OT (again): MODEM to MODEM?

[ ... ]

Fine when everything is close by. However, he is talking about an hour's drive. 10BaseT (Cat-5) cable has a limit of just a few hundred feet -- with *good* quality wire. Thinnet goes a bit farther, Thicknet is better than that. But *none* of those will handle distances represented by a 1-hour drive -- unless he has to drive around a lake, while the wires can just hop a back fence. :-)

Enjoy, DoN.

Ah! I missed that part. PPP is about the way to go, or PCAnyWhere.

Gunner

"At the core of liberalism is the spoiled child - miserable, as all spoiled children are, unsatisfied, demanding, ill-disciplined, despotic and useless. Liberalism is a philosphy of sniveling brats." -- P.J. O'Rourke

In article , asdf wrote: :Robert Nichols wrote: :> In article , :> Mike Bartman wrote: :> : :> :A null-modem lets you connect two of the same kinds of equipment :> :together because it crosses the transmit and recieve wires. This :> :connects one piece of equipment's transmit pin to the other's recieve :> :pin, and vice versa. All the other wires, such as the chassis ground, :> :signal ground, ring, etc. are straight through connections. :> :> That's not true. There are several status signals that must be cross :> connected if you want anything resembling the common, old communications :> software (Kermit, Wincom, Minicom, Zmodem, etc.) to work: :> :> Outputs: Data Terminal Ready, Request to Send :> Inputs: Data Set Ready, Clear to Send, Ring Indicator :> :> If you just wire those straight through, you'll have outputs fighting :> other outputs, and pairs of floating inputs connected. No matter what :> you do, you cannot exactly simulate what the software will see when :> talking to an actual modem. The hardware just won't allow it. :> :> On the original question about connecting two modems together -- :> Most modems today use an electronic relay to connect to the :> telephone line. You'll need to supply a DC bias current to allow :> those components to operate properly. Without that bias current :> your connection will be intermittent at best. Been there, done :> that, got tee shirt. :> : :Three wires are all you need for software flow control :or no flow control, the only reason for the others are :for hardware flow conrol. Quite often you can get away :without flow control on todays fast computers.

True if you configure the software to ignore the status signals, but it certainly won't be a good simulation of an actual modem-to-modem connection, which is what the OP wanted. In any case, the statement, "All the other wires, such as the chassis ground, signal ground, ring, etc. are straight through connections" is a recipe for trouble.

Since I've actually done it, more than once, I know it works. There may well be situations in which it won't work, but using PC hardware with MS-DOS as the OS, on *slow* processors (10 Mhz 286 for instance), and running Kermit, or couple of other terminal emulators/file transfer programs (including one "serial network" link program that worked a bit like NFS for MS-DOS as far as a user was concerned), I had no problems at any speed up to 115K.

You are right that this won't be a simulation of a modem-to-modem link, but as was pointed out, there isn't really any way to do that simply other than to use modems and a couple of phone connections. If you could rig up a true "modem simulator" you wouldn't need to be asking here about it. ;-)

These days I don't even own a modem for my desktops, and my newest MB only came with one serial connector (9 pin). Same MB refers to 3.5" floppies as "legacy" devices, and doesn't support two of them (so much for my old 5.25" drive... ;-). Ethernet LAN is much faster and more flexible for local connections, and the internet extends it as needed, so I haven't tried to hook two machines together with serial cables in the last decade or so.

-- Mike B.

---------------------------------------------------------------- To reply via e-mail, remove the 'foolie.' from the address. I'm getting sick of all the SPAM...

----------------------------------------------------------------

On 7 Sep 2004 01:48:21 -0400, snipped-for-privacy@d-and-d.com (DoN. Nichols) vaguely proposed a theory ......and in reply I say!:

remove ns from my header address to reply via email

AND I have to lay the cable

***************************************************** I know I am wrong about just about everything. So I am not going to listen when I am told I am wrong about the things I know I am right about.

Thanks for the terminology correction. It's been a long time since I had to care about this stuff...haven't made up a cable since the late

80s, and haven't programmed at the UART level since before that. The memory is the second thing to go...I forget what the first is. ;-)

-- Mike B.

-- Mike B.

'04 FLSTCI

---------------------------------------------------------------- To reply via e-mail, remove the 'foolie.' from the address. I'm getting sick of all the SPAM...

----------------------------------------------------------------

A pair of SDSL modems will get you 15,000 to 20,000 feet at a couple megabits over conventional phone cabling. They're pretty cheap on ebay as well. Apparently its a turnkey deal, the modems find each other and autonegiotiate- the modems connect by ethernet to the users network.

Gregm

You can't just connect those to the phone line. They need a special circuit that the phone company has to set up.

Apparently they work just fine with nothing connecting them except plain old wire. You might be thinking of ADSL modems- different beasts. SDSL modems particularly support a peer to peer connection.

Gregm

On Mon, 06 Sep 2004 21:46:32 +0800, Old Nick vaguely proposed a theory ......and in reply I say!:

remove ns from my header address to reply via email

OK. Radmin, bless their hard little hearts, will not support me as far as Win networking goes. I think this is very shortsighted, and I may well use VNC, because at least there I will not _pay_ for a lack of any help.....but anyway.

