.. -... -- .. -... -- .. -... -- .. -... -- .. -... -- .. -... -- News about IBM Terminals that use EBCDIC Character Encoding .. -... -- .. -... -- .. -... -- .. -... -- .. -... -- .. -... -- Newsgroups: comp.terminals Path: cs.utk.edu!emory!europa.eng.gtefsd.com!MathWorks.Com !transfer.stratus.com!xylogics.com!Xylogics.COM!carlson Date: 14 Mar 1994 19:08:46 GMT Organization: Xylogics Incorporated Message-ID: <2m2cnu$6n3@newhub.xylogics.com> References: Keywords: IBM 3270 terminal-emulation specification From: carlson@Xylogics.COM (James Carlson) Subject: Re: IBM 3270 specifications In article , tmichl@cosy.sbg.ac.at (Thomas Michlmayr) writes: |> Hi all! |> |> I am trying to get information about the IBM 3270 |> terminal emulation. I would need some specifications |> about how 3270-terminals are addressed by the host. |> |> If anyone's got some info, please send me some |> pointers, as I could not find any info (FAQs, ...). Start with getting the "IBM 3270 Information Display System Data Stream Programmer's Reference," catalog number GA23-0059. It should be about USD$10.50. If you think you're just going to bang out an emulation like you might for VT100 .... Whew, do you have a surprise coming! ;-} -- James Carlson Tel: +1 617 272 8140 Annex Software Support / Xylogics, Inc. +1 800 225 3317 53 Third Avenue / Burlington MA 01803-4491 Fax: +1 617 272 2618 .. -... -- .. -... -- .. -... -- .. -... -- .. -... -- .. -... -- ////////////////////////////////////////////////////////////////////////////// Newsgroups: comp.protocols.ibm Path: cs.utk.edu!gatech!asuvax!ncar!destroyer!news.itd.umich.edu !condor.hosp.med.umich.edu From: h198@hosp.med.umich.edu (Gene Mangum) Subject: Re: 3278 Protocol Date: Wed, 30 Jun 1993 09:38 EDT Organization: Univ of Michigan Hospitals Lines: 19 Distribution: world Message-ID: <19930630093839.h198@condor.hosp.med.umich.edu> References: <20qnn0$lm4@usenet.INS.CWRU.Edu> NNTP-Posting-Host: condor.hosp.med.umich.edu X-Newsreader: FTPNuz (DOS) v1.0a In Article <20qnn0$lm4@usenet.INS.CWRU.Edu> "ab515@cleveland.Freenet.Edu (James Waler)" says: > > Does anyone have the reference for the 3278 protocol? I am > interested in programming my own routines and I need the > screen control descriptions. > > Any help would be appreciated > Jim Waler > The IBM manual you need is "3270 Data Stream Programmer's Reference" with IBM manual number GA23-0059. +=================================+=========================================+ | Gene Mangum | OS/2 ver 2 - Increased productivity | | Univ of Michigan Hospitals | through the use of a | | h198@hosp.med.umich.edu | "dead" operating system. | +=================================+=========================================+ .. -... -- .. -... -- .. -... -- .. -... -- .. -... -- .. -... -- Path: cs.utk.edu!not-for-mail Newsgroups: comp.terminals Followup-To: comp.terminals Organization: University of Tennessee, Knoxville--Dept. of Computer Science References: <3odjcq$r2q@VTC.TACOM.Army.Mil> Message-ID: <3oj4seINNjj6@duncan.cs.utk.edu> Date: 7 May 1995 14:51:58 -0400 Summary: a harder problem than you think Keywords: IBM, EBCDIC, 5250 From: shuford@cs.utk.edu (Richard Shuford) Subject: Re: TN5250 A question was asked: > > Can anyone give me a termcap entry for 5250 emulation? > Thanks in advance to anyone who can help. > > Bill Adair This is a harder problem than you think. The IBM 5250 is an EBCDIC-speaking terminal. Unix systems like to communicate using ASCII. These are two quite different code sets. Your Unix "termcap" handler assumes the use of ASCII; therefore, strictly speaking, there is not and CAN NEVER BE a termcap entry for "5250". (Or for "3270", the other common EBCDIC terminal type.) (By the way, people I talk to audibly pronounce ASCII as "ASS key". EBCDIC I usually hear pronounced as "EBB suh dik", although in some places you will hear "EBB kuh dik".) Now, given this, clever people in various localities have devised ways to do "protocol conversion", which places a translator between the host computer and the terminal. The host speaks its own language, and the terminal speaks its own language, and the protocol converter does a complicated dance in between to make the conversation understandable at both ends. So in using your 5250 terminal, you need to use (or obtain) some type of protocol converter. If your site already has one, your system administrator can tell you what type of ASCII-speaking terminal it emulates. If you need to obtain the merchandise, there are several vendors. One recommended by people in this newsgroup is OpenConnect Systems Inc. Suite 800 2711 LBJ Freeway Dallas, Texas 75234 USA Phone: +1 214/484-5200 Fax: +1 214/484-6100 E-Mail: -- ...Richard S. Shuford | "Speak up for those who cannot speak up for ...shuford@cs.utk.edu | themselves, for the rights of all those who ...Info-Stratus contact| are destitute." Proverbs 31:8 .. -... -- .. -... -- .. -... -- .. -... -- .. -... -- .. -... -- Newsgroups: comp.terminals Path: cs.utk.edu!gatech!howland.reston.ans.net!europa.eng.gtefsd.com !MathWorks.Com!transfer.stratus.com!xylogics.com!Xylogics.COM!carlson Organization: Xylogics Incorporated Message-ID: <2slfvn$bt8@newhub.xylogics.com> References: <2slel0$6ea@news.moneng.mei.com> Date: 2 Jun 1994 20:38:15 GMT From: carlson@Xylogics.COM (James Carlson) Subject: Re: Need IBM 3270 Termcap entry In article <2slel0$6ea@news.moneng.mei.com>, | patd@elvis.moneng.mei.com (Patrick Davies) writes: | |> Kieu Hua (khua@husc10.harvard.edu) wrote: |> : |> : I am looking for an Unix termcap entry for IBM3270 terminal. If you |> : have it, can you kindly send a copy to me? |> |> I am unfamiliar with the IBM 3270 terminal, but if it is at all backward |> compatible with other IBM terminals, maybe one or more of the following |> will help: IBM 3270 is not an ASCII terminal -- there is no possible termcap entry for it. It is not at all related to IBM character-oriented terminals. "3270" is the name of a class of terminal devices which use a packet- oriented protocol (called, naturally enough, the "3270 data stream") for commands and use EBCDIC for character codes. It's more like a network device than it is like a regular character terminal, such as a VT100. -- James Carlson Tel: +1 617 272 8140 Annex Software Support / Xylogics, Inc. +1 800 225 3317 53 Third Avenue / Burlington MA 01803-4491 Fax: +1 617 272 2618 .. -... -- .. -... -- .. -... -- .. -... -- .. -... -- .. -... -- Newsgroups: comp.terminals Path: cs.utk.edu!emory!swrinde!howland.reston.ans.net!vixen.cso.uiuc.edu !owens.slip.uiuc.edu!jbn Organization: University of Illinois at Urbana Message-ID: References: <2slel0$6ea@news.moneng.mei.com> <2slfvn$bt8@newhub.xylogics.com> Date: 3 Jun 94 18:12:46 GMT From: jbn@uiuc.edu (J.B. Nicholson-Owens) Subject: Re: Need IBM 3270 Termcap entry carlson@Xylogics.COM (James Carlson) writes: > >IBM 3270 is not an ASCII terminal -- there is no possible termcap entry >for it. It is not at all related to IBM character-oriented terminals. Yep, and it's also worth mentioning that "tn3270" (a telnet command that makes 3270 datastreams usable from a ASCII character-oriented terminal) and 100% compatibles are freely available for a number of systems (including C source for Unix, Macintosh and MS-DOS). Check "archie" for details by searching for "3270". .. -... -- .. -... -- .. -... -- .. -... -- .. -... -- .. -... -- Here are some technical articles in hardcopy publications, which relate to IBM 3270 support under Unix: Marsh, Bob. "How Important is TN3270 to You?" \Network Computing/, May 1993, pages 144-148. Nasr, Alex. "TN3270: An Interoperability Option." \3TECH: The 3Com Technical Journal/, Winter 1992, pages 51-59. Shein, Barry. "The Telnet Protocol." \ConneXions/, October 1989, pges 32-38. .. -... -- .. -... -- .. -... -- .. -... -- .. -... -- .. -... -- ////////////////////////////////////////////////////////////////////////////// Newsgroups: comp.protocols.misc Path: cs.utk.edu!gatech!swrinde!sdd.hp.com!portal!cup.portal.com!TRADER From: TRADER@cup.portal.com (Paul Vincent McGinnis) Message-ID: <84664@cup.portal.com> Date: Sat, 3 Jul 93 23:56:00 PDT Organization: The Portal System (TM) References: <21129fINNg0p@infoac.rmi.de> Subject: Re: Need INFO: BSC 2780 Terminalemulation In a recent message, afassl@infoac.rmi.de (Andreas Fassl) asked for information about the 2780 protocol. 2780 refers to the IBM 2780 Remote Job Entry (RJE) bisynchronous device. A more recent (relatively speaking...) device is the IBM 3780, which is described in detail in IBM manual number GA27-3063 ("Component Information for the IBM 3780"). If you are just interested in IBM bisynchronous (BSC) communications, the manual to refer to is IBM manual number GA27-3004 ("General Information - Binary Synchronus Communications"). Paul McGinnis / TRADER@cup.portal.com .. -... -- .. -... -- .. -... -- .. -... -- .. -... -- .. -... -- Newsgroups: comp.protocols.iso,comp.protocols.ibm Path: utkcs2!stc06.ctd.ornl.gov!cs.utk.edu!gatech!www.nntp.primenet.com !nntp.primenet.com!europa.clark.net!cpk-news-hub1.bbnplanet.com !news.bbnplanet.com!news.sprintlink.net!news-peer.sprintlink.net !news-pull.sprintlink.net!news-dc-10.sprintlink.net !news1.rcsntx.swbell.net!usenet Date: Tue, 01 Apr 1997 15:13:03 -0800 Organization: Southwestern Bell Internet Services Message-ID: <3341967E.3462@swbell.net> References: <333C0238.3C70@rz.uni-karlsruhe.de> To: ur6q@rz.uni-karlsruhe.de From: Gary Franco Subject: Re: IBM 2780 Protocol Thomas Grittner wrote: > > I'm trying to connect an older Computer (Nixdorf 886x) via the serial > interface to a PC. > One of the possible Protocols is the old IBM 2780. > Can anybody tell me, where I can find the specs of the protocol. The exact document specification number that you require is: GA27-3005 2780 DATA TRANSMISSION TERMINAL COMPONENT DESCRIPTION This Component Description manual describes the principles of operation of the IBM 2780 Data Transmission Terminal. The data-link control characters, code structures, timeouts, and throughput rates associated with the Binary Synchronous method of transmission are described. The communications facilities, data sets, and special features available for this Tele-processing terminal are also discussed. The retail price of this manual in the U.S is $3.45US You will also require a synchronous adapter that supports Bi-Sync protocol, IBM has a couple of varieties available. If you require to move to 3780 protocol then make sure you obtain an adapter that will support BiSync as well as SDLC as the 3780 goes both ways. I will be out of town for a week so if you wish to respond send me e-mail at . Good Luck, Gary .. -... -- .. -... -- .. -... -- .. -... -- .. -... -- .. -... -- Newsgroups: comp.terminals Path: cs.utk.edu!emory!news-feed-2.peachnet.edu!umn.edu!math.fu-berlin.de !zib-berlin.de!fauern!rz.unibw-muenchen.de!claude Organization: University of the Federal Armed Forces Munich Lines: 39 Message-ID: References: <94Mar9.145425est.18223@orasis.vis.toronto.edu> NNTP-Posting-Host: bauv106.bauv.unibw-muenchen.de Date: 10 Mar 1994 07:10:50 GMT From: claude@bauv.unibw-muenchen.de (Claude Frantz) Subject: Re: >Wanted: IBM 3178 termcap mann@vis.toronto.edu (Richard Mann) writes: | | Maybe a Termcap isn't any use ... Do you mean 3278 ? The 3270 series of terminal are screen-oriented ones using local editing. UNIX can only handle character-oriented terminals, i.e. every keystroke generates one or more characters which are sent immediately. Most of the keys of a 3270 do only change information on your terminal without sending anything. Only few keys send information to the control unit or to the line. A termcap entry for a 3270 terminal is useless because it cannot be used with UNIX. > I just discovered that this terminal has a Coax connector on the back. This model is entended to be connected to a control unit like 3274. This is an IBM proprietary interface for such connections. > It says something about X.25 network protocol, whatever that is. Some models of this series of terminals can be connected to networks using the CCITT X.25 protocol. Ask IBM for further details. > Count on IBM to make something > non-standard when everyone else uses RS 232! RS-232 is not the only standard of this world. Further it is a very bad standard because it is very imprecise. I can report many problems occuring because each manufacturer has it specific view on this RS-232 standard. Further UNIX is not the only operating system. | Richard. -- Claude F. (claude@bauv106.bauv.unibw-muenchen.de) This message may contain opinions which are not shared by my employer. The facts can speak for themselves. .. -... -- .. -... -- .. -... -- .. -... -- .. -... -- .. -... -- Newsgroups: comp.terminals Path: cs.utk.edu!gatech!howland.reston.ans.net!agate!news.duke.edu!zombie.ncsc.mil!news.mathworks.com!transfer.stratus.com!xylogics.com!Xylogics.COM!carlson Organization: Xylogics Incorporated Message-ID: <3hd261$le8@newhub.xylogics.com> References: <3h7v1m$19jm@sernews.raleigh.ibm.com> NNTP-Posting-Host: newhub.xylogics.com Date: 9 Feb 1995 12:35:13 GMT From: carlson@Xylogics.COM (James Carlson) Subject: Re: IBM 3270ish emulation through termcap/terminfo? In article <3h7v1m$19jm@sernews.raleigh.ibm.com>, Dick Locke writes: |> |> I could not find a FAQ for this group. I am looking for |> (now don't laugh) termcap/terminfo entries or any information |> related to 3270 terminal support to be used when going from |> a mainframe to a Unix box. We have basic telnet access in |> line mode, but we're looking for full-screen support (Curses?). |> |> Yes, I work at IBM ;-) and I can't find this any where. |> Are there special hardware requirements or is this just |> unrealistic based on protocol or mainframe limitations? This is not an easy thing to get working. It's much easier to get a Unix system to talk to an IBM system than the other way around. The problem is that the 3270 protocol is block-oriented (only certain keys send AID sequences to the host, the rest are buffered locally), whereas nearly all terminals in the Unix world are character-oriented (all keys send codes to the host, and are interpreted by the application itself). Since, when you press 'a' on your 3278, nothing is sent to the IBM host, there's nothing to forward along to the Unix host. If this is your own application you're trying to support, it'll probably be easier to put an RPC interface between the engine