jte@psuvax1.psu.edu.UUCP (09/02/87)
I have having a little trouble with uugetty and dialout uucp. I have one 2400bps modem, and I wish to have both dialin and dialout via the same line. The only problem is when I setup both niether work and the SD and RD lights on my modem go steady and sytem preformance drops visably. I am running on a ATT 3B2/300 UNIX System V 3.0. Any help would be greatly appricated. --Jon # jte@psuvax1.psu.edu <- ARPA Phone: 237-1901 (Home) # # jte@psuvax1 <- BITNET Looks like Reaganomics # # <ihnp4,burdvax,purdue>!psuvax1!jte <- UUCP worked! #
randy@chinet.UUCP (09/02/87)
In article <2896@psuvax1.psu.edu> jte@psuvax1.psu.edu (Jon Eckhardt) writes: >I have having a little trouble with uugetty and dialout uucp. >I have one 2400bps modem, and I wish to have both dialin and >dialout via the same line. If you are using a Hayes compatible modem, it can't be done without some hardware hacks. >The only problem is when I setup >both niether work and the SD and RD lights on my modem go >steady and sytem preformance drops visably. You probably have the modem set up to echo commands or status info. The getty sees this as login info, sends prompts to the modem, which thinks it is commands, and on and on. Be sure to set the modem up completely dumb when using it as a login line. -- that's the biz, sweetheart..... Randy Suess ..!ihnp4!chinet!randy
garrett@udel.EDU (Joel Garrett) (09/02/87)
On a line similar to this, I have been trying to use a Courier 2400 as a dialout modem on a 7300 running v3.5 and have been having the same problem (the receive and send lights just stay on and I can't get any place). I used the Administrative Hardware setup menu to configure the rs232 port for 2400 baud and when I couldn't get it to use the HAYES option for an attached modem, I tried just telling the config menu that there was a computer attached to the port and the port was to be used for outgoing calls only. This config works fine when you connect the rs232 port to a 3b2/400 and set the baud rate to 9600. Is there anything that I am missing? The phone manager manual wasn't much help. Thanks, Joel Garrett garrett@udel.edu
tim@amdcad.AMD.COM (Tim Olson) (09/02/87)
In article <1463@chinet.UUCP> randy@chinet.UUCP (Randy Suess) writes: +----- |In article <2896@psuvax1.psu.edu> jte@psuvax1.psu.edu (Jon Eckhardt) writes: |>I have having a little trouble with uugetty and dialout uucp. |>I have one 2400bps modem, and I wish to have both dialin and |>dialout via the same line. | | If you are using a Hayes compatible modem, it can't be done without | some hardware hacks. +----- This is not true. What you need to do is substitute a new getty for that line which understands the lock files created by the dial-out programs (uucp, cu, tip...) and stops looking at the line until the lock file disappears. I have been using such a program with a Hayes modem with no problems. The program is a stripped down version of a more general-purpose getty posted to comp.unix.sources a few months back. -- Tim Olson Advanced Micro Devices (tim@amdcad.amd.com)
randy@chinet.UUCP (09/03/87)
|In article <18171@amdcad.AMD.COM> tim@amdcad.UUCP (Tim Olson) writes: ||In article <1463@chinet.UUCP> randy@chinet.UUCP (Randy Suess [me]) writes: || || || If you are using a Hayes compatible modem, it can't be done without || some hardware hacks. (in reference to modems and 3b2's) || | |This is not true. What you need to do is substitute a new getty for |that line which understands the lock files created by the dial-out |programs (uucp, cu, tip...) and stops looking at the line until the lock |file disappears. I have been using such a program with a Hayes modem |with no problems. The program is a stripped down version of a more |general-purpose getty posted to comp.unix.sources a few months back. | How do you handle the problems of the 3b2 port drivers requiring CD (pin 8) to be hi to dial out, and the needs of the 3b2 to see CD go hi and lo for proper login/logout/hangup? -- that's the biz, sweetheart..... Randy Suess ..!ihnp4!chinet!randy
eravin@dasys1.UUCP (Ed Ravin) (09/03/87)