I _did_ this before. I must have worked out how! But I can't now.

I went to my old man's place, and managed to get the two modems talking direct. My Dad sat and made helpful suggestions but was largely out of the picture.

I'm caught. I had the MODEMs talking great quite often, but could not network.

Using DIN. My machine would connect, but then say "Verifying user name and password" then say "the computer you are connected to does not support networking" and hang up.

What's missing? Where do I set up a username and password _receipt_ system on the Old Man's machine?

Stage 1.

If I use my little dialler/answerer programme, the DIN Server won't start (something is using ghe modem) and if the DIN Server is running, my programme won't. So I can't use my little proggie to make my Dad's modem answer, if I want the server runnning to allow me to connect.

I am pretty sure what I did before was to give him a DIN connection that said ats0=2. The modem would dial '1" then fail, but get the ats0=2, so I could ring and his modem would answer.

Stage 2

Of course that doesn't work over the direct connection. So I used ata on my old man's machine instead. Same setup. Modem fails, but then tries to answer.

I tried using the homebuilt dialler/answer on my machine, sending a call signal while Dad's was "answering". This would connect the modems, but when I tried to use RADMIN to connect it would immediately fail.

So I quickly "send" using DIN on my machine, and the modems would _sometimes_ talk. But it was very dodgy, because the ata idea only seemed to work sometimes on my Dad's machine.

Then when I used DIN, I got the final problem.

***************************************************** I know I am wrong about just about everything. So I am not going to listen when I am told I am wrong about the things I know I am right about.

I don't know DIN, but I *think* that the problem is that neither machine is configured to be a server. I don't know *how* do that on a Windows box, though I find it easy on various unix flavors. (Unix is

*designed* for remote logins, Windows is mostly designed to be a client machine only.)

Though, there are programs (called "backdoors") which are often installed by virii to allow the bad guys to log into and use your computer. If you are *positive* that neither his or your machine will ever be on the internet, I guess that you could put one of those to use, but I would worry about how easy it would make others to get in if you are ever connected to the internet.

Good Luck, DoN.

On 9 Sep 2004 01:40:41 -0400, snipped-for-privacy@d-and-d.com (DoN. Nichols) vaguely proposed a theory ......and in reply I say!:

remove ns from my header address to reply via email

As I said, I did this once. I just can't remember how. But it must be doable!

It's a bit confusing, as the RADMIN and VNC programmes call themselves servers, biy then I need to have the Win DIN Server running to get them to work. So I'm not sure at what level everythong is working (or not :- I don't know DIN, but I *think* that the problem is that neither

***************************************************** I know I am wrong about just about everything. So I am not going to listen when I am told I am wrong about the things I know I am right about.

From

----- There are a number of commercial hardware devices, often referred to as PPP or terminal servers, that can do this task. It is also possible to run the RAS service on Windows NT, assuming that you have a spare Windows NT machine, or the dial-up networking server service that can be loaded on Windows 95, part of the Plus pack.

These alternatives are not very palatable. Hardware servers cost two to three thousand dollars, and my experience with the Microsoft RAS services has taught me that it is impossible to support more than a single modem on the things, let alone run it at the highest speed a PC serial port can support.

Fortunately, a stable, fast, and cheap PPP server can be built with FreeBSD Unix.

-----

Seems to agree with the previous answer about *n*x... Since you only want one modem, you are looking for Microsoft's "server service".

Shows how to install and configure the server service on 95 and 98, complete with screenshots. Should get you going... Or more deeply entangled in this quest...

There is also a similar server service, and much more, in 2K and XP. You can even run your own mail server and sendmail - but enabling any of it installs about a hundred megs of junk, forces you to use password login, and put an unacceptable load on my aging system.

Loren

Having read thru this thread, it seems to me that you're getting nowhere. A few suggestions:

1. Get a 2nd phone line at one end, at least, and a cell phone or 2. That'll allow talking while playing computer comms on the fixed line. A CDMA phone works pretty well in most areas and a friend has hers plugged straight into her laptop via a cable, connects to her ISP without any hassles. We do the same thing off a ship on occasion. Alternatively find somewhere with 2 phone lines, assemble all the gear there and stay in one place testing it until it all works.

1. Get a competent o/s on a cheap machine or 2 and ditch Windows, network (say) 2 linux boxes together and use them to do the comms. Since you're dialing into one, you really only need one at your father's place, set up to accept serial comms and start ppp, with your father's pc connected to the linux box via ethernet. You can still run vnc to get the desktop exported if you want or just smb mount the filesystem on the linux box if all you really want to do is manipulate/update files. Better still smb mount the linux filesystem (or part of it) on the Windoze box.

2. Connect to the internet, buy a Netgear firewall/router and set it up to allow telnet access to you. At 9 from Harris Technology, including 802.11g, it's a bargain. I'm using one right now with my Mac laptop running a unix variant connected along with computers running XP, Win2K, Solaris etc.