on the Unix side and a client on the IBM side which creates the 3270 data stream. If you're just trying to run vi or emacs, use ex or pull the file over and use XEDIT. It's much less grief. If it's someone else's application, I think you're on your own. Try to find a programmatic interface to it and write a front end. All of that aside, give a call to OpenConnect at +1 214 484 5200. They're the experts at this. Good luck ... --- James Carlson Tel: +1 617 272 8140 Annex Software Support / Xylogics, Inc. +1 800 225 3317 53 Third Avenue / Burlington MA 01803-4491 Fax: +1 617 272 2618 .. -... -- .. -... -- .. -... -- .. -... -- .. -... -- .. -... -- Newsgroups: comp.os.vms,comp.sys.dec Path: utkcs2!darwin.sura.net!zaphod.mps.ohio-state.edu!sol.ctr.columbia.edu !destroyer!cs.ubc.ca!bcsystems!ewilts Message-ID: <1992Oct30.083025.1153@galaxy.gov.bc.ca> References: Reply-To: EWILTS@GALAXY.GOV.BC.CA Organization: BC Systems Corporation Date: 30 Oct 1992 08:30:25 -0800 From: ewilts@galaxy.gov.bc.ca (Ed Wilts) Subject: Re: vt200 on IBM 3270 terminal In article , FILIPSEN_BMF@jt.dk (Bent Filipsen) writes: > Is it possible to run VT200 from an IBM 3270 terminal? > Gateway? Emulating? Many vendors have tried to do this with varying degrees of unsuccess. Emulating a character-echo terminal on a block-mode device is like trying to fit a square plug into a round hole. We use the STX product around here - it barely fits the bill from what I've heard although I've not personally worked with it. We experimented with TNVT100 which is the opposite of TN3270 - in this case using using a 3270 to perform VT100 emulation. TNVT100 is freely distributable but testing here found it unsuitable for our purposes (ie worse than STX). > And what is the best solution? Actually, the best approach seems to be put inteligent workstations on people's desks and then use the appropriate software packages depending on which host you're going to. The other approach (not too well taken by most IBM people) is to put a VT100 terminal on the user's desk and then use TN3270 when you're connected to the IBM system. -- Ed Wilts, BC Systems Corp., 4000 Seymour Place, Victoria, B.C., Canada, V8X 4S8 EWilts@Galaxy.Gov.BC.CA | Ed.Wilts@BCSystems.Gov.BC.CA | (604) 389-3430 .. -... -- .. -... -- .. -... -- .. -... -- .. -... -- .. -... -- Newsgroups: comp.os.vms,comp.sys.dec Path: utkcs2!darwin.sura.net!wupost!sdd.hp.com!caen!destroyer!ncar!noao !arizona!arizona.edu!telcom.arizona.edu!leonard Message-ID: <1992Oct30.115253.3915@arizona.edu> Date: 30 Oct 92 18:52:51 GMT References: <1992Oct30.083025.1153@galaxy.gov.bc.ca> Reply-To: Leonard@Arizona.EDU Organization: University of Arizona Telecommunications From: leonard@telcom.arizona.edu (Aaron Leonard) Subject: Re: vt200 on IBM 3270 terminal In article <1992Oct30.083025.1153@galaxy.gov.bc.ca>, ewilts@galaxy.gov.bc.ca (Ed Wilts) writes: | | In article , FILIPSEN_BMF@jt.dk (Bent Filipsen) writes: | > Is it possible to run VT200 from an IBM 3270 terminal? | > Gateway? Emulating? | | Many vendors have tried to do this with varying degrees of unsuccess. | Emulating a character-echo terminal on a block-mode device is like trying to | fit a square plug into a round hole. | | We use the STX product around here - it barely fits the bill from what I've | heard although I've not personally worked with it. We experimented with | TNVT100 which is the opposite of TN3270 - in this case using using a 3270 to | perform VT100 emulation. TNVT100 is freely distributable but testing here | found it unsuitable for our purposes (ie worse than STX). | | > And what is the best solution? | | Actually, the best approach seems to be put inteligent workstations on | people's desks and then use the appropriate software packages depending on | which host you're going to. The other approach (not too well taken by most | IBM people) is to put a VT100 terminal on the user's desk and then use TN3270 | when you're connected to the IBM system. There are other solutions to allowing 3270s fullscreen access to VT200-type applications. One that we looked at was to use ITT controllers instead of 3274s. These controllers had RS232 asynch lines plugged into them in addition to the normal IBM synch line. The controller gave the 3270 user an option to connect to the mainframe in 3270 mode, or to connect to the asynch lines in VT200 mode, in which case the controller performs the VT200 emulation on the terminal's behalf. The asynch lines can then be connected in milking-machine fashion to terminal server ports, allowing the user to connect to your Ethernet-connected host. I found (talking as a VAX guy) that the VT200 emulation provided by the ITT controller was just fine. EDT, TPU, etc. all looked great. (Of course, the user has to cope with a different key mapping.) A similar approach is taken by the CXM04 option to DEC's DECserver 550. In this case, the 3270 twinax lines are plugged into the DECserver, which is Ethernet-attached and which provides the VT200 emulation. The DECserver is milking-machine attached via twinax to a 3274 controller, and if the 3270 user selects 3270 mode, the DECserver passes the 3270 traffic on to the IBM controller, allowing the user to connect to the mainframe in normal fashion. I *think* that someone like McData may make a product whereby you replace your 3274s with their controllers, which can then tunnel LAT or TELNET in SNA, and then de-encapsulate the stuff via a gateway onto the Ethernet. This saves you running asynch RS232 or Ethernet to all your controller sites, but obviously swapping out all your controllers may be a nontrivial expense. You have a tradeoff between cost and functionality. The 3274/SNA blockmode approach is just too dumb (or smart, depending on your bias :-) to allow 3274-attached terminals to do a tolerable full-duplex, character-mode terminal emulation. So, if you want to provide real asynch terminal emulation to users with physical 3270s on their desks, you'll have to run Ethernet or asynch RS232 to each 3274 - which can be a huge cost in a large distributed SNA network. Or you can go with a single gateway between your SNA and your Ethernet network, which may be cheaper in the short run, but won't give your users true VT200 emulation. In the long run, of course, I'd recommend that you pitch all your SNA crap and your 3274s and your 3270s, give your users Macs or PCs or VAXstations, route TCP/IP everywhere, and let your users use TN3270 to access your mainframe and TELNET to access everyone else. Unfortunately, as J. M. Keynes pointed out, "in the long run, we're all dead." Cheers, Aaron Aaron Leonard (AL104), University of Arizona Network Operations, Tucson AZ 85721 "It's not a bug, it's a form of flow control." - Jerry Leichter on why [then] crash-prone Unix [was] a suitable platform for NSFNET core routers. (circa 1989) .. -... -- .. -... -- .. -... -- .. -... -- .. -... -- .. -... -- Newsgroups: alt.internet.services Path: cs.utk.edu!emory!wupost!usc!elroy.jpl.nasa.gov!ucla-cs!ucla-mic !MVS.OAC.UCLA.EDU!CSP1DWD Message-ID: <19930105015720CSP1DWD@MVS.OAC.UCLA.EDU> Sender: MVS NNTP News Reader Nntp-Posting-Host: mvs.oac.ucla.edu References: <1992Dec31.231922.25926@news.acns.nwu.edu> Date: 5 Jan 93 01:57:36 PST From: CSP1DWD@MVS.OAC.UCLA.EDU (CSP1DWD) Subject: Re: VT100 EMULATION THRU VM? In article <1992Dec31.231922.25926@news.acns.nwu.edu>, Albert-Lunde@nwu.edu (Albert Lunde) writes: >ANET *is* a well written commercial application. I doubt anyone can do >much better, but I would encourage people to avoid vt100->3270->vt100 >telnet paths, because of the inherent limits of the situation. Absolutely, if you can avoid having to do vt100 emulation on a 3270 terminal do so by all means... but if you must, and many sites are still stuck with old IBM-3270 terminals and mainframes then two packages exist, Arty Ecock's (at CUNY) Tnvt100 for VM and my own (at UCLA) package Xtelnet for MVS. They both deal with the generic problem of emulating vt100 terminals in very creative ways and the result while far from perfect is eminently useable. -- Denis .. -... -- .. -... -- .. -... -- .. -... -- .. -... -- .. -... -- Newsgroups: vmsnet.networks.tcp-ip.cmu-tek Path: cs.utk.edu!gatech!swrinde!sdd.hp.com!spool.mu.edu!uwm.edu!wupost !crcnis1.unl.edu!moe.ksu.ksu.edu!mccall!cmu-tek-newsgate!list Message-ID: <568@moretonp.demon.co.uk> Reply-To: peter@moretonp.demon.co.uk Organization: The Internet Return-Path: Resent-Date: Fri, 26 Feb 93 16:16:10 MEZ Resent-From: CMU-OpenVMS-IP-Request@DRYCAS.CLUB.CC.CMU.EDU Resent-Message-ID: <9303020856.AA15469@depot.cis.ksu.edu> Errors-To: CMU-OpenVMS-IP-Request@DRYCAS.CLUB.CC.CMU.EDU Resent-Reply-To: CMU-OpenVMS-IP@DRYCAS.CLUB.CC.CMU.EDU X-Mailer: PCElm 3.01 (1.3 gt) From: Peter S Moreton Date: Fri, 26 Feb 93 16:16:10 MEZ Subject: Re: Shareware Telnet-3270 The only Telnet-3270 in the public domain that I know of is the NCSA 'TN3270' product for IBM PC. It has been hacked by Clarkson university, who distribute TN3270 as 'CUTCP'. I have used both for TCP connectivity to IBM AS/400's (why, why?) and they work fine. Source code in microsoft 'C' is provided. .. -... -- .. -... -- .. -... -- .. -... -- .. -... -- .. -... -- Newsgroups: comp.protocols.ibm Path: cs.utk.edu!darwin.sura.net!udel!news.intercon.com!psinntp!shearson.com !newshost!wbrand From: wbrand@krishna.shearson.com (Willy Brandsdorfer) Subject: Re: vt100 emulation for MVS/ESA In-Reply-To: pm@ks.id.ethz.ch's message of Tue, 23 Mar 1993 13:23:09 GMT Message-ID: Sender: news@shearson.com (News) Organization: Lehman Brothers References: Date: Wed, 31 Mar 1993 16:52:59 GMT Lines: 29 In article pm@ks.id.ethz.ch (Peter Meyer) writes: We are running MVS/ESA on a IBM 3090 host and we are in desperate need of a vt100 emulation package which is able to talk to IBM 3270 terminals. Basically it should be straight the opposite of the 'ordinary' tn3270 emulation package available for UNIX, PC's and MAC's, that is, it should emulate a vt100 on an IBM3270 terminal. Contact: Tuebner and Associates Inc, P.O. Box 1994 Stillwater, Oklahoma 74076 Voice +1 405/624-2254 Fax +1 405/624-3010 Excellent product. We tested it and were amazed at how well it worked. ---------------------------------------------------------------------- William Brandsdorfer | UUCP: !uunet!shearson.com!wbrand Lehman Brothers | INET: wbrand@shearson.com 388 Greenwich St. | Voice: (212) 464-3835 New York, N.Y. 10013 | ---------------------------------------------------------------------- .. -... -- .. -... -- .. -... -- .. -... -- .. -... -- .. -... -- Newsgroups: comp.protocols.ibm Path: cs.utk.edu!gatech!howland.reston.ans.net!darwin.sura.net!haven.umd.edu !uunet!pau!splinter!robc From: robc@splinter.synnet.com (Rob Ciampa) Subject: Re: 3270 Protocol Conversion Message-ID: Sender: news@pau.synnet.com (News Administration) Organization: Synernetics Inc References: Date: Thu, 1 Jul 1993 16:43:09 GMT Lines: 24 In article , Dick.libby@ncrssd.stpaul.ncr.com (Dick Libby) writes: |> |> My company is looking for a Unix protocol conversion package that |> converts 3270 terminals to VTxxx. We require line-at-a-time |> operation on input (not character-at-a-time), although a package |> that supports both character- and line-at-a-time would be |> acceptable. We are interested in either co-marketing or OEMing the |> package. |> Check out Brixton Systems. They have a number of packages that allow for 3270 to VTxxx conversion. Contact them at: Brixton Systems 185 Alewife Brook Parkway Cambridge, MA 02138 Phone: (617) 661-6262 ---------------------------- Robert A. Ciampa (robc@synnet.com) Synernetics 85 Rangeway Road North Billerica, MA 01862 (508) 670-5488 .. -... -- .. -... -- .. -... -- .. -... -- .. -... -- .. -... -- Newsgroups: comp.terminals,comp.sys.ibm.pc.hardware Path: cs.utk.edu!darwin.sura.net!europa.eng.gtefsd.com!uunet!psgrain!percy !agora!tedm From: tedm@agora.rain.com (Ted Mittelstaedt) Subject: Re: Talking to mainframe via terminal emulator board? Message-ID: Organization: Open Communications Forum References: <1993Sep21.031038.25558@iitmax.iit.edu> Distribution: usa Date: Wed, 22 Sep 1993 18:32:01 GMT Lines: 48 Xref: cs.utk.edu comp.terminals:1256 comp.sys.ibm.pc.hardware:84070 In article <1993Sep21.031038.25558@iitmax.iit.edu> thssamj@iitmax.iit.edu (Aditya M. Jani) writes: > >I am trying to talk to a vm system thru a PC emulator >board. Is it possible to write data to the address or port >which the 3270 emulator board uses so that data can be sent >to the vm mainframe system as if it had been typed from a keyboard.? > >The emulator board type is IRMA/3IRM >The software is DCA's e78plus 3270 terminal emulator. > >I think this may possible because 'send.exe', a program to transfer >file to the main frame works even when the emulator s/w is not running but >connection to the mainframe is established. >That means send.exe talks directly to the address to which the board is >mapped. > >Is ther some s/w to find out what address data is being written to? >Does somebody know what address the board uses (does not use com ports)? > These boards typically occupy an address right above the VGA video buffers. Try using debug to examine memory addresses ce00-dfff right after boot-up and you should see it. You can also use the "u" (unassemble) command of debug on send.exe and look for "in" or "out" instructions in the code. Remember, the IBM mainframes don't care for ASCII, they expect to be fed a diet of ECBDIC. IBM's 3270/5250 emulation software and boards give you the interfacing instructions to write programs to communicate with their mainframes. What they basically do is their software runs as a TSR on the PC, and your program reads a pointer out of an unused interrupt vector that contains the address of a data structure that contains input and output buffers, along with some control memory locations. Your program places data into these locations and then toggles bits that triggers the TSR to send the data to the host. You might consider looking into their products and scrapping the IRMA/DCA stuff instead of spending a lot of time disassembling their programs. Ted Mittelstaedt tedm@agora.rain.com .. -... -- .. -... -- .. -... -- .. -... -- .. -... -- .. -... -- Newsgroups: comp.protocols.tcp-ip.ibmpc Path: cs.utk.edu!emory!news-feed-2.peachnet.edu!news-feed-1.peachnet.edu !gatech!usenet.ins.cwru.edu!eff!news.kei.com!news.oc.com!matt Message-ID: <1993Dec5.232028.27062@ra.oc.com> Organization: OpenConnect Systems, Dallas, TX References: <7547629022428@skeeve.mcs.anl.gov> <1993Dec2.010958.25389@exu.ericsson.se> Date: Sun, 5 Dec 1993 23:20:28 GMT From: matt@ra.oc.com (Matthew Lyle) Subject: Re: TN5250 ?? sims@sndsn1.sedalia.sinet.slb.com (David Sims) writes: >> In <7547629022428@skeeve.mcs.anl.gov>, dauns@skeeve.mcs.anl.gov >> (Frank Dauns) writes: >> > >> >Does anyone know of a IBM-5250 terminal emulator for the PC similar >> >to TN3270 perhaps ? >> > >> >Thanks in advance > >Open Connect Systems of Farmers Branch (Dallas), Texas sells such a beast. >Although I haven't any personal experience with it, I have seen it at >various trade shows. I believe it runs with PC-TCP from FTP Software Inc. We're in Dallas. is a good email address to request information from. -- Matthew Lyle matt@oc.com matt@utdallas.bitnet OpenConnect System, Dallas, Texas (214) 888-0474 .. -... -- .. -... -- .. -... -- .. -... -- .. -... -- .. -... -- Newsgroups: alt.winsock,bit.listserv.win3-l,comp.protocols.tcp-ip.ibmpc Path: cs.utk.edu!emory!europa.eng.gtefsd.com!howland.reston.ans.net!torn !news.ccs.queensu.ca!rymerson!Jim.Rymerson Message-ID: Keywords: TN3270 for windows Lines: 27 Sender: news@knot.ccs.queensu.ca (Netnews control) Organization: Queen's University X-Newsreader: Trumpet for Windows [Version 1.0 Rev Final Beta #3] References: <1993Dec21.165434.27262@hal.depaul.edu> Date: Tue, 21 Dec 1993 18:28:06 GMT From: Jim.Rymerson@qucdnadm.QueensU.CA (Jim Rymerson) Subject: Re: QWS3270 version 2.3 In article <1993Dec21.165434.27262@hal.depaul.edu> Jobin Ephrem writes: >From: Jobin Ephrem >Subject: Re: QWS3270 version 2.3 >Keywords: TN3270 for windows >Date: Tue, 21 Dec 1993 16:54:34 GMT >In article Jim.Rymerson@qucdnadm.QueensU.CA (Jim Rymerson) writes: >> >>Version 2.3a of QWS3270 a winsock compliant TN3270 for Windows is now >>available via anonymous ftp from ftp.ccs.queensu.ca in the pub/msdos/tcpip >>directory. This release includes the following: > Does Qws3270 take ip addresses at the command line? Is it possible to use it > with Hgopher as the tn3270 viewer? If not is there any other client that > can be used? >jobin. >acsjce@orion.depaul.edu QWS3270 will accept an IP address on the command line. It also will accept the port number from the command line (23 is assumed if not given). Several others have used it with Hgopher as a tn3270 viewer. Jim. .. -... -- .. -... -- .. -... -- .. -... -- .. -... -- .. -... -- Path: stc06.ctd.ornl.gov!fnnews.fnal.gov!uwm.edu!math.ohio-state.edu !howland.reston.ans.net!newsfeed.internetmci.com!news.mathworks.com !zombie.ncsc.mil!nntp.coast.net!torn!kone!news.ccs.queensu.ca!JIM!jgs From: jgs@fox.nstn.ca (Jim Rymerson) Newsgroups: comp.protocols.tcp-ip.ibmpc,bit.listserv.tn3270-l Subject: Re: QWS3270 Date: Sat, 16 Mar 1996 08:04:56 LOCAL Organization: Jolly Giant Software Lines: 34 Message-ID: References: <4i25pm$37tc@ns1-1.CC.Lehigh.EDU> In article , dave@software-europe.co.uk writes: > >>Date: Wed, 13 Mar 96 08:54:43 PDT >> >> Anyone using QWS 3270, freeware or shareware versions? >> >> Any problems under Win 3.1 and are they stable under WIN95? >I used QWS3270 for a while, mainly because it was the only >package I could find. However, I have moved to the Chameleon >package from Netmanage. I moved for a number of reasons, not >least of which was the inability to do file transfers with >QWS3270. >I agree with the other posters comments that it is a clean, >efficient way of accessing 3270 hosts, it just hadn't got >enough functionality for my needs. In terms of problems >I didn't hit a single bug while I used it and I used it >extensively for a few weeks. >The file-transfer issue may now be resolved as this was >one of the major areas targetted for future development >of QWS3270. IND$FILE file-transfer support is currently under development for QWS3270 EXTRA. I hope that this support will be in place by the end of March 1996. Jim. .. -... -- .. -... -- .. -... -- .. -... -- .. -... -- .. -... -- Newsgroups: comp.os.ms-windows.networking.tcp-ip Path: utkcs2!stc06.ctd.ornl.gov!cs.utk.edu!news.msfc.nasa.gov!news.sgi.com !news.mathworks.com!news-peer.gsl.net!news.gsl.net!portc01.blue.aol.com !newsstand.cit.cornell.edu!news.acsu.buffalo.edu!news.drenet.dnd.ca !crc-news.doc.ca!nott!kone!news.ccs.queensu.ca!Jim1!jgs From: jgs@fox.nstn.ca (Jim Rymerson) Subject: Re: tn3270 for win95 Date: Mon, 30 Sep 1996 08:19:35 LOCAL Organization: Jolly Giant Software Message-ID: In article <324EDE8D.5105@cluster.engr.subr.edu> Justin Young writes: > > Howdy, where can I get a tn3270 emulator for windows 95/nt (32-bit > winsock support)? Preferably freeware or cheapware. You may want to look at QWS3270 PLUS. A DEMO of version 2.6 can be found in the /pub/pc/win3/winsock directory at ftp.winsite.com and the /pub/simtelnet/win3/inet directory on FTP.Simtel.Net as Q3270V26.ZIP. The demo is just a 16 bit version but the full product ships with both a 16 bit version for Win 3.1 and a 32 bit version for Win 95 and Win NT. The register.txt file that is part of the zip file contains the current pricing info. .. -... -- .. -... -- .. -... -- .. -... -- .. -... -- .. -... -- Newsgroups: comp.dcom.modems,comp.sys.ibm.pc.hardware.comm Path: cs.utk.edu!gatech!swrinde!hookup!nic.hookup.net!xenitec!zswamp!geoff Message-ID: References: <2li0gv$1414@nwg.nectec.or.th> Organization: Izot's Swamp Date: Tue, 08 Mar 1994 21:51:25 EST From: geoff@zswamp.UUCP (Geoffrey Welsh) Subject: Re: Fossil drivers? roland@nwg.nectec.or.th (Roland Peter Sauermann) writes: > Geoffrey Welsh (geoff@zswamp.UUCP) wrote: > > Of course, a FOSSIL is useless and will only get in the way if your > > communications software doesn't make use of it. > > > Geoffrey Welsh geoff@zswamp.uucp, [xenitec.on.ca|m2xenix.psg.com]!zswamp!ge > > > > Have you heard of any IBM-3270 DOS term emul./comm packages that support > a fossil driver like X00? Most of the 3270 emulators that I've seen use an SNA card in the PC or a LAN connection to an SNA or SAA gateway; I have not seen serial ports used for 3270 emulation. My own experience with using IBM mainframes over the modem: I used a VT-100 terminal and either the SIM3270 or Yale emulation packages on the host, and that was several years ago. Sorry. -- Geoffrey Welsh geoff@zswamp.uucp, [xenitec.on.ca|m2xenix.psg.com]!zswamp!geoff "Recipes are for those who have a preconceived idea of what their food is supposed to taste like" - Ron Scoular .. -... -- .. -... -- .. -... -- .. -... -- .. -... -- .. -... -- Newsgroups: comp.sys.hp,comp.terminals,comp.terminals.tty5620,comp.misc Path: cs.utk.edu!emory!europa.eng.gtefsd.com!MathWorks.Com!news.kei.com !bloom-beacon.mit.edu!apollo.hp.com!hpwin052!gerryw Message-ID: Date: Fri, 1 Apr 1994 16:13:35 GMT References: Nntp-Posting-Host: hpbrak42.uksr.hp.com Organization: HP UK Professional Services X-Newsreader: TIN [version 1.2 PL0.8] Followup-To: comp.sys.hp,comp.terminals,comp.terminals.tty5620,comp.misc From: gerryw@hpbrak42.uksr.hp.com (Gerry WINSOR) Subject: Re: IBM AS400 5250 Emulation Jim Dunn (dunn@vet.vet.purdue.edu) wrote: : ========================================================= : Customer has IBM AS 400 (GROAN!) set up to support TCP/IP : and connected to Ethernet LAN. : Customer has population of users who have AT&T ANSI : terminals on their respective desks. They do NOT have : PCs with Windows or X terminals. Just ANSI terminals. : Users with ANSI terminals now log into HP 9000/8xx : and then establish telnet session to AS 400. But : they need to be able to emulate IBM 5250 display so : keyboard mapping make sense for AS 400 applications. : Need to locate a product that will run on HP 9000/8xx : and provide 5250 emulation to AS 400. Would prefer : that solution operate in telnet-TCP/IP-Ethernet : environment, but will am interested in any functional : alternative. : ======================================================== StarWare have a series of products which'll allow HP9000->AS/400 connectivity BUT only over SNA, ie HP 802.5, HP X.25 or HP SDLC interfaces. It does provide 5250 emulation, and more. StarWare can be contacted at: StarWare Inc., 2150 Shattuck Avenue, Suite 204, BERKELY, CA 94704 Phone: (510) 704-2000 FAX: (510) 704-2001 EMail: sales@starware.com OpenConnect Systems have recently announced that their OCS II Server is available on HP-UX 9.x. The OCS TELNET & TN Servers for OCS II / HP9000: "Allows all ASCII based terminals within a TCP/IP network to emulate an IBM 5250 via standard Telnet protocol." (their words not mine). This assumes that you have a TN5250 Server on the AS/400 ..... since the 5250 emulation, off the HP is via Telnet to the AS/400. IBM's AS/400 TCP/IP datasheet says: "The AS/400 TELNET offering will provide Client and Server support for the default line mode terminal emulation as well as full screen support for 525x and 327x terminal types. Existing applications of the AS/400 can provide full screen 525X and 327X modes while customer applications will have to be provided for server line mode support." So if the AS/400 does have full TCP/IP support, it SHOULD have TN5250 Server facilities. The OCS TN Client software should therefore do the trick. OpenConnect Systems are at: OpenConnect Systems Inc., Dallas, Texas Phone: (214) 484-5200 FAX: (214) 484-6100 EMail: info@oc.com Hope this helps, GerryW -- ____________________________________________________________________________ ************** **** /_ _ **** "Please Sir ! *** / //_/ *** Can I have a new brain, this one is **** / **** overloaded." - Larson ************** Gerry Winsor HPDESK: Gerry WINSOR/HP8005/OA Hewlett-Packard Ltd UnixMail: gerryw@hpbrak42.uksr.hp.com Mailstop J1 X.400 : WINSOR,Gerry/HP8005,OA/HP/GB/GOLD 400/HP Cain Road, Bracknell Telephone: (+44) 344 362384 Berks. RG12 1HN. Fax: (+44) 344 362199 United Kingdom Telnet: 316-2384 VoiceMail: 316-2375 .. -... -- .. -... -- .. -... -- .. -... -- .. -... -- .. -... -- Another possible source of help regarding 5250-type terminals would be the mailing list MIDRANGE-L. Note that this list covers a very wide range of topics surrounding the IBM System/36, System/38, and AS/400--not just terminals. Subscribe by sending a request like the following (in the body of the message) subscribe midrange-l to As with any general mailing list, you would be polite to follow the discussion for a few days before crying out for help with terminals. http://publib.boulder.ibm.com/doc_link/en_US/a_doc_lib/3270hcon/hconugd/About_3270Host.htm .. -... -- .. -... -- .. -... -- .. -... -- .. -... -- .. -... -- Newsgroups: comp.os.vms,comp.sys.dec Path: utkcs2!ornl!rsg1.er.usgs.gov!fmsrl7!destroyer!sol.ctr.columbia.edu!usc !sdd.hp.com!elroy.jpl.nasa.gov!decwrl!pa.dec.com!nntpd2.cxo.dec.com !nntpd.lkg.dec.com!smaug.enet.dec.com!garrod Message-ID: <1992Nov1.014156.20925@nntpd.lkg.dec.com> Sender: usenet@nntpd.lkg.dec.com (USENET News System) Organization: Digital Equipment Corporation References: Date: Sun, 1 Nov 1992 00:36:39 GMT From: garrod@smaug.enet.dec.com (Dave Garrod) Subject: Re: vt200 on IBM 3270 terminal In article , FILIPSEN_BMF@jt.dk (Bent Filipsen) writes... > > Hi, > > Is it possible to run VT200 from an IBM 3270 terminal? > Gateway? Emulating? > > And what is the best solution? Digital is shipping a product that does this. It's called the DEC SNA Domain Gateway-CT together with MEP-VT220 for 3270 Application Services The Domain Gateway is channel attached to the IBM system and emulates a PU type 5 ie it looks like another VTAM to the SNA network. MEP-VT220 runs on a VMS system and provides VT220 emulation for 3270 terminals and PCs emulating 3270s in the wide area SNA network. NO application level software is required on the IBM system. The VT emulator appears as a Cross Domain APPL to the SNA network. If you require further information in this product set please feel free to contact me, I'm the engineering manager at Digital responsible for this product set. Regards Dave Garrod .. -... -- .. -... -- .. -... -- .. -... -- .. -... -- .. -... -- Path: utkcs2!memstvx1!ukma!wupost!gumby!destroyer!uunet!portal!cpc Newsgroups: comp.os.vms,comp.sys.dec Subject: Re: vt200 on IBM 3270 terminal Message-ID: <1992Nov1.160724.3692@unix.portal.com> From: cpc@shell.portal.com (Chris Cebelenski) Date: Sun, 1 Nov 1992 16:07:24 GMT Sender: news@unix.portal.com References: Organization: Portal Communications Company--408/973-9111 (voice) 408/725-0561 Nntp-Posting-Host: jobe We're using DEC's Gateway CT and an as-yet unreleased emulator called MEP-RTE. This allows VT200 terminals on standard SNA/3270 terminals. It's NOT 100%, but seems to be useful for limited applications. Chris Cebelenski .. -... -- .. -... -- .. -... -- .. -... -- .. -... -- .. -... -- Newsgroups: comp.os.vms,comp.sys.dec Path: utkcs2!ornl!rsg1.er.usgs.gov!ukma!netsys!decwrl!pa.dec.com !nntpd2.cxo.dec.com!nntpd.lkg.dec.com!smaug.enet.dec.com!garrod Message-ID: <1992Nov4.235810.11757@nntpd.lkg.dec.com> Sender: usenet@nntpd.lkg.dec.com (USENET News System) Organization: Digital Equipment Corporation References: Date: Wed, 4 Nov 1992 22:56:49 GMT From: garrod@smaug.enet.dec.com (Dave Garrod) Subject: Re: vt200 on IBM 3270 terminal In article , FILIPSEN_BMF@jt.dk (Bent Filipsen) writes... > > Is it possible to run VT200 from an IBM 3270 terminal? > Gateway? Emulating? > > And what is the best solution? > > Sincerely > Bent Filipsen > Jutland Telephone Company > e-mail Bent.Filipsen@jt.dk Digital Equipment Corporation offers a solution to this problem. I attach information on the products we offer to address this problem. Integrata MEP-VT220 for 3270 Application Services CONTACT Your local Digital Office HIGHLIGHTS o Provides full-screen access to UNMODIFIED VAX/VMS character cell applications from IBM SNA 3270 Terminals o VT220 Emulation o Utilizes the recently announced DEC SNA Domain Gateway for Channel Transport so that no mainframe application is required. o Available November 1992 Product Description MEP-VT220 for 3270 Application Services is a full-screen VT emulator which works with the recently announced DEC SNA Domain Gateway-CT to provide interactive access to UNMODIFIED VAX/VMS character cell applications. MEP-VT220 is a software product which provides remote 3270 terminals in an IBM SNA environment with the ability to access full screen VAX/VMS applications including ALL-IN-1, existing VMS-based customer applications, and VMS utilities