Computer History Nostalgia.

Post #417,123 by static 2/13/17 3:32:10 PM 2/13/17 3:32:10 PM Reply	Computer History Nostalgia. ESR has put together a piece about what computing knowledge was once common but has now faded into obscurity. http://www.catb.org/esr/faqs/things-every-hacker-once-knew/ Wade.
Post #417,124 by Andrew Grygus 2/13/17 4:34:25 PM 2/13/17 4:34:25 PM Reply	I still deal with a lot of that stuff. Medical Testing Machines, for instance, are just starting to convert from RS232 to Ethernet. They have a long life, so RS232 will still be here for some time. I still keep all my tools and stocks of connectors current. Most of my Medical clients have transitioned from terminals to terminal emulation on PCs and network cabling, but some still have terminals. I use the same WECO connections over Cat5e for both serial and network, which makes things quite flexible. I watched several terminal emulation developers with good products go out of business because they refused to support the Wyse50 and Wise60 terminal protocols all small businesses used - and many still use. DEC was all they'd support, and that faded rather suddenly. I just configured a Linux server today with a Digi portserver providing 8 serial ports over an Ethernet connection. Today, the most common use for those serial ports is for scales, printers and test equipment, not terminals. I still have an external modem, through which I send faxes (remember those?). I often deal with ASCII and HEX, but never had much to do with Octal except a few unusual cases (I don't remember what they were). I never did have much to do with minicomputers - except for an IBM, when I absolutely infuriated the IT guy in charge when I installed terminal emulation (over coax in those day) in a couple of PCs while he was on vacation.
Post #417,125 by Andrew Grygus 2/13/17 5:20:13 PM 2/13/17 5:20:13 PM Reply	Damn! Responding to this post brought bad luck. I just got a call from a client I'd hoped never to hear from again - pleading for me to come down and configure a modem for remote printing from an SCO Unix system.
Post #417,126 by malraux 2/13/17 6:43:37 PM 2/13/17 6:43:37 PM Reply	:-) LRPD: "Fun is fun to have." Regards, -scott Welcome to Rivendell, Mr. Anderson.
Post #417,130 by Andrew Grygus 2/13/17 10:00:39 PM 2/13/17 10:00:40 PM Reply	Just Lovely The new modem seems configured the same as the dead one, but it's not sending reports. To go farther I need cooperation from the software vendor. Software vendor says, "They are no longer our client. They stopped making support payments almost 3 years ago".
Post #417,131 by boxley 2/13/17 10:17:43 PM 2/13/17 10:17:43 PM Reply	you have a copy of littlebiglan around? used to be useful in redirecting serial IO to files for diagnostics. It would redirect the interupt "Science is the belief in the ignorance of the experts" – Richard Feynman
Post #417,133 by Andrew Grygus 2/13/17 11:41:20 PM 2/13/17 11:41:20 PM Reply	I suspect it isn't that complicated. I suspect the software's not sending to the port the modem is on, but only the software vendor can really tell me as they have the port list. Believe it or don't, this rig has 49 serial ports. My port list is probably obsolete, I haven't dealt with this outfit for several years. Only the software vendor can change the ports the software uses. I think a bit of extortion is in order here.
Post #417,145 by drook 2/14/17 8:11:55 AM 2/14/17 8:11:55 AM Reply	If your port list is out of date, they must have moved something Can you just move the device from port to port til you find the one that works? -- Drew
Post #417,149 by Andrew Grygus 2/14/17 12:23:57 PM 2/14/17 12:23:57 PM Reply	Even if I knew how to make the software send . . . . . with 48 possible ports that would take a long time, and if there was actually a problem with the software, it would be inconclusive. No, I need the software vendor to tell me what port they are sending to, and to trgger a send - or we may need to move it to a different port, which only they can do. I do local hardware support for a couple of LIS (Laboratory Information System) vendors. They have been entirely uninterested in me knowing anything about the software, and I'm happy with that. The client is just going to have to pony up for a year's support.
Post #417,161 by crazy 2/14/17 6:25:10 PM 2/14/17 6:25:10 PM Reply	Symlink them
Post #417,162 by Andrew Grygus 2/14/17 7:50:34 PM 2/14/17 7:50:34 PM Reply	My relationship with the software companies I work with . . . . . has bounds I have no interest in stepping beyond. I work coordinating with them and keep my hands off their stuff. It is up to them to tell me what port to use, up to me to make the hardware work as they need. I'm especially not going to step out of bounds for a client who is chronically bad pay. If the client wants this fixed, she's going to have to come to terms with the software publisher. The software people have expressed willingness to work with me on this issue, but they deserve to be paid for their efforts. By extortion is the only way this client pays.