I see people are trying to set up 2400 baud modems with uugetty and the like. Here's a quick summary of my experiences with a few Unix systems and a few modems: First things first: when you get your Hayes-clone modem from the factory, it is configured so a PC user will be able to use it with the least amount of pain. This means that the modem is set to echo input when in command mode (so the PC user can see what they're typing), to give you messages like OK or RING when the phone rings (so the PC user will know what is happening), and to ignore DTR as an enable command (in case the PC cable isn't wired properly). All three of these features MUST BE DISABLED if a Hayes modem is to work properly as a DIAL-IN modem. On the original Hayes SmartModem 1200, you must adjust the DIP switches behind the front panel. On the 2400 baud modems, you must turn off these parameters via commands over the serial port and then send the command to "burn in" the configuration settings (save them in the non-volatile RAM of the modem so that they will be the default power-up parameters). (Why must these features be disabled? You'll get the "getty-war" if you leave local echo on, as the modem echoes the getty message and the getty tries to log itself in. The "RING" message when the phone rings will also fool getty and this will interfere with the login procedure. Telling the modem to be enabled/disabled by DTR is optional, but hangup commands (stty 0) won't work without it, and the connection is much less reliable if the computer is unable to hang up and reset it.) The above covers setting up a modem for INCOMING CALLS only. If you want to set up a Hayes-beastie for OUTGOING CALLS ONLY, this is also possible. The only thing you really have to do is to turn off auto-answer in the modem. This is done by the DIP switch on the original Hayes 1200 modems, or by ATA0 and the burn-in command on the smarter, faster modems. Now you can set up your "chat script" or "acucap" or "modemcap" or whatever dial-out language your system uses. You might want to disable the "+++" feature that lets the modem be commanded out of on-line mode and into command mode, to prevent accidents. Other than that, there should be no problems. You should leave the status messages like "OK" and "NO CARRIER" and (especially) "CONNECT" ENABLED so that your chat script can tell if the modem is alive and if the phone call succeeded. If you've been patient enough to read this far, then you're probably also patient enough to be able to set up a Hayes-critter for INCOMING AND OUTGOING calls. First problem: support for bidirectional modems in many Unix systems is non-existent. You must either have a version of "uugetty" that understands the lock files used by UUCP and can create them when a person logs in on the modem, or a version of "uucp" and "cu" that is smart enough to disable the getty on a port that it wishes to dial out on. The Unix PC uses the latter method, and the 3b2/3b5 Unix comes with the former. For a Hayes-thingie modem, use the "uugetty -r -t 60" options. The "-r" tells uugetty not to consider the port open until a character is received. This is because a port is usually considered open when DCD is sensed. But for an outgoing modem, you must configure the modem so that DCD is always on, otherwise you wouldn't be able to connect to it to tell it to dial. So uugetty with the "-r" option will wait until someone types something (and it can't be the modem typing in the word "RING"!) before it will create the lock file, effectively locking out uucp or cu from using the port. Second problem: we can't dial out effectively if the status messages like "CONNECT" are turned off. We had to turn them off so that we wouldn't get fouled up by the "RING" message when the modem was used for incoming calls, but now we need them back. My solution was to use the "DTR reset causes modem to switch to power-up default" option of my Hayes 2400 clone. I programmed the modem to be an INCOMING modem, with status messages turned off. Then I created a chat script that would turn the status messages back on before the phone number was dialed when an outgoing call was made. If the call connected, the dialer would receive the "CONNECT" message from the modem. When the call was hung up, the modem would reset to its power-up defaults, which would turn the status messages back off again. --
tim@amdcad.AMD.COM (Tim Olson) (09/03/87)