1. Look at some of the Lantronix equipment. They used to sell a single port terminal server (LRS 1) which worked well; you plugged a modem in and the LRS box did all the serial-ppp-serial conversions for you. We use heaps of Lantronix gear everywhere for serial to ip translations, networking instruments on ships etc. Never waste time with dodgy s/ware when you can use cheap h/ware. Of course that assumes you put any sort of value on your time.

2. Waste an incredible amount of time/energy pursuing your current path in using unsuitable o/s each end with a user who really just wants something to work easily.

I suspect you'll classify this along with the replies you got from the PC lists etc who probably told you not to do what you're trying to do, which is not what you want to hear. Personally I'd go for option 2 as being easy to test, cheap to do and reliable if connecting to the internet really is totally out of the question. As for running your own s/ware to switch between voice & data then expect other stuff to work - heh. Lotsa luck.

FWIW I've been running mixed o/s and h/ware for nearly 30 years and recently spent a w/end debugging a sick database 3 states away from where I lived via dialup modem. The target machine was a unix (Sun Microsystems Solaris) machine however, not a Windoze POS, and all I needed was a CLI not a GUI. You can do what you want with Windoze but it either requires the server version of the o/s or 3rd party s/ware. Lotta pain when there are easier solutions.

Have fun. Since this is a metalworking n/g and I'm very time poor I really don't want to know. If I wasn't using Dejanews I'd kf this thread. As it is I only read it because DoN posted and he knows what he's talking about WRT computers and so worth reading. Reading between the lines, what do you reckon he's suggesting..........?

PDW

On Thu, 09 Sep 2004 12:13:21 -0700, Loren Amelang vaguely proposed a theory ......and in reply I say!:

remove ns from my header address to reply via email

Thanks for the link. I have been looking and found a couple, but none yet has "shown the light". IT's also difficult in that I have to get to my Dad's place before I try again, or ring him and "put him through the ringer" (pun intended). So I want to make as sure as poss I can get a result.

...say no more.

***************************************************** I know I am wrong about just about everything. So I am not going to listen when I am told I am wrong about the things I know I am right about.

On 9 Sep 2004 19:34:56 -0700, peter_d snipped-for-privacy@hotmail.com (Peter Wiley) vaguely proposed a theory ......and in reply I say!:

remove ns from my header address to reply via email

You are right. You are not saying what I want to hear.

I have had Win working this way. There is a way.

It may not be the easiest.

But - there is no way I am going to suggest my father use Unix, or Linux. - please do not judge me if I will not impose a network involvinh multiple computers and operating systems on my 88 year old father so I can try to make life easier for him. - I actually think that DUN connection in Win98 is quite simple...if you know how. And I don't. Ditching the horse will not help.

Actually I am getting somewhere. I just came to a problem.

AGAIN you are wrong. How _could_ they? They are there to supposeedly support exactly what I am trying to do. What I do not want to hear is hints, and sneering, and from people who "well I know, but you'll have to drag it out of me". They are smug and there for the sake of their egos.

You on the other hand are proposing preposterous and judgmental possibilities, when I know that Win98 can do DUN.

***************************************************** I know I am wrong about just about everything. So I am not going to listen when I am told I am wrong about the things I know I am right about.

On 9 Sep 2004 19:34:56 -0700, peter_d snipped-for-privacy@hotmail.com (Peter Wiley) vaguely proposed a theory ......and in reply I say!:

remove ns from my header address to reply via email

I don;t know why you bothered, but your reply bothered me so I am bothering.

In your reading, either I misposted or you missed. I _have_ had my own programme working fine to do what I wanted. It was only when it tried to work on the server side that it failed.

***************************************************** I know I am wrong about just about everything. So I am not going to listen when I am told I am wrong about the things I know I am right about.

On 9 Sep 2004 19:34:56 -0700, peter_d snipped-for-privacy@hotmail.com (Peter Wiley) vaguely proposed a theory ......and in reply I say!:

remove ns from my header address to reply via email

Oh Yeah! Of course. Now I recognise the name.

I will asume that you are winding me up, as the tone of both of your Johnnie-come-lately replies to my problems, in two cases now, leaves me with a much more unpleasant opinion otherwise.

If you are as pressed for time and your input is as critical to metalworking as you claim, then simply ignore my posts and stop wasting both of our lives.

***************************************************** I know I am wrong about just about everything. So I am not going to listen when I am told I am wrong about the things I know I am right about.

On Mon, 06 Sep 2004 17:53:21 GMT, "DaveK" vaguely proposed a theory ......and in reply I say!:

remove ns from my header address to reply via email

A belated question for you (I had some trouble with MY pc). What _happens_ when the ring tone goes down the line? I have found out that the Dial Up Server will not act as a Server until it answers a real ring, so my idea of direct connect does not work, even though I get the modems to talk.

Is there a way to make the MODE think that it;s getting a ringing tone, thus sending a message to the DUS to 'answer" even though I am not online. I assume it's not just a matter of recording a ring tone and playing it back :-<

***************************************************** I know I am wrong about just about everything. So I am not going to listen when I am told I am wrong about the things I know I am right about.