Post #417,166 by crazy 2/15/17 2:08:07 AM 2/15/17 2:08:07 AM Reply	Sigh Get cash to walk through the door. Get a check check in advance to be cashed on successful resolution. Is the system owned by the end user? Is the software legal?
Post #417,167 by static 2/15/17 4:15:13 AM 2/15/17 4:15:14 AM Reply	An admirable position. And clearly it works for you. Especially for known bad payers. :-) I can imagine someone is having a shouting match with their accountant. Wade.
Post #417,127 by malraux 2/13/17 6:44:59 PM 2/13/17 6:44:59 PM Reply	I used to deal with that stuff, thankfully no more. I remember once we had to solder a special pin-flipped version of an RS232 connector to talk to some weird LAN switches that needed to have spanning tree turned off. We carted an ancient terminal around to the various wiring closets to do the deed. Regards, -scott Welcome to Rivendell, Mr. Anderson.
Post #417,136 by static 2/14/17 3:32:27 AM 2/14/17 3:32:27 AM Reply	My favourite RS-232 stories. I had to make a cable to a HP Paintjet that some nong had purchased with a serial port instead of a parallel port. It needed all the signals because it couldn't buffer much data and had to tell the PC when to stop sending. After months of attempts with a simple cable, and quantities of wasted paper, I found the manual and followed the instructions. Worked perfectly after that. And then there were the cases where you two serial terminal devices needed to see each others' control signals. The idiot comms guys kept giving me cables where the control lines were just reflected back at each end. This did not make for reliable communication! Again, I made a proper null modem cable over the objections of the comms guys and things were a lot more reliable after that! Wade.
Post #417,139 by Andrew Grygus 2/14/17 4:33:30 AM 2/14/17 4:33:30 AM Reply	And then there was the guy . . . . . who claimed the RS in RS232 stood for Radio Shack.
Post #417,140 by hnick 2/14/17 5:57:35 AM 2/14/17 5:57:35 AM Reply	You knew him too?? Guy actually worked for Radio Shack (sort of). Summer of 1986, I think, company I worked for did some protocol converter work for Tandy. I had to go to Texas to show what we had and see if it needed any tweaks. I was talking to one of their big shots and made an off handed comment like "20 years and it's still a recommended standard?" and he came back with "Oh no! That RS stands for Radio Shack!" He was serious as a heart attack. "Religion, n. A daughter of Hope and Fear, explaining to Ignorance the nature of the Unknowable." ~ AMBROSE BIERCE (1842-1914)
Post #417,132 by a6l6e6x 2/13/17 11:29:16 PM 2/13/17 11:29:16 PM Reply	God, that makes me feel old! :) As a student in 1959/60, I first programmed the TX-0 (one of a kind computer, the first with transistors and ferrite core memory made at MIT Lincoln Labs). That machine was the progenitor of DEC's PDP-1 which I got to play with as an MIT employee in 1961. They were 18 bit machines, also using 6-bit codes for characters. It made sense to use octal, i.e. 3 bit groupings. This code written was for an assembler. I knew the group of guys that hung around the PDP-1 and coined the term "hack". Many eventually became DEC employees. As a student I got to use (actually submit card decks) for the IBM 709, a vacuum tube machine. The code in this case was written in Fortran II. My first job as an MIT employee was to re-design some equipment we had connected to the IBM 709 to work with the IBM 7090 which was a transistorized version of the IBM 709. These were 36-bit machines and again used 6 bit codes for characters. Again it made sense to use octal. It wasn't until I started to work for IBM in 1965 on the IBM 360 (model 67) that I had to think in bytes and hexadecimal. Anyway, this predates the cited nostalgia! :) Alex "There is a cult of ignorance in the United States, and there has always been. The strain of anti-intellectualism has been a constant thread winding its way through our political and cultural life, nurtured by the false notion that democracy means that "my ignorance is just as good as your knowledge." -- Isaac Asimov
Post #417,137 by static 2/14/17 3:34:39 AM 2/14/17 3:34:39 AM Reply	Octal and early processors. I kind of skipped over octal when learning computer hardware and went straight to hexadecimal, largely because home microcomputers and their microprocessors were documented in hex. Many years after I learnt how to hand assemble for the Z80, I discovered the op-code layout made a lot more sense when written in octal. Wade.

Welcome to IWETHEY!