In article <1469@chinet.UUCP> randy@chinet.UUCP (Randy Suess) writes: +----- | How do you handle the problems of the 3b2 port drivers | requiring CD (pin 8) to be hi to dial out, and the needs | of the 3b2 to see CD go hi and lo for proper login/logout/hangup? +----- You configure the port as a direct-connect port, not a modem-connected port. It should then ignore CD (if it doesn't, how do you connect terminals directly to the 3b2?). The revised getty program then runs immediately -- not just when the carrier is detected. This getty sits on the line, watching for characters coming in from the modem as well as looking for lock files. -- Tim Olson Advanced Micro Devices (tim@amdcad.amd.com)
ayac071@ut-ngp.UUCP (William T. Douglass) (09/03/87)
In article <18179@amdcad.AMD.COM> tim@amdcad.UUCP (Tim Olson) writes: >In article <1469@chinet.UUCP> randy@chinet.UUCP (Randy Suess) writes: >| How do you handle the problems of the 3b2 port drivers >| requiring CD (pin 8) to be hi to dial out, and the needs >| of the 3b2 to see CD go hi and lo for proper login/logout/hangup? >You configure the port as a direct-connect port, not a modem-connected >port. It should then ignore CD (if it doesn't, how do you connect >terminals directly to the 3b2?). I thought that ignoring CD on a modem line would cause the process to stay running if the user should hang-up without logging out proper. Is this not the case? Or does it vary from system to system? Bill Douglass ayac071@ngp.UUCP
randy@chinet.UUCP (Randy Suess) (09/04/87)
In article <18179@amdcad.AMD.COM> tim@amdcad.UUCP (Tim Olson) writes: >In article <1469@chinet.UUCP> randy@chinet.UUCP (Randy Suess) writes: >+----- >| How do you handle the problems of the 3b2 port drivers >| requiring CD (pin 8) to be hi to dial out, and the needs >| of the 3b2 to see CD go hi and lo for proper login/logout/hangup? >+----- > >You configure the port as a direct-connect port, not a modem-connected >port. It should then ignore CD (if it doesn't, how do you connect >terminals directly to the 3b2?). Ok, that's what I thought. That is completely unacceptable for modem dial-in use. What happens when a modem user loses their line for some reason and another person calls in on the same one? Because CD is not working, the shell is still active on that line, and the new user is right in the middle of the old users login. Makes things kinda rough when the original user was su'd to root and doing sysadmin thingies. I hope no one is running dialups on a port configured for direct connect.. -- that's the biz, sweetheart..... Randy Suess ..!ihnp4!chinet!randy
tim@amdcad.AMD.COM (Tim Olson) (09/04/87)
In article <6127@ut-ngp.UUCP> ayac071@ngp.UUCP (Bill Douglass) writes: +----- |I thought that ignoring CD on a modem line would cause the process to stay |running if the user should hang-up without logging out proper. Is this not |the case? Or does it vary from system to system? +----- Remember that we are talking about running our own "special" getty on this line. One of the last things that this getty does after it has validated a login and is about to exec the shell for the dial-in user is to perform an ioctl on the device, setting HUPCL and resetting CLOCAL. Therefore, if CD then drops, the login shell will terminate normally. -- Tim Olson Advanced Micro Devices (tim@amdcad. > mware!
jay@splut.UUCP (Jay Maynard) (09/04/87)
In article <1463@chinet.UUCP>, randy@chinet.UUCP (Randy Suess) writes: > In article <2896@psuvax1.psu.edu> jte@psuvax1.psu.edu (Jon Eckhardt) writes: > >I have having a little trouble with uugetty and dialout uucp. > >I have one 2400bps modem, and I wish to have both dialin and > >dialout via the same line. > > If you are using a Hayes compatible modem, it can't be done without > some hardware hacks. Not quite...I'm using a single Courier 2400 on my AT-clone for both dialin and dialout. The secret? Beats me. I put up Microport System V/AT, and it works. They did the serial driver right (one of a few things): they provided three minor devices for each hardware tty port. tty0 is intended for direct connections and dialout, ttym0 is intended for login-only use, and ttyM0 is like ttym0 except that an open on ttyM0 will wait if tty0 is open. If ttyM0 is open, then tty0 won't open. Hence, you can use ttyM0 for a getty port, and tty0 as the uucp device. The arrangement works great. -- Jay Maynard, K5ZC...>splut!< | uucp: hoptoad!academ!uhnix1!nuchat!splut!jay "Don't ask ME about Unix... | (or sun!housun!nuchat) CI$: 71036,1603 I speak SNA!" | internet: beats me GEnie: JAYMAYNARD The opinions herein are shared by neither of my cats, much less anyone else.