such as VMS Mail. MEP-VT220 for 3270 Application Services is a worldwide Digital Distributed and Supported (DDS) product from Integrata, AG, Tuebingen Germany. It is based on Integrata's MEP-RTE product which has a successful market presence in Europe, and has been ported to use the recently announced DEC SNA Domain Gateway and DEC SNA 3270 Application Services products for communication and connectivity between the Digital and IBM environments. FEATURES o Full-screen access to UNMODIFIED VAX/VMS applications from remote SNA networked 3270 terminals and PCs o Provides VT220 emulation (with video attributes such as reverse video and blinking mapped to bold) o Preserves customer investment in SNA terminal network while permitting flexible multi-vendor integration o Provides customers with a cost effective alternative to expensive mainframe applications development/deployment o Utilizes DEC SNA Domain Gateway. No requirement for IBM mainframe application and expensive CPU cycles o Provides maximum flexibility for remote users without the need for dedicated terminals, special hardware, re-wiring, or parallel network installation and support While most character cell applications can be accessed using MEP-VT220 (including ALL-IN-1) the lack of a block mode editor is a limitation given the inherent differences between VMS character cell and block mode terminals and editors. PRICING/ORDERING INFORMATION Model No. Description U.S. List QL-06CA*-AA MEP-VT220 for 3270 $11,429 Application Services Traditional Base QL-06CAA-3B MEP-VT220 for 3270 1,714 Application Services Concurrent User Users who want the Domain Gateway/Integrata solution should purchase the following: o DESNB-C* Domain Gateway Systems Package (includes DEC ChannelServer II hardware), or QP-LBRAA-01 Domain Gateway Software upgrade package o MEP-VT220 for 3270 Application Services QL-06CA*-AA "Traditional" Capacity, or QL-06CAA-3B Concurrent User o DEC SNA 3270 Application Services Runtime license QL-MKKA9-J* "Clusterwide" Capacity, or QL-MKKAA-3B Concurrent User RESOURCES Please refer to the following for additional information about the 3270 connectivity product family: - Software Product Description 44.00.xx ---------- IBM is a registered trademark of International Business Machines Corporation .. -... -- .. -... -- .. -... -- .. -... -- .. -... -- .. -... -- Newsgroups: alt.folklore.computers Path: cs.utk.edu!emory!swrinde!ihnp4.ucsd.edu!agate!blanket.mitre.org !linus.mitre.org!newsflash.mitre.org!mwunix!jcmorris Organization: The MITRE Corporation Message-ID: References: <2ppjf1$mbs@juniper..almaden.ibm.com> <1994Apr30.150315.27131@mnemosyne.cs.du.edu> Date: 5 May 94 12:57:38 GMT From: jcmorris@mwunix.mitre.org (Joe Morris) Subject: Re: Ahhh EBCDIC - remember that? bernie@fantasyfarm.com (Bernie Cosell) writes: >Amen. I haven't looked at the state of EBCDIC recently [like in the last >20 or so years] but it used to be barely-close. Does EBCDIC still have >things like record-mark, "lozenge", circle-C and circle-D and other strange >goodies? [Those last two, circle-c and circle-D, were the bane of my >life for a while while I was trying to implement a 2741->ASCIIuniverse >terminal driver back in the early days of the ARPAnet]. Is that really a fair complaint? The 2741 wasn't an EBCDIC device. It was closer in concept to a 5-level Baudot terminal: it used a 6-bit code with explicit upshift/downshift commands. The closest it came to the EBCDIC universe was the option to have a keyboard (and typeball) which delivered the EBCDIC glyphs and used the keyboard layout of other, EBCDIC-oriented terminals, but the line code certainly wasn't EBCDIC. Circle-C was documented to represent the EOT control code, which is well-defined in both the ASCII and EBCDIC universes. Circle-D was EOA (end-of-address) which is defined in neither, so there wasn't any "standard" way of representing it if you were forced to map the 274x line code to either ASCII or EBCDIC. Users who joined the computer scene in recent years never had the dubious pleasure of trying to work with networks where each vendor used a different code, and many terminals from a single vendor would use incompatible codes. One of the best (worst?) examples of what this could result in was the Coursewriter (early Computer-Aided Instruction) product which was originally written to run on the IBM 1401, using IBM 1050 terminals. When it was later ported to run under DOS/360 and later OS/360, and updated to use terminals popular in the late 1960s the porting team found that so much of the original logic was dependent on the data base having the 1050 line codes that they left it that way: no matter what terminal type (ASCII, 274x, or whatever) you were using the data was translated into 1050 line code as it arrived. This made all sorts of problems when someone had to go into the system to maintain or modify it...and I speak from bitter experience. Oh yes...the other characters such as "lozenge" and "record mark" were never to my knowledge officially assigned code points in the EBCDIC address space, a fate shared by several other glyphs which were defined in BCD. The card codes for lozenge (12-4-8) and record mark (0-2-8) mapped to EBCDIC "<" and "\" respectively. Joe Morris / MITRE .. -... -- .. -... -- .. -... -- .. -... -- .. -... -- .. -... -- Newsgroups: alt.folklore.computers Path: cs.utk.edu!emory!swrinde!ihnp4.ucsd.edu!agate!blanket.mitre.org !linus.mitre.org!newsflash.mitre.org!mwunix!jcmorris Date: 6 May 94 12:02:48 GMT Organization: The MITRE Corporation Message-ID: References: <1285@chac.win.net> <2q38qa$un2@inca.gate.net> <1994May2.162525.15818@jarvis.csri.toronto.edu> NNTP-Posting-Host: mwunix.mitre.org From: jcmorris@mwunix.mitre.org (Joe Morris) Subject: Re: Ahhh EBCDIC - remember that? alderson@netcom.com (Richard M. Alderson III) writes: >In any case, the peculiar placement of brackets and braces is as nothing when >compared to the treatment of the EBCDIC cent sign (for non-US folks, it's that >lowercase "c" with a slash or vertical line through it which is used for >monetary amounts less than $1.00--$0.02 doesn't really cut it :-). Am I missing something? What is so peculiar about the assignment of hex 4A as the codepoint for the cent sign in EBCDIC? It falls quite comfortably into the pattern that assigns the uppercase alpha characters, decimal digits, and common special characters in the address space in a way that allows them to be uniquely identified by the low six bits. And in any case the cent sign isn't directly translatable to ASCII, since it doesn't appear in the set of 7-bit ASCII glyphs. Joe Morris / MITRE .. -... -- .. -... -- .. -... -- .. -... -- .. -... -- .. -... -- Path: cs.utk.edu!stc06.CTD.ORNL.GOV!fnnews.fnal.gov!uwm.edu!vixen.cso.uiuc.edu !howland.reston.ans.net!gatech!newsfeed.pitt.edu!uunet!news.uiowa.edu!news Newsgroups: alt.os.multics,alt.folklore.computers Date: 11 Jan 1995 15:10:12 GMT Organization: University of Iowa, Iowa City, IA, USA Message-ID: <3f0sck$ke2@nexus.uiowa.edu> References: <199501101927.AA11327@world.std.com> NNTP-Posting-Host: pyrite.cs.uiowa.edu From: jones@pyrite.cs.uiowa.edu (Douglas W. Jones,201H MLH,3193350740,3193382879) Subject: Re: old mainframes & text processing From article <199501101927.AA11327@world.std.com>, by Bob_Frankston@frankston.com: > Doug Jones asked "How many others out there have had to write software (text > editors, formatters, or other supporting stuff) so they could edit or print > their thesis?" > > There were two character sets for the IBM Selectric terminals. Actually, > character set is the wrong term, they were rotate and tilt codes for the > typeball. IBM had a device called the mag card selectric which would print > off a magnetic strip. There was also the communicating Mag Card Selectric > that used a different, "Correspondence", code from Multics. The Communicating Mag Card Selectric used "Word Processing EBCDIC" a code I know all too well. It had a passing similarity to EBCDIC, except that, at any instant, only 88 printing characters were defined, depending on what "logical typeball" was mounted. Thus, there weren't just two character sets, there were as many different IBM WP EBCDIC character sets as there were typeballs! Furthermore, WP EBCDIC was translated to tilt-rotate codes by the communicating selectric as it recorded what it received on mag cards. This was adding insult to injury if your eventual print device wasn't the selectric. I used the damned machine to write my thesis to mag cards, then I carried the mag cards downstairs to the word processing center where they had this great ink jet printer. The ink jet printer had a nice big internal character set, but you had to tell it what logical typeball to emulate when it read from the mag cards. Gack! But, the output was very nice looking. > At some point, I wrote code to translate the Multics output so I could > create the mag card and print in the proportional font. This is just what I did with PLATO's internal 6 bit code. I wrote the translation and protocol conversion code on the internal microprocessor of a PLATO smart terminal (the second generation PLATO IV terminal, with plasma display panel and an internal Intel 8080). I had to unpack strings of incoming 21 bit PLATO frames (3 six-bit characters, plus parity, framing and whatnot), interpret shift codes, translate to IBM WP EBCDIC, do some macro substitution to make overstrike sequences to fake up some Plato characters, and handle the BISYNC protocol demanded by the RS232 link to the communicating selectric. I still have the code, somewhere in my office. Flow control was not a minor issue at 60 21 bit PLATO frames per second incoming and 1200 baud outgoing. Doug Jones jones@cs.uiowa.edu .. -... -- .. -... -- .. -... -- .. -... -- .. -... -- .. -... -- Newsgroups: comp.dcom.telecom Path: cs.utk.edu!martha.utk.edu!darwin.sura.net!news.larc.nasa.gov !lll-winken.llnl.gov!telecom-request X-Administrivia-To: telecom-request@eecs.nwu.edu X-Telecom-Digest: Volume 15, Issue 27, Message 7 of 14 X-Submissions-To: telecom@eecs.nwu.edu Message-ID: Organization: Tansin A. Darcos & Company, Silver Spring, MD USA Date: Wed, 11 Jan 1995 14:08:00 EST From: Paul Robinson Subject: Re: Need an EBCDIC Spec The specifications for EBCDIC are available in many IBM 370 Mainframe publications, as IBM is the inventor and primary user for the EBCDIC specification. If you have a college near you that has an IBM or equivalent mainframe and still teaches assembly language, their book store should have the Gold Card, which is a small pamphlet listing various assembly instructions and a list of the EBCDIC character set and the translation of the equivalent to ASCII. Most programming manuals for IBM Mainframe languages will include it, and if your local university or college has books on IBM mainframe themes, one of them will probably include a listing of the translation table, which is in the public domain. You may also want to take a look at: ftp://ds.internic.net/rfc/rfc1345.txt Which contains a list of many character sets, and will probably include the listing for EBCDIC. .. -... -- .. -... -- .. -... -- .. -... -- .. -... -- .. -... -- Entry from the "Multics Glossary" http://www.best.com/~thvv/mgs.html Selectric : IBM mechanism for typewriters and computer terminals. The character set was on a plastic "golf ball" that struck the paper through a typewriter ribbon. This mechanism was used in the IBM 1050 and 2741 terminals, in various third-party terminals, and in the console typewriter of the 6180. There were three grades of this mechanism OEM'd by IBM, light, medium, and heavy duty; the 6180 console had a heavy duty model. .. -... -- .. -... -- .. -... -- .. -... -- .. -... -- .. -... -- Newsgroups: comp.terminals Path: cs.utk.edu!gatech!swrinde!ihnp4.ucsd.edu!ames!cronkite.cisco.com !mojo.cisco.com!user Followup-To: comp.terminals,comp.protocols.tcp-ip, comp.os.ms-windows.networking.tcp-ip,bit.listserv.ibmtcp-l Date: 16 Jun 1994 17:28:39 GMT Organization: cisco Systems Message-ID: References: <2tpu1q$aps@nwfocus.wa.com> NNTP-Posting-Host: mojo.cisco.com From: ganzhorn@cisco.com (Charles Ganzhorn) Subject: Re: IBM 3270 emulation using TCP? In article <2tpu1q$aps@nwfocus.wa.com>, wadeware@halcyon.com wrote: > > I know nothing about 3270 emulation but that is seems way too complicated. > Our small network in Seattle wants to do 3270 emulation with a machine in > Dallas. > > Question: Using a leased line and routers, is there a way to achieve that > connection and emulation using TCP/IP rather than installing a 3270 > controller board in our server and connecting that way? I use Chameleon > TCP/IP for Windows, and see that it has a 3270 emulator....that where I > got this bright idea.... With a combination of tn3270 on the client and some sort of TCP/IP stack/gateway to the mainframe, you can do this via IP. The functionality of the 3270 screen depends greatly on whose tn3270 you use. Evaluate the 3270 functions that are necessary for you and use that as your guide to picking the tn3270. If all you really care about is getting a screen, that's relatively easy. Local printing tends to be a bit harder. Charles. -- Charles Ganzhorn Email: ganzhorn@cisco.com Consulting Engineer cisco Systems One Appletree Square, Suite 1452 8009 34th Avenue South Phone: 612-851-8310 Bloomington, MN 55425 FAX: 612-851-8311 .. -... -- .. -... -- .. -... -- .. -... -- .. -... -- .. -... -- Newsgroups: comp.terminals,comp.protocols.tcp-ip, comp.os.ms-windows.networking.tcp-ip,bit.listserv.ibmtcp-l Path: cs.utk.edu!news.msfc.nasa.gov!europa.eng.gtefsd.com!MathWorks.Com !transfer.stratus.com!xylogics.com!Xylogics.COM!carlson From: carlson@Xylogics.COM (James Carlson) Subject: Re: IBM 3270 emulation using TCP? Date: 16 Jun 1994 17:38:12 GMT Organization: Xylogics Incorporated Lines: 23 Distribution: world Message-ID: <2tq2m4$e7t@newhub.xylogics.com> References: <2tpu1q$aps@nwfocus.wa.com> Reply-To: carlson@xylogics.com NNTP-Posting-Host: steam.xylogics.com Xref: cs.utk.edu comp.terminals:2497 comp.protocols.tcp-ip:29081 comp.os.ms-windows.networking.tcp-ip:1488 In article <2tpu1q$aps@nwfocus.wa.com>, wadeware@halcyon.com writes: |> |> I know nothing about 3270 emulation but that is seems way too complicated. |> |> Our small network in Seattle wants to do 3270 emulation with a machine in |> Dallas. |> |> Question: Using a leased line and routers, is