I get the noise burst every 10-15 secs. I also can download a 200K file on the same line and not get a single error so it doesn't seem like your normal noisy line. Later, Greg Cs
#103394From: Steve AhlstromJan 13, 1988 11:37 PM
Damn! It looks more and more like a CTS modem problem. Your symptoms are identical to mine, Don, and Jim's. The weird thing is that it happens only on CIS nodes and ONLY at 2400. Guess it's time to start looking for a new modem.
#103439From: Ariel ButlerJan 14, 1988 9:31 AM
I missed the symptoms of this problem. If you describe them, I'll see if I see the same thing on the Supra 2400. Ariel
#103503From: Ted HaskellJan 14, 1988 4:10 PM
Steve, I don't think we should blame the CTS modem. I've been using one on CIS for many months now, and have not had a single problem at 2400 baud until CIS changed my 2400 baud access node here in Boston. One of the problems is the modem likes to hang up on CIS just as the carrier is coming on the line with the message "NO CARRIER". It seems to be caused by the way CIS answers the line. There appears to be a double line polarity reversal, which the modem detects as a line disconnect and therefore hangs up. I can circumvent this symptom by manually dialing the phone, waiting for the tone, and typing "ATD" which means go off-hook without dialing a phone number. Now the really annoying part is the last three weeks I have not been able to use 2400 baud for a download. It works fine for co or message reading. I fact I was using it for CTV for quite a while with no problem. But every time I try to download, I get to about the 70K point and the whole thing just dies. I'm using CIS-B protocol, by the way. I think the download problem is a little bit nastier because I have noticed that there are frequent network interruptions at 2400 baud, you see it as a momentary delay in transmission, but the "NETWORK INTERRUPTION" message has such a long timeout that you're not aware of the problems. I hope you can get CIS to do something, as I get error-free downloads at 1200 baud with the CTS, and I can use the CTS at 2400 with other services without any of these symptoms. Ted
#103523From: Greg WhiteJan 14, 1988 7:44 PM
Steve, see my message #103521 to Don. GREG