randy@chinet.UUCP (Randy Suess) (09/04/87)
In article <18180@amdcad.AMD.COM> tim@amdcad.UUCP (Tim Olson) writes: >In article <6127@ut-ngp.UUCP> ayac071@ngp.UUCP (Bill Douglass) writes: >+----- >|I thought that ignoring CD on a modem line would cause the process to stay >|running if the user should hang-up without logging out proper. Is this not > >Remember that we are talking about running our own "special" getty on >this line. >Therefore, if CD then drops, the login shell will terminate normally. Ah! Then you *do* have CD on your Hayes modem set to follow carrier. How, then, does uugetty dial out? It requires CD to be hi to open the port for dialing. -- that's the biz, sweetheart..... Randy Suess ..!ihnp4!chinet!randy
dave@westmark.UUCP (Dave Levenson) (09/05/87)
In article <1469@chinet.UUCP>, randy@chinet.UUCP (Randy Suess) writes: > How do you handle the problems of the 3b2 port drivers > requiring CD (pin 8) to be hi to dial out, and the needs > of the 3b2 to see CD go hi and lo for proper login/logout/hangup? This problem exists on other versions of UNIX, also. The best solution I have found is to use the Microcom AX/... series of modems. (Others may also have the feature I like). These modems have three options regarding the carrier-detect lead: 1) CD reports the true state of carrier. 2) CD always asserted. 3) CD mostly asserted, but lowered for 1 second on disconnect. Using option 3, your UNIX system can find CD true and dial out any time. It can answer incoming calls and accept logins (we use uugetty). If the connection is dropped, the modem lowers carrier-detect long enough for UNIX to see it, generate a SIGHUP, and clean up after the session that just ended. It then re-asserts CD and dial-in/dial-out operation is restored. -- Dave Levenson Westmark, Inc. A node for news. Warren, NJ USA {rutgers | clyde | mtune | ihnp4}!westmark!dave
randy@chinet.UUCP (Randy Suess) (09/05/87)
In article <1187@dasys1.UUCP> eravin@dasys1.UUCP (Ed Ravin) writes:
]I see people are trying to set up 2400 baud modems with uugetty and the
]like. Here's a quick summary of my experiences with a few Unix systems
]and a few modems:
]
.
.
]For a Hayes-thingie modem, use the "uugetty -r -t 60" options. The
]"-r" tells uugetty not to consider the port open until a character is
]received. This is because a port is usually considered open when DCD
]is sensed. But for an outgoing modem, you must configure the modem so
]that DCD is always on, otherwise you wouldn't be able to connect to it
]to tell it to dial.
I seem to be missing something with all this discussion of
bi-directional modem use on 3b's. As the above states, you have
to leave CD tied hi at all times for dial-out. What happens on
dial-in when the carrier is inadvertently dropped? The system
never sees the carrier go away, as CD is tied hi. So, another
person calling in is dropped into the middle of the previous
callers login. Could be real embarrassing. Has this been
discussed at all or have I just missed the answer?
--
that's the biz, sweetheart.....
Randy Suess
..!ihnp4!chinet!randy
tim@amdcad.AMD.COM (Tim Olson) (09/06/87)
In article <1478@chinet.UUCP> randy@chinet.UUCP (Randy Suess) writes: +----- | Ah! Then you *do* have CD on your Hayes modem set to follow | carrier. How, then, does uugetty dial out? It requires CD to | be hi to open the port for dialing. +----- Yes, the modem has switch 6 up, which asserts CD only when a carrier is present. However, while the getty is on the modem line, it is configured as "local" (so that the getty can send the modem initialization sequences) and any other dialer program (uucp, etc) can dial out. I may have forgotten to mention that I am *not* running this on an AT&T 3bxxx, it is on an IBM RT-PC, running AIX (SYSV). The uugetty code posted a while back seemed pretty general, though; I had no problems making it work on the RT. -- Tim Olson Advanced Micro Devices (tim@amdcad.amd.com)
randy@chinet.UUCP (Randy Suess) (09/06/87)
In article <152@westmark.UUCP> dave@westmark.UUCP (Dave Levenson) writes: ]In article <1469@chinet.UUCP>, randy@chinet.UUCP (Randy Suess) writes: ]> How do you handle the problems of the 3b2 port drivers ]> requiring CD (pin 8) to be hi to dial out, and the needs ]> of the 3b2 to see CD go hi and lo for proper login/logout/hangup? ] ]solution I have found is to use the Microcom AX/... series of ] 3) CD mostly asserted, but lowered for 1 second on disconnect. ] Aha! A modem after my own heart. I designed a box that sits between my Hayes and 3b2 that does basically the same thing. It consists of 2 one-shots in series between CD on the modem and the 3b2. Been working for two years. Thanks for the info. -- that's the biz, sweetheart..... Randy Suess ..!ihnp4!chinet!randy