there a way to achieve that |> connection and emulation using TCP/IP rather than installing a 3270 |> controller board in our server and connecting that way? I use |> Chameleon TCP/IP for Windows, and see that it has a 3270 emulator |> ....that where I got this bright idea.... Sure -- it's called tn3270. You'll need to have a protocol converter somewhere at the other site which knows tn3270 protocol. Give OpenConnect Systems, Inc, a call at +1 214 484-5200 or drop them a line at support@oc.com. -- James Carlson Tel: +1 617 272 8140 Annex Software Support / Xylogics, Inc. +1 800 225 3317 53 Third Avenue / Burlington MA 01803-4491 Fax: +1 617 272 2618 .. -... -- .. -... -- .. -... -- .. -... -- .. -... -- .. -... -- Path: cs.utk.edu!emory!swrinde!howland.reston.ans.net!europa.eng.gtefsd.com !MathWorks.Com!transfer.stratus.com!xylogics.com!Xylogics.COM!carlson From: carlson@Xylogics.COM (James Carlson) Newsgroups: comp.terminals Subject: Re: vt100 numeric keypad keys? Date: 14 Jul 1994 19:36:10 GMT Organization: Xylogics Incorporated Lines: 56 Distribution: world Message-ID: <30443a$1s2@newhub.xylogics.com> References: Reply-To: carlson@xylogics.com NNTP-Posting-Host: steam.xylogics.com Keywords: keypad numeric In article , filthaut@tanger.hef.kun.nl (Frank Filthaut) writes: |> Hi, |> |> I'm using the `tn3270' program to logon to the IBM from my |> vt100 terminal. I get most things ok, except that I would like |> very much to use the whole of my numeric keypad. I know that I |> can access the 1-9 keys using ^[Oq - ^[Oy sequences. But can |> anyone anyone tell me how to address the other keys over there? |> I'm no expert on the use of terminals (but nevertheless interested), |> so excuse me if this is a stupid question. Address those keys? I'm not sure what you mean. These keys can send the following codes (depending on mode): Key Keypad-Numeric-Mode Keypad-Application-Mode VT52 0 0 SS3 p ESC ? p 1 1 SS3 q ESC ? q 2 2 SS3 r ESC ? r 3 3 SS3 s ESC ? s 4 4 SS3 t ESC ? t 5 5 SS3 u ESC ? u 6 6 SS3 v ESC ? v 7 7 SS3 w ESC ? w 8 8 SS3 x ESC ? x 9 9 SS3 y ESC ? y - - SS3 m ESC ? m , , SS3 l ESC ? l . . SS3 n ESC ? n Enter CR SS3 M ESC ? M And the PF keys: Key VT100 Mode VT52 Mode PF1 SS3 P ESC P PF2 SS3 Q ESC Q PF3 SS3 R ESC R PF4 SS3 S ESC S (Where SS3 is ESC O (1B 4F) in 7 bit mode or 8F in 8 bit mode.) To map those keys into your tn3270 session, you'd edit the /etc/map3270 file on your system and set up entries for each key. For example, you might start of by doing this: # Assign IBM PF1 on keypad '1' key. pfk1 = '\EOq'; # Assign IBM PA1 on DEC's PF1 key. pa1 = '\EOP'; The mapping is arbitrary ... and I don't know how to guess at what mapping you'd like to have ... -- James Carlson Tel: +1 617 272 8140 Annex Software Support / Xylogics, Inc. +1 800 225 3317 53 Third Avenue / Burlington MA 01803-4491 Fax: +1 617 272 2618 .. -... -- .. -... -- .. -... -- .. -... -- .. -... -- .. -... -- Path: cs.utk.edu!gatech!newsfeed.pitt.edu!uunet!newstf01.cr1.aol.com !newsbf01.news.aol.com!not-for-mail From: ronzayas@aol.com (RonZAYAS) Newsgroups: comp.terminals Subject: Re: AS/400 Terminal Emulation Date: 18 Oct 1994 20:58:06 -0400 Organization: America Online, Inc. (1-800-827-6364) Lines: 12 Sender: news@newsbf01.news.aol.com Message-ID: <381quu$ec9@newsbf01.news.aol.com> References: <380c28$87u@gulfa.ods.gulfnet.kw> NNTP-Posting-Host: newsbf01.news.aol.com In article <380c28$87u@gulfa.ods.gulfnet.kw>, marting@access.ods.gulfnet.kw writes: <> There are two companies that I know of in addition to Irma and Rumba. They are Andrew Corp which makes the Emerald emulator (800-776-6174) and WRQ (Walker, Richer, Quinn) that has a 5250 emulator (800-872-2829). Hope this helps, Ron .. -... -- .. -... -- .. -... -- .. -... -- .. -... -- .. -... -- DCA (Digital Communications Associates) merged with Attachmate Corporation in December 1994. DCA products such as IRMA3, IRMAkey/3270, MacIRMA, INTERCOM, and Crosstalk are now sold by Attachmate. Attachmate Corporation 3617 131st Ave SE Bellevue, WA 98006 voice: +1 206/644-4010 WATS: 800/426-6283 URL: http://www.atm.com/ .. -... -- .. -... -- .. -... -- .. -... -- .. -... -- .. -... -- Path: cs.utk.edu!stc06.CTD.ORNL.GOV!fnnews.fnal.gov!mp.cs.niu.edu !vixen.cso.uiuc.edu!howland.reston.ans.net!EU.net!Germany.EU.net !Munich.Germany.EU.net!ecrc!news.forth.gr!csi.forth.gr!dialhawk!ace From: ace@ics.forth.gr (Andreas C. Enotiadis) Newsgroups: comp.terminals Subject: Re: AS/400 Terminal Emulation Date: Mon, 24 Oct 1994 17:47:43 Organization: InterNetwork Ltd Lines: 32 Message-ID: References: <380c28$87u@gulfa.ods.gulfnet.kw> <381quu$ec9@newsbf01.news.aol.com> NNTP-Posting-Host: dialhawk.praxis.forth.gr X-Newsreader: Trumpet for Windows [Version 1.0 Rev B] In article <381quu$ec9@newsbf01.news.aol.com>, ronzayas@aol.com (RonZAYAS) writes: > >Subject: Re: AS/400 Terminal Emulation >Date: 18 Oct 1994 20:58:06 -0400 >In article <380c28$87u@gulfa.ods.gulfnet.kw>, >marting@access.ods.gulfnet.kw writes: ><emulation under windows?>> >There are two companies that I know of in addition to Irma and Rumba. >They are Andrew Corp which makes the Emerald emulator (800-776-6174) and >WRQ that has a 5250 emulator (800-872-2829). >Hope this helps, >Ron Turbosoft Australia also has an excellent (and cheaper) product for 5250. Contact matt@atlas.abccomp.oz.au for details [or http://www.ttwin.com/] eard however that the AS-400 is difficult to emulate b'coz of the HW solutions. > You want a cheap (~ $35) ASCII serial terminal (synch or async; RS423 or > RS232, or RS422/RS485). Almost any terminal that was *NOT* connected to > an IBM or clone (Amdahl, Sperry/Univac, Wang, etc) will work just fine. > In fact, I saw a set of old Data General terminals and Prime Computer > terminals the other day that would work just fine (stick them in TVI-925 > or TVI-950 mode and any UNIX won't have a problem). I C. Think IBM used to have a pretty big market share over here in Norway though, so it's just so much easier to come by IBM terminals. I think most terminals just end up in the trash, doesn't seem to be a marked for them over here. However I would love to have a couple connected to my linux box. Any one from Norway know where to get terminals that work with linux? Regards, Per ////////////////////////////////////////////////////////////////////////////// Newsgroups: comp.sys.ibm.as400.misc Organization: T-Online Date: Wed, 11 Aug 1999 20:23:28 +0200 From: Timothy Subject: Telnet to UNIX/LINUX Hi, if i start a telnet session between my AS/400 and a LINUX or a AIX, the password is visible during login. I can surpress it by typing PF6 before entering the password but How to it by default? I dont have the problem when i start a telnet from a PC or other Linux. Thanks in advance Timothy ////////////////////////////////////////////////////////////////////////////// Newsgroups: comp.sys.ibm.as400.misc References: <7osehj$6v4$1@news00.btx.dtag.de> Message-ID: <37B1EFE5.61199EDE@ods.ods.net> Date: Wed, 11 Aug 1999 16:49:26 -0500 From: Scott Klement Subject: Re: Telnet to UNIX/LINUX > if i start a telnet session between my AS/400 and a LINUX or a AIX, > the password is visible during login. > I can suppress it by typing PF6 before > entering the password but How to it by default? AFAIK, you can't. The AS/400 deals with terminal I/O as a full screen at a time. This differs from PCs and unix machines in that they tend to do terminal I/O one character at a time. When you telnet to your unix box from a PC, every time you hit a key, its immediately transmitted to the unix box, which then sends back a character to print on your screen. [Archiver's note: this is called "full-duplex" operation.] Therefore, when you type a password, it simply doesn't send anything back. The AS/400, however doesn't receive control from the terminal until you press enter, (or one of the various function keys.) To simulate a telnet session, the terminal echos back all the characters that you type. When you press enter, it sends them to the AS/400. The AS/400 then sends them to the unix box, and receives characters echoed back. It will then replace what you typed with what was echoed. Since the echo from the unix box isnt happening as you type, it can't selectively suppress echoing of the password characters. (I hope that explanation makes sense!?) At any rate, I'd suggest just trying to use F6, this tells the AS/400 not to echo back anything until you hit enter again. Its an extra keypress, and a bit hard to remember, but its about the best you can do :) You'll probably notice other issues, also relating to the screen-at-a-time thing. They include things like using the cursor keys to control movement in a text editor, or trying to get the last command you typed by hitting the up arrow, etc. If you've got the option, I'd suggest using a PC to telnet to your unix box :-) HTH ////////////////////////////////////////////////////////////////////////////// Newsgroups: comp.terminals Message-ID: <01bf2326$fba22460$6ea8a8c0@shiner> References: <37F9B09F.272004C1@lhpca.univ-lyon1.fr> <3NbS3.5$oa2.1011@iad-read.news.verio.net> <3819492D.DA3CDBF2@lhpca.univ-lyon1.fr> NNTP-Posting-Host: 12.73.225.53 Date: 30 Oct 1999 20:38:42 GMT Organization: AT&T WorldNet Services From: Larry Liska Subject: Re: IBM 3192 termnials under Linux OS Sylvain Oranger wrote in article <3819492D.DA3CDBF2@lhpca.univ-lyon1.fr>... > > jrwz@servtech.com wrote: > > ... > > Won't work. .... unless you want to go to great expense to buy a remote > > controller device that is capable of bridging the two..... > > ...... > > OK, i've to buy the remote controller 3174. If you buy a 3174 controller, you are still most likely throwing good money down the toilet. If you do go ahead and buy one, then at least make sure that they throw in the control program microcode disc(s)-- without the microcode you are utterly, absolutely screwed. IBM used to license the microcode seperately from the physical 3174 box, and I assume that they still do...? The terminals are only capable of talking to a terminal controller. But, once you get a terminal controller, you are then faced with the issue of getting your Linux box to talk to the terminal controller. This is definitely more doable then talking to the terminals directly, but, that is like saying it is comparatively "easier" to climb Mount St. Helens with a broken leg versus climbing Mt. Everest with a broken leg.... My knowledge of Coax and local attachment is pretty limited. So, I'll discuss SNA attachment here. (Perhaps the non-SNA terminal rules are simpler, but then you also have to deal with esoteric coax hardware that isn't documented. SNA is terribly complex, but at least you can find reasonably complete documentation for it.) Under traditional SNA, the relationships between devices are very rigid. Terminals are total slaves. The session is owned and driven by the host. IBM terminals are not independent and autonomous like a VT-320 is. So, to stand a chance to get an IBM terminal to talk to a Linux box, you would need some device or driver package on the Linux side that would fill the role played by the mainframe. On top of that, SNA is just the networking protocol. The terminal speaks 3270 datastream protocol, which is totally alien to ASCII-based async hosts. So there are two challenges: writing an SNA emulator device driver just so Linux and the terminal controller/terminal can talk, and then writing an async-to-3270DS protocol converter so that the conversation will be meaningful. That said, you might want to take a look at Black Box. (http://www.blackbox.com/). They sell lots of interface equipment, including protocol converters, etc. They might have a device that will work for you. (They didn't a few years ago, but maybe they do now.) It has been a few years since I've played with IBM terminals, SNA, etc. So, be forewarned :-) Here is what I remember about good old traditional SNA: The mainframe runs a program called "VTAM", virtual telecommunications access method. Logically, this is called a "PU 5" device. The front end processor runs NCP, "Network control program", and is logically a "PU 4" device. The 3174 terminal controllers run "3174 Control Program" and are "PU 2" devices. The terminals themselves didn't count-- they were just IDs in the terminal controller. Initially, before you log in, VTAM "owns" your terminal. Depending on how VTAM is configured, you can have it automatically put you into an application, or you can have it show you a chooser menu. Note that, from VTAM's point of view, OS's are applications. On the system I used to work on, we would be taked straight into the VM/CMS "application" on connect. So when you go into an app, VTAM gives up the connection and hands it off to the new app. But, VTAM is still controlling data flow. Here is an attempt at ASCII art: ================[ cut here ]================ (view this in a Courier font) Apps <======> VTAM (PU5) | Mainframe ----------------------------------- | rest of the world NCP (PU4) / \ / \ / \ 3174 3174 terminal controllers (PU2) /|\ /|\ / | \ / | \ X X X X X X terminals ================[ cut here ]================ On some mainframes, like the 9370 that I used to work with, the NCP component was bundled in, and was not a seperate device. Despite all of the above, I think that trying to write a "mainframe emulator" might be an amusing project-- but it would also be extremely challenging and time consuming. There are about 10 different manuals you will need to buy from IBM. I don't have numbers handy, but here are some names: 3270 Datastream Programmer's Ref, SNA Formats Ref, 3174 Functional Specification, various VTAM references, Sessions Between Logical Units (talks about session initiation and management), SDLC Protocol Ref (or Fundamentals?) Regards, lcl ////////////////////////////////////////////////////////////////////////////// Message-ID: <01bf2329$30124c00$6ea8a8c0@shiner> References: Date: 30 Oct 1999 20:54:28 GMT Organization: AT&T WorldNet Services Newsgroups: comp.terminals From: Larry Liska Subject: Re: Looking for 5250 telnet protocol Carl Huff wrote in article ... > I am looking for a document describing the 5250 protocol that rides on > telnet. > ..... You need a bunch of RFCs, then you may also need some IBM documents, if you don't already have access to them. RFCs: ========= RFC1205 5250 Telnet Interface **** this is the major one RFC854 Telnet Protocol Specification