From: "Van Snyder" To: "Stan Paddock" Subject: Re: I'm Van Date: Monday, August 20, 2007 11:39 AM On Tue, 2007-08-14 at 18:52 -0700, Stan Paddock wrote: > Hi Van, > > I now wish I had kept some of the decks I had when I worked for > IBM in 1968. How was I to know I would want them some 40 years later? > > I have found some decks on-line but at the current time, my PC is un- > able to punch cards. > > I am going to build an interface to fix that problem. I think it is > easier to interface to a keypunch than our IBM 1401. (I have done that > before) Mike Cheponis and Ed Thelen are thinking of the same project. > I am also considering writing a simple assembler to run on the PC to > conserve punched cards. I have an Autocoder assembler that runs on a PC under Windows or Linux. > If I feel real crazy, I will write an IBM 1401 simulator for the pc. Bob Supnik has written a family of simulators called SimH (Simulation History). See http://simh.trailing-edge.com. Ron Mak has written a graphical wrapper (in Windows) to my Autocoder and Simh. It's called ROPE (Ron's Own Programming Environment). I can send you Linux executable, or Fortran 90 source code (compiles with gnu g95, gnu gfortran, NAG f95, Lahey/Fujitsu lf95, Intel ifort, probably others as well). > By now you must realize that I am retired and have time on my hands. > > I remember a simple loader deck where you wrote one instruction per > card for simple programs. By chance is this in your library? This is one of the formats used by the diagnostic decks. The first card consists of ",0080121001". Additional cards other than the last consist of "Lxxxyyy1001" where xxx and yyy are addresses on the card and the area to be loaded, respectively. Text to be loaded starts in cc 12. In the diagnostics, xxx is always 039, but that's an unimportant detail. The diagnostics also have comments starting in cc 40, ID in 73-77, blank in 78, and sequence in 79-80. The last card has "N000000Bxxx" or "/xxxyyy" in 1-11 or 1-7. Later diagnostics use a condensed format, with the first card ",008015,022029,033033N 1001" and subsequent cards other than the last beginning "Laaabbb,cccddd,eeefff,ggghhh1001" with text to be loaded in 33-72. The last card begins "/xxx080". ID and sequence are the same as in the one-per-card format. SPS uses yet another loader format. This has text to be loaded in 1-38, a 7-character load instruction in 56, 4-character instructions in 63, 67, 71, 39, 43, 47 and 51, and "1" in 55 if the instructions in 39 etc are used. The last card has /xxx080 in column 56. The three-card clear-core-and-bootstrap is ,008015,022026,030034,041,045,053,0570731026 L072116,110106,105117B101/I99,027A074028)027B0010270B026/0991,001/001117I0 ,008015,022029,036039,043047/039036 ,051,055,056,063N,067071,075,0011056 Autocoder uses yet another format. This is described on page 51 of http://www.bitsavers.org/pdf/ibm/14xx/C24-3319-0_1401_tapeAutocod.pdf. The tape format is described on the same page. I've attached a file of bootstrap loaders for Autocoder-format decks. Mike Cheponis wrote a loader that works like one he remembers from the 1440. See http://www.ed- thelen.org/1401Project/MC-1401_3_card_Loader.html. Instead of using set-word-mark instructions, it uses the same word-mark character as is used on tape (0-5-8). My comments on it are attached in Cheponis.txt > Is your library on-line? Unfortunately, no, but I'm happy to send you anything you want from it. Al Kossow has a copy of my 1401 web as of the time my employer prohibited me from publishing web pages (and I haven't bothered to sign up for another one from somebody else) at http://www.bitsavers.org/1401/index.html. It's not completely up to date. Someday I need to send in my application to the Procrastinators' Club. > How many IBM 1401s do you think are still in operational condition? Haus zur Geschichte der IBM Datenverarbeitung (House of the history of IBM data processing) in Sindelfingen, Germany (just south of Stuttgart) has an operational 1401 system. Their 1401 has 16k memory, five 729 tape drives, a 1407 console, a 1311 disk drive, 1402 and 1403. They also have lots of other operational IBM equipment, including a 650. Computer History Museum (cool address: 1401 Shoreline Blvd., Sunnyvale, CA) has a semi-operational 1401. It has 16k memory, five tape drives (some 729-II and some 729-IV - only one of the tape drives is presently operational), 1402 and 1403. No console or disk drive. It is currently under restoration and doesn't work entirely reliably. It was purchased from Arnold Schweinsberg, who had stored it in a garage in northern Germany for 18 years. Paul Pierce has a 1401 that he says worked the last time he plugged it in, but he hasn't run it for several years. He has a large collection of other computers, including two 709's. The Australian Computer Museum Society has a 1401, but I don't think it's in operational condition. Contact John Deane . I understand there is a non-operational one sitting in the lobby of the New Zealand National Railway offices. Best regards, Van > Stan Paddock > San Jose, CA > Van Snyder wrote: > Stan: > > I'm Van. I've been working with CHM, HzG, SPC, and anybody > else who'se > interested to preserve 1401 software. The biggest item still > on my > unsatisfied wish list is COBOL, preferably a source tape, but > an > operational tape would be welcome. > > Best regards, > Van Snyder > > PS, there's also a Van Gardner working on the 1401 at CHM. > >