randy@chinet.UUCP (Randy Suess) (09/06/87)
In article <124@splut.UUCP> jay@splut.UUCP (Jay Maynard) writes: ]In article <1463@chinet.UUCP>, randy@chinet.UUCP (Randy Suess) writes: ]> In article <2896@psuvax1.psu.edu> jte@psuvax1.psu.edu (Jon Eckhardt) writes: ]> ]> If you are using a Hayes compatible modem, it can't be done without ]> some hardware hacks. ] ]Not quite...I'm using a single Courier 2400 on my AT-clone for both dialin ]and dialout. ]The secret? Beats me. I put up Microport System V/AT, and it works. They did This is true. The original discussion was re: 3b1's and 2's. The Microport kernel has drivers that will open a port without CD being hi, but 3b?'s won't. -- that's the biz, sweetheart..... Randy Suess ..!ihnp4!chinet!randy
jhc@mtune.ATT.COM (Jonathan Clark) (09/06/87)
In article <1485@chinet.UUCP> randy@chinet.UUCP (Randy Suess) writes: > I seem to be missing something with all this discussion of > bi-directional modem use on 3b's. As the above states, you have > to leave CD tied hi at all times for dial-out. What happens on > dial-in when the carrier is inadvertently dropped? The system > never sees the carrier go away, as CD is tied hi. So, another > person calling in is dropped into the middle of the previous > callers login. Could be real embarrassing. Has this been > discussed at all or have I just missed the answer? I don't think that the real answer to this has come out yet, so here goes: The 'latest-and-greatest' HDB solves this entire problem by allowing one to specify an optional NDELAY open() and to toggle CLOCAL and part of the chat script on the communication line for dialout. One would then leave DCD alone, so that an incoming call would be answered correctly when DCD went high. On dial-out, one locks the line, does an NDELAY open(), sets CLOCAL (uugetty wakes up at this point, finds a lock file, says Oops! and goes to sleep), and does the dialling, turning CLOCAL off at some appropriate point, like when DCD has come up. I don't know who thought of this, but massive kudos to them. And no I don't know how anyone gets hold of this version, it's difficult enough for me! It's bound to be in a release sooner or later. -- Jonathan Clark [NAC,attmail]!mtune!jhc An Englishman never enjoys himself except for some noble purpose.
kai@ihlpa.ATT.COM (Irwin) (09/09/87)
In article <1478@chinet.UUCP>, randy@chinet.UUCP (Randy Suess) writes: > In article <18180@amdcad.AMD.COM> tim@amdcad.UUCP (Tim Olson) writes: > >In article <6127@ut-ngp.UUCP> ayac071@ngp.UUCP (Bill Douglass) writes: > >+----- > >|I thought that ignoring CD on a modem line would cause the process to stay > >|running if the user should hang-up without logging out proper. Is this not > >Therefore, if CD then drops, the login shell will terminate normally. > Ah! Then you *do* have CD on your Hayes modem set to follow > carrier. How, then, does uugetty dial out? It requires CD to > Randy Suess > ..!ihnp4!chinet!randy I always thought the REAL solution to this problem (the one I recommend to any one that asks) is to just buy Penril (non Hayes compat) modems or Penril dialers that work just fine. Also Rixons work well too, but Hayes are a pain in the A%! to deal with on a UNIX box! And bidirectional Datakit ports are just about as pleasent! Ken A. Irwin (former victim of the chroot, pay your $50 for the real thing CHINET) AT&T Bell Labs, Naperville IHP 1A332 (312) 416-4485
jfh@killer.UUCP (The Beach Bum) (09/09/87)
In article <1478@chinet.UUCP>, randy@chinet.UUCP (Randy Suess) writes: > In article <18180@amdcad.AMD.COM> tim@amdcad.UUCP (Tim Olson) writes: > >In article <6127@ut-ngp.UUCP> ayac071@ngp.UUCP (Bill Douglass) writes: > >+----- > >|I thought that ignoring CD on a modem line would cause the process to stay > >|running if the user should hang-up without logging out proper. Is this not > > > >Remember that we are talking about running our own "special" getty on > >this line. > >Therefore, if CD then drops, the login shell will terminate normally. > > Ah! Then you *do* have CD on your Hayes modem set to follow > carrier. How, then, does uugetty dial out? It requires CD to > be hi to open the port for dialing. I have been following this one since these guys started at each others throats, so now it's my turn. If this guys is so smart as to realize there is a problem with the modem having CD follow the carrier, can't he realize the solution? Only when a user is on the line must CLOCAL be turned off. After the user is gone, CLOCAL can be turned back on. The `getty' can do it, or the communications programs can handle it before the port is opened. Hopefully this won't evolve back into the earlier Catch-22 discussion. It can be done, I've done it and so have plenty of other bodies. John. -- John F. Haugh II HECI Exploration Co. Inc. UUCP: ...!ihnp4!killer!jfh 11910 Greenville Ave, Suite 600 "Don't Have an Oil Well?" Dallas, TX. 75243 " ... Then Buy One!" (214) 231-0993