RFC855 Telnet Option Specifications RFC856 Telnet Binary Transmission RFC1091 Telnet Terminal-Type Option RFC885 Telnet End of Record Option Also, I think there are some newer RFCs out, such as "5250E", and "5250 Current Practices." You may want to look around a good RFC index. IBM Docs: ============ SA21-9247-6 IBM 5250 Information Display System Functions Reference Manual You will also want to get the Functional Specifications manuals for at least one of these: 5294/5394/5494 Controllers. Sorry, I forget the document numbers for these. Also, you might want to get some AS/400 documentation, like info on "DSPT" (display station passthrough), and other terminal-related stuff. Hope this helps, lcl ////////////////////////////////////////////////////////////////////////////// ARCHIVER'S NOTE: Other notes on using "telnet" with IBM terminal types are stored behind the "telnet news" link. ////////////////////////////////////////////////////////////////////////////// Newsgroups: alt.folklore.computers Date: Wed, 26 Sep 2001 13:04:47 GMT Message-ID: References: From: Jay Maynard Subject: Re: 3270 protocol On Wed, 26 Sep 2001 11:31:32 GMT, Anne & Lynn Wheeler wrote: > >a couple days ago somebody sent me a question on 3270 protocol which I >didn't know the answer to ... but it jog'ed some memory cells that >i've been trying to remember what were the terms used for the >3272/3277 and 3274/3278/9 protocol. I have some vague recollection >that one of the terms was CUT and may refer to the 3272/3277 protocol >... but I can't remember the other term (and/or even sure CUT is one >of the terms). Close. CUT is the protocol used by dumb 327x devices (control-unit terminal). I don't recall that the 3277 version was called anything besides "type A" (referring to the customization parameter you had to set; when customizing a 3274, you set up a number of type A ports for 3277s, and a number of type B ports for 3278s). (Or maybe it's the other way around.) The other protocol you're thinking of is DFT (distributed function terminal), which offloaded a bunch of intelligence to the downstream device. The device's microcode was downloaded from storage in the controller. The first such device was the 3290. ////////////////////////////////////////////////////////////////////////////// Message-ID: <0241ADC16FCB3D46BAE4C694A161CC57EE872A@DISmxOLY02> Date: Mon, 25 Feb 2002 17:34:53 -0800 From: Chuck Moore Hercules ( http://www.conmicro.cx/hercules/ ) is software that emulates either an IBM 370, 390, or ESA machine instruction set. It allows a person to run an IBM operating system (e.g. MVS, VM, OSVS1, VSE) within a Windows or Linux host. The Hercules emulator takes an incoming telnet for a specific port defined in the Hercules configuration/startup file (say, for example, a telnet to port 9876) and maps it to a device number that is defined (for example, 009 or 01F) to the specific IBM operating system. From an IBM operating system, the incoming device is directly attached. ////////////////////////////////////////////////////////////////////////////// Newsgroups: comp.terminals References: <3CDD891C.F661539@yahoo.com> Message-ID: Organization: Mayday Technology Ltd Date: Sun, 12 May 2002 08:10:07 +0100 From: Robert de Bath Subject: Re: IBM 3192 help On Sat, 11 May 2002, Sark wrote: > I just recieved a very nice looking IBM 3192 terminal. I want to connect > it to my Linux box. Can anyone tell me how I can connect this? it has a > coax port on the back, as well as what appears to be a serial or printer > port. Any help would be appreciated. In short, you can't. In long: 1) The terminal uses block mode, Unix (and therefor Linux) uses character mode as a rule. 2) The coax port is an IBM SNA style connection. To use it with a PC you would need a _very_ expensive interface card that probably doesn't have a Linux driver anyway. -- Rob. (Robert de Bath ) ////////////////////////////////////////////////////////////////////////////// Newsgroups: comp.terminals References: <3CDD891C.F661539@yahoo.com> <25be338e69c48092@mayday.cix.co.uk> Message-ID: Organization: Harvard University, Cambridge, Massachusetts Date: 14 May 2002 13:00:05 GMT From: Chip Coldwell Subject: Re: IBM 3192 help Robert de Bath wrote: > On 14 May 2002, Chip Coldwell wrote: >> >> Are the coax/twinax interface running some SDLC variant (I realize >> that they probably run either the 3270 or 5250 protocols higher up in >> the protocol stack)? > > Sorry I've only _used_ these systems, File transfers between an AIX 6150 > and a mainframe and typing stuff into a 3270 (ish) terminal from a crib > sheet. I never wanted to understand it, as the so-called mainframe was > on its way out 'cause its annual maintenance was more than the price > of an equivalent Unix machine (Sequent IIRC). > > You might try: > http://www.planetmvs.com/links/linux.html > http://www.linux-sna.org/ Thanks for the links. I also found http://www.gcom.com/home/products/gpk.html which appears to be hardware+software solution for connecting a Linux box to an SNA network (a commercial alternative to this? http://www.linux-sna.org/ I'm guessing that the hardware is nothing more or less than a synchronous serial card (such as the Cyclades PC300) that you can run SDLC/HDLC on top of. Their page at http://www.gcom.com/home/products/sna_spec.html also had some tantalizing hints. I wonder if one bought a 3174 controller (although IBM has dropped this line, there's usually a couple for sale on eBay at any given time), could it be connected to a Linux box via SDLC running on one of these synchronous serial cards? That would solve the original poster's problem: plug the 3192 into the 3174, and the 3174 into the Linux box, which would have to run both SDLC and some higher-level SNA protocols. A great hack would be to set all this up and then run Hercules (http://www.conmicro.cx/hercules/) on the Linux box to emulate an S/370. So you would have a real, hardware IBM 3270-class terminal talking to an emulated mainframe .... BTW, I read in Andrew Tanenbaum's computer networks book that the origin of the seven layer OSI reference model for networks is, in fact, IBM SNA. Apparently there's approximately a one-to-one correspondence between the OSI layers and SNA layers. No such correspondence can be claimed for TCP/IP, even though that's the context in which you normally run into that OSI model. -- Charles M. "Chip" Coldwell "Turn on, log in, tune out" ////////////////////////////////////////////////////////////////////////////// From coldwell@frank.harvard.invalid Thu May 16 17:20:19 2002 Newsgroups: comp.terminals References: <3CDD891C.F661539@yahoo.com> Message-ID: Organization: Harvard University, Cambridge, Massachusetts Date: 14 May 2002 00:25:23 GMT From: Chip Coldwell Subject: Re: IBM 3192 help Robert de Bath wrote: > 2) The coax port is an IBM SNA style connection. To use it with a PC > you would need a _very_ expensive interface card that probably > doesn't have a Linux driver anyway. I have some questions regarding the protocol run on these coax/twinax interfaces that maybe some of the mainframe gurus reading this NG can answer. From some reading I've been doing lately, I get the impression that HDLC was derived from SDLC which was one of the common link-layer protocols run on SNA. Is this correct? Are the coax/twinax interface running some SDLC variant (I realize that they probably run either the 3270 or 5250 protocols higher up in the protocol stack)? Chip -- Charles M. "Chip" Coldwell "Turn on, log in, tune out" ////////////////////////////////////////////////////////////////////////////// From rd11821@mayday.cix.co.uk Thu May 16 17:20:25 2002 Newsgroups: comp.terminals References: <3CDD891C.F661539@yahoo.com> Message-ID: <25be338e69c48092@mayday.cix.co.uk> Organization: Mayday Technology Ltd Date: Tue, 14 May 2002 09:18:09 +0100 From: Robert de Bath Subject: Re: IBM 3192 help On 14 May 2002, Chip Coldwell wrote: > Robert de Bath wrote: > > 2) The coax port is an IBM SNA style connection. To use it with a PC > > you would need a _very_ expensive interface card that probably > > doesn't have a Linux driver anyway. > > I have some questions regarding the protocol run on these coax/twinax > interfaces that maybe some of the mainframe gurus reading this NG can > answer. > > From some reading I've been doing lately, I get the impression that > HDLC was derived from SDLC which was one of the common link-layer > protocols run on SNA. Is this correct? > > Are the coax/twinax interface running some SDLC variant (I realize > that they probably run either the 3270 or 5250 protocols higher up in > the protocol stack)? Sorry I've only _used_ these systems, File transfers between an AIX 6150 and a mainframe and typing stuff into a 3270 (ish) terminal from a crib sheet. I never wanted to understand it as the so called mainframe was on it's way out 'cause it's annual maintenance was more than the price of an equlivent Unix machine (Sequent IIRC). You might try: http://www.planetmvs.com/links/linux.html http://www.linux-sna.org/ There's also a lot of stuff over at www.ibm.com of course. -- Rob. (Robert de Bath ) ////////////////////////////////////////////////////////////////////////////// 2005-07-09 AIX-centric documentation "Host Connection Program (HCON) Facility" http://publib.boulder.ibm.com/doc_link/en_US/a_doc_lib/3270hcon/hconugd/About_3270Host.htm ////////////////////////////////////////////////////////////////////////////// From coldwell@frank.harvard.invalid Thu May 16 17:20:31 2002 Newsgroups: comp.terminals References: <3CDD891C.F661539@yahoo.com> <25be338e69c48092@mayday.cix.co.uk> Message-ID: Organization: Harvard University, Cambridge, Massachusetts Date: 14 May 2002 13:00:05 GMT From: Chip Coldwell Subject: Re: IBM 3192 help Thanks for the links. I also found http://www.gcom.com/home/products/gpk.html which appears to be hardware+software solution for connecting a Linux box to an SNA network (a commercial alternative to www.linux-sna.org?). I'm guessing that the hardware is nothing more or less than a synchronous serial card (such as the Cyclades PC300) that you can run SDLC/HDLC on top of. Their page at http://www.gcom.com/home/products/sna_spec.html also had some tantalizing hints. I wonder if one bought a 3174 controller (although IBM has dropped this line, there's usually a couple for sale on eBay at any given time), could it be connected to a Linux box via SDLC running on one of these synchronous serial cards? That would solve the original poster's problem: plug the 3192 into the 3174, and the 3174 into the Linux box, which would have to run both SDLC and some higher-level SNA protocols. A great hack would be to set all this up and then run Hercules (http://www.conmicro.cx/hercules/) on the Linux box to emulate an S/370. So you would have a real, hardware IBM 3270-class terminal talking to an emulated mainframe .... BTW, I read in Andrew Tanenbaum's computer networks book that the origin of the seven layer OSI reference model for networks is, in fact, IBM SNA. Apparently there's approximately a one-to-one correspondence between the OSI layers and SNA layers. No such correspondence can be claimed for TCP/IP, even though that's the context in which you normally run into that OSI model. -- Charles M. "Chip" Coldwell "Turn on, log in, tune out" ////////////////////////////////////////////////////////////////////////////// From larryliska@worldnet.att.net Thu May 16 17:20:38 2002 Newsgroups: comp.terminals References: <3CDD891C.F661539@yahoo.com> <25be338e69c48092@mayday.cix.co.uk> Message-ID: Organization: AT&T Worldnet Date: Thu, 16 May 2002 05:00:41 GMT From: Larry Liska Subject: Re: IBM 3192 help Chip Coldwell wrote in message news:abr1ol$t5t$1@news.fas.harvard.edu... > Thanks for the links. I also found > > http://www.gcom.com/home/products/gpk.html > > which appears to be hardware+software solution for connecting a Linux > box to an SNA network ....... > ..... Interesting site but I don't think that will help--then again maybe it will. Before going any further let me issue a disclaimer: it has been many years since I've played with IBM stuff so read the following with the appropriate amount of skepticism... Here is a partial post from the Hercules listserv archives: =======[ cut here ]======= Message: 8 Date: Wed, 7 Jun 2000 11:36:59 -0400 From: "Hall, Ken (ECSS)" Subject: RE: Re: 3276 question (slightly OT) History time. (Expanding on what others have already replied.) I'm not sure what you mean by "don't have a control unit". All *REAL" 327x terminals connect via coax to a 3274/3174 type device. Those came in various flavors: BSC local, BSC Remote, X.25 (IIRC), SNA Local, and SNA Remote. The terminals (within the real IBM product line) are all the same, only the controller changes. To this day, you need at least one BSC Local 3x74 to connect consoles to a real S/390, except for a few CPU's that had built-in 3274-type devices, such as the 4381. Those required no external 3x74, but the functionality was still there, and you genned the devices as local BSC 3274-attached terminals. The 3x74 loads microcode that handles almost all management of the display. That way, the terminals could be made (relatively) cheaply. EXTREMELY stupid terminals (called CUT mode terminals) were completely managed by the CU, but later, more elaborate devices such as the 3290 plasma panel had some smarts of their own, and were known as DFT (Distributed Function Terminal) devices. The main difference from a support viewpoint was that they required additional microcode on the CU, and could do tricks like juggle multiple virtual terminal sessions over a single coax, appearing to the OS as discrete devices. An IRMA card, and similar cards from IBM and others, pretends to be either a CUT or DFT mode terminal on a control unit. That's all. ..... =======[ cut here ]======= There is the issue of the coax wire protocol. Check out this site-- the IBM networking site. There are some papers that discuss the coax protocol: http://www.networking.ibm.com/support/techtips.nsf/3174tips?OpenView Also, back in 1999 or 2000, there was a thread, again on the Hercules listserv IIRC, discussing terminals and controllers. Someone mentioned that not all coax adapter cards for the PC could be made to operate "in reverse"-- that is, make the PC mimic a control unit. I seem to recall that someone said that only one card on the market, using a