randy@chinet.UUCP (Randy Suess) (09/10/87)
In article <1509@killer.UUCP> jfh@killer.UUCP (The Beach Bum) writes: > >If this guys is so smart as to realize there is a problem with the modem I don't recall anyone saying that they were "so smart". >having CD follow the carrier, can't he realize the solution? Only when a >user is on the line must CLOCAL be turned off. After the user is gone, "After the user is gone"? If CD is set always hi, (which it must be for a 3b2 to dial out on the line. This is, after all, what we were talking about. A binary only 3b2. All the stock comm programs on the 3b2, uucp, cu, etc. use the driver in such a way that CD must be hi to open the port. ) how does the system know the "user is gone"? >It can be done, I've done >it and so have plenty of other bodies. Please, tell me what I must do to my 3b2 for this to work. I'm just a dummy. -- that's the biz, sweetheart..... Randy Suess ..!ihnp4!chinet!randy
eravin@dasys1.UUCP (Ed Ravin) (09/11/87)
In article <1485@chinet.UUCP> randy@chinet.UUCP (Randy Suess) writes: (parts of my message about DCD sensing on modems is quoted) > > I seem to be missing something with all this discussion of > bi-directional modem use on 3b's. As the above states, you have > to leave CD tied hi at all times for dial-out. What happens on > dial-in when the carrier is inadvertently dropped? ( can someone else log into the session, will the system sense the hangup, etc) Oops. Come to think of it, I once wondered about that, but never had the problem. Lord knows I've seemingly had every other possible problem with modems and uugetty. By the way, this was all on a 3b5 using uugetty and BNU (HDB) UUCP. The reason I didn't discuss that problem is that I didn't have it: and the reason I don't think that problem occured to me is that the modems used here toggle DCD low whenever someone hangs up. I know that when I dial out and the other side dumps me, "cu" reports "Disconnected" and exits on its own. It must be sensing DCD, what else? And it never would have been able to dial out without DCD present. So if your modem is smart enough to toggle DCD low whenever the other side hangs up (loses carrier), the system will always be able to sense a hangup and act appropriate CoOrganit
randy@chinet.UUCP (Randy Suess) (09/11/87)
In article <1252@dasys1.UUCP> eravin@dasys1.UUCP (Ed Ravin) writes: > >Oops. Come to think of it, I once wondered about that, but never had the >problem. > >The reason I didn't discuss that problem is that I didn't have it: and the >reason I don't think that problem occured to me is that the modems used here >toggle DCD low whenever someone hangs up. Well, that's it then. You are fortunate to have *real* intelligent modems, whereas those of us with Hayes compatibles have just smart-ass modems. They don't have the sense to toggle CD when carrier goes away when it is set hi by switch or command. More than likely, that is the reason others have no problem with their systems on dial-in/out. I had to add an extra box to make the Hayes do the proper thing. Thanks for the clarification. -- that's the biz, sweetheart..... Randy Suess ..!ihnp4!chinet!randy
russell@imtec.co.uk (Russell Brown) (09/14/87)
In article <152@westmark.UUCP> dave@westmark.UUCP (Dave Levenson) writes: >In article <1469@chinet.UUCP>, randy@chinet.UUCP (Randy Suess) writes: >> How do you handle the problems of the 3b2 port drivers >> requiring CD (pin 8) to be hi to dial out, and the needs >> of the 3b2 to see CD go hi and lo for proper login/logout/hangup? > >This problem exists on other versions of UNIX, also. The best >solution I have found is to use the Microcom AX/... series of >modems. (Others may also have the feature I like). > 3) CD mostly asserted, but lowered for 1 second on disconnect. Steebek Modems (or Doughty) also have this sensible feature. They're sold under the Steebek name and badge engineered by everyone and his brother in the U.K. (the Quatro is quite popular, zillions of features, multiple speeds and the neat DCD trick). (no advertisment intended. Just a satisfied user :-) -- ------------------------------------------------------------------------ | Russell Brown | Voice: +44 733 234433 | | Imtec Computers Ltd. | UUCP: ...!mcvax!ukc!{inset,stc}! | | 18 Manasty Road, Orton Southgate | imtec!russell | | Peterborough, Cambs, PE2 0UP, UK | EMAIL: russell@imtec.co.uk | ------------------------------------------------------------------------