specific chipset (was it the Madge card???) could do this, but I'm not sure about this. I'm trying to find this post in my archive but I'm not having luck at the moment.. Here is one snippit, but it isn't the article I was thinking of: > > The answer to the second question is yes it is possible to connect a crt > > to a PC Coax adapter card provided you have access to the microcode > > assembler, loader for the particular card you are using. That makes it > > not practical. > > The microcode on the card would have to be re-written to be a source > > device, not a destination device. And another: =======[ cut here ]======= Message: 3 Date: Wed, 15 Nov 2000 09:14:01 -0500 From: "Hall, Ken (ECSS)" Subject: RE: Re: 3270 Terminals + Laptop: How? (snip) > I think DFT vs. CUT is usually related to how the control unit is > connected--locally attached vs. SNA attached. Can anyone answer if > local terminal controlers run their terminals as DFT terminals? Or, > is attachment unrelated to the "mode" of the terminal? > DFT vs. CUT is strictly related to the protocol between the controller and the terminal. Any 3x74 with the proper microcode will support either kind of terminal. The 3290 plasma panel needed an additional "feature diskette" to download microcode. Other DFT devices had the support built-in. > SNA terminal emulators, such as DynaComm/Elite, NetSoft's (now part > of NetManage) products, Attachmate, etc. are DTF emulators. IBM's > PCOM has two versions-- a SNA DFT emulation, and the entry level CUT > mode emulation that must be used with a coax card. Most emulators that use either the IRMA or IBM 3270 card (IBM card better) will do either CUT or DFT. DFT is preferred because it reduces the load on the controller and adds support for multiple sessions, but on local non-SNA controllers, each session ties up a port. So a 3274-41C, for example, can do either 32 single-session terminals, or 7 four session terminals, plus four singles, or any combination that adds up to 32 total sessions. The only restriction is that on a 3274, port ZERO MUST be a CUT mode device to allow customizing of the controller. That means you can't have eight four session terminals on a single 3274, but you could have 7 fours, 1 three, and the single CUT device on port zero. In any case, this is all customized at the CONTROLLER LEVEL when the 3x74 is set up. Specific ports are assigned to specific (ranges of) device addresses. With remote-attach 3x74's it's a little different, but I think there's still a limit on concurrent sessions, and I believe the restriction on port zero was eliminated with the 3174. > Someone else mentioned finding a 3174 with ethernet. They also have > ones with token ring. I don't know this for a fact, but I STRONGLY > suspect that the ones with the net connections MUST be SNA-attached. > (the low level net protocol, on both token ring and ethernet, is > 802.2 LLC.) So, you will need a SNA protocol stack that knows how to > mimic a PU5 host... The other attachment method uses RS-232 --a > serial connection. SDLC is the only serial protocol I know these > things used. I'm not sure if SDLC was only used for SNA-attached > controllers. Anyone know what serial protocol was used to talk to > non-SNA-attached controllers? Maybe BiSync??? > Yes, bisync. You would need an FEP of some kind, or software to emulate one. I agree that I don't think they made bisync token-ring attached 3174's. The token-ring devices were intended to talk SNA to a 3745 with token ring adapter. SDLC is the line-level protocol that SNA uses, but the term "SNA" covers a lot more than that. Remember, you can't run an MVS console on a "remotely" attached 3x74. Consoles have to be on LOCAL (channel attached) BISYNC (NON-SNA) controllers like a 3274-41C. =======[ cut here ]======= I'm not sure if anyone has gotten something like this working, but here is my thoughts on it: 1) you need to hook the terminal to a 3x74 controller, or a controller emulator. 2) If you DIY, then you will need to grok the coax wire protocol. If you use a PC coax card then you don't need to worry about this. But then you are faced with the challenge of making your PC coax card a "master", not a "slave", and possibly you might need to dork around with the microcode, as alluded to in the post pasted in above. Good luck :-/. 3) Once you get past the terminal-to-controller issue, then you will need to deal with the linux-to-controller issue. If going against a real 3x74 controller this will be another big step. This is where the SNA stuff in your earlier post might come in handy...? 4) One other thing-- IBM terminals are page oriented and block mode. So you will need to write a terminal driver shim (using curses?) to fake scrolling, etc. 5) yet one more other thing-- many of the 3270 family terminals depended on microcode downloads from the controller when the terminal was first powered up. Without this, the terminal was effectively a boat anchor. This is terminal microcode, not coax card microcode. I think that the 3x74 controllers had terminal microcode on their floppy disks, so if you get a 3x74 controller off of ebay you might be set. If you go with a total DIY solution you might be faced with figuring out how to push terminal microcode down to your terminal... 6) It seems that you'd be dealing with at least one gnarly protocol no matter which way you go. If you get an ethernet or SDLC-attached 3174, then you need to use SNA-Linux, or the "gcom.com" products to make your linux box look like a host. If you get a bisync controller then you need to figure out how to talk bisync, then you need to figure out how to mimic the local attachment protocol. If you DIY your coax connection, then you have to figure out how to mimic a controller well enough to fool the terminal. The bright side here--at least you don't have to make your linux box look like a real host in this case..... For fun, check out these links too: http://www.networking.ibm.com/314/314prod.html http://www.mid-atlanticresearch.com/3174.shtm http://www.lordsnet.com/UsedIBM3174Summ.htm HTH, lcl ////////////////////////////////////////////////////////////////////////////// Newsgroups: comp.terminals References: <3CDD891C.F661539@yahoo.com> <25be338e69c48092@mayday.cix.co.uk> Message-ID: Organization: Harvard University, Cambridge, Massachusetts Date: 16 May 2002 15:28:45 GMT From: Chip Coldwell Subject: Re: IBM 3192 help Larry Liska wrote: > Very useful and interesting posting, thanks. I found the online documentation from Big Blue itself at http://publib.boulder.ibm.com/cgi-bin/bookmgr/LIBRARY Put "3174" into the search field and you get a pretty complete set of docs for this product line. These documents also discuss the differences between CUT and DFT terminals in the context of how they are connected to the establishment controllers. > All *REAL" 327x terminals connect via coax to a 3274/3174 type > device. Those came in various flavors: BSC local, BSC Remote, > X.25 (IIRC), SNA Local, and SNA Remote. A number of the 3174 controllers were designed to be "telco attached", which seems to mean that they ran SDLC on a synchronous serial port which could be connected via a CSU/DSU to a T1 line (for example) to connect a remote office to the mothership. This is probably the "SNA Remote" flavor mentioned in the above quote. HDLC, an SDLC derivative is still a very popular data link layer for T1 lines that connect non-SNA networks, and, for this reason, one which is included in the stock Linux kernels these days. As far as I can tell, these days the practical difference between SDLC and HDLC is nil; IBM calls it SDLC and everyone else calls it HDLC (much like Hewlett-Packard and HPIB, which everyone else calls GPIB). > 1) you need to hook the terminal to a 3x74 controller, or a controller > emulator. > > 2) If you DIY, then you will need to grok the coax wire protocol. I believe you're right about this. My notion was that if you have to grok a link-layer protocol, it's easier to use SDLC, for which there is already a Linux implementation, than to try to fathom the coax protocol (but thanks for the pointer to the specification at IBM), especially in light of the fact that the establishment controller might be downloading microcode to the terminal. > 3) Once you get past the terminal-to-controller issue, then you will need to > deal with the linux-to-controller issue. If going against a real 3x74 > controller this will be another big step. This is where the SNA stuff in > your earlier post might come in handy...? Indeed. I would bet (but I certainly don't know for sure) that most of the "telco attached" establishment controllers probably want to talk to FEPs (PU4s), not directly to the mainframe host (PU5). So whatever the controller is talking to on the Linux box would have to look like an FEP, i.e. it's going to have to run VTAM. I've been digging through the Linux-SNA docs to try to see if it implements VTAM, but so far it's not clear to me (I have essentially zero experience with SNA, so I'm on the steep part of the learning curve here.) > 4) One other thing-- IBM terminals are page oriented and block mode. > So you will need to write a terminal driver shim (using curses?) to fake > scrolling, etc. Definitely. But if the terminal is talking to a real instance of OS/390 running on Hercules, then presumably it will implement the 3270 data stream and it won't be necessary to write it myself. > 5) yet one more other thing-- many of the 3270 family terminals depended on > microcode downloads from the controller when the terminal was first > powered up. Without this, the terminal was effectively a boat anchor. I hadn't realized this, but it reinforces my opinion that the low impedance path to connecting one of these terminals to a Linux box is to get a "telco attached" establishment controller and talk to the Linux box using SDLC on a synchronous serial port instead of trying to directly connect to a coax adapter in the Linux box. > 6) It seems that you'd be dealing with at least one gnarly protocol no > matter which way you go. If you get an ethernet or SDLC-attached 3174, > then you need to use SNA-Linux, or the "gcom.com" products to make your > Linux box look like a host. Agreed, although there is some hope that you wouldn't have to do any of the protocol implementation yourself. Linux already does SDLC, and Linux-SNA may be able to do VTAM over SDLC on some synchronous serial port adapter that you can get a vendor supplied device driver for (e.g. this gcom stuff or a Cyclades PC300). Presumably, an OS/390 instance running on Hercules would talk the 3270 data stream, so block vs character mode issues won't come up as long as you're satisfied with only talking to Hercules from your IBM terminal. The things I'm still fuzzy on are: 1. Does Linux-SNA run VTAM? 2. Is there some way of getting Hercules to use a Linux-SNA network device? Or even the underlying SDLC network device (so that the emulated PU5 would run both VTAM and NCP)? The worst case scenario would be if the establishment controller insists on talking to an FEP, and a PU5 (i.e. Hercules) is not an acceptable substitute. In that case, the problem becomes one of writing an FEP emulator on Linux (re-implementing VTAM?), which involves, as you said, grokking a new gnarly protocol (one infamous for its complexity). I don't think buying a real FEP is a viable alternative since they want to be channel attached, so here we go again with some new piece of hardware for which we need a driver and protocol stack. > HTH, Helps alot. Thanks again. Chip -- Charles M. "Chip" Coldwell "Turn on, log in, tune out" ////////////////////////////////////////////////////////////////////////////// Newsgroups: comp.terminals References: <3CDD891C.F661539@yahoo.com> <25be338e69c48092@mayday.cix.co.uk> Message-ID: <8mSE8.3675$D41.329394@bgtnsc05-news.ops.worldnet.att.net> Date: Thu, 16 May 2002 17:53:08 GMT From: Larry Liska Subject: Re: IBM 3192 help Chip Coldwell wrote in message news:ac0j7d$2pj$1@news.fas.harvard.edu... > I found the online documentation from Big Blue itself at > http://publib.boulder.ibm.com/cgi-bin/bookmgr/LIBRARY > Put "3174" into the search field and you get a pretty complete set of > docs for this product line. Cool, thanks for the link, I'll check it out. > As far as I can tell, these days the > practical difference between SDLC and HDLC is nil; IBM calls it SDLC > and everyone else calls it HDLC (much like Hewlett-Packard and HPIB, > which everyone else calls GPIB). Wasn't the name HDLC given by the ISO committee? Maybe IBM had a trademark on "SDLC"? (who knows...?) > Indeed. I would bet (but I certainly don't know for sure) that most > of the "telco attached" establishment controllers probably want to > talk to FEPs (PU4s), not directly to the mainframe host (PU5). So > whatever the controller is talking to on the Linux box would have to > look like an FEP, i.e. it's going to have to run VTAM. I've been > digging through the Linux-SNA docs to try to see if it implements > VTAM, but so far it's not clear to me (I have essentially zero > experience with SNA, so I'm on the steep part of the learning curve > here.) VTAM was an application, at least on VM/CMS. I think it is a separate app/job on MVS systems too. IIRC VTAM is what implements the PU5. I have heard the PU4 layer described as the NCP (network control program), and I think this is what ran in the FEP. My experience was solely on a 9370 "baby" box which didn't have a FEP-- I guess the FEP portion was self contained or there was a special version of VTAM for these boxen or something.... As for whether someone has implemented a VTAM lookalike for Linux I'm not sure. If you find that you need to write a VTAM workalike, then here are some interesting manuals to search for on IBM's docs site: 1) SNA Formats (GA27-3136) Discusses the SNA frames for various conversations, typical sense codes, etc. If you know someone with the Microsoft SNA Server (now called Host Integration Server), look at the CD-- Microsoft used to include a PDF of this manual on the CD. 2) 3274 Control Unit Description and Programmer's Guide (GA23-0061) ** This may be out of print by now. There is a version of this for the 3174 as well (GA23-0218) The 3274 manual had a chapter about the SNA connection and conversation between the 3274 and the PU5--I'm pretty sure the 3174 version does as well. 3) Sessions Between Logical Units (GC20-1868-2) This may or may not be helpful. It describes what happens when LUs talk to each other. The interesting part here will be the description of what heappens when the terminal talks to VTAM's SSCP. > > > 4) One other thing-- IBM terminals are page oriented and block mode. So > > you will need to write a terminal driver shim (using curses?) to fake > > scrolling, etc. > > Definitely. But if the terminal is talking to a real instance of > OS/390 running on Hercules, then presumably it will implement the 3270 > data stream and it won't be necessary to write it myself. True, true. I guess I was thinking along the lines of "hey, I want to be able to do console stuff and play NetHack on my linux box using this groovy 3270 terminal." I'm out of my area of expertise here. Folks have written CTC emulators for hercules. I'm wondering if that will help get the controller traffic to/from your OS running on herc? > it reinforces my opinion that the low > impedance path to connecting one of these terminals to a Linux box is > to get a "telco attached" establishment controller and talk to the > Linux box using SDLC on a synchronous serial port instead of trying to > directly connect to a coax adapter in the Linux box. I concur. 5 years ago this would have been a tougher decision, but since you can find controllers on eBay for cheap these days, this is the only way to go... > The things I'm still fuzzy on are: > > 1. Does Linux-SNA run VTAM? Sorry, don't know. I looked at Linux-SNA 3 or 4 years ago but I haven't kept up... > 2. Is there some way of getting Hercules to use a Linux-SNA network > device? Or even the underlying SDLC network device (so that the > emulated PU5 would run both VTAM and NCP)? Again, I don't know. I mentioned the CTC (channel-to-channel) project above. Folks have been using this to get TCP/IP connectivity into their OS/390 systems running on Hercules. If the 3174 eventually ends up talking to a channel, then perhaps from the host side at least you could use the CTC. On the linux side you would then need to pipe between the CTC and the HDLC driver. Will this work? I don't know. (I'm waaaay out of my league here.) > The worst case scenario would be if the establishment controller > insists on talking to an FEP, and a PU5 (i.e. Hercules) is not an > acceptable substitute. The PU5 is not the host itself, but the VTAM (or other network control app) running on the host. Of course if you have OS/390 then you have VTAM most likely... Wait-- please clarify something: are you wanting to use your 3270 terminal to talk to OS/390 on Hercules (as you alluded to above), or do you want to be able to use the 3270 as a generic terminal on your linux box? If you are trying to do OS/390 stuff, then there is another thing to consider. I am not sure but I think the newer generation of control unit--the one that replaced the 3174-- could also talk AnyNet in addition to SNA/SDLC. (Not to belabor the point, but I don't know this for sure...) Anynet could go over TCP/IP, couldn't it? If so then perhaps you could use the existing TCP/IP-over-CTC emulator-to-hercules, and use this to talk AnyNet to your new generation controler. This begs the question-- can one find a newer controller on eBay at geek-affordable prices? Anyway, just a thought. It looks like you have a groovy project ahead of you. If you make progress on this please post updates. I know that the topic of using 3270s on linux boxen has come up periodically on this newsgroup, and it was a hot hot topic on the hercules listserv back in 2000. If you can get this to work, a lot of folks would be interested. Good luck, and have fun!!! lcl ////////////////////////////////////////////////////////////////////////////// Newsgroups: comp.terminals References: <3CDD891C.F661539@yahoo.com> <25be338e69c48092@mayday.cix.co.uk> <8mSE8.3675$D41.329394@bgtnsc05-news.ops.worldnet.att.net> Message-ID: Organization: Harvard University, Cambridge, Massachusetts Date: 16 May 2002 19:29:17 GMT From: Chip Coldwell Subject: Re: IBM 3192 help Larry Liska wrote: > Chip Coldwell wrote in message > news:ac0j7d$2pj$1@news.fas.harvard.edu... >> >> As far as I can tell, these days the >> practical difference between SDLC and HDLC is nil; IBM calls it SDLC >> and everyone else calls it HDLC (much like Hewlett-Packard and HPIB, >> which everyone else calls GPIB). > > Wasn't the name HDLC given by the ISO committee? Maybe IBM had a trademark > on "SDLC"? (who knows...?) Yeah, it's probably the same old story: open standard implementation of formerly company-proprietary standard gets a new name. HPIB and GPIB is the same story. > VTAM was an application, at least on VM/CMS. I think it is a separate > app/job on MVS systems too. IIRC VTAM is what implements the PU5. I have > heard the PU4 layer described as the NCP (network control program), and I > think this is what ran in the FEP. Yup, it looks like I was confusing VTAM and NCP. So what I meant to say is that probably the telco attached establishment controllers want to talk to an FEP, or more precisely the NCP running on an FEP. So the (corrected) picture is 3270 terminal --coax-- 3174 controller --SDLC-- ??? and my guess is that ??? had better look like an FEP/NCP to the 3174. Here are some possibilities for ??? 1. Linux-SNA talking through synchronous serial port 2. NCP on Hercules talking through Linux-SNA talking through ssp 3. NCP on Hercules talking directly through ssp 4. VTAM Hercules talking to something I write (NCP) talking through ssp my "worst case scenario" is number 4, although that's the one which affords the greatest opporunity to learn something about SNA ;->. > If you find that you need to write a VTAM workalike, then here are some > interesting manuals to search for on IBM's docs site: Thanks for the pointers. > Again, I don't know. I mentioned the CTC (channel-to-channel) project > above. Folks have been using this to get TCP/IP connectivity into their > OS/390 systems running on Hercules. If the 3174 eventually ends up talking > to a channel, then perhaps from the host side at least you could use the > CTC. On the linux side you would then need to pipe between the CTC and the > HDLC driver. Will this work? I don't know. (I'm waaaay out of my league > here.) Well, there are flavors of 3174 that talk Token Ring, and certainly Linux can talk Token Ring, so if CTC can run on Token Ring that might work in place of HDLC. > Wait-- please clarify something: are you wanting to use your 3270 terminal > to talk to OS/390 on Hercules (as you alluded to above), or do you want to > be able to use the 3270 as a generic terminal on your linux box? The former; I don't think it would work very well as a generic terminal because of its preference for block-mode transactions. > If you are trying to do OS/390 stuff, then there is another thing to > consider. I am not sure but I think the newer generation of control > unit--the one that replaced the 3174-- could also talk AnyNet in addition to > SNA/SDLC. (Not to belabor the point, but I don't know this for sure...) > Anynet could go over TCP/IP, couldn't it? If so then perhaps you could use > the existing TCP/IP-over-CTC emulator-to-hercules, and use this to talk > AnyNet to your new generation controler. This begs the question-- can one > find a newer controller on eBay at geek-affordable prices? I dunno ... new 3274s ain't cheap. > It looks like you have a groovy project ahead of you. If you make progress > on this please post updates. I know that the topic of using 3270s on linux > boxen has come up periodically on this newsgroup, and it was a hot hot topic > on the hercules listserv back in 2000. If you can get this to work, a lot > of folks would be interested. Just wait till I figure out how to hang a punch card reader on my Linux box .... Chip -- Charles M. "Chip" Coldwell "Turn on, log in, tune out" ////////////////////////////////////////////////////////////////////////////// Of possible interest: Hercules (IBM mainframe) emulator http://www.conmicro.cx/hercules/ also see discussion group http://groups.yahoo.com/group/hercules-390/ (see Links section for other Hercules groups). ////////////////////////////////////////////////////////////////////////////// Newsgroups: comp.terminals References: <1281bef.0301190305.358b62f@posting.google.com> Message-ID: <93b98db41ef90039@mayday.cix.co.uk> Organization: ntl Cablemodem News Service Date: Mon, 20 Jan 2003 07:43:26 +0000 From: Robert de Bath Subject: Re: IBM 3476 Infowindow On 19 Jan 2003, William Fargo wrote: > > Hi, > > I just want to know if it's possible to plug a IBM 3476 terminal on > a PC to have a simple terminal under linux. > > I don't know which card or special device I must buy, or which > terminal emulator under linux I must use. If it exists. "The InfoWindow series 3476 is a 14-inch, flat screen, smudge-resistant, monochrome monitor that attaches to the IBM System/36, IBM System/38 and IBM AS400 processors. The IBM 3476 attaches remotely to these processors via the appropriate IBM 5294 or 5394 Remote Control Unit." This means it's a block-mode terminal that connects using a "Twinax" connection. Probility of getting it working without special hardware = 0. Probility of using it as a generic unix terminal even with special hardware and software = 0. There is a project somewhere to connect SNA and Twinax devices to Linux but I don't remember any URL. > If it's impossible to use this terminal for linux, I can use the > keyboard or the monitor for a PC (with special adaptors of course) !. Using a non-multisync monitor on a PC is a PITA and a keyboard adapter, if such thing exists, will cost a lot more than a PC keyboard. -- Rob. (Robert de Bath ) Google Homepage: http://www.google.com/search?btnI&q=Robert+de+Bath ////////////////////////////////////////////////////////////////////////////// Newsgroups: comp.terminals Message-ID: <9fea29e6.0301231210.30143fdb@posting.google.com> References: Organization: Forvus Research Date: 23 Jan 2003 12:10:17 -0800 From: Jeff Schultz Subject: Re: 3472 help mike_booher@secura.net (Mike) wrote in message news:... > > Looking for feedback... > We recently had a 3472 alternate mainframe console/terminal go down in > a "blaze of glory" (smelled pretty bad). I would like to replace the > two alternate consoles/terminals -(models 3472-4 control module and > monitor P/N's 16f1200/39f7573 InfoWindow) - Is there an option to use > a PC based console with a PCI card that can emulate these types of > scrolling consoles? I have found 3270 coax adapters that are PCI > cards but I don't think that is the going to behave the same as the > 3472's. Am I correct with that assumption? > > More info / Summary of query: Three (1 master & 2 alternate 3472) > terminals connected via coax to 3174 controller. Would like to > replace with a PC based solution that would allow for a larger display > and eliminate the dependancy on refurbed terminals that only last a > year and are hard to read cause they go fuzzy. > Any help with this would be very helpful. Thanks. Mike, Forvus Research builds and sell a PCI 3270 coax adapter that will replace your 3472's. It comes with the emulator software, but also works with emulators from other's such as IBM and Rumba. We also have versions for the new Low Profile PCI slots, and even a USB version. Check out our web site at http://www.forvus.com/ or call us at +1 919/954-0063. Jeff Schultz Forvus Research Inc. ////////////////////////////////////////////////////////////////////////////// Newsgroups: comp.unix.solaris References: Message-ID: Date: Tue, 23 Sep 2003 17:15:17 -0500 From: Alex Moore Subject: Re: not tn3270, but 3278 ? (SUMMARY!) On Tue, 23 Sep 2003 17:20:07 -0400, Dragan Cvetkovic wrote: ... > From online man page for x3270 > (http://www.geocities.com/SiliconValley/Peaks/7814/man/x3270-man.html): > > x3270 provides limited support for 5250 (AS/400) emulation. This is not > native 5250 terminal emulation, but rather for use with hosts which > allow a 3270 terminal to emulate a 5250. A 5250 keymap is defined, > which maps the keyboard function keys to the 5250 emulators PA/PF key > sequences, as well as redefining the Return key to use the special > FieldExit action. The command line for 5250 support is: > > x3270 -model 2 -keymap 5250 > > > HTH, Dragan Thanks. I will download the source and see how it works. Alex Moore ////////////////////////////////////////////////////////////////////////////// Newsgroups: comp.sys.ibm.as400.misc References: Message-ID: <37e3643f.0402011845.78217254@posting.google.com> Date: Sun, 1 Feb 2004 18:45:48 -0800 From: marty Subject: Re: Webfacing all or nothing? Didjit wrote in message news:... > > Question. My shop is in the process of rewriting one of our main apps. > Unfortunately, we are doing this in phases, so we cannot just rip the > app up all at once. > > Question, is it possible to Webface only certain screens > and if the screen has not been webfaced, then fall back into 5250? > We want this to be seamless to the users, IE one client. > > I know 5250 can be streamed to a browser, so I'm guessing Websphere > would need to be configured to know what to serve up to the user > (webfaced screen, or 5250 screen). Is this possible? > > Thanks > Chris The scenerio that you are talking about is more characteristic of a Windows Screen scaper such as Seaguls's GUI. Basically if you choose not to Webface a 5250 screen in IBM's Webfacing you will simply get a not-too-friendly error message in the web browser when you attempt to access this screen. If you must use Webfacing, then roll out the Webfaced screens at the same time as the new 5250 screens. Note that webfacing is a HTML screen-scraping technology. If you don't have a functional RPG (Cobol, ect) program and a DDS defined display file, then webfacing is useless. If you don't start messing too much with generated code, basic webfacing conversion is very fast. Once the product is configured and staff is trained, it is possible to convert literally dozens of screen an hour. There are other products that claim "on the fly" conversion. Maybe you should look into these. Note that whatever product you use, there is a twenty year gap between 5250 and HTML technologies. 5250 screens generally do not map very well to web pages. The results are often awkward or confusing to both 5250 and web users. regards, marty //////////////////////////////////////////////////////////////////////////////